Tuesday, May 24, 2011

Does Intelligent Design Make Testable Predictions? Part 2

Could ID predict this scenario? If it happened would the explanatory filter fail to detect ID?

In the first part of this series I expressed doubt about Intelligent Design’s ability to make hard predictions; after all, ID covers a large range of opinions from Sir John Polkinghorne’s fruitful evolutionism to Beyond our Ken Ham’s mature creation, a creation created “just like that” 6000 years ago. It seems that ID can be used to underwrite so many creation scenarios that its predictive content is seriously compromised. I say this as an ID creationist myself, so I am not saying it from an anti-ID stance. In fact I say it because I am anxious to make all due allowance for the problems of prediction that arise naturally from an ontology that starts with the assumption of some complex presumably self-explaining entity whose nature is very alien to human kind; hardly a sound basis for successful prediction - and we thought predicting human behaviour was difficult enough. So perhaps the lack of predictability in ID is less a bug than it is an expected feature of the ontology it posits. Let’s face it, ID is going to struggle to be hard science.

So where does this leave the predictions in Uncommon Descent’s post here? Well, ID does become a harder science if, as seems to be the case in the community UD represents, ID is de-facto anti-evolutionism. A de-facto anti-evolutionism is liable to contradict certain evolutionary scenarios and will presumably impact expected biological and paleontological observations. But even if we assume an anti-evolutionary position that still leaves us with an open ended range of ID candidates running from Old Earth Creationism to fundamentalist views like that of Beyond our Ken Ham. Anti-evolutionists, then, will claim that they know what hasn’t happened, but they have no consensus on what has happened and consequently they are going to be vague about just what they expect the fossil record to reveal; all they know is that it’s not going to reveal evolution, at least as currently understood. Nevertheless, when ID slips over into anti-evolutionism it at least appears to be saying something with a little more content than “God did it – somehow”.

Anyway, without further ado here are the first four predictions in UD’s list, each followed by my own comment. I was hoping to get through all predictions in one post, but as there are a total of 14 predictions it looks as though it’s going to be a long hot summer.

ID Prediction 01: ID predicts that the Universe had a beginning.

My Comment: I can’t see why this follows. In his book “The Intelligent Universe” Sir Fred Hoyle moots the idea that life is intelligently designed and yet as a lifelong supporter of Steady State Theory he obviously didn’t think his version of ID implied a universe with a beginning. Neither do I see this as a prediction of ID in its most abstracted form. However, I agree that it is a prediction of mainstream Christian Theism; but Christian theism isn’t ID in its most abstracted form.

ID Prediction 02: ID predicts an increase (and not a decrease), as science progresses, in the number of finely-tuned parameters pertinent to the laws and constants of physics.

My Comment: I have a feeling that this prediction arises out of a misconception I have occasionally seen on UD that physical laws can’t carry information: The naive argument runs as follows: Laws equate to “necessity” and therefore the patterns they generate have a probability of 1. A probability of 1 entails zero information. The upshot is that many anti-evolutionists cannot see how the patterns implied by physical laws can be a receptacle of information. The only kind of information they recognize, then, are configurations that appear to have been patched in “by hand” rather than generated by physical laws. Moreover, Dembski’s explanatory filter, although valid in circumstances within the cosmic context, has the unfortunate side effect of making it look as though Law and Disorder causes obviate the need for Intelligent Design; the filter doesn’t readily raise the question of the Intelligent origin of Law and Disorder but instead invokes a regress whereby Law and Disorder generates Law and Disorder.

All in all this amounts to a double whammy against the efficacy of physical laws, for not only do they appear to be information-less, but the all important explanatory filter fails to explicitly acknowledge that laws themselves may be intelligently contrived. Thus, on UD there is tendency for only “hand patched” information to register on their radar as having ID origins. Hence in UD’s view the greater the level of overt hand patching the firmer the case for ID. Anti-evolutionists, then, expect more hand patched configuring rather than less. In my view high levels of “hand patching” does not necessarily follow from ID. If it really is possible to generate life using law and disorder, then given the level of Intelligence envisaged by ID creationists it follows that that intelligence is well able choose an elegant set of laws with a minimum of adjustable variables. Whether elegant laws can actually generate information I have considered on the following web pages:


However, I must add a disclaimer: None of this necessarily means that Physical Law is the means by which  a creative intelligence has introduced information into our world. The UDers may be right and life isn’t a product of law and disorder, but has, in fact, been directly patched in by Divine Intelligence in the form of a second creative dispensation – this is why I like to keep tabs on UD in case they come up with something; they just might. Unlike some people who call them “IDiots” I do have respect for their abilities.

ID prediction 03: ID predicts the presence of specified complexity in living systems.

