Monday, February 01, 2010

 

Natural Methods


Larry Moran has a post up about Methodological Naturalism (MN) and a talk given at the Darwin Conference in Toronto last November by Maarten Boudry, a graduate student in the Department of Philosophy at the University of Ghent. Larry gives the abstract for the talk, in which Boudry gives a fair enough summary of MN:

Creationists and intelligent design proponents have previously claimed that the commitment of evolutionists to naturalism and materialism constitutes a philosophical prejudice on their side, because it rules out any kind of supernatural causes by fiat. In response to these charges, some philosophers and scientists have argued that science is only committed to something they call methodological naturalism: Science does not deal with supernatural causes and explanations, but that does not mean that the latter do not exist. However, there has been some philosophical discussion about the correct understanding of MN. The principle of MN is often conceived of as an intrinsic and self-imposed limitation of science, as something that is part and parcel of the scientific enterprise by definition. According to this view (Intrinsic MN or IMN) - which is defended by people like Eugenie Scott, Michael Ruse and Robert Pennock and has been adopted in the ruling of Judge John E. Jones III in the Kitzmiller vs. Dover case - science is simply not equipped to deal with the supernatural and therefore has no authority on the issue. It is clear that this depiction of science and MN offers some perspectives for reconciling science and religion. Not surprisingly, IMN is often embraced by those sympathetic to religion, or by those who wish to alleviate the sometimes heated opposition between the two.
Boudry seeks to argue that IMN does not offer a sound rationale for the rejection of supernatural explanations and, instead, should be seen as:

... a provisory and empirically grounded commitment of scientists to naturalistic causes and explanations, which is in principle revocable by future scientific findings (Qualified MN or QMN). In this view, MN is justified as a methodological guideline by virtue of the dividends of naturalistic explanation and the consistent failure of supernatural explanations in the history of science.
Boudry cites four arguments for IMN ("the argument from the definition of science, the argument from lawful regularity, the science stopper argument, and the argument from procedural necessity") that he thinks fail but, of course, without more, there is little that I can say about his claims.

Boudry also states that ID Creationists have exploited IMN, citing Philip Johnson's rhetorical claim that, if science is about following the evidence wherever it leads, scientists should not exclude a priori the possibility of discovering evidence for the supernatural and concludes that IMN is actually grist to the ID mill. This claim is, I think, historically incorrect, in that Johnson was making this claim in support of "theistic science" long before anyone actually formulated MN as a response to Johnson. Nor should we, I think, give up a helpful explanation of the actual practice of science just because creationists will exploit it. After all, what discussion of science, particularly in popular forums, haven't creationists twisted and tried to exploit?

My main question would be why are IMN and QMN incompatible? I would agree that the dividends of naturalistic explanation and the consistent failure of supernatural explanations in the history of science are good grounds for adopting MN. Why wouldn't it be good grounds for defining science as including MN? Sure, maybe someday, someway, we will come to change our definition of science but, in the absence of some plausible scenario of why we might do so, why should we worry about that?

That's about all I can say about Boudry's claims but Larry's claims are another matter. Larry makes two basic assertions:

1. If God exists, then there's no reason why supernatural explanations can't be a legitimate part of science.
The problem with this is that any proposed cause that has unknown properties and that acts with no known regularity is uninvestigatable. Suppose I was to propose that the timing of the radioactive decay of any uranium atom in the universe is controlled by a force I call "grumpus." However, I maintain that grumpus cannot be detected by any scientific instrument and does not act by any pattern of regularity. How could anyone test whether grumpus exists?

Now maybe some people claim that grumpus does act, at least sometimes, with regularity and others claim that they can detect grumpus. But if you show that there is no pattern in the decay of uranium atoms and the claimed detection of grumpus was in error, all you've done is proven those people's claims are in error, not that grumpus doesn't exist or that it doesn't act in ways we can't detect.

Of course, you can reject grumpus as a scientific concept on the basis that there is no empiric evidence for it ... but that's just MN.

These problems exists in spades with an infinite, omnipotent and omniscient being with unknown motives and aims.

2. How do draw the line between methodological naturalism and philosophical naturalism? Obviously, there's no difference for an atheist; in fact, the distinction seems rather silly. If supernatural explanations are never found to be necessary in explaining the natural world then doesn't it make sense to conclude that fairies and Santa Claus don't exist?
Not everything that it "makes sense" to conclude is science. It is reasonable to conclude that politicians are liars and/or crooks based on past experience. That is not (without more) a scientific result. Science implies a systematic study of an empiric phenomena under conditions as controlled as possible to eliminate other possible factors than those being studied. Larry's vague "sense" that the world works in ways that make fairies and Santa impossible is not a scientific result and, given a paucity of terms to describe "Larry's "sense," I'd call it "philosophy."

