What Philosophers Believe: Results from the 2020 PhilPapers Survey
Results from the 2020 PhilPapers survey, with responses from nearly 1,800 philosophers (mainly from North America, Europe, and Australasia), to questions on a variety of philosophical subjects and problems, have now been published.
In their commentary on the survey, David Bourget (Western University) and David Chalmers (NYU) explain the its value:
Surveys like this can play at least three roles within philosophy. First, today’s sociology is tomorrow’s history, and these results
may be of some use to future historians of philosophy. Second, philosophers often appeal to sociological claims about the distributions of views among philosophers, for example in justifying which views should be taken seriously, and it makes sense for these claims to be well-grounded. Third, if philosophy has any tendency to converge to the truth, then philosophers’ views might provide some guidance about the truth of philosophical views. It is not clear whether philosophy tends to converge to the truth, so we don’t make the third claim about guidance, but surveys can clearly play the first two roles in philosophical practice.
The survey asked 40 “main” questions and 60 “additional” questions.
Here are the results of the main questions:
Bourget and Chalmers note that the results reported above cannot be directly compared to the 2009 results, owing to the change in the population surveyed. Still, they noted some of the largest shifts in responses among those who took both surveys. These include swings:
- towards not switching in trolley problem (bystander)
- towards non-classical logic
- towards non-cognitivism in moral judgment
- away from invariantism regarding knowledge claims
- towards Platonism about abstract objects
- away from “no free will”.
See their write-up for a more detailed longitudinal analysis of philosophers’ views.
Here are a few results from the newly added questions:
- 81.7% think abortion is sometimes permissible
- 75.1% think capital punishment is impermissible
- The most popular view of consciousness is functionalism with 33% of respondents supporting it
- 51.3% accept some version of the extended mind thesis
- 50.9% endorse revising gender categories
- 40.4% endorse eliminating race categories
- 64.2% think human genetic engineering is sometimes permissible
- 44.9% would choose immortality, 41.3% would not, while 13.6 selected “other” as a response
- Naturalist realism edged out non-naturalist realism and constructivism as the most popular metaethics
- Almost 30% support capitalism, but 53% support socialism, while close to 20% choose “other”
- 56% think there is “a lot” of philosophical knowledge and 32.5% think there is “a little,” while paradox did not stop 3.6% from saying there is “none”
- Respondents were about evenly divided on whether time travel is metaphysically possible
- Only 2% thought particles could be conscious
Bourget and Chalmers also include information about which answers to different questions are correlated. For the complete set of results and additional analysis, go here.
Discussion welcome.
Note: The post was edited to make clear that the 2020 data cannot be directly compared to the 2009 data, owing to changes in the survey population.
Cool! Will we get to see the individual surveys of each participant at some point (as was the case with the earlier surveys)? Partially just out of curiosity and partially because I think it would be interesting to look at the stability of particular philosophers’ views over time (I see there’s an aggregate comparison available, but it would be interesting to see whose views have changed).
Respondents had the option of making their views public on PhilPeople. For those that agreed, their views were made public as part of their profiles this morning. Later we will release a searchable list of people with public views.
Professors Bourget and Chalmers don’t include in their list of valuable roles of this research ‘providing interesting and diverting browsing fodder for philosophical dilettantes’, but they have certainly achieved this too, so they have my gratitude.
Question – maybe I am totally off-base here, but people often seem to me to speak and write as if ‘Two Dogmas’ debunked the analytic / synthetic distinction. But this survey and the 2009 one both indicate that a substantial majority of contemporary philosophers accept this distinction. Am I just mistaken? (entirely possible); if not what’s up?
I think it depends on the circles you run in. True, there is a certain strand of naturalism, inspired by Quine that rejects the A/S distinction as “unclear” or “broken backed”, whatever that means. But there have always been resisters (right back to Grice and Strawson “In defense of a dogma”). My sense is that with the rise of work on Carnap and the history of analytic philosophy, many people now think that Quine’s criticisms don’t touch Carnapian style analyticity (see the work by Friedman, Richardson, Creath etc.) and also that a more Fregean approach to analyticity is salvageable against Quine’s objections, esp. if confirmation holism is problematic (see e.g, Sober’s paper “Quine’s two dogmas”). So, my sense is that many people who work in traditional epistemology/metaphysics/language accept it, and now also increasing numbers of philosophers of science who work broadly in the neo-Carnapian tradition. Other people just defer to their colleagues.
