I want to thank the organizers of this conference for inviting me here today, although I do rather suspect their motives. This is the very definition of hell for a philosopher — to be invited to a conference, to sit through 2 and a half days of really interesting presentations taking voluminous notes, a lot of which have marginalia attached to them like "no no no!" and "???", and at the very end to be told, "Okay, right! 15 minutes: vent your spleen all you want." Well, there is only so much I can vent. But I am going to try to get together a few basic thoughts I have had and offer them in some kind of coherent form.
What I want to try to do is give you a basic blueprint for respectability. If we make the assumption (and there are lots of people who would question this assumption, but I will make it for the purposes of this talk) that ID theory seriously wishes to become a respectable scientific theory, then I will tell you how to do it. If you follow my 4 simple steps to scientific respectability, you will get what you want: scientific respect, research funds, access to science classrooms, and so on, and so forth. It is actually fairly simple — all you have to do is follow the 4 steps. So what are they?
Step 1: IntelligibilityWell, the first step is a little complicated, but it involves intelligibility. I prefer not to talk about naturalism, but rather intelligibility. I think the one thing that scientists cannot compromise on — one of their most fundamental philosophical principles — is a commitment to intelligible causal factors. That is to say, they will not accept, in principle, explanations that make reference to causal factors that cannot be explained by human reason. Why does this matter? Well, it matters because it makes a big difference as to whether or not ID theory can be done within the context of science. I should add, by the way, that I fully grant that this is, in some sense, a philosophical assumption. It may be wrong. If you guys would like an admission that any of your favorite versions of creationism or ID theory could perhaps be correct, you have it from me. As a representative of the scientific orthodoxy, I will admit that they could, perhaps, be correct — but that is not a sufficient reason to believe them.
So can we practice ID theory as a science if we buy this notion of intelligibility? The answer is that it depends on what you mean when you talk about the designer. I have heard speakers here at this conference point out how it is necessary to specify the nature of the designer, and then some people in the audience say that it seems a bit unfair to require ID theorists to come up with a clear notion of what the designer is like. Well, this is not a tangential issue; it is a fundamental issue. An unwillingness to talk about this is going to cripple at the outset any attempt to make ID theory a scientific theory. And here is why. (Although I am not a theologian and I am aware of the fact that this is a horrendously complicated theological dispute, I am going to simplify egregiously.)
There is a large continuum of theories about God's nature, and on this continuum there are 2 basic endpoints. On one endpoint, you have a view of God as an intrinsically mysterious agent. Human reason is simply incapable of penetrating into the mysterious God's motives, mechanisms, and the like. On the other end of the continuum, there is God as a rational God, a God whose motives and mechanisms are analogous to those of human intelligence (a phrase that came up in an earlier talk). In other words, a rational God is a God that we can understand in some important sense of that word.
If we are talking about a rational God, I think that it is perfectly okay, in principle, to include theological hypotheses as part of a scientific theory. Now, for my evolutionist colleagues who are getting uncomfortable at this point, I would like to point out that there are going to be a lot of people out there who are not willing to accept what goes along with making God rational in this sense. But if you do, it is possible to include some kind of theology in scientific theory — Spinoza, for example, had a theological account that is not inconsistent with the sorts of claims scientists would normally want to make. But if you cite a mysterious God, you are inserting a factor in your explanation that is in principle inexplicable.
Paul Nelson just gave you the example of a professor's suddenly getting up and walking around a conference table as something that is not explained. Here is a good example of a fundamental confusion. Intelligibility is not a question of whether or not we have explained something, it is a question of whether or not it is explainable — is it, in principle, subject to explanation? I think someone's getting up and walking around the table is perfectly explainable. There are epistemic problems to be sure — it is not a trivial exercise to explain it. However, it is hardly inexplicable, and that is what makes it fair game for a scientific explanation. (As an aside, I should say that some of the confusion about whether ID theory is testable but false or untestable has to do with equivocal notions about the nature of the designer.)
