"The relevant Stanford Encyclopedia of Philosophy entry was first published on October 3, 2005, received a substantive revision on February 3, 2015, and a minor correction in Winter 2023. Thus, the last significant content update dates to 2015. The most recent PhilPapers survey, however, was conducted in 2020."
I've seen someone criticize the article by interpreting you as making the claim that there's been a dramatic shift the definition of realism philosophers favor in that period of five years. First, I think it could be read this way but I was skeptical you meant to suggest this, so some rewording might help clarify.
Second, note that there was also a 2009 PhilPapers survey where the realism rate was 56.4%:
It is probably worth mentioning and engaging with this as well.
I do think there's been a shift, but I think it's occurred over a few years. Finally, regarding the update of 2015 to the SEP entry: I don't know how these updates are made, but SEP articles typically have only one author, and as a result may reflect the perspective of that author even if their definition is less popular.
As Lance Bush has pointed out, it's not hard to argue that moral naturalism is in fact a form of moral anti-realism, should one use Mackie's definition of this term (as done in this post), and if one accepts this then the majority of philosophers would in fact be anti-realists.
I don't recall arguing that moral naturalism is not a form of moral realism; I accept the contemporary convention that moral realism is the position that there are stance-independent moral facts. What I argue is that moral naturalism often (though not always) dispenses with or struggles to account for features moral non-naturalists sometimes consider non-negotiable, primarily having to do with irreducible normativity, authority, and related notions of "clout," "oomph," and so on. If I've said otherwise in the past then that's not a position I still endorse.
I've also suggested that a better dividing line for the central dispute would be between non-naturalist moral realists and everyone else, and I believe Mike Huemer has expressed similar sentiments in interviews.
I think these are legitimate concerns, but I'm not convinced by the overall analysis for three reasons. For one thing, it's not obvious that "leaning towards" moral realism should be seen as necessarily distinct from actually being a moral realist; the terminology is ambiguous, and it seems to me there's a broad range of confidence levels where both terms would be appropriate and fail to signal anything meaningfully distinct. I can think of many things where, if given the option, I would probably pick "lean towards X," but would also happily choose "affirms X" if that was the only option. So while it may not be legitimate to lump in every "leans towards" into the category of moral realist, it's also not legitimate to automatically exclude them.
Secondly, while I'm no expert on exactly how professional philosophers conceptualize moral realism, my own experiences with the literature lead me to think it's very unlikely that a sizeable number of those surveyed are thinking of the (as you say, somewhat outdated) definition that doesn't reference stance-independence. Especially since, as you say, that's almost never the way it's presented nowadays.
So for these two reasons, while I can see someone saying we can't be sure that most philosophers are moral realists, I don't think it's at all warranted to make the positive claim that they aren't. And thirdly, I don't think it really matters - even if moral realism is merely the preferred option of the majority of philosophers, or if it's the accepted position of a solid third, then a wide range of critical claims anti-realists makes (that it's obviously absurd, incoherent, a relic of a bygone age, etc) are just as overblown as they would be if realists had hit the 50% mark unambiguously.
Re: your first point—They were given the option to accept realism and chose not to.
If you leaned towards the position, but didn't accept it, and were given both options, why would you pick the "accept" position? If you accepted it, why would you pick the "leans towards" position? Whatever overlap these positions per se is not very relevant when BOTH options were given.
"I would probably pick "lean towards X," but would also happily choose "affirms X" if that was the only option"
This quote just makes me think you think they weren't given both options?
Well sure, I get that they had the option to pick either. But the point is that offering a "leans towards" option in addition to an "accepts" option generally splits up the distribution of confidences differently - if that "leans towards" option is available, then the cut-off for what counts as "accepting" goes up in people's minds. And that means many people who decide to pick "leans towards" out of epistemic humility would otherwise have legitimately chosen "accepts." I don't think this is super controversial when it comes to other polls, like asking for whether you support some candidate or bill or whatever.