The distinction between the two seems to me to be pretty easy to recognize (though, as a plethora of philosophers have discovered, to their dismay, pretty hard to define). Someone who is doing rigorous science and (as rigorous science demands) reporting only the clear empiric results of her observations and experiments and an interpretation of those results in light of a well-worked-out theoretical structure (that is empirically testable in many ways) without overstating her results and conclusions is probably doing science and will automatically be practicing MN. Someone who is making broad conclusions with little or no reference to specific evidence (especially when appealing to "sense") is almost certainly doing philosophy.
.

Comments:
"If supernatural explanations are never found to be necessary in explaining the natural world then doesn't it make sense to conclude that fairies and Santa Claus don't exist?"

I think Larry isn't referring to some vague and ill-defined "sense" like spider man's spider sense but good logic necessary for science. If something is never found to be a necessary part of the explanation for anything it makes scientific sense to conclude it does not exist.

Take your grumpus example; I conclude it does not exist because despite than the fact that you posit it's existence there is no evidence for it. One could make up non-testable pretend things by the dozens (hundreds if I was feeling really inspired), and rejecting these would be good methodological naturalism, just like you said. However, I feel it wouldn't just be a philosophical rejection but a scientific one.

I found your politician analogy rather flawed. That assertion is obviously testable.

"Science implies a systematic study of an empiric phenomena under conditions as controlled as possible to eliminate other possible factors than those being studied."

Yup, apply those scientific principles to Santa and the Tooth Fairy. I have studied the phenomena, and concluded that while they do exist in culture, it makes no scientific sense to conclude that they exist in reality. I could start with the laws of physics precluding Santa's xmas eve escapades... Scientifically testable and falsifiable ideas, both of them.
 
Or, to be a bit more succinct, I think this is a scientific position, not a philosophical one:

If you're going to propose something that sounds like you just made some shit up, you have to prove it to me, I don't have to disprove it to you.
 
I have studied the phenomena, and concluded that while they do exist in culture, it makes no scientific sense to conclude that they exist in reality.

Good. Where did you publish your paper for peer review?

I could start with the laws of physics precluding Santa's xmas eve escapades... Scientifically testable and falsifiable ideas, both of them.

Now where is your detailed empiric study showing that the laws of physics always hold?

The point here is that there is an activity that we call "science" that is precise and operates by certain methods. Simply appealing to "the laws of physics" does not make your musings "science," any more than Stephen Meyer appealing to the "inference to the best explanation" makes ID "science." Nor is the fact that good logic is necessary for science mean that all logic, even when it appeals to the "facts" produced by science, is science.
 
P.S. I found your politician analogy rather flawed. That assertion is obviously testable.

That's why I said "without more." Until you do the testing any conclusion that is drawn is not "science."
 
I'm coming around to the idea that we make too much of the idea of evidence/no evidence, truth/falsity, existence/non-existence. In the case of methodological naturalism I'd go for the idea of 'merit' rather than 'truth' - where merit is the rather more fuzzy idea of 'potential utility'.

So in the case of the posited grumpus - does the idea have merit? If the concept of grumpus can explain more than just the original radioactive decay and if it can fit in to the wider body of scientific knowledge then it has some potential utility. It might not yet be accepted as 'truth', but it has some utility in explaining what we see by other means. If we find out how grumpus works then it has greater merit still.

If we find no use for the concept of grumpus, or we find a more useful explanation for radioactive decay, then grumps has no merit and is excluded from the scientific method. Substitute 'phlogiston' for 'grumpus' to see how this works in science.

On the other hand, if we decide that grumpus does not exist (because there is no support for the consequences of its possible existence) then it is excluded from our philosophy.

So the difference between methodological naturalism and philosophical naturalism is one of potential utility versus potential existence.

In the case of god/gods/faeries/ghosts - scientists ask if the 'supernatural' concepts have some use? If so they may exist. Philosophers ask if the 'supernatural being' exists what would the consequences be? People of faith or superstion ask what they should do because the 'supernatural' beings exist.

I would argue that scientists do not exclude the supernatural because of their beliefs. Scientists will consider any concept that has potential utility. It's just that as science has developed a more detailed understanding of the world around us the standard of validation of new concepts has become much more rigorous.
 
So the difference between methodological naturalism and philosophical naturalism is one of potential utility versus potential existence.

That may be a helpful way to approach it, though I'll have to think about it some more.
 
I had a longer comment brewing, but James McGrath in the second comment at your Sandwalk link pretty much nailed it for me.

