The Normative Error Theorist Cannot Avoid Self-Defeat

The Normative Error Theorist Cannot Avoid Self-Defeat (PENULTIMATE DRAFT forthcoming in the Australasian Journal of Philosophy. Do not cite without permission) Abstract Many philosophers have noted that normative error theorists appear to be committed to saying ‘Error theory is true, but I have no reason to believe it’, which seems paradoxical. In defense of error theory, some have claimed that the word ‘reason’ in that statement is ambiguous between ‘normative reason’ and a purely descriptive sense of ‘reason’ that the error theorist can accept. I argue, however, that there is no descriptive sense of ‘reason’ that can prevent the paradox from re-emerging. Moreover, these implications of error theory probably provide good grounds for rejecting the view. 1. Introduction Normative error theory, the view that normative judgments attribute properties that are never instantiated, faces a persistent objection. The normative error theorist is apparently committed to saying that he has no reason to accept his own position, which seems self-defeating. One response, urged by Stan Husi [2013], Jonas Olson [2014], and Christopher Cowie [2016], distinguishes between normative reasons and some other sense of ‘reason’ that the error theorist can accept. If epistemic reasons are of the latter kind, then the error theorist can say that he has epistemic reasons to believe that there are no normative reasons. My primary task here is to argue that this strategy cannot succeed. If epistemic reasons are understood to be purely descriptive, then the problem reemerges, I contend. Can the error theorist simply embrace self-defeat, then? Bart Streumer [2013, 2018] takes that line, arguing that a position is no less likely to be true in virtue of being self-defeating in this way. I think this gambit is unpromising, but I will not seek to decisively refute Streumer’s position here. Rather, I will show that if the error theorist is to escape the self-defeat objection, then it must be through biting the bullet and acknowledging these apparently paradoxical implications. 2. The Self-Defeat Argument Normative error theory is the view that ‘normative judgments are beliefs that ascribe normative properties, even though such properties do not exist’ [Streumer 2013: 194]. Normative error theory—henceforth, simply ‘error theory’—is thus inconsistent with non-cognitivism, according to which normative judgments lack propositional content, and with various forms of realism, constructivism and subjectivism, all of which concur that normative properties of some sort are instantiated. Examples of thoroughgoing normative error theorists are hard to find. Many philosophers are suspicious categorical normativity, but leave room for instrumental normativity. J.L. Mackie, for example, argues that morality’s supposed “objective prescriptivity” is unacceptably ‘queer’, but he does not reject subjective, desire-based forms of value or normativity [1977: 15-25, 38-42]. So although Mackie is famous for his defense of moral error theory, he is not an error theorist in this stronger sense. As we shall see, Husi [2013], Olson [2014] and Cowie [2016], who are also skeptical of categorical normativity, sometimes say things that seem at odds with error theory. Streumer [2013, 2017] appears to be an exception, but with the ironic twist that his defense of error theory entails that no one, including himself, can fully believe it. About that, more later. Error theory does not have to be widely accepted to be worthy of philosophical attention, however. Few people accept Pyrrhonian skepticism. Arguably, no one can accept it. Nonetheless, radical skepticism is an appropriate subject of philosophical curiosity. We are rightly interested in finding the strongest grounds for rejecting it even if we know that it must be false. Likewise, people who are not skeptical of normative properties might want to know why positing them is unavoidable. One feature of error theory seems problematic: it is hard to see how it could appear reasonable from the point of view of the person who accepts it. After all, if error theory is true, then there are no reasons for doing, thinking, feeling or believing anything; hence there are no reasons to believe error theory. A parallel issue can raised for normative skepticism, the view that we have no reason to think that normative properties are instantiated. Diverse philosophers have noticed the problem for these views: Husi: ‘Skepticism [about normative reasons] appears to be cutting off the very justificatory branch it sits upon, seeking to engage [in] a dialectical enterprise while denying its currency’ [2013: 429]. Streumer: ‘The property of being a reason for belief, in the sense of a consideration that counts in favor