FALSIFIABILITY AND THE SCIENTIFIC STATUS OF [ID] INTENTIONAL DESIGN

By Rabbi Dr. Dovid Gottlieb

There is a reason widely used to dismiss intelligent design [ID] and other purported examples of pseudo-science. The reason is this:

A proposition is scientifically meaningful only if it is falsifiable – that is, only if some observation can show that it is false.

The intuition behind falsifiability is that theories are created to explain phenomena, and are tested by predicting new observations. If those predictions are found to be correct, the theory has passed a test and is worthy of tentative acceptance and further investigation. If any prediction is found to be incorrect the theory is false and is discarded and we search for a new theory.

ID violates falsifiability because we do not know the motivation of the designer, so we cannot predict what it will do. So no observation could rule out ID. The same argument is used to show that religious propositions across the board are not falsifiable – since we do not know God’s intentions, nothing that happens could rule out God as the cause of various phenomena.

Falsifiability and its supporting intuition have multiple faults. The one that I wish to describe here is that there are counter-examples.

COUNTER-EXAMPLES TO FALSIFIABILITY

Rudolph Carnap objected that according to falsifiability the proposition: “There is a unicorn somewhere is space-time” is not scientifically meaningful since there is no observation that would show that all of space-time lacks unicorns. But a single observation could verify the proposition – just find one unicorn. Any proposition that can be verified by observation deserves to be regarded as scientifically meaningful. 

For a contemporary example, consider the fact that certain models of the big bang predict the creation of magnetic monopoles. As Roger Penrose points out, there is currently a scientific effort to find one. If that effort succeeds, we will know they exist. If that effort fails we will know nothing – perhaps they exist and we have not found them. Thus the hypothesis that they exist cannot be falsified, and yet that hypothesis is certainly scientifically meaningful. Penrose uses this example explicitly as a counter-example to falsifiability.

Now let’s consider parallel verification of ID. Even if ID is unfalsifiable, there might be unmistakable positive evidence of design. This is the foundation of SETI – the search for extra-terrestrial intelligence. Imagine finding the first one hundred places of the decimal expansion of pi in radio waves from a distant galaxy. That would clinch the existence of ETI. Even if we have no information about the intentions of those intelligent agents, and so we cannot make any predictions about what they will do next, this does not disqualify the conclusion that intelligent agents produced those pi radio waves. [This point is recognized by Derek Parfit in Appendix A of On What Matters.]

A BETTER MODEL OF THE SCIENTIFIC METHOD

The intuition supporting falsifiability ignores conclusive verification [as pointed out above]; it also ignores evidence that supports or opposes a proposition without conclusively verifying or falsifying it. Furthermore, I remember hearing Richard Rorty say at a convention of the American Philosophical Association [in the 1970s] that every scientific theory ever accepted had known counter-examples. Theory change means rejecting one theory with its counter-examples in favor of another theory with its counter-examples. And no one knows how we do that – no one has a good description of the relevant factors for that decision.

So falsifiability should not rule out ID [or anything else]. Still, it does seem right to require some kind of link between legitimate science and observation. Perhaps this is what we have in mind:

A proposition is scientifically meaningful only if observation can make some difference to its credibility – some observation must count for or against the proposition.

But this will be useless to rule out ID and religious propositions because it will be impossible to show that no observation affects their credibility. Just as the example of pi shows the possibility of verification of a case of ID, so too for many religious claims. Propositions from many faiths could receive verification from observation – just imagine the arrival of a messiah complete with all the predictions of a particular faith!

The key here is that something may be observed that could only be reasonably explained on the basis of ID or some religious proposition. The existence of such a possible observation shows that the condition above is satisfied by ID and religious propositions.  

HOW TO FORMULATE AND DEFEND ID

Suppose we are trying to explain some phenomenon. We have a list of non-design factors relevant to explaining it. We have reason to believe that no combination of those factors, can succeed in explaining it. We have reason to believe the same holds for any other non-design factors of which we are presently unaware. The phenomenon is similar to known products of intelligent design. Then we have some reason to explain the phenomenon by appealing to intelligent design.

The strength of the reason in the conclusion is proportionate to the strength of the premises. Let’s look at two examples.

SETI and the decimal expansion of pi. What do we know of galaxies and inter-galactic space? Gravity, black holes, Hawking radiation, collisions of stars, novae explosions, nuclear fusion, and so on. None of those factors, nor any combination, is going to produce the decimal expansion of pi. Are there other unknown non-design forces in galaxies and space? Undoubtedly – dark matter and dark energy and perhaps many more. Would they be able to produce pi? It is hard to say for sure, but we can say that if they can, perhaps we should have seen signs of their operation. Since we have not, we are reasonably sure that the missing factors will not explain it. And since we produce pi, we know that intelligent agents do so. Thus the reason to accept ID [=ETI] to explain receiving pi is very strong.

By contrast, try using ID to explain the existence of frogs. Do we have reason to think that the non-design factors of which we are unaware cannot explain the existence of frogs? Certainly not: As Stuart Kauffman and co. keep reminding us, there may be many causes of self-organization. So, even if there is some analogy between frogs and human products, the support for ID in explaining frogs is very slim at best.

Now suppose we have accepted a premise that the existence of frogs is to be explained by some particular factor or by ID.  In effect, the list of non-design factors has been reduced to that one factor. Then evidence against the success of that factor will be support for ID. For many, that factor is evolution. In the context of the premise, evidence against evolution is evidence in favor of ID.

The premise itself might be defended by pointing out that at present there is no third alternative even proposed. Of course, we cannot know what new theories will be invented. But that is true in all investigations, so it cannot be a reason to reject the tentative acceptance of ID if evolution is discredited.

One focus of the debate at present is the explanation of new information. It is claimed that there is no known non-design factor that creates new useful information. This is a perfectly legitimate investigation. It cannot be avoided by appeal to falsifiability.

The bottom line: Lack of falsifiability is no reason to rule out a theory as scientifically meaningless. ID is a meaningful scientific theory that must be evaluated on normal scientific grounds. Let the scientific debate on ID proceed!