A rational God has clear, practical consequences for a scientific theory. Suppose you posit a rational God and then assume that God designed the traits of organisms to maximize ________ (you are going to have to fill in the blank because I do not know what your particular rational God would want to maximize; I personally would tend to say something like "adaptiveness within a particular selective environment", but that is just me). Whatever goes in this blank, it seems we can then formulate a null hypothesis and say, "We expect God to be at least as good as a human engineer would be in designing traits to maximize ________." Any trait that seems poorly designed from a human perspective would then represent a prima facie problem for an ID theorist. What you certainly cannot do in this kind of situation is to argue that it is simply a mystery why God created this trait in this particular way, because then all you are doing is reverting to a mysterious God.
I cannot really convince you right now why it is a good idea to buy into intelligibility. I would quickly say something like this, though: the consequences of not buying into this are far worse than you might think. To take what Winston Churchill said about democracy and apply it to neo- Darwinism, "Neo-Darwinism seems like a really bad theory — until you consider the alternatives".
Step 2: Internal CritiqueThe second step has to do with internal critique. You really can learn a lot about somebody by the people with whom they choose to associate. ID theory wants to be a "big tent" movement, but to fail to critique highly divergent arguments of colleagues who happen to share the same conclusion is tacitly to accept them. ID theory is not going to be a scientific discipline until it takes a clear stand on some major methodological issues. You simply cannot have a scientific discipline that talks about evolution but does not take a clear stand on the age of the earth or on common descent with modification! A scientific discipline shares, at least to some large extent, a certain common core of questions and methodologies. ID theory has no core methodology or theoretical commitments, and thus it is not yet a discipline (scientific or otherwise). Perhaps this can be remedied, but not as long some of your adherents persist in making the ambiguity of your own positions a defensive virtue, as when a critique is deflected with, "Well, you know, we don't all say that." If ID theorists themselves do not take a consistent stand, they are certainly not entitled to complain about the imprecision of their critics over the very same points!
So it seems as though there are 2 alternatives here: either you can institute a thorough, rigorous system of internal critique and try to develop a consensus along some of these issues, or you can fragment into separate groups, each group having a relatively unified approach to these kinds of things. Either will accomplish what I have called "internal critique".
Step 3: External CritiqueThe third step has to do with external critique — with stepping away from over-reliance on critiquing your opponents. When I teach basic philosophy classes, I have my students write argumentative papers. I want them to understand that critical ability is extremely important — it is something that they hone in a philosophy class, if nothing else. But I also want them to understand that there is a fundamental difference between critiquing your opponent and demonstrating your own position. This is a fundamental distinction that ID theorists need to take more seriously.
There are lots of reasons to keep the two activities clearly separate. First, it is always easier to be negative than positive. Given that any complicated theory is going to have anomalous data, anybody can find interesting cases to harp on. This does not really prove a whole lot. Second, it is just too tempting to engage in distortions of what your opponent says (straw man arguments). This is particularly true if you believe deeply in what you're saying — this makes it extremely difficult to be fair to your opponent, who you "know" is completely wrong. I do not see a whole lot of evidence, to be perfectly honest with you, that there has been a change in some of these practices since the bad old creationist times. Third, it may very well be that you are implying a false dichotomy. The underlying assumption when you substitute critique for demonstration is that, if my opponent is wrong, then I must be right. But there may well be a third alternative. Suppose it turns out that neo-Darwinism, whatever that means (people define this term in lots of different ways for different purposes), is wrong. It does not necessarily follow that ID theory is right, unless those are the only 2 alternatives, and believe me, there are more. Finally, it draws attention away from a basic practical fact (that it is practical makes it no less important). Nobody is going to abandon an extremely fruitful scientific theory until there is a viable alternative that does as much or more. It is just not going to happen. I have been in a position of espousing a theory on the outside of orthodoxy, and I understand that it can be frustrating, but at some point you have to shift away from whipping up the crowd by complaining about your lot versus the orthodoxy to coming up with a theory that works.
