Testing for Design

Question by the Intelligent Design and Evolution Awareness Center (the IDEA Center):
Does intelligent design make predictions? Is it testable?

Answer according to the IDEA Center:
Yes.

“Intelligent design theory predicts:
1) that we will find specified complexity in biology. One special easily detectable form of specified complexity is irreducible complexity. We can test design by trying to reverse engineer biological structures to determine if there is an “irreducible core.” Intelligent design also makes other predictions, such as
2) rapid appearance of complexity in the fossil record,
3) re-usage of similar parts in different organisms, and
4) function for biological structures.
Each of these predictions may be tested–and have been confirmed through testing!”

ATA Answer:
Yes and no, but the no is bigger.

The IDEA Center is a student-targeted initiative co-founded by Casey Luskin, now of the Discovery Institute. It encourages high school and tertiary students to set up their own IDEA clubs and talk about intelligent design. I use the present tense because it’s never officially folded, but there has been no visible activity by IDEA since a screening of The Privileged Planet in 2007.

I’m not simply flogging a dead horse (with regard to IDEA and intelligent design itself) because the question applies to the now more openly advocated hypothesis of divine design. It concerns a major aspect of claims by any creationist group that their version qualifies as science: is it falsifiable? What does it tell us that we don’t already know? What would disprove it?

I’ll work through the points in IDEA’s answer.

1) Two kinds of “specified complexity” crop up in creationist arguments: irreducible complexity, and “complex specified information” such as that found in DNA.

The discovery of irreducible complexity in a lifeform might go a long way toward proving design, though this hasn’t happened so far. Every purported example of irreducible complexity has been hypothetically reduced, in other words a possible evolutionary pathway to its current state has been found. On the other hand, the absence of real examples doesn’t mean nothing is designed. Irreducible complexity is only useful to this issue at all if it’s actually found.

The term “complex specified information” subtly begs the question, implying that someone specified it. While there’s certainly complex information in DNA, this is predicted and indeed required by evolutionary theory as well. The instructions for building body parts, the essential products of mutation and natural selection, have to be stored somewhere.

There’s a larger example of begging the question here. Of course the design hypothesis will predict the very thing it was created to explain: the complexity of life, including its means of reproduction and information storage. The important predictions we get from science are those with new, as yet unknown information – those we have to go out and test, not just refer to the same information that spawned the hypothesis.

For instance, we have one less chromosome pair than our closest ape relatives. Losing a pair outright is catastrophic, so the only way our species could have arisen from apes with one less chromosome pair is if two pairs fused. Therefore the end-markers for chromosomes should appear in the middle of just one of our pairs. On investigation, they showed up in pair #2. If they hadn’t, or had appeared in more than one pair, evolutionary theory as it exists now would have been falsified.

2) Again, the rapid appearance of complexity is a non-unique and retrospective prediction by the design hypothesis. Evolutionary theory allows for it as well, for a given value of “rapid”.

The most famous example is the “Cambrian explosion”, which took not less than two million of the fifteen million years now known as the Cambrian period. That’s as long as homo habilis took to evolve through homo erectus into us, homo sapiens. Given that it’s also roughly the point where the animal kingdom itself began, one would expect tremendous diversification.

A prediction certain design proponents might make which is not shared by evolutionary theory is not the rapid appearance but the literally instantaneous appearance of complexity. Like irreducible complexity, this would be earthshaking if found, but until then its apparent absence tells us nothing.

3) Similar features in creatures which are not at all closely related is another thing evolutionary theory also predicts in unremarkable hindsight, via convergent evolution. For example, across the animal kingdom we see eyes at all stages of development because eyes are undeniably useful. Any creature which begins to develop sight has an immediate advantage over the blind.

4) “Function for biological structures” mostly refers to DNA, and the idea that it’s all useful. The existence of “junk DNA”, that which has no effect on the creature which carries it, does not obviously indicate a design purpose, so the prediction of design is that practical effects will eventually be traced to most or all of it.

Given that we don’t know the purpose of a lot of DNA, new purposes will surely be found for some of it, but what’s the failure standard here? What percentage of DNA, if confirmed to be absolutely useless, would disprove design? There’s no sensible answer. Design isn’t falsifiable this way either.

In the broad view, we have 1. a bunch of different ways evolution could at any moment be debunked or disproved (and, notably, hasn’t been) and 2. no good way divine design could ever have potentially been disproved. It makes obvious predictions, all right, but it can’t truly be tested. Hopefully I’ve managed to demonstrate the difference a little more for you.

– SmartLX