Moral antirealists concede too much to moral realists
The moral realist Eric Sampson and the moral antirealist Jonas Olson had a discussion about moral realism last year on The Analytic Christian. You can find their discussion here.
At the time, I wrote a comment in response to the video on YouTube and also wrote a Facebook post raising some of my concerns with the discussion. It is precisely such content that prompted me to create a blog. A random comment here or there has little reach, and what reach it does have is inconsistent. Hence the need for a central location where I can respond to this sort of content.
A few meta-comments before I get into what I said at the time.
First, I make a point of reliably responding to such content in order to illustrate that my fulminations against moral realism and the way the dispute between realists and antirealists is characterized does not reflect the state of the dialectic among nonphilosophers. Instead, many of the same concerns and issues arise in contexts in which everyone involved is a professional philosopher.
Second, I think it is important to focus on this sort of content, and not just published academic content, because it reveals angles and elements of philosophical thought that don’t make their way into print. The demands and expectations of published philosophy differ from what people will say in conversational contexts. Of course, that doesn’t mean what they say in such contexts is better or more accurate. Speaking rather than writing is more spontaneous and one is more likely to say something they don’t really mean, or to word themselves in a less careful way than they would if they were sitting down to write. On the other hand, passing one’s thoughts through the sieve of eventual publication may encourage a degree of caution and hedging that isn’t quite as omnipresent a force when having a conversation. In that respect, conversations like these may accurately reflect features of how philosophers think that wouldn’t make their way into print. There is also the fact that certain thoughts and positions may only emerge in a dialogue. Other people’s questions or thoughts may bring to the fore insights and comments that wouldn’t emerge in an article. And when we speak to different audiences, we may rephrase ourselves in ways that differ: perhaps they are more clear, perhaps they are less saddled with technical jargon, perhaps you’ll land on just the right example to illustrate a point, and so on.
As such, I see immense value in exploring what philosophers say outside of peer-reviewed publications, and I think it’s a shame such work isn’t recognized and appreciated as much as formal published responses are.
I’ll share my earlier comments in full. On Facebook, I posted this (I’ve reformatted some of this to make it read more cleanly on here, but the text itself is the same):
1.0 My Facebook post about the video
In a recent discussion between two philosophers who specialize in metaethics, we see a clear case of normative entanglement expressed by Olson:
“Here’s one possibility. The idea might be that someone says roughly what Eric said before that you know take any old moral proposition like ‘it’s wrong to torture people just for fun.’ We know that. Okay that seems extremely obvious to most of us. If the appeal is to the general consensus and to how strongly, how convinced we are about the truth of this proposition…then I think, in that case, an evolutionary explanation might be useful. Because the evolutionary explanation predicts that we’ll believe in that proposition and that we will feel confident about it.”
---
Focus on the claim “take any old moral proposition like ‘it’s wrong to torture people just for fun.’ We know that. Okay that seems extremely obvious to most of us.”
“It’s wrong to torture people just for fun” is a normative moral claim. There is no explicit metaethical position expressed by it. There may be an implicit one, but unless we know what implicit metaethical presuppositions (if any) are presumed by someone responding to this, it’s unclear what it would even mean for someone to say that this “seems extremely obvious” for the simple reason that the first-order normative claim could mean different things to different people.
I’m a moral antirealist. I don’t think there are stance-independent moral facts. I don’t need to talk about whether this claim is “obvious” or “true,” I can simply tell you that I disapprove of torturing people for fun, I think it’s wrong, it’s inconsistent with my standards, I’m strongly committed to this, I find torture repugnant, and so on. If I were asked:
“Do you think it’s wrong to torture people just for fun?”
…how am I supposed to respond? This is presented as a yes or no question, but it’s underspecified: is the question presuming realism or not? Why not make this explicit?
If I were to say “no,” this would be misleading, since it in no way indicates that I think there are stance-independent normative facts. This is because the metaethical presupposition, “It’s a stance-independent moral fact that it is morally wrong to torture people just for fun,” isn’t made explicit, but could be retroactively assumed so as to insist that if I agree, I agree with realism; while if I disagree, this can be taken to mean I disagree with the normative claim. Without making the metaethical presupposition explicit, the question is effectively a trap that allows moral realists to depict antirealists as either conceding to the realist or as a terrible person who doesn’t object to whatever horrible action they come up with.
Think about what someone would mean if they said “It’s morally wrong to torture people just for fun.”
(1) Even if they were a moral realist, they wouldn’t just be saying “I think there are stance-independent moral facts,” they’d *also* be taking a normative stance on the moral status of torturing people just for fun: that they believe it’s morally wrong.
(2) They wouldn’t even just be expressing a normative moral stance on the moral status of torturing people for fun (that they believe it is morally wrong). They’d also be pragmatically implying their personal opposition to torturing people just for fun.