Step 4: Novel, Testable HypothesesAnd that gives me a segue into my last and probably most important step, that ID theory really has to start generating novel, testable hypotheses (and then testing them). Now, I'm a philosopher, so I am perfectly aware of all the dilemmas that arise if you try to make testability some sort of touchstone for science. I am not saying that. I am not a logical positivist. But I think it is clearly true, however you want to characterize it, that testability is a critically important element of successful scientific theories. Therefore, if ID theory is going to be a successful scientific theory, it must generate testable hypothesis — and note here I am also saying testable hypotheses (I am not talking about the theory's being testable itself. That is a can of worms I would rather not open right now).
The basic reason for this is pretty simple. We are clever and creative explainers. It is very easy for us to sit in a dark room and convince ourselves, based on a priori principles and data that are already in front of us, that we have the correct explanation. The only way we can really know whether or not our explanation is right is if we make novel predictions and then go out and see if they are met. That is the function of testability. As far as I can tell, ID theory just does not do this. There might be, maybe, some very minimal claims you can make about the heuristic value of ID theory — one of the talks, for example, postulated that perhaps the design heuristic is akin to notions such as beauty in evaluating scientific theories. Even if I were to grant that, this claim does not put ID anywhere close to the status of a developed scientific theory. We do not teach beauty in elementary science classes as a way of talking about scientific theories, and for very good reason. It is, at best, an interesting subtlety you get into when you talk about the philosophical aspects of science. It is not a fundamental theory in itself.
I think that nothing shows the difference between ID theory and its orthodox opponents better than a comparison of professional conferences. If I go to the annual meeting of the Society for the Study of Evolution, and I bring a bunch of graduate students who want to do research in evolution, I can point them in any direction and they can go and find hundreds of seminars going on, each of which has all kinds of evolutionary ideas, testable hypotheses, new methodologies based on evolutionary reasoning. They can go out there, they can find research projects, they can go back to the lab, and they can go to work. This ID conference, on the other hand, really does not have a single session presenting testable hypotheses. There was one that, in a way, sort of, kind of, made allusions to testable consequences, but I can say firsthand that there was nothing discussed in this session that neo-Darwinism had not already predicted. So there is really no practical import to ID theory that I can see.
Now, in case people are tempted to say, "Ah, well, give us some time, we'll figure this out," let me suggest that there clearly are testable consequences right now. Any relatively bright graduate student should be able to sit down with a pencil and a piece of paper and come up with some for you pretty quickly, and then you have got a research project. In case you cannot find anybody like that, allow me to offer your first testable hypothesis. A lot of ID theorists (and again we run into the ambiguity problem) have absolutely no problem accepting the theory of population genetics as a microevolutionary theory. But they do not like to accept macroevolution. They believe that there are certain natural kinds, or species, or whatever, and there are boundaries in between the kinds that cannot be overcome. In other words, population genetics works, but it cannot push allele change past these postulated boundaries between kinds. The existence of such boundaries is not a prediction of population genetics, and there is nothing in population genetic theory that would even lead one to suspect such bounded change. So, if this ID view is correct, there should be populations left and right that meet all the population-genetic criteria predicting evolution, but that are in fact not undergoing evolution. It should be a relatively straightforward job to go out and confirm experimentally the existence of these populations.
Now, let me just insert a caveat here because I worry about having my words taken out of context. The populations you find have to meet all the population genetic criteria. There are cases one could point to that might seem on superficial examination to meet all the criteria, but really do not. I have in mind something like sickle cell trait. When people first discovered sickle cell trait and did population genetics on it, it was a bit of a mystery why the sickle cell allele was so common in the population. People knew that in its homozygous form it was highly deleterious, so they calculated the selection coefficients based on this and said, "This is very strange — there are a lot more of these alleles floating around in the population than we would expect." This puzzle led to an investigation to see what other factors were at work in the population. As it turns out, the sickle cell allele confers an advantage in the heterozygous condition, so people who have only one copy of the sickle cell allele are more resistant to malaria, and in certain parts of the world that is a very good thing. After discovering this, the geneticists adjusted their selection coefficients and re-did the calculations, and it works perfectly. So I am not talking about superficially analyzed populations that appear not to undergo evolutionary change. I am talking about a situation in which you have excellent data on the population genetic variables and they simply do not add up.