Preposterous Universe

 
GPB

One of the great things about the internet (one of the things, anyway; great or not is up for dispute) is that I can surround myself by only those voices that I want to hear. For example, I tend not to spend much time lingering at The Corner at National Review Online; I'll trust others to point to anything interesting that goes on over there.

Impearls has just pointed to one such thing: a short dialogue between Peter Robinson and John Derbyshire on the imminent launch of Gravity Probe B, a satellite designed to test general relativity. This satellite has a long and checkered history; scientists started working on the idea in the late Sixties, and it has since grown over-budget and out of control, but has been kept afloat by the persistence of its supporters in lobbying Congress. It is infamous for suffering delays in being launched, and we are in the midst of one such delay as I type this; the current plan is to launch on the morning of April 20th.

The Robinson/Derbyshire discussion was more interesting than I would have thought. (I've honestly never read anything by these guys before, although their names are familiar from rants by people I do read.) Derbyshire clearly knows something about the science, and gives an essentially-correct account of the usefulness of testing general relativity. (Okay, a quick lookup on the web: Derbyshire is the author of a popular-level book on the Riemann hypothesis, and also a self-described racist and homophobe. Takes all kinds, I guess.)

Robinson, on the other hand, is pretty much in the dark, but in a very revealing way. In the very first post in the exchange, he expresses surprise that there would be any such thing as experimental relativity:

Test Einstein? When I first learned about all this a few days ago, the idea shocked me. Einstein, I'd always assumed, must already have been proven correct. But apparently not, or at least not completely. In any event, full marks to the old man for providing us with a theory that meets Karl Popper's test of falsifiability, which is a lot more than Freud or, as best I can make it out, Darwin ever did.

Whew, where to begin? Robinson seems to have deep-seated misconceptions about how science works, ones that are very revealing in the context of debates about evolution. It is interesting to see Popper's criterion of falsifiability brandished with such confidence, when he clearly misses the most fundamental point of Popper's philosophy: the reason why "falsifiability" is a necessary quality is because it is impossible to "prove" any scientific theory correct. You can do all the experiments you like, and have them be perfectly consistent with your theory, but then someone tomorrow does a new experiment which isn't.

Not only can we not prove Einstein's theory correct, nobody believes that it actually is perfectly correct. It's not consistent with quantum mechanics (see long-promised post that hasn't yet been written), at the very least. But it could easily break down much sooner than we might expect from considerations of quantum gravity, which is why we need new tests in every regime. General relativity isn't sacrosanct, it's something people try to modify all the time. I've done it myself. (Another great thing about the internet -- the effortlessness of shameless plugs.)

It was also amusing to see Freud and Darwin mixed together. Of course Darwin's theory is eminently falsifiable, in a million possible ways. There could simply be no evidence whatsoever for alteration in the fossil record through time, or species could never appear or disappear, or we could demonstrably pass on developed characteristics to our offspring, or the necessary timescales for evolution could be much larger than the age of the Earth or the universe. None of these is true, though. Here we see one of the big reasons why it is so frustrating to argue with anti-evolutionists. Real scientists know perfectly well that all of our theories are approximations, and that every theory should be tested, and that no theory is perfect; but they are comfortable with the existence of different levels of approximate truth, and understand that evolution and the Big Bang and general relativity can be "true" in a way that doesn't mean they can't be improved upon. Anti-evolutionists cling to a more Manichean view of scientific knowledge, in which evolution isn't all that great because it's "just a theory." Like what isn't?

Meanwhile, hopefully Gravity Probe B will be successfully launched by the time you read this. They have constructed the world's most precise gyroscopes, to look for a subtle effect known as "frame-dragging." To be honest, the experiment will only yield modest improvements over existing limits on deviations from general relativity; that's why an anonymous physicist was opining that the only possible outcomes are agreement with GR, or nobody believing the experiment. Maybe; we'll have to see. This particular satellite cost a lot of money, much of which was appropriated outside the conventional peer-review process. If anything, the lesson worth emphasizing is that prioritization of scientific projects is a task best left to scientists, not to politicians. Once a satellite is on the launch pad, however, we can all cross our fingers and hope for some impressive new results.

Update: It's been launched! First science results are expected late in 2005, if all goes well.

~ posted by Sean Carroll at 10:12 PM ~ link ~ ~ ~