A slightly more common yet vague view, I conjecture, is this: there is an analytic-synthetic distinction, but it plays little to no explanatory role in philosophy, or much less of one that some philosophical movements (e.g. positivists, proponents of conceptual analysis) would have it. That would approximate Putnam’s view and Timothy Williamson’s view, I take it. Some would argue this even approximates Quine’s view, at least late in his career. (See Hylton 2021: https://onlinelibrary.wiley.com/doi/full/10.1111/nous.12321). This is a situation where these ‘yes’/’no’ poll questions understandably have their limits, although I don’t intend that as a criticism of them. I guess one could have answered ‘other’ for this view, but people can interpret that in different ways.
(Edit: meant in reply to Rollo Burgess)
If I remember correctly, Quine’s argument crucially depends on the tacit claim that necessity (as in necessary truths) reduces to analyticity. So any attempt to explain analyticity by necessity is circular.
If you don’t accept the idea that necessity reduces to analyticity, you are not faced with the circularity objection, and that might be the reason why people these days accept the analytic/synthetic distinction.
Interesting results again in many ways. Some of the results seem pretty puzzling, though.
For example, 62.1% of respondents support moral realism in meta-ethics—yet 50.2% support naturalism in metaphilosophy and only 31.1% support non-naturalism.
This seems odd on its face, given that many (and probably most, though by no means all) moral realists are non-naturalists about moral properties. Perhaps most respondents are attracted to naturalist forms of moral realism, then? It would be really neat to find out about these kinds of details in future iterations of the survey.
hi marcus — those details are on the survey website, e.g. at https://survey2020.philpeople.org/survey/results/4866 (under correlations) and also https://survey2020.philpeople.org/survey/results/correlations?category=Philosophical%20Questions&variable=Morality%3A%20naturalist%20realism. to a first approximation, moral realists are fairly evenly divided on moral naturalism vs. non-naturalism.
Cool, thanks so much – those kinds of comparisons are fantastic!
Honest question: do you (=the person reading this) have to google these positions to learn what they are? I’ve been in grad school for 7 years, and have hardly heard of any of them. Is there some kind of quick and dirty reference guide somewhere for what these positions are? I took the survey last year and selected “other” for almost every answer since I didn’t know what the options really meant (slightly alienating to experience, oh well).
Also, I’m not looking for SEP articles or an introduction to the debates, but a list of positions with a one line description and the field they are located in.
I’ve been in graduate school for two years, and with each of these questions I am either familiar with the debate or feel confident that I am able to piece together in extremely elementary terms what the debate is about; in the latter cases this is mostly based on the connotations of the provided potential responses (the significant exception being the sleeping beauty problem, which I recognize by name only). I am also markedly less confident in my ability to adequately reconstruct the question about environmental ethics. I hope this honest response is helpful.
This may still be too much detail for what you want, but a good dictionary of philosophy, such as Audi’s The Cambridge Dictionary of Philosophy, has something closer to one line descriptions of the positions rather than full blown introductory essays a la the SEP. What you might really be asking is did Chalmers or Bourget include one line statements somewhere – or shouldn’t they have included them – with the survey? I suspect it would be a lot more work to get one line statements right, rather than just let folks fill in labels for positions in slightly different ways.
Thanks for this 🙂
I would say that there is a huge bias in the survey towards a very specific strain of analytic philosophy. I was in your situation re familiarity for the duration of my PhD, but this year, with a single semester’s worth of TAing a history of Analytic philosophy course behind me, I’m familiar with most of them. And my PhD was in Analytic philosophy! But just not B&C’s subfield within it.
*Some* questions might not be familiar to those outside analytic philosophy, but I find it very odd to say that there is a *huge* bias here. Consider these questions (I am not reproducing them verbatim):
‘Is there a priori knowledge?’, ‘Are you nominalist or Platonist about abstract objects?’, ‘Is the mind material?’, ‘Is there free will?’, ‘Is there a mind-independent external world?’, ‘Are you an empiricist or a rationalist about knowledge?’ I suppose Hegel is *not* an analytic philosopher, but he would clearly understand them.