I don't think this is a problem. We can just say what we mean by "Accepts moral realism" is whatever the cut-off is in people's minds when they're presented with the options. For the subset that lean towards realism but would have picked "Accepts" if that were the only option, we can just say that's not what we mean by "Accepts moral realism". We only care about the people who accept it when all options are presented BECAUSE they considered the less confident positions.
I mean, that's fine - but then only like 12% of philosophers should be called anti-realists, right? In which case the margin is even better for realists!
"Better" only if you believe the margin is meaningful e.g. has some epistemic import to our own metaethical beliefs (Lance Bush incoming). The only point being made is that the majority for realism disappears, not that the plurality disappears or that antirealism has a higher "Accepts" when you un-collapse the two.
This is a good point, I'm more persuaded by the question of whether respondents actually interpreted 'moral realism' as referring to stance-independence. Some quick observations from a glance at the survey data:
253 respondents who said yes to Moral Realism also said yes to non-naturalism. I don't think non-naturalism is *strictly speaking* incompatible with subjectivism (for lack of a better grouping term than 'antirealist views that aren't noncog and error theory'), I take it to be a fair assumption that a non-naturalist will not be a subjectivist, if they understand that view properly. There are 642 pro realism respondents to that question, so we can be relatively confident that at least a third of that portion would be realists in the stance-independent sense.
There are 284 naturalist respondents. This is harder, and it seems to me you could interpret subjectivism as a species of naturalism. Given that, I think we should probably cede this other portion as indeterminate whether they should be understood as properly pro or con.
I'll also flag another interesting quirk in the data, of the fifty respondents to the realism question that accept error theory, only 9 accept moral realism. Now, the thing is, I think there is an important sense in which error theorists *are* moral realists. They think moral language expresses stance independent facts. They just think all those facts are false! So part of me wants to say this is evidence for your thesis. On the other hand, we tend to include in the moral realism thesis that there are at least some true moral facts. In that case of course the error theorist is definitely not a moral realist. Still, just another interesting point.
Theres also 100 realists who deny cognitivism!
The consequences of the naturalism/non-naturalism point will change depending on how many of those respondents were a 'lean towards' answer vs a committed answer, i.e. if most of the committed realists were non-naturalists, and we *should* think that non-naturalists are almost always realists, that would give us good reason to think at least that number was more accurate. Anyway, good post!
All good points. I'm going to take these specifics into consideration and propose that we conduct a targeted survey to more accurately address what philosophers think about metaethics.
Re: your first point, if we only count philosophers who *accept* either realism or anti-realism, realists outnumber anti-realists by more than three to one. (Roughly 37.4% are realists and roughly 11.6% are anti-realists.) So, the margin in favor of realism gets even larger. This is definitely not good news for anti-realism.
I agree with Both Sides that most metaethicists ca. 2020 would have understood, by “realism,” a view committed to stance-independence or something similar. The SEP article you cite is indeed outdated, since it was written before Street, Enoch, Shafer-Landau, and Huemer became major players in the debate, and these philosophers all understand realism in terms of stance-independence or something similar.
If we're just focused on ratios it's not good news, but people commonly appeal to there being a majority of philosophers in favor of realism as evidence for the view. As the proportion of the total declines, their relative difference to antirealists becomes less relevant. If it were 1% antirealists and 5% realists, for instance, there'd be five times as many realists as antirealists, but they'd hardly be in a position to point to this as evidence of realism.
I also suspect most philosophers would think moral realism implies stance-independence, but "most" just means "more than 50%." Even if it's 80-90%, a good 10-20% interpreting "realism" differently would attenuate the total proportion of realists and their ratio relative to antirealists. The latter is especially important: if lots of those realists would be categorized as antirealists, this can shift the ratios by quite a lot even if it's just 10% who interpret realism as not requiring stance-independence.
