Today we considered the question of the role of the social sciences. Social sciences sometimes need to allow exceptions to their laws. Hard, or natural sciences don’t. Does this make the hard sciences better?
Alexander Rosenberg thinks so. He criticizes economics for two reasons. The first is that it is ‘extremal’. All theories have core hypotheses and auxiliary hypotheses. But extremal theories have just one core hypothesis, and they apply it to everything. Extremal economists think that everyone acts to maximize utility, just as extremal Darwinists think that every organism maximizes its survivability. There’s nothing wrong with extremal theories, but they are hard to falsify. That’s because if the economist does not get the results he expects, he has to chose between changing his auxiliary hypothesis and giving up his core assumption which means giving up his whole theory. It’s always going to be easier to modify an auxiliary hypothesis. The result is that economists should be humble, because their views are hard to give up. They should be worried about the fact that they do not make good predictions.
The reason that economics doesn’t make good predictions, Rosenberg thinks, is that it is based on a mistaken view of human nature. Economists think that human beings are maximizers. We’re not! Economists should recognize that they are actually doing abstract math, not studying people.
Fritz Machlup has a much more optimistic view. He thinks that social sciences are almost equal to hard sciences. His innovative argument has to do with his distinction between the real world, the laboratory world, and theories. He thinks that if we recognize that social scientists can’t really test things in labs, we can see that social science is about equal to hard science. Both of them are poor at making predictions about the real world, and both are good in theory. Hard sciences shine in the lab - but social sciences don’t have a lab, so it’s not a fair point of comparison. And indeed, if we count all the adjustments to lab equipment as false predictions, hard sciences aren’t even that good in the lab.
The issue between these two philosophers is reductionism. Rosenberg thinks that we should reduce terms in the social sciences to simpler, more basic terms. Machlup thinks the social sciences do valuable work on their own. At the end of class, we considered the possibility that some things are multiply realizable, meaning that there are many ways to make up certain things. Prison wardens discover that there are uncountably many ways to make a weapon. If a physicist tried to study all the weapons, he would be looking at a huge number of different physical things. But doesn’t that show that when we reduce our social sciences we can lose track of what is important, which is in this case the multiply-realizable concept ‘weapon’?
Every week I'll add some information about what we discussed in class.