No, there’s a huge bias. Most of these are nonsensical outside of analytic philosophy
Huh no. These would be perfectly fine questions for someone who stopped reading philosophy after 1800.
Shot in the dark: most of these positions/arguments were addressed in my undergrad degree, I could trace them to specific classes, but I did notice that some professors omitted a “zoomed out” overview, but focused more on specific arguments or thinkers, rather than where it fit with other thinkers of its time or in history. So perhaps, I just encountered 1-2 professors who were especially keen to present the positions in contrast to each other, and historically, but that it would be possible to study for a long time and not encounter the content and arguments in the way that this survey presents. (I am also thinking of a single professor who contributed to half of my current philosophical interests, but had I not run into this person by chance, I would have had a completely different philosophical path).
A much welcome distraction!
virtue ethics: 37%, Consequentialism: 30%, Deontology: 32%. A bad day for situationists, deontologists, and consequentialists!
I didn’t realize that our profession is currently dominated by people born in the 1970s. Go Generation X!
I have been looking forward to this. Good stuff for a variety of reasons. Thanks! I do have one confusion, however. On the aggregated results page, I see my responses. But, on my PhilPeople page, I am missing true contradictions, spacetime, properties, politics, concepts, and others. It’s not a big deal. I am just wondering whether anyone else has any potential fixes or similar experiences.
Sorry to hear about the glitch. If you send me your Philpeople profile link i can check what is going on.
Hi David, you also seem to have a bug in your JS code. When trying to POST to the database to update the morality response, I get a 500 response and a TypeError. See attached.
Thanks we will look into this!
Interesting. On the footbridge, the majority said “do nothing”, on the trolley the majority said the opposite. This seems contradictory to me – in both cases your action or lack thereof is directly responsible for harming one person and saving five (or not).
Yes, that was interesting to me also. In fact, the most interesting part (to me) of this survey are the various correlations to be made among the questions.
How many in the survey hold contradictory beliefs? How many of us (humans in general) hold contradictory beliefs?
The whole survey is pretty fascinating.
I think there is a moral difference between redirecting a lethal threat to minimize its harm, and initiating a lethal threat. Whether that is true or not, there is no contradiction in saying don’t push, but do switch.
Yes, that is a very common combination of responses. The two cases are explicitly meant to elicit that combination together, actually: the most common explanation of this is the doctrine of double effect/intention: in the standard trolley problem, the harm to one is a side effect of your plan to save the five, while in the footbridge problem the harm to one is part of your plan.
If I switched the tracks, and the one person tied to the other track freed himself and got out of the way, I would be ecstatic, since I don’t intend to harm him, just to save the five.
If I push the guy off the footbridge and he moves out of the way and the five die, I would have failed, since I pushed him off *so that* he would get hit by the train and stop it.
Maybe intention is morally significant.
That is just the trolley problem, i.e., how to reconcile those two judgments given that as you say both involve harming one and saving five.
One big issue with the question, though, is that the switching case, the usual question is whether it is permissible, whereas the question on the survey makes it sound like it’s about what you are required to do. Permissibility is enough for the trolley problem to get off the ground, and even more people might be willing to switch if the question was merely about whether it is permissible to do so.
Great article about a great study that gets better each time. I’m excited to see some results relating to non-western philosophies in future renditions of this work.
The sociological criteria is follow the money and exclude the excluded ones? I didn’t see references to African or Latin America researches. Do they make reasearches or take positions in philosophy? I think we do. But, of course, third world isn’t part of the world at all.
It’d be wonderful if a future version of this survey at least included a section that featured questions about these positions, or responses to the questions presently on the survey from the philosophical traditions in these regions of the world (Asian philosophies included). It’s clear from the comments to this entry that many people (myself included) believe the survey has pedagogical value. The inclusion of non-Western philosophical questions, traditions, and positions in surveys like these could become one more avenue for their introduction to people within and outside of the discipline, especially given that Western-educated philosophers and laypeople seem to constitute the majority of this survey’s respondents (and are its primary targets).
