Tuesday, January 03, 2006

Is Methodological Naturalism Warranted?

I got this message from a colleague who is a devout Christian:

If the American science community agrees with [Judge John Jones III in the Dover case] then American science is not the search for truth. Instead it is some sort of search for the best explanation that fits some preconceived notions. Whether the notions are religious or secular, theistic or atheistic, I think that attitude will ultimately cripple scientific endeavor, just like religious restrictions did in the middle ages. Yes, atheistic assumptions have advanced science quite well but can you prove they are necessary to science and take precedence over ultimate truth?

I find so much to disagree with in just one paragraph!

First, science isn't the search for truth, and I don't know any scientists or philosophers of science who think it is. Science is about modelling the natural world with models that are necessarily imperfect. Newton's laws are good approximations, not truth; they were refined by Einstein to better, but still incomplete, models.

Second, I don't think science has "preconceived notions", at least not in the way my colleague means. Science uses a certain method of inquiry, to be sure, and that method consists of pieces such as hypotheses, experiments, theories, testability, repeatability, peer review, and a reliance upon evidence. Praying for revelations or appeals to dogma are not part of the scientific method, but neither are bingo, choral music, or playing frisbee. Could there be important aspects of the world that only prayer could reveal? Or only bingo could reveal? Perhaps, but I personally don't see any at the moment. Nothing is preventing religious scientists from getting information through revelation; science simply demands that there also be some hard evidence for any resulting claims. Ramanujan claimed he got revelations from the Hindu goddess Namagiri, but that didn't stop mathematicians from appreciating his results, finding his mistakes, and providing proofs for the correct results.

As Stephen Weinberg has remarked, "The fact that Newton and Michael Faraday and other scientists of the past were deeply religious shows that religious skepticism is not a prejudice that governed science from the beginning, but a lesson that has been learned through centures of experience in the study of nature."

Third, I don't think science makes "atheistic assumptions". I have yet to open a science textbook and find the statement "There are no gods". Rather, science is non-theistic; that is, it proceeds without making claims about gods or lack thereof. I think the record shows pretty clearly that when religion enters the practice of science, the result is really bad science. To see just one example, look at the grotesque Bible Codes phenomenon.

Finally, I don't see any need to prove that methodological naturalism is necessary to science. I am perfectly willing to argue empirically that methodological naturalism has served science well, and those advocating adding supernatural causation to science have yet to show any benefit to science. When they do, give me a call.


Lippard said...

One question (discussed at Vic Reppert's "Dangerous Idea" blog, among other places) is whether methodological naturalism could result in a conclusion that there is a God, or if it really does (as Jones seems to say) *prevent* inferences to the existence of supernatural entities. (Vic's blog has been critical of Jones' decision for suffering the same failings as Overton's decision in McLean v. Arkansas, as criticized by philosophers of science such as Larry Laudan and Philip Quinn.)

I'm inclined to think that, if the right sort of repeatable data were seen (in particular, data which involved interactive communication with an entity that claimed to be able to make omnipotent interventions and have all knowledge, and was able to back it up), it could yield a conclusion that supernatural entities exist.

Anonymous said...

On the subject of "methodological naturalism," I recommend this interview with Barbara Forrest, who said that:

Science, however, is a naturalistic enterprise. Scientists cannot appeal to supernatural explanations because there is neither a methodology for testing them nor an epistemology for knowing the supernatural. Science has a naturalistic methodology, known less controversially as “scientific method.” That simply means that scientists seek natural explanations for natural phenomena. Science also has an epistemology, namely, the use of human sensory faculties to collect empirical data about the world and the use of our rational faculties to draw conclusions and construct explanations of this data. This is the only successful way to do science, and the pragmatic success of this naturalistic methodology is the only reason scientists use it. There is neither a conspiracy by scientists to prohibit “alternative explanations” nor an arbitrary commitment to naturalism, as ID proponents charge. Scientists use this naturalistic methodology because it works. Period.


Jeffrey Shallit said...

