We've had a few discussions here, and there have been several more elsewhere, on a strange idea: that scientists automatically exclude the supernatural from science because of some close-minded metaphysical bigotry, that for instance there is an a priori determination that any explanation that mentions the words "ghosts" must be wrong and excluded from science. I disagree. We practice methodological naturalism, and all that matters is whether we have an operational toolkit to evaluate a claim. I don't dismiss "ghosts" (or Intelligent Design) because I have some prejudice against supernatural beings, but because those beings are so poorly described and delimited by their proponents that I have no way to evaluate them—and if these proponents want to be taken seriously, they must make the effort to establish clear definitions, criteria, and procedures for their study, something Intelligent Design creationists have steadfastly refused to do.
Now Brian Leiter has ripped into this topic at length. One of the interesting points there is that this represents a common strategy of trying to muddy the waters and pretend that science is religion, and that religion is science, and therefore religiously-motivated babble, like Intelligent Design creationism, is on an equal footing with evolutionary biology.
I assumed that he [VanDyke] --like all the others who peddle Intelligent Design--might be making a non-trivial point, namely, that methodological naturalism was genuinely a priori, i.e., a dogma immune from and indifferent to the empirical evidence, and thus on a par, epistemically, with supernaturalism. If that were true, then we would have an argument for saying that evolutionary biology, with its genuinely a priori commitment to methodological naturalism, was indistinguishable from supernaturalism. Alas, it is not true that methodological naturalism is a priori in the only sense that is relevant.
I've also written up an anecdote from 1999, Scientific bias and the Void-Of-Course Moon. Scientists are willing to consider even the most ridiculous hypotheses, if they are stated with sufficient clarity that they can actually assess them. Even if the idea is as absurd as astrology.