Normative antirealism is not self-undermining
Twitter Tuesday #37
1.0 Normative domains
Moral antirealism is restricted to the moral domain. The moral domain is the domain of potential facts concerning the moral rightness or wrongness of actions. Moral antirealism only involves a denial of stance-independent moral facts; it does not involve a denial that there are any other types of normative facts. Moral antirealism is consistent with realism towards any other normative domains, such as aesthetics and epistemology.
Normative antirealism is a more expansive position that rejects the notion that there are any stance-independent normative facts in any domains, including aesthetics and epistemology.
Critics of moral realism sometimes maintain that the same sorts of considerations that would motivate endorsement of moral realism also apply to epistemic norms, and, as such, a consistent moral realist ought to endorse epistemic realism. These critics maintain that this poses a problem: if you deny that there are epistemic facts, then this undermines any argument for such a position, because anyone considering such an argument would have no epistemic reasons for accepting the conclusion of the argument as true.
2.0 Error theory & self-defeat
Consider this objection to moral error theory from Rowland (2013):
(1) According to the moral error theory, there are no categorical normative reasons.
(2) If there are no categorical normative reasons, then there are no epistemic
reasons for belief.
(3) But there are epistemic reasons for belief.
(4) So there are categorical normative reasons (2, 3).
(5) So the error theory is false (1, 4)
There’s nothing fundamentally wrong with this argument. There are a number of ways an error theorist might resist it, but it’s worth noting a few things.
As an aside, I’m not sure a moral error theorist is committed to the view that there are no categorical normative reasons. They might think that there are categorical normative reasons, but that none of them are moral. This isn’t really important for the purposes of this argument, though. Maybe that’s how error theorists construe their own position. Or if it isn’t, maybe it’s something one could argue they are committed to. In the latter case, if we assume the error theorist is in some way committed based on considerations that don’t appear in this argument to a broader normative error theory, according to which there are no categorical normative reasons in any domain, then they would be committed to endorsing that there are no categorical normative reasons. It still seems a bit weird to me to say that there are no categorical normative reasons according to error theory, but whatever, maybe I’m missing something or unfamiliar with the context or phrasing. I would’ve thought an error theorist thinks there’s no categorical moral normative reasons in particular, but I digress.
Either way, the second and more important point is that this is an objection against moral error theory in particular, not an objection to moral antirealism. Error theory is one specific form of moral antirealism. Even if one could show that there are problems with error theory, this wouldn’t mean that moral antirealism is false. That being said, this argument doesn’t really threaten moral antirealism much. Moral error theory buys into the semantic thesis endorsed by moral realists; it’s at least possible in principle that a moral error theorist wouldn’t endorse an analogous semantic thesis when it comes to epistemic normativity, which could break the parity premise of some kinds of companions in guilt arguments. However, that’s a weird position to take, would be hard to defend, and, my present goal isn’t to defend moral error theory, anyway. Moral error theory has enough capable defenders as it is; they hardly need me.
Another issue with this argument is that it drops the use of “categorical” in the second premise:
(2) If there are no categorical normative reasons, then there are no epistemic
reasons for belief.
This might look like a serious oversight. After all, couldn’t someone believe that epistemic reasons aren’t categorical, and that therefore there could be epistemic reasons even if they weren’t categorical normative reasons?
They could. And that makes this premise look a bit misleading if it isn’t contextualized against the backdrop of the presumptive commitments of error theorists in particular. Insofar as an error theorist is committed to the view that there are no categorically normative reasons, and that if there were epistemic reasons, they’d have to be categorically normative reasons, then it follows that there are no epistemic reasons.
One might reason in this way. Error theorists agree with realists that the correct account of moral judgments is that they are propositions that purport to describe stance-independent (we can say “categorical” here) moral facts. However, there are no such facts, so such judgments are false. If this same line of reasoning is extended on pain of consistency to epistemic judgments, then these, too, are stance-independent normative judgments that are likewise subject to an error theory. However, if the correct account of epistemic reasons for belief was that they must be stance-independent (or categorical), and there are no such facts of this kind, then the category “epistemic reasons” is an empty set: there are no “epistemic reasons” as such.
In other words, insofar as a moral error theorist concedes to the parity premise and accepts epistemic error theory, and insofar as this commits them to the view that “epistemic reasons” just are a kind of categorical normative reason, then, if there are no categorical normative reasons, then there are no epistemic reasons.
…And if the error theorist grants that one ought only accept an argument if one has an epistemic reason to do so, but there are no epistemic reasons, then one has no epistemic reason to accept an argument for moral error theory: doing so would commit one to the view that there are no epistemic reasons!
