I miss Sir Karl!

Do you ever hit one of those points when you are just weary of some of the things with which it is your task to deal? I hit one of those; possibly the mountain of grading has something to do with it. But some of the common problems my students have run into are rather close to some of the problems I keep bumping into in discussions of science among the public at large.

Let me explain. My Philosophy of Science students have been reading Karl Popper and trying to get clear on how he drew the line between science and the other stuff. To help them do this, I’ve been having them do a group task where they take a suspicious hypothesis or cluster of hypotheses and see if there’s any reasonable way to test them. The Popperian line, of course, is that a test of an hypothesis is an attempt to falsify it. So the trick is to figure out some things that you ought to be able to observe under test conditions if the hypothesis is true, then set up the test conditions and see if you observe them.

But my students get distracted by the suspiciousness of the hypotheses. Give them a scenario with claims about alien abduction and whether or not there’s a plausible way to test the claims, they’ll go right to aliens → pseudo-science → untestable. In other words, if the hypothesis is one they’d reflexively reject, they tend to reject it immediately based on its implausibility rather than finding a way to get empirical evidence that would give them good grounds for rejecting it.

They forget Popper’s point that real science is often more implausible than pseudo-science. Seriously, put the implications of quantum mechanical theory up against the daily horoscope next to the comics. But QM makes testable predictions – there are particular outcomes that just wouldn’t fit with the theory. On the other hand, it’s pretty hard to construct a day that couldn’t somehow be reconciled with the “predictions” of the newspaper horoscope. If we get what’s distinctive about science, says Popper, we see that plausibility is not the essential thing; falsifiability is.

(Yeah, I know, there are shortcomings with Popper’s picture of science. We get to those in the course, soon. But it would please me if they understood Popper’s picture before we take it apart.)

This is where my students’ muddle with Popper shines some light on popular debates about science that are starting to get old (like the Intelligent Design vs. evolution slugfest). But the connection didn’t really click into place until I read this post at Generalizations, etc. There, Peter writes:

And why do they [the “intelligent design people”] get under my skin? Because I'm quaking in my boots that they will succeed in demolishing my precious covert religion of Scientific Materialism? Not likely. The problem is that I like science. It's an extraordinarily effective method for wringing useful information out of our vast and bewildering universe. Do you catch my drift? The problem here is that science is a method while "intelligent design" is a conclusion. If "intelligent design" were to enter the biology curriculum, the method would be circumvented to favor a conclusion that is dubious at best.

Part of the public discourse about evolution vs. ID has been whether it’s more plausible that the variety of life on this planet emerged from the operation of natural selection or from the action of an intelligent designer. But, see, plausibility is not what makes something science. Rather, the deal with science is that it uses a method of testing. You don’t get the conclusions — even the plausible ones — for free. You only get what you can support with the method. Popper said you get the very best support when you try to knock your hypothesis out of the water and the experimental results keep lining up as if the hypothesis were true rather than false. (I gotta think Popper was a fun bofriend.) And, even if it’s really well supported, you may have to set your conclusion aside in the light of new evidence that falsifies your theory. (Newtonian physics, anyone?) On the other hand, a theory that seems implausible and takes a hit when early attempts to test it are not helpful, like cold fusion, always have a shot at making a comeback if serious application of the method comes out in the theory’s favor.

So, if the public (or at least my students) could get clear on the distinction between what one believes and what would count as good, scientific reasons for believing it (grounded in the scientific method), maybe some apparent disagreements would be cleared up and I could feel less bone-tired than I do right now.

Edited to add: Ed Brayton gives an excellent analysis of an unfalsifiable hypothesis and why there is no conceivable way it could be falsified. Doesn't mean it couldn't be true, or that one couldn't believe it. But you can't use scientific evidence or methodology as your justification for believing it.


Technorati tags: scientific method, Popper.