Philpapers Survey Results

The results of the philpapers survey are here. I was surprised by how low the proportion of compatibilists turned out to be: 59%. Libertarianism, no free will and 'other' don't differ significantly from each other, which is also surprising. Thoughts? In particular, what is the 'other' view which is so popular? Some kind of Double-style subjectivism? Or is this the result of over-scrupulousness on the part of semi-compatibilists? C'mon guys, John wrote the chapter on compatibilism in the Four Views book...

Comments

[Josh Shepherd]

Interesting stuff, especially when viewed on the 'fine' setting for results, and when viewed in terms of the AOS (which might explain what's going on with the 'other' voters).

Since God is right below Free Will, I couldn't resist checking those results against one another. And although this hardly shows much, it is interesting that two of the areas which reverse the atheism trend - phil of religion and medieval - see a significant upswing in libertarians as well, while areas which get above average atheist or leaning atheist votes - phil of biology, cog sci, and decision theory, for example - also have far fewer libertarians. One also sees a divergence from average in these groups on naturalism in metaphilosophy, and physicalism in mind.

Of course no direct relationship need be inferred. The correlation is vague, since the results are still vague, and you might think that the correlation is somewhat incidental, occurring more for reasons related to the subject matter, interests, use of terms and assumptions of those in these specific areas. Normative ethics respondents, for example, have higher than average atheist votes, but stay the same as average on the Free Will issue, while Continental Philosophy folk reverse the trend on naturalism and physicalism, but stay about average on Free Will.

Good grief I really should get back to work.

[tony]

Neil, if you switch to the "fine" view of the results, you get the break down of all the results classed as other. For the target faculty, the top four in that group are:
Agnostic/undecided (4%)
The question is too unclear to answer (2.7%)
Accept another alternative (1.7%)
Insufficiently familiar with the issue (1.7%)
Accept an intermediate view (1%)

So you really only have 2-3% taking some kind of unidentified other position; most people just seem unfamiliar or undecided.

[tony]

Er, top 5 in that group. I can count, I promise.

[Neal Tognazzini]

Other potentially interesting results:

1) 53% of those who self-identify as philosophers of action are compatibilists, whereas compatibilists account for 66% of meta-ethicists, 68% of normative ethicists, and only 55% of metaphysicians.

2) The 9 philosophers of action who took the metasurvey predicted with remarkable accuracy the number of respondents who would fall into each category.

[Kip]

"C'mon guys, John wrote the chapter on compatibilism in the Four Views book..."

LOL

[Eddy Nahmias]

Unlike Neil, I was surprised by how *high* the proportion of compatibilists was. If we look at the relative proportions of just compatibilists, libertarians, and skeptics among professional philosophers (taking out the "other" responses), they are, respectively: 69.5%, 16%, and 14%. That is, more than twice as many people accept the "compatibilist conception" of free will than the "libertarian conception" of free will (forgive me for using these terms if you agree with van Inwagen that doing so is a mistake).

One reason I was surprised by this number of compatibilists is that in the large-scale study Thomas Nadelhoffer, Trevor Kvaran, and I have done, when we look at professional philosophers and take out the "other" responses, we found: 60% compatibilists, 21% libertarians, and 19% skeptics. Also, see here: http://gfp.typepad.com/the_garden_of_forking_pat/2006/10/counting_heads.html

In our data we also see the same trend you see in the PhilPapers data: that the relative proportion of compatibilists goes up as you move from undergrads to grads to faculty.

Clear evidence that philosophical training can indeed help people discover the truth ;-)

[Tamler Sommers]

Most shocking to me is that Eddy just used a winking emoticon.

(Note: This comment should not be taken completely seriously. If only there were a way to denote that on a blog...)

[Neil]

It may be that my surprise reflects my Australian background: I suspect that the proportion of Australian philosophers who are not compatibilists is less than 10% (I will ask one of the Davids whether it is possible to filter results by nationality).

[Kip]

"In our data we also see the same trend you see in the PhilPapers data: that the relative proportion of compatibilists goes up as you move from undergrads to grads to faculty."

Clearly, philosophers are corrupting our youth.

[Neil]

Kip, Eddy, there is an alternative explanation (due to Stephen Stich, IIRC): philosophical education works as a filter, sieving out those with intuitions that diverge from those of the profession (presumably by rewarding those with the 'right' intuitions).

[djc]

Eddy: If you filter results by "All faculty or PhD" rather than "Target faculty", the percentage of compatibilists is a few percent lower, although still higher than your figure.

Neil: We'll add geographical filtering at some point, most likely next month.

[Daniel Cohen]

I have a worry about how the free will question was posed. Are 'compatibilism' and 'no free will' mutually exclusive? One might think that free will and determinism are compatible but that we have no free will for any number of other reasons. For instance, a compatibilist might think that determinism is necessary for free will and that given the likely truth of quantum mechanics we therefore have no free will. Or a compatibilist might think that free will requires something impossible like self-creation, and that we therefore lack free will. (So, it's not clear to me that Galen Strawson isn't both a free will denier as well as a compatibilist.)