My Comment: For something to have “specified complexity” two conditions must be met. Firstly, the object must be complex in configuration – this rules out highly ordered configurations like crystalline periodicities. Secondly, the object must be highly improbable - this rules out highly disordered configurations because high disorder, by virtue of its overwhelming representation in the space of possibilities, is, assuming equal a-priori probabilities, very probable. Hence objects with specified complexity are objects that are both complex and at the same time members of a very improbable class. In absolute terms it is clear that organic structures fulfill both these conditions – they are complex and yet ordered enough to be very far removed from absolute disorder. So in one sense the above “prediction” is less a prediction than it is a manifestly obvious feature of life. I must emphasise, however, that I have just said “in absolute terms” – that is, a la Fred Hoyle’s Junkyard Jumbo “explosion” where it is assumed that each and every possible configuration of elements is equally possible thus entailing an absolutely minute probability of a Jumbo Jet forming during an explosion.

But if as evolutionists are implying life is a product of the “self organization” implicit in the laws of physics, then in terms of conditional probabilities the probability of life is much greater than it would be in absolute terms. The probability of life conditioned on the Laws of Physics can be represented as Prob(Life|Laws of Physics). Evolutionists are effectively telling us that the value of Prob(Life|Laws of Physics) is great enough for there to be a realistic chance of the formation of life in the history of the universe. Thus, the “specified complexity” of life, given the Laws of Physics, is correspondingly reduced. Now, as far as the UD community is concerned, specified complexity is used to detect Intelligent Design; it is used in the explanatory filter to decide whether a configuration is a product of intelligence or law and disorder. The trouble is, as I have pointed out above, the explanatory filter does not explicitly acknowledge the present of intelligence if a configuration can be explained, with a reasonable probability, as the outcome of law and disorder. Hence, if Prob(Life|Laws of Physics) is sufficiently high UDers are liable to lose sight of a cosmic intelligence behind the universe. Consequently, they favour the view that Prob(Life|Laws of Physics) is extremely low, which is the mathematical way of expressing their strong anti-evolutionist line; that is, they maintain that the laws of physics are not a sufficient condition for life thus entailing the high specified complexity of organic structures even given the laws of physics.

In order to convert the conditional probability Prob(Life|Laws of Physics) into the absolute probability Prob(Life), we need to know Prob(Laws of Physics). Trouble is, we haven’t got a clue what this value is and in fact we have little idea how to calculate it, if indeed it is an intelligible concept. In spite of some speculations about how physical laws might change across a multiverse no quantification is currently possible because there is no known Law and Disorder regime that generates law and disorder regimes. However, we can perhaps get an inkling of an answer to this question if we remind ourselves of an important mathematical abstraction: Namely, of all possible platonic patterns of behavior the ordered behavior described by the laws of physics is clearly an extreme rarity. Hence on this basis, and assuming equal a-priori probabilities, Prob(Laws of Physics) looks to be very low. This implies that even if Prob(Life|Laws of Physics) is relatively high, then the absolute probability of life, Prob(Life), is very low. So as far as specified complexity is concerned life has a very low absolute probability whether or not it is implicit in the laws of physics. That means that the above UD “prediction” has more the character of a logical truism in as much as it is true whether we are evolutionists or not. Therefore this “prediction” does not necessarily express an anti-evolutionist view.

ID prediction 04: ID predicts that, as scientific research progresses, biological complexity will be seen to increase over time, and information will have a more and more central role in the governing of life’s operations.

My comment: By “increase over time” I think the author actually means human research time. If so then this seems a rather trivial prediction given that as far the extent of our knowledge about organisms is concerned the only way is up!
But I’m going to use this opportunity as an excuse to talk about an increase in organic complexity over paleontological time; a prediction that is not peculiar to anti-evolutionist ID but also follows if Prob(Life|Laws of Physics) has realistic values - that is, evolution itself predicts an increase in biological complexity over time. Let me explain.
It is clear that not all biological organisms are of equal complexity. Here I am not so much talking about configurational complexity but the more abstract computational complexity measured by computation time. For example, it is fairly self evident that metazoans must be preceded by precursor structures; very likely single cell organisms. Single cell organisms in turn will be preceded by basic chemical constituents. In short any organic structure is preceded by a minimum number of precursor conditions that follow in a strict stratified sequence. This stratified sequence forms a kind “critical path” in the manufacture of the structure. The longer this path, the more computationally complex the structure is because the minimum time needed to arrive at the structure  is an increasing function of this sequence length.
If evolution is a kind of diffusion motion through configuration space then the precursor structures will necessarily be reached first. This diffusion motion causes a slow population growth in organic configuration space where the stratification sequence ensures that the “lower levels” will be populated before the “higher levels”. Thus it follows that biological complexity will be seen to increase over time. One way to think about this result is thermodynamically: The Earth is in a kind of morphological disequilibrium; as time increases the shuffling of happenstance entail a kind of diffusion motion which ensures the progressive population of morphological slots over time. Those morphological slots that require a sequence of morphological stages to be followed before they are arrived at will be populated later rather than sooner. The result is that biological complexity, computationally speaking, increases over time.

No comments: