Courtesy of Teppo at orgtheory.net, a certain physicist dumping on the social sciences. I realize this is a tired critique, but it is really persistent, so it either hasn’t been properly answered by social scientists (unpossible!) or it pokes us at a weak spot in the armor. Or both.
Feynman’s main criticism here is that the social sciences don’t produce laws. I think that is a very peculiar argument, because I don’t think that having a law in hand is the proof of science. The pursuit of laws is certainly important, at least to a positivist like me, but God knows every social scientist in the world would love to discover some universal law for all societies. We do, however, have some stuff that works pretty well within particular scopes. Which makes Feynman’s argument sort of like saying that you can sprint all you want, but if don’t win a medal at the Olympics, you’re not actually running.
It’s also my belief that the social sciences are fundamentally a hell of a lot more difficult than the physical ones. I’m not saying that that makes sociologists smarter than physicists – the opposite is probably true in general – but it does mean that we’ve got a much harder task. We don’t get to smash people into their component parts in a lab environment, for one thing, and even if we could (and could get IRB approval), we probably wouldn’t get a $5 billion grant to do it. Add in that social scientists actually have to abide by ethical codes, and well, it’s just not that simple. It’s like meteorology – we’re studying a vast, complex system that is certainly governed by laws, but isn’t really comprehensible to humans through that particular lens. Except if we want to fly a instrument-laden plane into our hurricanes, we have to get its permission first.
However, much like meteorologists, I think that the increasing power of computer simulations will pave the way for better prediction. I just hope it comes in my lifetime, because I am rather Feynmanesque in my views of most of the methods we use.