Which gives me an opportunity to compliment you on your links. I use some of them regularly.
 
Wait! I've got it!

Methodological naturalism: Assumes natural explanations

Philosophical naturalism: Includes supernatural explanations so that, when none are found, that lack of evidence for the supernatural can be used to justify their position on religion.
 
"Now where is your detailed empiric study showing that the laws of physics always hold? "

If they don't hold, then obviously they aren't the laws of physics, are they?

You still don't answer my point that it is a scientific position to ask for proof. That it makes sense to ask for proof. If you claim Santa, I don't have to publish a peer-reviewed paper that says Bullshit. YOU have to publish a peer-reviewed paper that proves Santa.

Where's your peer-reviewed paper proving Grumpus? Where is your peer-reviewed paper scientifically proving any supernatural concept? Could it be that believing these don't exist is a scientific position, not a philosophical one? You would seem to imply that it is.

If there is no legitimate scientific endeavor to prove the supernatural it makes rigorous scientific sense to be skeptical of it's existence. Again, the burden of proof is on the positive assertion: "Until you do the testing any conclusion that is drawn is not "science." The default position of science is skepticism. Science is a process, not just a collection of results, and I think you are skating around this. Is the scientific method philosophical naturalism?

"Nor is the fact that good logic is necessary for science mean that all logic, even when it appeals to the "facts" produced by science, is science."

Are you seriously arguing that applying logic to the facts produced by science isn't science? So do you believe natural experiments aren't really scientific endeavors? Theoretical physics? Well, maybe I'd give you string theory...

I also think it was a bit of a cheap shot to use "appeals" instead of "applies" and to put "facts" in scare quotes. Do you not believe in these "facts" produced by science? Do I overuse rhetorical questions?

You differentiate between good logic and all logic. I see your point, some things might be a bit of a logical gray area, but most is fairly cut and dried, especially when applied to science. I know the IDers and similar ilk like to use the word a lot, and we should all be cautious of seeing it thrown around. But the peer-review process is based on a logical scrutiny of methods, evidence, and interpretations of evidence. Logical interpretations of evidence.

I doubt I will change your mind, or you mine, but I really see no difference between MN and PN. The things that are supposedly outside the domain of MN bear all the hallmarks of just being made up.
 
Stupid Blogger make me break this up. Part 1:

If they don't hold, then obviously they aren't the laws of physics, are they?

Five words:

Hume, the problem of induction.

You still don't answer my point that it is a scientific position to ask for proof. That it makes sense to ask for proof. If you claim Santa, I don't have to publish a peer-reviewed paper that says Bullshit. YOU have to publish a peer-reviewed paper that proves Santa.

Yes, to ask for "proof" (not a good word) is appropriate for science. As I said, rejecting "grumpus" as a scientific premise is appropriate as part on MN. To say that the lack of scientific evidence, when never rigorously sought, is grounds for rejecting any and all ideas is, at the least, philosophically naive.

Where's your peer-reviewed paper proving Grumpus? Where is your peer-reviewed paper scientifically proving any supernatural concept?

I'm not claiming it is scientific.

Could it be that believing these don't exist is a scientific position, not a philosophical one? You would seem to imply that it is.

Nope. Not "believing" ... without specific evidence ... is not scientific.

If there is no legitimate scientific endeavor to prove the supernatural it makes rigorous scientific sense to be skeptical of it's existence.

That's just a statement of MN.

Again, the burden of proof is on the positive assertion: "Until you do the testing any conclusion that is drawn is not "science."

Yes, that's MN (or close enough). But that also goes for the assertion that there is no Santa.

The default position of science is skepticism. Science is a process, not just a collection of results, and I think you are skating around this. Is the scientific method philosophical naturalism?

But skepticism is not grounds for justified "belief." Or else the belief that evolution was not possible was scientifically justified before Darwin (and a thousand other examples). The point of MN is that "justified scientific belief" is open ended and related to our ability to test our beliefs empirically.
 
Part II:

"Nor is the fact that good logic is necessary for science mean that all logic, even when it appeals to the "facts" produced by science, is science."

Are you seriously arguing that applying logic to the facts produced by science isn't science? So do you believe natural experiments aren't really scientific endeavors? Theoretical physics? Well, maybe I'd give you string theory...


Until you give me a reason to distinguish the claim that IDers make, that they are applying logic to scientific facts, from your claims to be doing the same thing (other than "they're wrong," which, of course, they'll say about you), no, I would not agree that that science is simply applying "logic" to scientific facts. Much of what science tells us about the world is not "logical" (as humans define it). Science, while it necessarily uses logic, is better than that ... it's empiric.

