The debate about whether intelligent design (ID) can be science or not is still ongoing. In order to address this problem, a non-question begging account of science is required.
In some debates, I have seen science defined in terms of involving only naturalistic explanations. If this definition is accepted, then the usual suspects in the case of ID can be dismissed by definition. For example, if God is supposed to be the designer and God is supernatural, then this falls outside the realm of science. QED.
Of course this approach has two main weak points. First, the door is still open to natural designers. As such a form of ID involving a non-supernatural agency would still be (potentially) scientific. Second, to say that ID is not scientific based on this definition is not a substantial criticism of ID. It would, in fact, beg the question. It would be somewhat like saying that French literature is not literature, because literature must be written in English. If this were accepted, then French literature would not be literature-but only by definition and not on any substantial ground.
It could be replied that science is by its very nature concerned solely with natural phenomenon. This reply would, of course, ignore a substantial part of the history of science (such as Newton) but can be granted as a viable approach for the sake of argument. One might, for example argue that just as math deals with mathematical matter and literature deals with literary matters, science deals with natural matters. Supernatural matters belong to philosophy and religion.
In this case, most (but not all) forms of ID would be non-scientific. But, the obvious reply is that ID is being dismissed as non-scientific by fiat and (once again) by mere definition. It would, one might say, be more honest to say that supernatural ID theories are not part of the natural sciences. This is perfectly reasonable. So, one might ask, why not stop the discussion now and simply accept that ID is not scientific?
The obvious reply is that the classification of ID as non-scientific is more than just a matter of saying that supernatural theories do not fall under the natural sciences. Rather, to tar ID as unscientific is often an attack on such theories. The idea is not to say that they belong in a different field (as one might say that the matter of ethics belongs in philosophy and not chemistry) but that ID theories are somehow defective and perhaps even fundamentally irrational.
An obvious reply to this is to point out that some ID theories are irrational and unscientific in the most negative sense of the terms. This is true, but does not seem to warrant dismissing all ID theories as being unscientific in this highly negative sense.
Perhaps there is a more substantial way to assess whether ID can be considered scientific or not-without begging the question by simply relying on the fact that supernatural ID theories are, by definition, not natural theories.
One hallmark of scientific (and philosophical) reasoning is that a claim or theory is subject to rational evaluation and, most critically, testing. If a claim or theory cannot be tested, then it cannot be considered a scientific claim or hypothesis. Naturally, a claim might be beyond our means to test now (like a claim about the nature of the interior of a black hole or about the composition of Dark Matter) but as long as it is testable, then it can be considered scientific (to some degree-obviously I am cutting numerous corners here for the sake of brevity).
So, can ID theories and claims be tested? If a claim or theory is such that it cannot be tested, then the obvious answer is no. To use a rather absurd example, imagine that someone claimed God zapped the life, the universe and everything into existence 4,000 years ago. Further, God made it so that the universe seems much older to all possible tests-this is why scientists mistakenly think the universe is older. Obviously, this view cannot be tested in an meaningful sense (though it might be true). But this does not mean that no ID theory can be tested.
If someone claims that living creatures are the product of intelligent design, because of the evidence of complexity, then this can be tested. The test would be (to put it simply) to determine if the complexity of living creatures can result from something other than an intelligent designer. To the degree that such an alternative (or alternatives) is established, the hypothesis of the designer has been dis-proven. To the degree that the complexity can only be accounted for in terms of design, the hypothesis is supported. This certainly seems to be properly scientific.
At this time, the natural selection hypothesis seems to have the upper hand. It has been claimed that the complexity of living creatures can be explained in terms of a random process (one not guided by a designer) and a selection mechanism (natural and opposed to intelligent) selection. As such, by Occam’s razor, there is no reason to bring in an intelligent designer.
So, it could be reasonably held that ID can be scientific. It just so happens that the consensus in science today is that it is an implausible theory.
So, what about teaching ID in schools? That is a matter for another time.