of a belief, is a normative property. If the error theory is true, this property does not exist. The error theory therefore entails that there is no reason to believe the error theory’ [2013: 197; see also 2017: 155-188]. David Copp: ‘Paradoxically, if one thinks that an argument proves normative skepticism to be true, he cannot consistently hold that the argument justifies belief in normative skepticism’ [1995: 46–7]. Richard Joyce: ‘to question practical rationality is unintelligible—it is to ask for a reason while implying that no reason will be adequate’ [2001: 49–50]. Terence Cuneo: ‘If… epistemic nihilists hold that we do not have epistemic reasons to believe their position then their position is polemically toothless in the following sense: No one would make a rational mistake in rejecting it and no one would be epistemically praiseworthy for accepting it’ [2007: 117]. To be clear, the problem is not that error theory and normative skepticism are supposed to be logically contradictory. The issue is that something analogous to ‘Moore’s paradox’ arises. G.E. Moore [1993] famously noticed that statements like ‘It’s raining, but I don’t believe that it’s raining’ and ‘It’s not raining, but I believe that it is’ seem incoherent although they are not straightforwardly contradictory. The same is true of unspoken thoughts with such contents. Intuitively, whatever is amiss with these assertions and thoughts also seems to be wrong with ‘It’s raining, but I have no reason to believe that’ and so with ‘Error theory is true, but I have no reason to believe that.’ It is hard to see how a reasonable person could sincerely say such things, or think them privately. Philosophers have offered sundry accounts to explain what exactly is wrong with Moore paradoxical statements or the agents who say them—see, for example, Hintikka [1962] Williamson [1996], and Huemer [2007]. For overviews of the literature see Williams [2015a] and [2015b]. I do not intend to offer an account about what is wrong with such statements. But something intuitively is wrong with them; many philosophers with very different views recognize this. So for now I will proceed on the assumption that it is a bad feature of a philosophical position if accepting it commits us to self-defeating propositions. We are in a position to consider the following argument: Self-Defeat Argument The error theorist is committed to the self-defeating propositions, ‘Error theory is true, but I have no reason to believe that’. If adopting any philosophical position commits us to self-defeating propositions, then we should reject that position. C1: We should reject error theory. If we should reject error theory, then error theory is false. C2: Therefore, error theory is false. The argument does not end with C1 because we might have practical reasons to reject a true, and epistemically justified, philosophical position. For instance, a utilitarian could consistently think that utilitarianism is true while conceding that she has strong practical reasons to think that it is false. Note that the error theorist could not say anything similar about error theory. If any normative reasons are instantiated, including practical reasons to believe that error theory is either true or false, then error theory must be false. So (3) follows from the definition of error theory. For now, I am assuming that (2) is true, although we will consider Streumer’s challenge to it later. The weight of the argument rests on (1), the claim that error theory commits those who accept it to one or more self-defeating propositions. 3. Why the Error Theorist Cannot Avoid Self-Defeat In order to reject (1), the error theorist must be able to avoid saying that he has no reason to accept error theory. The only way to do this is to distinguish between normative reasons, which he must reject, and some other form of reasons—presumably, epistemic reasons—that he can accept. The basic problem for this reply is this: if epistemic reasons are to be consistent with error theory, then ‘epistemic reason’ must be understood in a deflationary way; however, the self-defeat paradox reoccurs when ‘epistemic reason’ are so understood. I see no way of threading this needle. Jonas Olson [2014] grapples with this problem in his extended defense of moral error theory, which seems motivated by a suspicion of normativity generally. Olson seems to offer two distinct answers to the self-defeat objection, although perhaps they amount to the same thing. The first is an appeal to reductionism. Following Mackie [1977], Olson claims that moral facts are unacceptably ‘queer’ because they entail an ‘irreducibly normative favoring relation’ [2014: 118]. He adds that ‘there need be nothing metaphysically queer about there being a reason for writers in English not to split