Miss ht psych:

When I deny that science has "preconceived notions", I am referring to the comments of my colleague. I meant, there is nothing in the scientific method that says one has to begin by saying "There are no gods".

Since science is a human endeavor, it goes without saying that individual scientists may have biases that affect their judgment.

I haven't read any Sandra Harding, but Gross and Levitt didn't have a high opinion of her in their book Higher Superstition.

Julie said...

An awful lot of the "logic and reason" purported to support prejudice is just plain illogical and unreasonable. Mistaking correlation for causation, or confusing the meaning of between-groups and within-groups variance, are examples of unreasonable assumptions (see some of the many critiques of The Bell Curve from within the biological and statistical communities).

Also, logic and reason, while important, aren't sufficient for drawing conclusions through the scientific method. They have to be backed up by observational and/or experimental evidence. "Thought experiments" can help you refine your approach, but they're not a substitute for a good set of research methods.

Anonymous said...

This is not directly on topic, but this discussion brings to mind a course I took on genetic algorithms, where one of my classmates was a true believer in creationism. He asserted at one point that increasing the mutation rate of the algorithms was always going to be a bad thing because ... well, he never really said why. But it was easy enough to test and simply show that, for some problems, higher mutations rates lead to quicker convergence.

He never brought that argument up again, but he did point out that genetic algorithms didn't work without someone to implement and design the algorithms, and without a computer for them to run on. His point was that *we* were playing the role of God in this case, and that was *our* decision to use genetic algorithms instead of, say, heuristically guided guesswork.

In the study of genetic algorithms --- or any kind of algorithm --- it is preposterous to consider within the framework of the algorithms questions like "Who created the algorithm?" or "Who created the computer the program is running on?". These are irrelevant to the question of the algorithms themselves.

The analogy with science (which he didn't make, but it is obvious) is that an "algorithm" like evolution clearly works as claimed, but there's no way that the study of evolution proper could ever answer questions like "Who thought of evolution?" or "Who created the universe in which it runs?" (there's an argument to be made that these are not even sensible questions, but I'll ignore). Such questions could be studied in a scientific way, but that way doesn't promise to jibe with the religious traditions of church-goers.

Anonymous said...

Some people might not agree, but my opinion is this:

I am only interested in naturalistic scientific theories because only naturalistic theories are useful. Only a theory which attempts to explain the universe in terms of causal processes that are understandable can ever give rise to any technology.

A supernatural "theory" by contrast cannot ever give rise to any technology. It is useless.

Of course, the first objection I'll probably hear is: but aren't a lot of scientific theories like the big bang useless according to this view? No, I don't think they are. While the big bang theory might not *directly* give rise to any technology, as a causal explanation it is capable of providing inductive input regarding other theories that may. If the big bang theory fits with our physics and with the observed data, then it inductively confirms our physics. If it does not, then either the big bang is wrong/incomplete, the data is wrong, or our physics is wrong/incomplete. A more complete and correct understanding of physics in turn allows us to do things like build faster CPUs or more practical fusion reactors. So the big bang is practical, and I think a good philosophical argument could be made that all naturalistic theories are of practical use in the same way-- either directly or indirectly.

Naturalistic theories make stuff work. This, IMHO, makes them more important and interesting than supernatural speculation.

Anonymous said...

If science isn't a search for truth, what is? Is there no search for truth? Or perhaps there is, but it is silly and misguided, that is, if we were sophisticated we wouldn't bother searching for the truth?

Science doesn't make atheistic assumptions, it just proceeds without making any assumptions about God? Well, you should know, that is what many atheists claim to be a definition of atheism (so they can claim that it is a natural state of affairs -- babies are born atheists in fact).

You seem to think math is a science? That is not what they think over at Panda's Thumb when they blast Dembski.

Jeffrey Shallit said...

Dear anonymous:

Why don't you tell us which human endeavors you think are atheistic? As far as I can see, most are non-theistic, including history, economics, art, music, physics, chemistry, mathematics, geology, and paleontology.

No, I don't think mathematics is a science. What made you think I did?