The PhilPapers survey results came out last week, and there’s already been a fair amount of discussion of them on the web. They seem to raise two big issues in particular: the obvious is something unsurprising, but that may never have been documented before: philosophers specializing in philosophy of religion, as a class, differ in their beliefs systematically from philosophers in other disciplines. The other is what these results mean for discussions of how we should let experts and peers influence our beliefs.

The Believers

People familiar with analytic philosophy shouldn’t be surprised how the God question turned out: 72.8% of the targeted faculty in general picked atheism, but 72.3% of the philosophers of religion picked theism. This result is easy enough to explain: mostly people go into philosophy of religion because they believe in God. Atheist analytic philosophers tend to see the god question, in contrast as a “settled issue” (as one of my undergraduate professors told me).

Before seriously discussing this issue, I should deal with a few rather silly things a couple people have said about the issue. One is Edward Feser’s post hinting that only someone blinded by their atheism would think that the stats are because theists are much more likely than other people to enter philosophy of religion, and that really, and we should think most philosophers of religion are theists because they’re experts. This is terribly silly for anyone who’s paid attention to what’s gone on in the philosophy world in the last few decades: lots of top phil. religion people disavow having been led to their position by arguments, and the phil religion revival of the last several decades was led by Alvin Plantinga, who’s probably best-known for arguing theists don’t need arguments to support their beliefs.

Similarly silly is Kelly James Clark’s claim that the fact that there are more theist grad students than professors is evidence of anti-theistic bias in hiring. Or: evidence of moderate demographic change between generations.

(Three side points: (1) Clark might be right that there’s other evidence of hiring bias (2) Leiter definitely is right that there’s nothing wrong with this in principle (3) Departments should be happy to hire theists with generally worthwhile things to say about religion, and shouldn’t be afraid to hire people with purely silly religious beliefs if it looks like it won’t impact their academic work.)

Some people–particularly Trent Dougherty and Alexander Pruss, have said not-stupid things about the raw, theism/atheism numbers. But what I want to focus on is something different: God isn’t the only question where there was a dramatic flip between faculty in general and religion people in particular. Here’s a run down –realize that in all these cases, “other” was a popular option, so even getting 50% of the votes means a lot:

59% of the target faculty said free will is compatible with determinism, but 57.4% of philosophers of religion said free will exists in a sense incompatible with determinism (libertarianism). 49.8% of the target faculty supported a naturalistic metaphilosophy, while 57.4% of philosophers of religion embraced metaphilosophical non-naturalism. 56.4% of the target faculty went for physicalism, while 63.8% of the philosophers of religion went for non-physicalism.

On other questions, there was no clear favorite in either group, but there were still smaller flips: rationalism did poorly among the target faculty, significantly better among philosophers of religion. Consequentialism beat virtue ethics among the target faculty, but this flipped among philosophers of religion. “Further fact” views of personal identity (a phrase I assume was taken to refer to the soul) finished last among the target faculty, first among philosophers of religion. The target faculty tended to say that in the teletransporter (new matter) case you survive, philosophers of religion tended to say you die. A-theories of time were a lot more popular among philosophers of religion. Finally, the target population tended to say that philosophical zombies are conceivable but not metaphyisically possible, but philosophers of religion tended to say they’re metaphysically possible.

There was also a set of questions where the target faculty and the philosophers of religion had a shared favorite view, but the win was more decisive among philosophers of religion by 10-25 percentage points. These views are: the existence of a priori knowledge, objective aesthetic value, the existence of the analytic-synthetic distinction, non-Humean laws of nature, moral realism, and moral cognitivism.

I would be surprised if there were any other sub-groups in the profession that differed that systematically from the general target faculty, with the exception of fields known to attract similar people as philosophy of religion (like Medieval philosophy).

Now, what does it mean? Some connections are obvious–since “God” is typically defined in moral terms, if you believe in God, you’re almost certainly going to want to be a moral realist (though only “almost”–I have a friend who’s a theist but a non-cognitist about morality). Proposed explanations for other connections are out there–Neil Levy thinks that theists tend to be libertarians because it’s perceived as a necessary part of solving the problem of evil.

I think there’s something deeper here, though. These stats may be indicative of the fact that the kinds of people who go into philosophy of religion–and tend to take religious thought seriously in general–tend to have a generalized dissatisfaction with the common stereotype of a naturalistic worldview: a dissatisfaction with the kind of free will that seems to make sense in that stereotype, a dissatisfaction with anything short of non-natural moral realism, a dissatisfaction with the idea that humans mere “clouds of atoms” (as C.S. Lewis once put it), and so on. And sometimes, you see the connection make explicitly running from these other things to theism: if morality is objective, or if the human mind is non-physical, then God must exist. It’s a horrible non-sequitur, but it gives you a glimpse of the mindset involved.

Peers and Experts

This ties in with my pre-results post on the survey. In Robin Hanson’s post on the subject, he was very proud to find out that the survey results were largely in line with his beliefs, and more generally discussed the possibility of non-experts using these survey results to guide their beliefs. This seems to radically misread the survey results insofar as it misses this about them: while on some questions, one side won in a landslide (non-skeptical external world realism, atheism, and scientific realism are the clearest cases, though you might also include the a-priori and classical logic) lots of issues were really close calls: (abstract objects, aesthetic value, empiricism/rationalism, normative ethics, proper names, teletransporter). If you believe, as Robin does, that in cases of disagreement we should incline towards agnosticism, you should think we should be agnostic about a lot of those philosophical questions.

Peter van Inwagen has used the fact that its hard to get philosophers to agree on much of anything as a point against views on disagreement like the one that Robin holds, and concludes that we shouldn’t feel bad about having strongly-held views on issues where we can’t get most other people to agree with us. At first glance, the results of the survey seem to support this position: the only cases where the “agnosticism” option got more than 10% of the votes were the rather obscure questions of Newcomb’s problem and theories of time. But interestingly, on the cases that were split, philosophers were a lot more likely to pick a “lean” option than an “accept” option. I’m willing to bet that the unpopularity of “agnosticism” is an artifact of survey design: to proclaim agnosticism, you had to find it in the drop-down list under “other,” where as the “lean” and “accept” options were right out there. Based on this, I would think that the actual behavior of philosophers, though certainly not consistent with a strong conciliatory approach to disagreement, is closer to it than you would think reading van Inwagen.

Something to ruminate over in the comments: was putting the “agnostic” option in the “other” category good or bad survey design? Does it reflect an understandable desire to make the survey sensitive to slight inclinings in philosophical views, or a foolish failure to take seriously the idea of being agnostic about major philosophical questions?