Sources: https://survey2020.philpeople.org/survey/design/population and https://survey2020.philpeople.org/survey/results/demographics
we would love to include questions about non-western philosophy. we discuss that and related issues about biases in the survey questions in our editors’ thoughts on the survey. we did try to come up with questions in non-western philosophy for the 2020 survey. the preliminary discussion document has some suggestions we received: e.g. “Yogācāra ‘mind only’: metaphysical or phenomenological?”, “Madhyamaka: skepticism, anti-realism, or quietism?”, “Confucian ethics: virtue ethics or something else?” “Better exemplification of Daoism: Laozi or Zhuangzi?”. in the end it seemed that these questions would be too unfamiliar to too much of the audience to work well in the current format where questions are asked of the whole population. one thing we are considering doing for future surveys is having some more specialized questions that go especially to specialists. that may make it easier to ask questions in these areas. in the meantime, all ideas, including suggestions for relevant questions we might ask on future surveys, are welcome!
p.s. the “editors’ thoughts” link should be https://survey2020.philpeople.org/survey/design/thoughts.
Thank you so much for your reply!
I think including specialized sections like the ones you mention in your comment are a wonderful idea, and I very much hope you receive good suggestions for the questions that might go into those sections in the future!
Some of the correlations are utterly fascinating. Here’s my favorite: among philosophers who reject the many-worlds interpretation of quantum theory, 40 out of 225 are opposed to first-trimester abortion. Among philosophers who accept or lean towards many-worlds, only 1 out of 96 is opposed. (For the record I’m not that one!)
I have absolutely no idea what could explain that. Probably it’s just a statistical fluke, but still…
My thoughts: more liberal people tend to like weirder ideas. Many-worlds is a weird idea, so more liberal people might tend to like it more?
If you kill the baby, there are many branches in which you don’t, which is a relief. 🙃
If you decide on whether to abort using a quantum-mechanically random process, then sure…
I mean you have your next book on a plate here. The Everett interpretation of the ethics of abortion.
there’s a strong correlation between accepting that newborn babies are conscious and accepting that propositions are structured entities. who knew?
[https://survey2020.philpeople.org/survey/results/correlations?category=Philosophical%20questions&variable=Other%20minds%3A%20newborn%20babies]
Looking further:
(1) many-worldsers tend to adhere to a broad collection of views settling around functionalism (the actual MW/functionalism correlation doesn’t get into the ‘highest correlations’ list, but you can see it in a bunch of proxies. And that’s not surprising given the role of functionalist accounts of high-level ontology in MW.
(2) functionalists are much less likely to oppose first-trimester abortion (11/287 vs 84/377).
That, plus small-number statistics, probably explains most of it. (I suspect there’s also some substructure in functionalism that does some of it.)
But what if anything explains the correlation between functionalism and that first order judgment about abortion?!
Theists are both far more likely to oppose abortion and more likely to reject functionalism.
So you think the correlation is driven by those who reject functionalism (and abortion) more than by those who accept it?
David/David: have you considered running some kind of multivariate cluster analysis on this results? Eyeballing the correlations, it looks like there might be something interesting to see.
Are the correlation data downloadable? There are a bunch of different cluster analyses that might be fun to run, and there are probably some pretty graphs to be made to visualise how debates fit together.
we did a factor analysis on the first survey — see “what do philosophers believe?”. there are limits to how meaningful those are, since they depend on the fairly arbitrary choice of questions, so we haven’t done one for 2020 so far, but we may eventually.
One result that was very surprising to me in the survey was on the question regarding whether individuals who shared a body of evidence could rationally disagree. Around 70% accepted or leaned toward the claim that they could while only around 20% accepted or leaned toward the claim that they could not. I’ve worked on this issue before, and my sense is that the issue is much more divided than this survey would suggest; if anything, it seems to me that the literature favors an anti-permissive position, rather than the permissive position that more people favored in the survey. Does anyone else who has worked on this question share my sense of surprise about this? Any hypotheses about what might explain this data?
This is kind of a normal occurrence. There are plenty of fields where there is a very standard view among people who are not working in the field, and it is much more contentious inside the field. Sometimes there is even a causal story that sounds plausible about why this is so – someone goes into the field because they don’t like the soft consensus that has developed among other philosophers.
