(Note: This post has been updated (1) as of 2:59 PM Eastern on 25 Nov. See below.)
In a previous post on this blog discussing string theory, I wrote:
I confess I've a secret wish and suspicion that string theory with its fundamental requirement of a 10-dimensional space-time universe (as opposed to our 4-dimensional one; and N.B., string theory doesn't merely predict a 10-dimensional space-time universe; it requires it), and its fundamental non-disprovability will turn out ultimately to be the bust some few physicists think it will turn out to be. I intuitively don't like its messy, complicated, exercise-in-masturbatory-mathematics feel, in the same way I intuitively dislike the probabilistic basis of quantum mechanics. (I'm in good company vis-à-vis the latter. Einstein was unhappy with quantum mechanics for the same reason. It was what provoked his famous, "I cannot believe God plays dice with the world.") In short, from this mathophobic layman's vantage point, string theory seems but a complexly elaborate sandbox in which theoretical physicists can play about for decades without getting bored while seeming to be engaged in something on the verge of revealing the true nature of the reality underlying all the physical phenomena of the universe, and the true nature of the reality underlying the very universe itself from its birth to its ultimate death, if indeed it can be said to have either.
Physicist Lawrence Krauss, a professor of physics and astronomy at Case Western Reserve University, has his doubts as well:
Krauss' book [Hiding in the Mirror] is subtitled The Mysterious Allure of Extra Dimensions as a polite way of saying String Theory Is for Suckers. String theory, he explains, has a catch: Unlike relativity and quantum mechanics, it can't be tested. That is, no one has been able to devise a feasible experiment for which string theory predicts measurable results any different from what the current wisdom already says would happen. Scientific Method 101 says that if you can't run a test that might disprove your theory, you can't claim it as fact.
I haven't read the book, but Slate's Paul Boutin has, and reports on it here.
Worth a look.
Update (2:59 PM Eastern on 25 Nov): To correct a minor gaffe in Mr. Boutin's piece, his,
Elegance is a term theorists apply to formulas, like E=mc2, which are simple and symmetrical yet have great scope and power. The concept has become so associated with string theory that Nova's three-hour 2003 series on the topic [PBS] was titled The Elegant Universe....
is, of course, quite misleading, not to say flat-out in error in this context. The PBS Nova series was titled The Elegant Universe because the host's (Brian Greene) best-selling book of the same title is the source on which the series is based; is, in fact, what initially provoked the series.