I don't think the SEP article is outdated so much as it was written by someone who happens to hold a minority stance on how to define realism. They've had every opportunity to update it since then, and SEP routinely does so. They just haven't updated it in this particular way.
There are philosophers who continue to push alternative characterizations of the metaethical landscape. I recently encountered someone who insists not only that relativism is not a metaethical position, but that this is what experts in metaethics agree to (though I'm not sure if they think it's a majority or not)
Given that the only two positions are realism and antirealism, all that finding 5% realists and 1% antirealists would reveal is that most philosophers do not know which is true but that amongst those with opinions one way or another realism seems much more plausible, which is a relevant statistic. But we don't live in that world, 37.4% is a lot of philosophers.
I’m not trying to credential-check you, but I’m curious if you work in professional philosophy. My experience in professional philosophy is that moral realism is definitely the dominant view. Though, of course, this is purely anecdotal. It also depends on what you mean by “moral realism.” The strongest forms of moral realism, on which moral truths are non-natural and irreducibly normative, seem to be, in my experience, pretty popular among metaethicists and considerably less popular among philosophers in other areas. But if you count less extreme views like moral naturalism and constructivism as realist, then altogether realism definitely seems to have the most support. At least in my experience.
That being said, philosophy is a slave to fashion. Which views get the most support tends to change pretty dramatically over the years. 30-40 years ago non-naturalist moral realism was considered crazy by a ton of philosophers before its revival. Given how the tides change so much in philosophy, I don’t think a view’s being the most popular at a given time says much in terms of whether it’s the best view.
No, that's not all that matters. Causality matters. We don't actually know if studying philosophy causes people to become moral realists. Them being a majority is even consistent with studying philosophy causing more people to become antirealists than realists.
As a point of clarification, you say:
"The relevant Stanford Encyclopedia of Philosophy entry was first published on October 3, 2005, received a substantive revision on February 3, 2015, and a minor correction in Winter 2023. Thus, the last significant content update dates to 2015. The most recent PhilPapers survey, however, was conducted in 2020."
I've seen someone criticize the article by interpreting you as making the claim that there's been a dramatic shift the definition of realism philosophers favor in that period of five years. First, I think it could be read this way but I was skeptical you meant to suggest this, so some rewording might help clarify.
Second, note that there was also a 2009 PhilPapers survey where the realism rate was 56.4%:
https://philpapers.org/surveys/results.pl
It is probably worth mentioning and engaging with this as well.
I do think there's been a shift, but I think it's occurred over a few years. Finally, regarding the update of 2015 to the SEP entry: I don't know how these updates are made, but SEP articles typically have only one author, and as a result may reflect the perspective of that author even if their definition is less popular.
Just FYI it's Eric Sampson, not Erik Sampson. Feel free to delete this!
Thanks!
As Lance Bush has pointed out, it's not hard to argue that moral naturalism is in fact a form of moral anti-realism, should one use Mackie's definition of this term (as done in this post), and if one accepts this then the majority of philosophers would in fact be anti-realists.
I don't recall arguing that moral naturalism is not a form of moral realism; I accept the contemporary convention that moral realism is the position that there are stance-independent moral facts. What I argue is that moral naturalism often (though not always) dispenses with or struggles to account for features moral non-naturalists sometimes consider non-negotiable, primarily having to do with irreducible normativity, authority, and related notions of "clout," "oomph," and so on. If I've said otherwise in the past then that's not a position I still endorse.
I've also suggested that a better dividing line for the central dispute would be between non-naturalist moral realists and everyone else, and I believe Mike Huemer has expressed similar sentiments in interviews.
I think these are legitimate concerns, but I'm not convinced by the overall analysis for three reasons. For one thing, it's not obvious that "leaning towards" moral realism should be seen as necessarily distinct from actually being a moral realist; the terminology is ambiguous, and it seems to me there's a broad range of confidence levels where both terms would be appropriate and fail to signal anything meaningfully distinct. I can think of many things where, if given the option, I would probably pick "lean towards X," but would also happily choose "affirms X" if that was the only option. So while it may not be legitimate to lump in every "leans towards" into the category of moral realist, it's also not legitimate to automatically exclude them.
