Del Ratzsch on methodological naturalism

TGL: In your review Design Theory and its Critics, you wrote that "If (perhaps for overwhelmingly good reasons) science is restricted (even just methodologically) to 'natural' explanatory and theoretical resources, then if there is a supernatural realm which does impinge upon the structure and/or operation of the 'natural' realm, then the world-picture generated by even the best science will unavoidably be either incomplete or else wrong on some points. Unless one assumes philosophical naturalism (that the natural constitutes the whole of reality) that will be the inescapable upshot of taking even mere methodological naturalism as an essential component of scientific procedure." This suggests that the distinction between the two forms of naturalism collapses, but there seems to be little awareness of the argument. Do you intend to develop it further?

DR: I have discussed it some elsewhere (e.g., in "Natural Theology, Methodological Naturalism, and 'Turtles all the way down'" (Faith and Philosophy, Vol 21 #4, October 2004, pp. 436-455)).

[...]

The basic problem with pre-stipulated conceptual/theoretical boundaries is that if reality itself happens to fall outside those boundaries, theorizing within the confines of those boundaries will inevitably generate either incompleteness or error. But methodological naturalism just is a stipulated prohibition on anything outside the 'natural' playing any conceptual role in scientific theorizing and explanation. If it turns out that reality chooses to ignore our restrictions (and why on earth shouldn't it?), then theorizing forbidden to cross those boundaries will inevitably be either incomplete or mistaken.

Here is an analogy. [All right - caught analogizing again.] Suppose that during the final pre-launch crew briefing for NASA's first manned mission to Mars, the head of NASA warns the crew of the dangers of starting public panics and instructs them to make no mention in any of their reports of aliens - regardless of what they happen to find on Mars. The restriction does make some sense. But suppose that the first thing the crew sees upon exiting their lander is an utterly undeniable Martian bulldozer. The question instantly arises: where did that come from? But the crew has a problem answering that question. Given the prohibition barring reference to aliens, the crew has only two options: (a) they can refrain from addressing the question, or (b) they can construct a theory of the chemical evolution of Martian bulldozers. But that means that their science of Mars will be either (a) woefully incomplete - leaving out perhaps the single most fascinating aspect of the mission - or (b) outrageously mistaken.

[...]

But even just methodological naturalism conjoined with aspirations for completeness has substantive implications. First, if one restricts science to the natural, then assumes that science can in principle get to all truth, then one has implicitly presupposed philosophical naturalism. But even if one merely stipulates methodological naturalism as essential to science, then assumes only that science is competent for all physical matters, or that what science (properly conducted in the long run) does generate concerning the physical realm will in principle be truth, then if the truth of the specific matter in question is non-natural, even the most excruciatingly proper naturalistic scientific deliverances on that matter may be wide of the mark, typically in exactly the way a science built on philosophical naturalism would be. For practical purposes, that comes close to importing philosophical naturalism into the structure of science.

So whether methodological naturalism has substantive philosophical implications (contrary to the common denial) or is philosophically neutral depends upon what it operates in tandem with. At the least, methodological naturalism makes the de facto assumption that there is an identifiable realm of reality which is on the scientifically relevant level functionally self-contained, and which is on that level functionally de-coupled from the supernatural. That assumption is neither obvious, trivial, nor - since it is an empirical universal negative - demonstrable.

But to actually answer your question, I may try to push it a bit further. But despite the above (and some other) reservations and qualifications, I think that methodological naturalism is a useful - perhaps even essential - provisional strategy, and one not to be lightly overridden.

TGL: Much has been made of the importance of methodological naturalism, particularly as definitive of what makes something science. What do you think of the arguments in its favour?

DR: Arguments for its value as a provisional strategy may be right. But even as a strategy, it has to be used with care. Over-rigid adherence can (as indicated earlier) have consequences for the self-corrective nature of science, and it can have other consequences (as noted just above) if care is not taken concerning what assumptions it is employed with.

Arguments for it will depend in part on exactly what methodological naturalism is, and more care is required there than is sometimes given. For instance, it is quite common to see methodological naturalism defined as a requirement that science be restricted just to natural concepts, resources, data, and theories, that being interpreted to mean that whether or not philosophical naturalism is true, science must proceed as if it is. (That, for instance, is the position of the National Center for Science Education - or at least of its director.) But the problem here is that (as Boyle pointed out three plus centuries ago) nature in a created universe might well - indeed most likely would - be very different from nature in a random, chance universe. Thus, the typical equating of a restriction to the natural with proceeding as if philosophical naturalism is true, turns out to beg some deeper questions.

Most of the actual arguments for methodological naturalism being a definitive, unchallengeable rule of science seem to me to be problematic. Very briefly, the three most common types of arguments are (1) arguments that anything non-natural is outside the realm of empirical detectability or testability, (2) arguments that allowing the non-natural into science is destructive in that it allows scientists to take the lazy way out in difficult scientific situations (simply saying "Well, God must have done that - no point in trying to figure it out", then wandering off to find the coffee pot) and (3) historical arguments claiming that the history of science has shown the bankruptcy of non-natural considerations in science. The first is the most prima facie plausible, but I think that there could be possible empirical cases in which the most reasonable conclusion would be that something supernatural was at work. (That's one of the cases I try to make in Nature, Design and Science.) Regarding the second, it is often the conviction that something is a product of design that keeps scientists in the hunt. Any company trying to reverse engineer a competitor's new computer model pays particular attention to puzzling components - refusing to give up trying to understand it precisely because they believe it to be a product of design. And of course historically most scientists took nature to be a product of design, and saw themselves as in effect reverse engineering nature - trying (as Kepler is alleged to have said) to think God's thoughts after him. The fundamental intelligibility of nature consequent upon its being designed by God was one of the key motivations underpinning the whole scientific project.

http://www.galilean-library.org/manuscript.php?postid=43816