Thursday, March 16, 2006

Paul Nelson is confused, or scientific methods are methodologically naturalistic

Nelson defends the proposition that the evolution of life on Earth cannot be explained by purely natural processes.

[I] argued that the naturalistic program of explanation hasn't even come close to satisfying its claims about explanatory sufficiency for events such as the origin of life, or even the origin of RNA.

Sahotra didn't disagree, or, to put it another way, there is no room for disagreement, if one wants to know the actual natural processes responsible for key events in the history of life. Questions like the origin of RNA, the origin of life, the origin of "orphan genes," and the other puzzles I mentioned … are open research problems. If Sahotra knew the answers, he didn't say; nor did Bill Wimsatt; nor did the two biologists on the faculty panel; nor did anyone in the very feisty, overflow audience. And that's because nobody knows.
We can summarize Nelson's argument: that science has yet to explain something suggests that it will never and can never explain such a thing. He then makes the remarkable claim that one cannot disagree with this statement. It is true that one cannot disagree with the antecedent (science has yet to explain some things), but it simply doesn't follow that this shows that those things can't be explained.

The fact that the things he lists are "open research programs" means that many people think that naturalistic explanations do exist for those things, and even have good candidates for what those explanations are. They are testing them using what we like to call "the scientific method." That method, by its nature, cannot address certain classes of hypotheses. It is an inherent limitation of science.

Nelson fails to appreciate this point, and I suspect he is also rather severely misinterpreting his opponent's position, and equivocating to do it. He argues that methodological naturalism is either irrelevant or blocks valid research:

Ask oneself a simple question. Suppose life actually were designed by a nonhuman intelligence [note that he doesn't specify how the intelligence effected the design, though this is what is at issue -TfK]-- would methodological naturalism allow us to discover that? If the answer is no, then methodological naturalism hinders scientific discovery and dictates the shape of reality as thoroughly as philosophical naturalism. If the answer is yes, then methodological naturalism is superfluous and says nothing more than that science should [must? -TfK] be empirical and testable.
Unfortunately for Nelson, the answer is: maybe. If space aliens designed life on earth by means of natural processes which are empirically testable, then yes, we can detect the effects of "nonhuman intelligence." So long and thanks for all the fish.

What we cannot detect is a supernatural force which changed the rules of the universe to effect its design. We can't, by the nature of science, test for such a thing using the scientific method. Methodological naturalism (I prefer "practical naturalism") is a practical consequence of science's empirical and testable method. By their nature, such a method cannot detect the supernatural.

Hence my "maybe" to Nelson's question, by means of which we recover the usefulness of the concept. It demarks what we can practically do with science from what we cannot practically do. The method of science cannot detect supernatural activities, so we don't try. Methodological naturalism isn't a barrier imposed, it's a description of what science can and can't do. It doesn't proscribe, it describes. As such, it is superfluous only to the extent that science would be as limited if we didn't talk about MN as it is when we do talk about it. As a scientist, I spend exactly no time wondering whether MN forbids a given experiment. I also spend exactly zero time wondering of I'm fulfilling a logical positivist's description of science or a Kuhnian approach.

Philosophical concepts don't hinder scientific progress. Practical considerations do, and sometimes, those practical considerations match philosophical concepts. If we follow Nelson's reasoning, philosophy may well be superfluous (at least to the practice of science).

But Nelson is a philosopher, debating philosophy. As such, the standard for superfluity must be substantially lowered. Methodological naturalism is not an imposed hindrance to the progress of science, it is an inherent limitation on the scientific method.