The Unraveling of String Theory

That lack of specificity hasn't slowed down the string folks. Maybe, they've argued, there really are an infinite number of universes--an idea that's currently in vogue among some astronomers as well--and some version of the theory describes each of them. That means any prediction, however outlandish, has a chance of being valid for at least one universe, and no prediction, however sensible, might be valid for all of them.

That sort of reasoning drives critics up the wall. It was bad enough, they say, when string theorists treated nonbelievers as though they were a little slow-witted. Now, it seems, at least some superstring advocates are ready to abandon the essential definition of science itself on the basis that string theory is too important to be hampered by old-fashioned notions of experimental proof.

And it is that absence of proof that is perhaps most damning. Physicists have a tolerance for theory; indeed, unless you were there to witness a phenomenon yourself--the Big Bang, say--it will always be, at some level, hypothetical. But the slow accretion of data and evidence eventually eliminates reasonable doubt. Not so--or at least not yet--with strings.

"It's fine to propose speculative ideas," says Woit, "but if they can't be tested, they're not science." To borrow the withering dismissal coined by the great physicist Wolfgang Pauli, they don't even rise to the level of being wrong. That, says Sean Carroll of the University of Chicago, who has worked on strings, is unfortunate. "I wish string theorists would take the goal of connecting to experiment more seriously," he says. "It's true that nobody has any good idea of how to test string theory, but who's to say someone won't wake up tomorrow morning and think of one? The reason so many people keep working on it is that, whatever its flaws, the theory is still more promising than any other approach we have."