Secondly, while I'm no expert on exactly how professional philosophers conceptualize moral realism, my own experiences with the literature lead me to think it's very unlikely that a sizeable number of those surveyed are thinking of the (as you say, somewhat outdated) definition that doesn't reference stance-independence. Especially since, as you say, that's almost never the way it's presented nowadays.
So for these two reasons, while I can see someone saying we can't be sure that most philosophers are moral realists, I don't think it's at all warranted to make the positive claim that they aren't. And thirdly, I don't think it really matters - even if moral realism is merely the preferred option of the majority of philosophers, or if it's the accepted position of a solid third, then a wide range of critical claims anti-realists makes (that it's obviously absurd, incoherent, a relic of a bygone age, etc) are just as overblown as they would be if realists had hit the 50% mark unambiguously.
Re: your first point—They were given the option to accept realism and chose not to.
If you leaned towards the position, but didn't accept it, and were given both options, why would you pick the "accept" position? If you accepted it, why would you pick the "leans towards" position? Whatever overlap these positions per se is not very relevant when BOTH options were given.
"I would probably pick "lean towards X," but would also happily choose "affirms X" if that was the only option"
This quote just makes me think you think they weren't given both options?
Well sure, I get that they had the option to pick either. But the point is that offering a "leans towards" option in addition to an "accepts" option generally splits up the distribution of confidences differently - if that "leans towards" option is available, then the cut-off for what counts as "accepting" goes up in people's minds. And that means many people who decide to pick "leans towards" out of epistemic humility would otherwise have legitimately chosen "accepts." I don't think this is super controversial when it comes to other polls, like asking for whether you support some candidate or bill or whatever.
I don't think this is a problem. We can just say what we mean by "Accepts moral realism" is whatever the cut-off is in people's minds when they're presented with the options. For the subset that lean towards realism but would have picked "Accepts" if that were the only option, we can just say that's not what we mean by "Accepts moral realism". We only care about the people who accept it when all options are presented BECAUSE they considered the less confident positions.
I mean, that's fine - but then only like 12% of philosophers should be called anti-realists, right? In which case the margin is even better for realists!
"Better" only if you believe the margin is meaningful e.g. has some epistemic import to our own metaethical beliefs (Lance Bush incoming). The only point being made is that the majority for realism disappears, not that the plurality disappears or that antirealism has a higher "Accepts" when you un-collapse the two.
This is a good point, I'm more persuaded by the question of whether respondents actually interpreted 'moral realism' as referring to stance-independence. Some quick observations from a glance at the survey data:
253 respondents who said yes to Moral Realism also said yes to non-naturalism. I don't think non-naturalism is *strictly speaking* incompatible with subjectivism (for lack of a better grouping term than 'antirealist views that aren't noncog and error theory'), I take it to be a fair assumption that a non-naturalist will not be a subjectivist, if they understand that view properly. There are 642 pro realism respondents to that question, so we can be relatively confident that at least a third of that portion would be realists in the stance-independent sense.
There are 284 naturalist respondents. This is harder, and it seems to me you could interpret subjectivism as a species of naturalism. Given that, I think we should probably cede this other portion as indeterminate whether they should be understood as properly pro or con.
I'll also flag another interesting quirk in the data, of the fifty respondents to the realism question that accept error theory, only 9 accept moral realism. Now, the thing is, I think there is an important sense in which error theorists *are* moral realists. They think moral language expresses stance independent facts. They just think all those facts are false! So part of me wants to say this is evidence for your thesis. On the other hand, we tend to include in the moral realism thesis that there are at least some true moral facts. In that case of course the error theorist is definitely not a moral realist. Still, just another interesting point.