Given the overlap between the options, it may be that there are more compatibilists than 59% or more free will sceptics than 12.2%.

[Kip]

Eddy's neuroticompatibilism is the best example of that possibility (as far as I know). Richard Double, at the end of The Non-Reality of Free Will, expresses similar concerns.

But they have worries based on psychology: compulsions, undue influence, etc.

Someone who adopts G. Strawson's view should not count as a compatibilist, as far as I am concerned. Strawson says that fw is impossible, that it is incompatible with indeterminism, and incompatible with determinism. It strikes me as very strange to say that the impossible free will is still, somehow, compatible with determinism.

Neil Levy often seems to write that way, though: he seems to have sympathies for both anti-realism about free will *and* for compatibilism, because he doesn't see determinism per se as any threat. But he knows his view much better than I do...

[Neil]

Kip, as I read him Neil Levy is a free will sceptic and a compatibilist. Some people object to this way of speaking, because they take "compatibilism" to be a success term. Neil Levy thins that because views which deny the existence of free will for reasons other than determinism have not been very prominent in the debate, "compatibilism" is indeterminate between "there is free will in at least one possible world in which determinism is true" and "there are no possible worlds in which there is free will, though there are possible worlds in which determinism is true and in which free will is not true for reasons that have no essential reference to this world". But I must admit that I cannot really claim to understand Neil Levy's view very thoroughly.

[Eddy Nahmias]

As far as I know, there has never been a compatibilist who denies that (at least some) humans have free will (though Neil may be the exception). If anyone knows of any other compatibilist skeptics, please let me know (so I stop making this claim in my publications!). From my perspective, this is incredible. It indicates an obsessive compulsion with determinism, as if that is the *only* possible thesis that could threaten human free will (hence the label: "The problem of free will and determinism"). Part of this may be because determinism is used by some as a cover-all term for *any* threat to free will (some contemporary skeptical scientists talk this way), rather than the specific thesis contemporary experts discuss. Part of it may be historical accident. In any case, it is a mistake.

Daniel and Kip note two ways one might be a skeptical compatibilist:

1. old-fashioned compatibilist who thinks free will requires determinism but indeterminism is true (Daniel, I don't think your other example works if it's meant to represent G. Strawson's reasoning--he is no compatibilist and his arguments rely on a Transfer principle)

2. compatibilist who thinks that the sciences of the mind show we lack free will (this is what I would be if I thought the sciences showed this, but I don't think they show it, though I do think they show we have limited free will, hence my neurotic compatibilism, or compatibilism by degrees).

But one could also think:
3. compatibilism is true but God's foreknowledge entails we have no free will (and he's got that foreknowledge independently of the possible truth of determinism).

4. compatibilism is true but epiphenomenalism or eliminativism about the relevant mental states is true, so we have no free will. (This may be the reasoning behind the skepticism of someone like Alex Rosenberg expressed in recent blog post--materialism entails eliminitivism which wipes out free will, even if free will could exist in deterministic but not materialist/eliminativist universe.)

I suspect the list goes on (e.g., compatibilism true but, independently, logical fatalism wipes out free will...)

But, since I don't think anyone actually is a compatibilist skeptic, I don't think the stats were likely to be skewed by that possibility. Hence, I take it that among professional philosophers, about 2/3 believe both (a) compatibilism and the *clearly independent* claim (b) humans have free will.

[djc]

We note in our discussion of the questions (scroll down to the free will question) that the options here aren't exclusive. But for what it's worth, only 3 out of 931 target faculty (although 23 out of 3226 in the overall population) selected "accept more than one". I recall that one of those had a comment saying that they accepted both compatibilism and libertarianism (we have libertarian free will but compatibilist free will is also possible).

Neil: Of the 117 Australian respondents overall, 60 favor compatibilism, 21 favor no free will, and 18 favor libertarianism. Of the 37 Australians among the target faculty, the figures are 23, 6, and 0 respectively.

[Kip]

It's worth remembering that, according to some research (which I think is in the right direction), compatibilists are more extroverted.

So, if the study was voluntary, we would probably expect more extroverted, and therefore compatibilist, people to respond. That might help explain the high compatibilist number. Either that, or philosophers are corrupting our youth.

[Mark Balaguer]

Regarding the idea of endorsing both compatibilism and libertarianism. Whether this is possible depends, of course, on how the two views are defined, but on one obvious way of proceeding, the two views are perfectly compatible. Suppose you thought that (a) the ordinary term 'free will' denotes some compatibilist kind of freedom, e.g., Humean freedom or whatever; and (b) we possess libertarian freedom (i.e., we make decisions that are undetermined, non-random, etc.--fill in all the requirements you think are necessary). Claim (a) looks sufficient for compatibilism because it entails that free will is compatible with determinism. Whether (b) is sufficient for libertarianism depends on whether you define that view as involving the idea that libertarian freedom captures the ordinary notion of free will. If you do, this person is no libertarian. But if you define libertarianism as the view that humans have libertarian freedom, then this person is a libertarian compatibilist.