The Chicago School of Philosophy

A defense

I recently watched for the second or third time a debate between David Friedman and Bob Murphy. The debate was about the Chicago school versus the Austrian school of economics. Friedman argued that our economic theories should be open to empirical tests, while Murphy argued that it should be based on purely logical derivations. He gave the analogy of geometry. One who proposed testing whether the square of the hypotenuse side is equal to the sum of the squares of the other two sides would be missing the point.

Friedman tells the story of his first ever published paper. It was about the size of nations. He presented it to George Stigler, the publisher of the paper, thinking that it was a good paper. Stigler rejected it on the grounds that there was no way to test it empirically. Friedman thought of a few ways of testing it empirically, concluded that it was empirically correct, wrote those up, and got published.

Friedman argues that this is how economics should be done. He cites the example of the minimum wage — there are possible assumptions on which the minimum wage benefits the poor. Friedman thinks that one needs an empirical test to figure out if those assumptions are, in fact, true.

I think we should have something like this for philosophy.

Now, I obviously don’t think that papers on platonism should be turned away because they don’t have empirical tests. However, I do think that empirical tests should be done much more in philosophy.

Nearly everything that’s discussed by philosophy is a-priori knowable. One can’t really empirically test nominalism. However, I think that there are a lot of cases where one can do some minimal empirical tests of their theory.

This is especially important in philosophy given another important insight by Friedman — philosophers don’t make very much progress. We’re still arguing about topics that have been debated for thousands of years. There is no idea in physics, for example, that lasts a thousand years unscathed.

This is because humans are very good at deluding themselves. Philosophers can come up with lots of clever arguments for very implausible conclusions. A lot of philosophy will come down to seemings — but people have very different seemings. There isn’t a great methodology in philosophy beyond ‘appeal to plausible-sounding principles.’ But one man’s modus ponens is another’s modus tollens — we should expect what sounds plausible to one to sound totally implausible to many others.

One might think that, because philosophical knowledge is a-priori, we can’t have any empirical tests of it. This is false! To consider one example, suppose every single philosopher finds some idea implausible. This, while a posteriori, should inform us that the idea is probably implausible.

To illustrate this methodology, I’ll apply it to my favorite philosophical view — utilitarianism. To be clear, the relevant pool of evidence is the things that one can deduce from facts about the world that should raise or lower our credence in utilitarianism.

If utilitarianism were correct, we’d expect a few things that we see.

1. People who are not ideologically committed to utilitarianism ending up finding utilitarian results about various cases irresistible. If utilitarianism were correct about what one should do in all cases, we’d expect that in cases where utilitarianism seems wrong it would turn out to be right. If this were so, we’d expect that, in fields like population ethics, non-utilitarians would end up finding the utilitarian results hard to resist. This turns out to be true — the repugnant conclusion is the best example of this.

2. More careful prolonged reflection would lead people to be more utilitarian. This is exactly what we see.

3. The early utilitarians would be far ahead of their time, having moral views that we discover to be true only centuries later. This is also what we see.

One thing counting against utilitarianism is that so many philosophers aren’t utilitarians. This is surprising on the hypothesis that utilitarianism is true.

I’d be in favor of more of this in philosophy. While imperfect, there are very often clever ways of testing philosophical views. Only rarely will the things about the world that one would expect if the view were true be the same as those if the view were false.