Theres also 100 realists who deny cognitivism!
The consequences of the naturalism/non-naturalism point will change depending on how many of those respondents were a 'lean towards' answer vs a committed answer, i.e. if most of the committed realists were non-naturalists, and we *should* think that non-naturalists are almost always realists, that would give us good reason to think at least that number was more accurate. Anyway, good post!
All good points. I'm going to take these specifics into consideration and propose that we conduct a targeted survey to more accurately address what philosophers think about metaethics.
Re: your first point, if we only count philosophers who *accept* either realism or anti-realism, realists outnumber anti-realists by more than three to one. (Roughly 37.4% are realists and roughly 11.6% are anti-realists.) So, the margin in favor of realism gets even larger. This is definitely not good news for anti-realism.
I agree with Both Sides that most metaethicists ca. 2020 would have understood, by “realism,” a view committed to stance-independence or something similar. The SEP article you cite is indeed outdated, since it was written before Street, Enoch, Shafer-Landau, and Huemer became major players in the debate, and these philosophers all understand realism in terms of stance-independence or something similar.
If we're just focused on ratios it's not good news, but people commonly appeal to there being a majority of philosophers in favor of realism as evidence for the view. As the proportion of the total declines, their relative difference to antirealists becomes less relevant. If it were 1% antirealists and 5% realists, for instance, there'd be five times as many realists as antirealists, but they'd hardly be in a position to point to this as evidence of realism.
I also suspect most philosophers would think moral realism implies stance-independence, but "most" just means "more than 50%." Even if it's 80-90%, a good 10-20% interpreting "realism" differently would attenuate the total proportion of realists and their ratio relative to antirealists. The latter is especially important: if lots of those realists would be categorized as antirealists, this can shift the ratios by quite a lot even if it's just 10% who interpret realism as not requiring stance-independence.
I don't think the SEP article is outdated so much as it was written by someone who happens to hold a minority stance on how to define realism. They've had every opportunity to update it since then, and SEP routinely does so. They just haven't updated it in this particular way.
There are philosophers who continue to push alternative characterizations of the metaethical landscape. I recently encountered someone who insists not only that relativism is not a metaethical position, but that this is what experts in metaethics agree to (though I'm not sure if they think it's a majority or not)
Given that the only two positions are realism and antirealism, all that finding 5% realists and 1% antirealists would reveal is that most philosophers do not know which is true but that amongst those with opinions one way or another realism seems much more plausible, which is a relevant statistic. But we don't live in that world, 37.4% is a lot of philosophers.
It's not that many and it isn't a very impressive number. Even 62% isn't impressive.
I’m not trying to credential-check you, but I’m curious if you work in professional philosophy. My experience in professional philosophy is that moral realism is definitely the dominant view. Though, of course, this is purely anecdotal. It also depends on what you mean by “moral realism.” The strongest forms of moral realism, on which moral truths are non-natural and irreducibly normative, seem to be, in my experience, pretty popular among metaethicists and considerably less popular among philosophers in other areas. But if you count less extreme views like moral naturalism and constructivism as realist, then altogether realism definitely seems to have the most support. At least in my experience.
That being said, philosophy is a slave to fashion. Which views get the most support tends to change pretty dramatically over the years. 30-40 years ago non-naturalist moral realism was considered crazy by a ton of philosophers before its revival. Given how the tides change so much in philosophy, I don’t think a view’s being the most popular at a given time says much in terms of whether it’s the best view.
I think this is a clumsy semantic point, all that matters for this discussion is that a minority of philosophers are moral antirealists
No, that's not all that matters. Causality matters. We don't actually know if studying philosophy causes people to become moral realists. Them being a majority is even consistent with studying philosophy causing more people to become antirealists than realists.
I don’t believe anyone is a moral realist. There are philosophers who think they are, but they really aren’t.