Now, consider an antirealist who says that they disagree that it’s “wrong to torture people just for fun.” They wouldn’t merely be denying moral realism, they’d also be denying the first-order normative claim that it’s wrong to torture people for fun (~1), they’d also be implying that they don’t object to torturing people just for fun.
It’s a package deal. So you are presented with two options when responding to this claim:
(1) Morally good realist response
If you agree with this claim, you affirm (a) there are stance-independent moral facts (b) one of those facts is that it’s morally wrong to torture people just for fun and (c) you personally find torturing people just for fun to be bad.(2) Evil monster antirealist response
If you disagree with this claim, you (a) deny there are stance-independent moral facts (b) do not endorse the first-order normative claim that it’s morally wrong to torture people just for fun and (c) you don’t personally find torturing people just for fun to be bad.
---
Moral philosophers discussing metaethics frequently embed metaethical claims inside normative claims, the response to which has overwhelming pragmatic implications for the standards and character of the person responding to those claims. This obscures our ability to think carefully about what ordinary people might think, by inappropriately grafting the metaethical positions under discussion to normative considerations. It also allows realists to inappropriately imply moral antirealists are bad people if they don’t make concessions to the realists.
This is an unfortunate aspect of the ongoing dialectic in metaethics that I hope philosophers will take note of and avoid.
I generally agree with what I said here. However, I’ll make one observation here. I find myself at odds with both moral realists and, in some ways, other moral antirealists. And I think part of the reason for this is that most other moral antirealists publishing in contemporary metaethics are operating within the analytic tradition. As such, they often accept many of the metaphilosophical presuppositions, and share the methodological framework, in which moral realists operate. As a result, their objections largely fall within and draw on those methods and presuppositions. Since I reject those methods and presuppositions, my objections often involve rejecting both realism and the presuppositions that led realists to arrive at those conclusions. Yet my rejection of these background assumptions often spillover into my stance towards other antirealist positions as well, leading me to reject those views. This is why, for instance, I also reject Streumer’s central thesis in Unbelievable Errors.
This puts me in the unfortunate position of having few philosophical allies. While I am a moral antirealist, that does not mean I align, metaphilosophically or methodologically, with other moral antirealists. In general, I don’t. This makes me somewhat of an outsider to the whole dialectic. It’s as though everyone else has agreed to play the game of metaethics by a certain set of rules. Realists and antirealists are rival teams both generally abide by these rules. So when someone comes along who doesn’t, they may be dismissed as failing to adequately address the problems they’re dealing with. One unfortunate consequence of this is that my objections may seem especially dismissive or hostile towards the whole enterprise of metaethics. I find this strange. I simply locate the mistake I think moral realists are making somewhere further down the chain of inference than the error theorist or the noncognitivist.
It’s not as though I am tossing out logic or anything that fundamental. My own views stem from traditions that fall well within the scope not merely of philosophy, but of analytic philosophy in particular. After all, much of what I have to say falls squarely in line with the sorts of objections Schiller, Wittgenstein, Quine, and Stich might take, along with anyone else sympathetic to e.g., pragmatism, or who takes the later Wittgenstein seriously, or who thinks experimental philosophy and cognitive science more generally has the potential to raise at least some challenge to philosophical methods and conclusions. It’s not as though I invented these sorts of objections. And yet, for whatever reason, few people publishing in contemporary metaethics seem to be championing these approaches. That’s going to change.
2.0 My YouTube comment on the video:
It’s great to see this conversation. I’m glad to see a strong proponent of realism and antirealism on this channel. Olson’s done great work on error theory and I appreciate his cautious approach. While I’m convinced of moral antirealism, it makes for more interesting discussion if there’s a variety of people who fall along various points in their level of confidence (even if I’d prefer everyone be an antirealist).
I had some criticisms of both Sampson and Olson.
(1) Sampson and Olson both assume ordinary people are moral realists. There’s no good evidence this is true, and quite a bit of evidence that it isn’t.
(2) Sampson doesn’t provide much in the way of substantive arguments for realism in his opening remarks. It’s not clear he offers much in the way of an argument at all. Note that Sampson presents a very robust form of realism that includes e.g, epistemic access to moral facts, yet does not present any arguments for this, either. It all comes down to little more than realism seeming true…to Sampson. Well, antirealism seems true to me. Is there any good reason for people to think realism is true if it doesn’t seem true to them?
(3) Sampson employs a common strategy moral realists use: a kind of epistemic piggybacking strategy. The idea is to say “we agree that A, B, and C is true, so why not D?” This move is a subtle burden-shifting maneuver. Rather than presenting a positive case for moral realism, the move is to suggest that one’s reasons for accepting moral realism are on par with one’s reasons for accepting other claims. Now, if one wants to object to realism, someone using this move can try to make you defend some other claim, such as the existence of time or the external world.