the infinitive since, in one sense of ‘reason’, this is just for there to be a rule of grammar according to which splitting the infinitive is inappropriate’ (2014: 121). The same goes for other convention-based reasons: ‘[T]he point here is that moral error theorists need not deny that there are standards of correctness in logic and reasoning because these reasons are reducible’ [2014: 138. Emphasis mine]. The same does not seem to be true of moral requirements, which Olson finds metaphysically suspicious. The problem is that reduction is not elimination. Indeed, reductionistic views preserve entities instead of eliminating them. To say that A’s are reducible to B’s is to affirm the existence of A’s as B’s. An eliminative materialist about the mind in the mold of Paul S. and Patricia M. Churchland [1998], who maintain that there is no mind or folk psychological states, must reject the view that the mind is reducible to the brain and that folk psychological states are reducible to brain states. That is because reductionism about the mind and folk psychological states leaves these things in the picture instead of eliminating them. The error theorist must reject the view that normativity is reducible to convention-based norms for the same reason. As Streumer correctly notes [2017: 154, footnote 23]: for it to be true that we [do in fact] believe the error theory, we do not need to believe that judgements about reasons for belief ascribe an irreducibly normative relation. We only need to believe that these judgements ascribe a normative relation. The error theorist could say ‘there is no normativity, there are only conventions.’ That is consistent with error theory, but it does nothing to defuse the worry about self-defeat. So it seems that error theory is either self-defeating or contradictory. If a reductive view of reasons is assumed, then the error theorist contradicts himself. If normative reasons are eliminated, then the view is self-defeating. Other passages suggest a different solution. Or perhaps we should charitably read them as clarifying Olson’s initial ‘reducibility’ proposal. Olson cites John Broome, who identifies a weak sense of normativity: ‘Catholicism requires you to abstain from eating meat on Fridays. This is a rule and it is incorrect to eat meat on Fridays. So Catholicism is normative in this sense’; although, Broome adds, its norms are not binding upon non-Catholics [Broome 2007: 162]. Sarah Stroud similarly distinguishes between genuine normativity and domain-relative normativity, which is only forceful within a domain, or set of standards, whose authority an agent need not accept [1998: 172-173; see also Parfit 2011: 144-146]. For Olson, ‘normative’ implies ‘authoritative’ so that ‘To say that some behavior is correct or incorrect relative to some norm, N, is not to say anything normative. It is merely to say something about what kind of behaviour is required, recommended or forbidden by N’ [2014: 120]. This suggests a way of understanding epistemic reasons non-normatively: Epistemic reasons are reasons relative to norm E, the fundamental epistemic norms. As Husi exclaims: ‘Rejecting a norm’s authority, in short, does not amount to rejecting the norm itself!’ [2014: 431]. He adds that ‘there is nothing paradoxical or incoherent in the notion that skeptics may use the basic norms in order to advance their position while simultaneously denying their authority’ [2014: 435]. Christopher Cowie says something similar when he notes that we can reject categorical reasons for belief while maintaining that ‘evidential support relations’ hold between some beliefs [2016: 119]. I understand Husi and Cowie to mean that the error theorist can consistently accept that there are epistemic reasons, meaning reasons relative to norm E, while denying that E is authoritative. In other words, to use Owen McLeod’s terminology, the error theorist is free to deny that we Just Plain Ought to believe what we epistemically ought to believe [2001: 269-280]. ‘Just Plain Ought’ is also known as ‘ought simpliciter’ or ‘all-things-considered ought.’ For a defense of the idea that there is a Just Plain Ought, see Case [2016]. For a critique of that claim, see D. Baker [2018]. The error theorist can say that he has an epistemic reason to believe error theory without saying anything contradictory or self-defeating. This response relies on the assumption that it is possible to reject the authority of the epistemic domain while continuing to accept that there are identifiable epistemic reasons. Husi seems to think that it is always possible to accept a domain without regarding it as authoritative, though it is far from clear that this is the case. That does seem to be true of the rules of games and the rules of grammar. We cannot credibly cite grammatical rules to criticize a poet who deliberately bends or breaks them for effect, or someone who is speaking a different language in which they do not apply. The same goes for legality. An anarchist can recognize that some actions are illegal without thinking that this must affect his practical deliberation. He is certainly not contradicting himself when he says that some actions are illegal, but that we may disregard the law’s recommendations. Other normative seem to be defined in part by their claim to authority. Morality is a case-in-point. Some philosophers take ‘moral overridingness’ to be central to the nature of morality. This means that moral requirements are authoritative in such a way that our reasons to abide by them always, or nearly always, outweigh competing considerations. Sarah Stroud writes that ‘any moral conception which cannot deliver overridingness has a special burden to bear’ [1998: 171]. Some philosophers have proposed counterexamples to moral overridingness—for example, see Philippa Foot [1978] and Michael Slote [1983]. But even the supposed counter-examples to moral overridingness suggest morality has some degree of authority that has to be overridden by weightier, non-moral considerations. Suppose that we found out that we never had all-things-considered most reason to take morality’s dictates seriously. This discovery, I think, would cast doubt on whether moral reasons, obligations, and permissions existed at all. A psychopath who knew what actions were proscribed and recommended by so-called ‘moral requirements’, but did not understand that he should take these requirements seriously, assigning them substantial, if not indefeasible, weight in his decision calculus, would have a defective understanding of morality. So the existence of moral reasons, properly understood, implies that these reasons have a non-trivial degree of normative authority. So contra Husi, rejecting the authority of moral norms is to reject the moral norms themselves. The same is true, I think, of epistemic reasons, requirements and permissions. The essence of epistemic norms is that they are genuinely authoritative over how we should think. They are not merely epistemically authoritative over what we ought to think, which would be trivial. Rather, they tell us how we Just Plain Ought to think—at least in ordinary circumstances. Pascal’s wager-like cases, in which practical reasons overwhelm epistemic reasons, might be an exception to ordinary circumstances. For an argument that epistemic reasons can be commensurable with practical reasons, see Booth [2014]. That is what distinguishes epistemic norms from arbitrary norms of thought, and makes them a worthy subject of philosophical investigation. We could imagine an epistemic counterpart of the psychopath who knows that the epistemic norms forbid undisciplined pattern-seeking and indulgence in confirmation bias, yet thinks that these norms are arbitrary, like the rules of a game he can opt out of playing any time. I think we would say that this person is rationally criticizable, in some non-trivial sense, for failing to respect these norms. If that is right, then part of being reasonable full stop, part of being responsive to what we Just Plain Ought to do and think, is having a disposition to treat epistemic reasons, like moral reasons, as authoritative. If it turned out that no standards have any real authority, then neither morality nor epistemic standards would exist. Husi’s claim that the error theorist can accept the epistemic norms while denying their authority is therefore dubious. Suppose for the sake of argument that Husi is right, that epistemic norms exist although they lack normative authority. The error theorist encounters another problem. He must think that the domain of epistemic normativity is not just one arbitrary set of standards among many. It is a domain that he identifies with, accepts, or adopts—at least while he is doing philosophy—although there is no ultimate reason why he should adopt it over the innumerable alternatives. Evan Tiffany uses the term ‘partisanship’ to describe this relation [2007: 244-45]: Just as one may be a partisan supporter of the Canucks over the Maple Leafs – perhaps even seeing support for the Leafs as a character flaw, admittedly non-culpable for those raised in greater Toronto – without thinking that there is some deep metaphysical truth backing up one’s partisanship; so too can one be similarly partisan toward, e.g., morality, prudence, or authenticity. If the error theorist were not partisan toward the epistemic domain, to use Tiffany’s terminology, then he could not have epistemic reasons for accepting error theory (and thus avoid self-defeat). By way of analogy, if I am playing baseball, then I could have baseball-related reasons to do