Falsifiability

In a post discussing conservative attempts to spin the UN bombing in Baghdad as a sign of terrorist desperation (via Yglesias), Josh Marshall said something that caught my eye.

I’m probably getting certain particulars of this wrong, but there’s a basic principle in scientific theory: an hypothesis, to be a real hypothesis, must be capable of disproof. In other words, for an hypothesis to be a valid basis for research, there must be some data which, if found to be true, would prove the hypothesis was false. Otherwise, there’s no way to test it.

Marshall is not wrong about this, but he’s a little fuzzy. The notion he’s referring to was common among the Logical Posivists, a species of mid-20th century anglo philosopher, and it held not just that an untestable hypothesis was invalid, but that it was meaningless–it literally bore no meaning. Karl Popper (who was closely aligned with the positivists, but broke with them over the issue of whether any hypothesis can be conclusively verified) is famous for advancing the notion of falsifiability: the idea that an hypothesis can never be taken as certain and verified, but that we can legitimately prefer certain hypotheses over others if they a) offer substantive predictions that can be falsified, and b) have not yet been falsified.

In short, the way we distinguish science from superstition, and good science from bad science, is that it includes benchmarks and predictions by which it can be proven wrong. It says, “I predict X; if X does not turn out to be the case, I’m wrong.”

What’s the point? Well, Marshall is referring to recent attempts by conservatives–see the “flypaper theory” and now the “they’re desperate” theory–to spin literally anything that happens in Iraq as good news. One begins to wonder what bad news would look like.

But I don’t think it’s spin. There is a tendency on the right, and I think it’s earnest, to respect ideological simplicity (sometimes known as “moral clarity”), the idea is that it’s a virtue to hold a few simple principles fast in the face of life’s inevitable cascade of confusing or ambiguous evidence. (Reagan is taken as the paradigm case; the right desperately wants to think of GW Bush this way.) Perhaps there is some virtue to this kind of principled simplicity, but it seems pretty clear that some conservatives use it as an excuse never to confront their own assumptions. You see it in Iraq. You see it in debates about punitive violence–‘Israeli reprisals haven’t stopped terrorism? launch bigger, harsher reprisals!’ You see it in debates about economic theory–‘tax cuts haven’t produced economic growth yet? launch bigger, deeper tax cuts!’

Not to point out the obvious, but: if any positive result reinforces your ideological positions, and any negative result reinforces your ideological positions, well… you’ve become insulated from any possible counter-evidence, any possible falsification or course correction, and you will blunder forward like a blind man.

One need not strain to imagine how such an administration would behave.