I also think it was a bit of a cheap shot to use "appeals" instead of "applies" and to put "facts" in scare quotes. Do you not believe in these "facts" produced by science?

No, claims are "appeals" until demonstrated. As to "facts," what they are is probably the hardest question in all of philosophy. Again, see Hume. I almost always scare quote facts for just that reason.

Do I overuse rhetorical questions?

Naw! I am very guilty of it myself. (BTW, I'm enjoying this.)

You differentiate between good logic and all logic. I see your point, some things might be a bit of a logical gray area, but most is fairly cut and dried, especially when applied to science. I know the IDers and similar ilk like to use the word a lot, and we should all be cautious of seeing it thrown around. But the peer-review process is based on a logical scrutiny of methods, evidence, and interpretations of evidence. Logical interpretations of evidence.

But that's what I want ... for "science" to be understood as the cut and dried bits and for the other arguments to be, if not anything I'd or you'd agree with an IDer about, at least something more objective than what they're doing.

I doubt I will change your mind, or you mine, but I really see no difference between MN and PN. The things that are supposedly outside the domain of MN bear all the hallmarks of just being made up.

I agree wholeheartedly with the conclusion ... any difference is about how we get there.
 
John,

I posted some comments over at Sandwalk that you may be interested in.

The short version is basicall:

1. Supernaturalist concepts are generally falsifiable, unless and until you contort them to make them unfalsifiabe---but you can do that with pretty much any hypothesis, even geocentrism, and that's not a bar to studying them scientifically.

2. In science we are not generally neutral toward unfalsifiable hypotheses. We are rightly quite skeptical of of them, because most of them are wrong---or worse, not even wrong.

3. We are extremely skeptical of unfalsifiable hypotheses that systematically resist falsification, even in principle, and most especially those that show signs of being contrived precisely to avoid falsification. We usually provisionally discard such theories as probably wrong, or worse; we are not neutral toward them.

4. Many central religious hypotheses are evidently false, in light of modern science. (E.g., the traditional substance dualist soul; it turns out the brain does what the soul was hypothesized to do). There are unfalsifiable variants of those, but they are unpersuasive and contrived.

The question is why we should treat supernaturalist/religious hypotheses differently from other unfalsifiable hypotheses, which we often regard as wrong or worse.

Like Larry, I don't think we should. The usual excuse for doing so is based on basic misrepresentations of what "supernatural" means, and what the significance of unfalsifiability is.


-- Paul W.
 
1. Supernaturalist concepts are generally falsifiable, unless and until you contort them to make them unfalsifiabe---but you can do that with pretty much any hypothesis, even geocentrism, and that's not a bar to studying them scientifically.

So, tell me how you can tell a "miraculous" point mutation from a "natural" one? If you can't, then all of evolution is open to "guidance" by an omniscient "designer." Science ignores that possibility without attempting to test it. That is a philosophical choice that is part of MN. Some religious claims are testable ... the age of the Earth, for example ... but many supernatural claims are not, including the Omphlaos "hypothesis" that God made the world with "the appearance of age." We reject those claims within science on the basis of philosophy, not testing.

2. In science we are not generally neutral toward unfalsifiable hypotheses. We are rightly quite skeptical of of them, because most of them are wrong---or worse, not even wrong.

I agree. That's MN, however.

3. We are extremely skeptical of unfalsifiable hypotheses that systematically resist falsification, even in principle, and most especially those that show signs of being contrived precisely to avoid falsification. We usually provisionally discard such theories as probably wrong, or worse; we are not neutral toward them.

Again, MN.

4. Many central religious hypotheses are evidently false, in light of modern science. (E.g., the traditional substance dualist soul; it turns out the brain does what the soul was hypothesized to do). There are unfalsifiable variants of those, but they are unpersuasive and contrived.

As a scientific proposition they are, but "unpersuasive and contrived" is an a priori judgment not an empiric result. Again, this is MN.

The question is why we should treat supernaturalist/religious hypotheses differently from other unfalsifiable hypotheses, which we often regard as wrong or worse.

Like Larry, I don't think we should. The usual excuse for doing so is based on basic misrepresentations of what "supernatural" means, and what the significance of unfalsifiability is.


"Supernatural" is just that set (quite possibly empty) of causes and phenomena that are not "natural." As I said over at Larry's, the harder problem is defining "natural."

One of the main problems with Popper's version of "falsifiability" is that it does not reliably distinguish science from non-science. What you are doing, in my opinion, is philosophy (in this case, the philosophy of science), not science itself ... which is not to say it is wrong, just not science.
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

. . . . .

Organizations

Links
How to Support Science Education
archives