Moral error theorists have a number of ways they can attempt to thread this needle. And perhaps they can succeed. But there’s a simpler way around this objection:
Reject normative error theory.
I think we shouldn’t grant the realist’s conception of normative reasons, and agree that if there were epistemic, moral, or other normative reasons, that they’d have to be categorical. Note, then, that even if some form of self-defeat objection worked against error theory, it wouldn’t necessarily work against other forms of antirealism. And one way around the self-defeat objection is to just not buy into the notion that having a categorical/stance-independent epistemic normative reason to accept the conclusion of an argument is needed for our arguments to succeed.
3.0 The necessity principle
Let’s grant that self-defeat arguments could perhaps cause some trouble for error theorists. Alright, does that mean they cause trouble for other antirealists?
No. Nevertheless, some people may believe that normative antirealism (the view that there are no stance-independent normative facts, including stance-independent epistemic facts) is in trouble. This is a Twitter Tuesday, so let’s get to the tweets. Here’s the first:
I agree with this. I think moral antirealists (previous posts make it clear they’re talking about moral realism) do tend to rely on arguments that are so general they throw out epistemic realism. I’m a normative antirealist, and I think this is a good thing.
The second is from (who else?) JPA:
I don’t think this is a hot take. It’s just a bad take. Evolutionary debunking arguments against moral normativity probably do extend to epistemic normativity.
Personally, I don’t think a capacity for distinctively moral cognition evolved. At most, perhaps a capacity for normative cognition evolved, though I’m not sure of that, either. It’s not at all clear to me that there is any way to isolate, from the perspective of natural selection shaping human cognition, moral norms from other norms (including epistemic norms), and hold that the former, but not the latter, are subject to debunking. One might think that natural selection shaped our moral inclinations to be adaptive, not necessarily truth-tracking; but if no capacity for distinctively moral cognition evolved, and moral and epistemic normativity are peas in the same cognitive pod, it’s not clear concerns about the truth-tracking capacities of one don’t extend to the other. There might be ways of preserving epistemic normativity. In fact, my coauthors and I wrote a whole paper on this you can find here. Personally, I don’t see good reasons to single out moral cognition as distinctively threatened by evolutionary concerns; I think all normative consideration is under suspicion. So, I agree with the first part of JPA’s take: that whatever debunking applies to morality also extends to epistemology.
The second part of this remark is the problem: that this is somehow self-undermining.
No, it isn’t.
There’s a couple ways of cashing out why this is a bad take. First, insofar as evolutionary debunking arguments threaten both moral and epistemic normativity, they only threaten distinctively realist conceptions of moral and epistemic normativity. The antirealist isn’t thereby deprived of some form of moral and epistemic normativity consistent with antirealism. Antirealists are not required to be error theorists.
Epistemic antirealism is the view that there are no stance-independent normative epistemic facts. That is, there are no facts about what epistemic practices you ought to employ that are true independent of your (or anyone else’s) goals or values. The absence of stance-independent epistemic facts of this kind doesn’t mean that there are no better or worse ways of acquiring true beliefs relative to the goal of acquiring true beliefs, or that one can’t know what these are, or value employing them. Thus, even if there were no stance-independent epistemic truths, this doesn’t bar us from employing truth-conducive epistemic standards. As such, it’s not at all clear how the epistemic antirealist would hold a self-undermining view unless epistemic realism were somehow epistemically required for, e.g., epistemic justification. But why would an epistemic antirealist hold that:
I am justified in rejecting epistemic realism.
The truth of epistemic realism is required for anything to be justified.
Let’s call (2) the epistemic-realism-as-necessary-for-justification, or necessity principle for short.
This is clearly an inconsistent pair of positions to hold, and someone who held this would be in trouble. Yet JPA and others seem to think we’re somehow obligated to endorse the necessity principle, even if we reject epistemic realism.
As the kid’s say, one philosopher’s modus ponens is another’s modus tollens. While one solution to the tension between epistemic antirealism and the necessity principle is to reject antirealism, there’s another, equally viable solution: reject the necessity principle.
What this shows is that epistemic antirealism is not self-undermining. To the extent that it is threatened, it is only threatened by the conjunction of both a commitment to epistemic antirealism and a commitment to the necessity principle. Since epistemic antirealism doesn’t commit one to the necessity principle, epistemic antirealism cannot be self-undermining. JPA is just wrong here. One would independently have to argue for the necessity principle in order to show that epistemic antirealism is undermined, but this would only result in the position being undermined by a conjunction: epistemic antirealism and the necessity principle. Epistemic realism would not, in and of itself, be self-undermining.