(4) Around 3:10 Sampson talks about why people care about realism. “If moral realism is true that would mean that you have moral obligations to help others and not to do bad things to them. So there are demands that are placed on your life that are independent of what you care about or what your goals are, so you can’t just get out of them by changing what you care about or what your goals are.” This is one of the most puzzling elements of realism, and one not adequately discussed. What on earth does Sampson mean, you can’t get out of them? Suppose moral realism is true. Suppose, also, that you know what the moral facts are. For instance, one of the facts is that you shouldn’t steal.
Now let’s suppose you steal, anyway. Setting aside the consequences of getting caught, being punished, and so on, all of which could occur even if moral realism were false, let’s just consider the consequences of violating the stance-independent moral rule for its own sake. Okay. What are those consequences?
As far as I can tell: none whatsoever. You can just steal and, even if there were some moral fact saying you ought not to, well, so what? Nothing happens if you steal. So when Sampson says you can’t just get out of these moral rules, what does that amount to? For all practical purposes, you can: you can simply ignore the rules and there are no consequences at all. Literally nothing in the universe changes. As far as I’m concerned, that means that, yes you can get out of the moral rules. Indeed, it’s unclear what there was to get out of in the first place.
(5) Olson concedes too much to Sampson. Why does Olson think Sampson’s Moorean move has force? Why does he think moral realism is “prima facie” plausible? What even is it that’s supposed to be plausible?
(6) I don’t think Olson should so willingly grant that stance-independent requirements are coherent (I prefer the term “intelligible”). I don’t think Sampson, or any other realist, could explain what stance-independent moral facts are. This is why many realists claim that the concepts they’re appealing to are primitive that cannot be further explained. In other words, it’s not like realists (or at least non-naturalist realists) are out there offering explanations of what they mean. They aren’t. Rather, they’re out there offering explanations for why they can’t explain what they mean.
Again, I endorse most of what I said here. Points (1) through (4) are concerns I’ve raised ad nauseum. Philosophers continue to presume that most people are moral realists even though the best available empirical evidence casts serious doubt on this claim. Moral realists continue to lean too heavily on their own intuitions, likely bolstered by the presumption that their intuitions are in good company. Realists often employ a common approach I see in philosophy: the opposite of guilt by association - one takes some view they expect another person to endorse or have difficulty denying it, analogize their view to that one, and then maintain that rejecting their view would require rejecting that other view.
Such epistemic grafting may prove successful for all I know; there’s nothing fundamentally illegitimate about such moves. But such moves represent, at best, indirect arguments for the position in question, and this should prompt caution: I am suspicious of a view when, rather than arguing for it directly on its own merits, one attempts to fuse its plausibility to some other view one expects one’s interlocutors to endorse (or at least not reject). The legitimacy of such moves is only as strong as the strength of the parity between the two views, and since that line is easy to blur, such moves strike me as a risk factor that could potential obscure the degree to which a given position is, in fact, plausible. In other words, if someone does not find an argument for God to be plausible, one can insist that denying the existence of God would require denying the utility of mathematics. Provided someone doesn’t want to do the latter, this may push them, like a good Bayesian, towards reduced confidence in their disbelief in God.
Yet such a shift is only warranted if the connection between God and the utility of mathematics really is a strong one. Any uncertainty at all about the connection between God and the utility of math could prompt one to assess the risk that they’d have to reject math to reject God as having some nonzero chance of probability. If it’s difficult to assess what that probability actually is, people may lean on the heuristic of it being substantive enough to influence their judgments, reducing their confidence in atheism, even if there is absolutely no reason at all why math could only have utility if God existed. In short, one can leverage the uncertainty of an obscure and indirect argument to artificially bolster support for some desired claim, like the existence of God, or moral realism. This may generalize to obscure and hard-to-evaluate arguments more generally. Generate a huge arsenal of obscure and complicated arguments for some thesis, and this could create an epistemic mirage that bolsters the plausibility of that position in the minds of anyone who doesn’t have the time and energy to dig into and thoroughly evaluate all those arguments. And most people probably don’t have the time to do so.
Lastly, I think the main takeaway from these comments, and the main point I want to illustrate, is that moral antirealists have, on my view, historically conceded far too much to realists. They often accept, or at least don’t object, to the claim that most people are moral realists, that moral realism “is intuitive,” that intuitions in favor of realism count in favor of it, that our language looks like its committed to realism, that people act “as if” realism were true, that there are lots of good arguments in favor of realism, that they themselves have realist intuitions, that realism is intelligible, and so on. I don’t think they should grant any of these claims.