I think you’ll find an even more dramatic version of this with free will. I don’t think the literature on free will has compatibilists outnumbering libertarians by 3-to-1. But I also suspect some libertarians work on free will because they don’t like the majority view of philosophers that compatibilism is right.
This is part of why I find this survey so valuable. Reading the journals can give you a rough sense of the division of views among experts on a topic, but only a survey can tell you what philosophers who aren’t experts think.
Fellow vegetarians and vegans, should I be heartened or disheartened by these results? I’m having difficulty knowing how to react. On the one hand, I have yet to hear a good argument in favor of omnivorism in normal circumstances. So I’m disappointed our numbers aren’t higher. On the other hand, I imagine the numbers would be much different only 20 years ago.
The numbers are much higher than in the general population but they don’t really support the idea that vegans are taking over philosophy.
Good question. I suspect it’s still the case that a great swath of philosophers “reason” with their bellies (habits, human prejudices, cravings) when they eat rather than with their minds, despite their residual pride to the contrary. The evidence is strong that a plant-based diet is best for human health, non-human animal treatment, and the planet. Factory farms and agribusiness contribute heavily to climate change, and cutting down on or eliminating meat is one of the significant personal responses to the climate emergency.
I suspect the numbers would have been much better had the question been worded differently. I am a longtime vegetarian on ethical grounds, but not because I think it is “impermissible” to eat meat in “ordinary circumstances”. So I answered that I “lean towards” omnivorism.
Yeah, to be honest, I think I struggled with this wording too, and I am pretty much a vegan. If “eat meat” means, “eat meat once”, then I think that’s probably permissible. Or if “meat” means “any kind of meat”, including oysters, then I think it’s probably permissible to do that. I guess I assumed that their intention was “eat meat in ordinary circumstances” means “eat factory farmed mammal and bird products regularly”.
“I have yet to hear a good argument in favor of omnivorism in normal circumstances.”
I kind of like the inefficacy objection to veganism. If you completely stop eating meat, you likely won’t make any impact on the number of animals caged, killed, etc. If you eat small amounts of meat, you probably won’t influence anything either. May as well go ahead and eat. It’s the same issue as the obligation to vote (probably won’t make a diff, so no obligation to vote), obligation to take shorter showers to fight climate change (probably won’t make a diff so no obligation), obligation to fight prevailing beauty norms for women by not wearing makeup, etc.
The voluminous literature on these topics suggests that this argument is decent–about as good as most arguments in philosophy–right?
Isn’t this why Kant said moral maxims must be universalisable? They make no sense unless they are attached to the belief that others should do likewise.
I actually eat meat but I think vegetarianism/veganism is often followed on categorical grounds. People may have other principles which they follow despite knowing full well it makes practically no difference, simply out of the belief that ‘if everyone did likewise, it would be a better world’.
For my part, I think it is wrong to eat meat but don’t consider it as unethical as other things I would never do (and indeed, some vegetarians/vegans might!) There’s no serious, objective reason for these convictions, they are just happenstance. Whether that is true of people generally, I don’t know.
I am a consequentialist, and I believe the objections you are referring to have been refuted, but I won’t get into my reasons here. However, I do grant that somebody buying meat once almost certainly has no impact on the total number of animals produced, and so is almost certainly not wrong. I noted this above. If people understood “eat meat under ordinary circumstances” to mean “eat meat once under ordinary circumstances”, then I would be somewhat cheered by the results of the survey. However, I don’t think that’s the natural reading of the phrase, and so I am not cheered.
I would appreciate it if you could point towards literature which refutes, in your view, the previously mentioned objections.
On the free will question — I am neither a compatibilist nor do I accept the position that there is no free will. But if one is a no-freedom compatibilist, which answer from the survey should one select?
I suspect the survey question assumes (for reasons I do not understand) that compatibilists must [?] believe in freedom. Neither incompatibilism nor compatilism commits one to any position on the existence of freedom.
easy: a no-free-will compatibilist should endorse both “no free will” and “compatibilism”. the survey makes it straightforward to endorse multiple options.