In other words, normative antirealism is only “self-undermining” conditional on the necessity of epistemic realism for justification. While someone might endorse this, it’s unclear why they’d have to. Much of what would prompt one to reject epistemic realism in the first place would also extend to the rationale for denying its necessity. A realist might scoff at the rejection of the necessity principle, insisting it’s obvious, or true, or whatever. But if the epistemic antirealist is prepared to reject epistemic realism, why think the antirealist would consider it a worrisome cost to reject the necessity principle, too?
The realist might insist this is some kind of extra bullet to bite: that one must bite two bullets instead of one, but this is all predicated on the notion that there is some presumption in favor of moral realism and, along with it, the necessity principle, such that to reject either is to pay a theoretical cost. The realist could then insist that you’d be paying two heavy costs, instead of just one. That’s double the bullets to bite!
Only, the antirealist can also deny that rejecting any of these positions involves biting any bullets. Sure, the realist might say, they can do this. One can reject anything! But…well, what? Surely there’s some presumption in favor of realism and the necessity principle? I don’t agree, and don’t think there are good arguments for this. Realists have insisted that their view is “commonsense” and “the default” and so on. This could apply to both moral realism and epistemic realism, and, arguably, may apply even more strongly to epistemic realism than moral realism.
4.0 Self-defeat objections to normative antirealism only work if you let them
Let’s take stock of all the positions the normative antirealist must allegedly “pay a cost” to deny.
The epistemic realist might think that there is some sort of presumption in favor of epistemic realism, and that one would pay a theoretical cost in denying this. And if one believes that there’s a presumption in favor of realism, one might also think there’s a presumption in favor of the necessity principle.
The moral realist may think there’s an analogous presumption in favor of moral realism, and perhaps its own moral form of the necessity principle (i.e., nothing is really right or wrong unless it’s stance-independently right or wrong; another common remark realists make).
The moral realist might start out only appearing to pay one cost (denying moral realism), but they end up paying four!
So normative antirealists are, on such a view, capable of denying moral realism only by paying a very heavy cost: they must reject epistemic realism, the epistemic necessity principle, moral realism, and the moral necessity principle. The antirealist is thereby denying four claims, each of which independently comes at a theoretical cost, not just one claim (denying moral realism). This would suggest that consistently maintaining moral realism while resisting the charge of being self-undermining comes at a very heavy cost.
Yet the normative antirealist can go further: they can also deny both necessity principles and deny that there’s any presumption in favor of moral realism. In other words, moral antirealists can and should embrace normative antirealism, but also deny that this requires them to pay any theoretical cost.
Note that these theoretical costs are not internal to the normative antirealist’s position. It isn’t an entailment of normative antirealist that it pays such costs. Facts about whether or not, and to what extent, one pays a theoretical cost are themselves substantive normative epistemic claims that are themselves (a) not an intrinsic feature of normative antirealism and (b) are subject to dispute.
I do dispute them! My primary focus in my academic research hasn’t even been to defend moral antirealism, but to argue that it’s not the case that most nonphilosophers are moral realists, or are implicitly committed to moral realism. I think the arguments in favor of folk realism are terrible. I wouldn’t be paying a cost to reject a presumption in favor of moral (or epistemic) realism, I’d be paying a cost to accept such a presumption! I’m not the first or only person to deny any presumption in favor of realism. See these takes
Franzén, N. (2024). The presumption of realism. Philosophical Studies, 1-22.
There is a fundamental problem with the dialectic surrounding normative realism and antirealism: narrative control. This is a phrase I recently heard Nathan Ormond employ. I take narrative control to be just that: control over the narrative surrounding some topic. For whatever reason, realists have helped themselves, for a very long time, to the notion that their position is “the default,” that it is the antirealist on the argumentative back foot, and that while we antirealists are allowed to exist under the benevolent eye of the realist, we must subordinate ourselves by paying the theoretical taxes they impose on us.
Well, I’m an American. We don’t like taxes. It’s in our blood. We’ve gone to war over this stuff. What I propose is that antirealists reject this narrative. Moral realists have never convincingly established a presumption in favor of their views. They’ve never shown that their view is “the default,” and they’ve never provided good evidence that most people are moral realists. We are not paying a cost in denying their ideas. The notion that we are lends credence to the idea that there is something preferable, or desirable-by-default about their ideas. But we can and we should reject this, and wrest the narrative control from realists. We should take all their presumptions, and all their claims to be “intuitive,” “commonsense,” “the default,” and “obvious,” bundle them up and dump them into the harbor.
References
Rowland, R. (2013). Moral error theory and the argument from epistemic reasons. The Journal of Ethics and Social Philosophy, 7(1), 1–24