things like steal a base, or throw the ball to the third baseman to prevent a player of the opposing team from stealing a base. If I am not currently playing baseball, then I would have no reasons to do any of these things (though there are such reasons—for players). Similarly, the error theorist, in order to have epistemic reasons to believe error theory, has to adopt, identify with, or be partisan toward, the epistemic domain. Now consider: Weak Normativity Argument The error theorist’s partisanship toward the epistemic domain either makes a normative difference for him or it does not. If it does not, then the error theorist remains committed to self-defeating propositions. If it does, then error theory is false. C1: Therefore, error theory is either self-defeating or false. 1 is analytically true. It is also easy to see why 3 is true. If partisanship toward a domain makes a normative difference for an agent, then it makes the agent have normative reasons she did not previously have, which is inconsistent with error theory. Someone might be tempted to reply as follows: ‘What the error theorist rejects is authoritative normative reasons, not instrumental, or non-categorical reasons.’ That response amounts to moving the goalpost. Error theory was defined at the outset as the view that ‘normative judgments are beliefs that ascribe normative properties, even though such properties do not exist.’ This rules out categorical and hypothetical reasons alike. Error theory has to be defined this narrowly so that it is distinct from normative subjectivism and instrumentalism, which are accounts of normativity, not views that eliminate it. The weight of this argument rests on 2, the claim that the error theorist remains in a self-defeating position if her partisanship toward the domain of epistemic reasons does not make a normative difference for her. Husi suggests that the error theorist can spell out this relationship in psychological terms. He distinguishes between ‘commitments’—‘are stable psychological dispositions endorsed by reflection to follow some norm’s dictates’—and ‘fetishes’, which are less well-integrated into the agent’s psychology [2014: 443]. The error theorist will be committed to the epistemic norms, Husi writes: ‘Without any confidence in the authority of the relevant norms, the realist may complain, the [error theorist]’s intention to stay committed to them anyway must emerge as an unsustainable fetish, a sort of rule-worship in its worst form’ [2014: 443]. The issue was never to explain how belief in error theory is a psychological possibility, however. It is rather to explain how error theory can seem reasonable from the first-person perspective. In this respect, commitments are not relevantly different than fetishes unless ‘endorsed by reflection’ smuggles in normative content. If it does, then the initial dilemma reoccurs; if it does not, then the error theorist’s commitment to error theory remains a fetish, albeit a deep-seated one. For we would at most have an explanation for the error theorist’s belief in error theory, with no sense of how it could seem reasonable from the first-person perspective. The self-defeat problem persists. An alternative response is for the error theorist to concede that he does not have any epistemic reasons to believe error theory; nevertheless, there are such reasons, meaning there are considerations that count as reasons relative to the epistemic standards, whether adopted by the error theorist or not. But does it makes sense to say that there are reasons that no one has or ever could have as the error theorist must say? I assume that if normative error theory is true, the instantiation of normative properties would be impossible. Consider some arbitrary normative domains: promoting the interests of your least favorite politician, doing the opposite of whatever prudence and morality require, or maximizing the number of plastic forks in the world. Are there really reasons of the promoting-the-interests-of-your-least-favorite-politician kind? In what does their existence consist—the mere stipulation that they exist? I am inclined to say that there are no such reasons, since we are always free to disregard them. If that is right, then there would be no epistemic reasons, either, if they are supposed to be analogous to deflationary ‘reasons’ of this sort. Finally, even if we concede that epistemic reasons can be like this, it is not clear that they could help defuse the self-defeat objection. Compare the following: Self-Defeat 1: Error theory is true, but there is no reason for anyone to believe that. Self-Defeat 2: Error theory is true, but there is no reason—of a kind anyone need take the least bit seriously, all things considered—for anyone to believe that. The error theorist who takes the response that we are now considering would reject Self-Defeat 1, but remains committed to Self-Defeat 2. After all, the error theorist, in order to avoid positing normative reasons, would have to maintain that epistemic reasons have no more authority than any of the absurd ‘reasons’ just mentioned. I think that if Self-Defeat 1 is problematic or paradoxical, then Self-Defeat 2 is as well. The error theorist’s having a reason to accept error theory only resolves the paradox if that reason is in some way binding on what he ought to think. The error theorist cannot avoid the self-defeat objection simply by appropriating the word ‘reason’ while stipulating that the reasons are normatively vacuous. Left available to the error theorist is a tu quoque reply. The error theorist might say something like this: Take whatever your account of the epistemic standards is. I predict that it will evaluate the reasons I have for believing in error theory as good reasons. For example, suppose that one of the epistemic rules is ‘avoid believing contradictions.’ I can point out that my view is consistent so that belief in it is permitted by this rule. If error theory receives a clean bill of health by your standards, then you have no basis for rejecting it. I can win the argument by showing that my opponent’s house cannot stand. The problem with this response is that there is no space for it to work. This reply rests on the notion that epistemic standards either constitute or entail normative standards or they provide us no reason, obligation, or permission to believe anything. If the interlocutor were to have an account of normative reasons capable of constituting or justifying epistemic standards, then in virtue of this account he would reject error theory. I consider responding to the arguments given in favor of error theory to be beyond the scope of my project. See Huemer [2005: 199-214] and Shafer-Landau [2003: 231-346] for responses of this kind. My purpose is to develop and defend the Self-Defeat Argument against error theory. For this, it suffices to point out that the error theorist must do more than show that his opponent’s house cannot stand. That will not rescue his own house. 4. Embrace Self-Defeat? I turn my attention to the second premise of the Self-Defeat Argument, which until now I have been taking for granted: if adopting a philosophical position commits us to self-defeating propositions, then we should reject that position. Actually, we could weaken that premise somewhat. It would be enough to show that we have at least some reason, though not necessarily dispositive reason, to reject self-defeating views, since even this is incompatible with error theory. Streumer boldly rejects even the weak formulation of this premise. According to him, error theory entails that there is no normative reason, to believe error theory, but that is no reason for thinking that error theory is false. Streumer further thinks that we must take ourselves to have reasons for our beliefs, and that in order to fully believe anything—meaning, roughly, that no one can believe it with complete conviction—we must understand the obvious implications of our belief. Since it is an obvious implication of error theory that there are no reasons to believe error theory, it follows that no one, including Streumer, can fully believe it. For a critique of Streumer’s claims about the believability of error theory, see Lillehammer and Möller [2015]. Streumer insists that this puzzling feature of error theory gives us no reason to think that it is false [2013: 201; see also 2017: 129-137]: Just as a theory can be true if we do not believe it, a theory can also be true if we cannot believe it. Of course, if we cannot believe a theory, we cannot sincerely say that this theory is true. But our inability to sincerely say that a theory is true does nothing to show that it is false. Whereas Husi [2013: 431] seeks a path to ‘faithful participation in the argumentative enterprise,’ Streumer seems unbothered by the thought that his own dialectical participation might not be entirely faithful: ‘Instead of showing that my arguments are unsound or that the error theory is false, this would merely show that I have insincerely put forward sound arguments and have insincerely told you the truth’ [2013: 210-11]. He rejects David Lewis’s “simple maxim of honesty [to] never put forward a philosophical theory that you yourself cannot believe in your least philosophical and most commonsensical moments” on the grounds that “There is no reason why the truth could not be beyond our grasp” [2013: 212; see also Lewis 1984: x]. Streumer might also, if he is not too worried about the company he keeps, quote Thrasymachus, who tells Socrates: ‘What difference does it make to you whether I believe it or not? Isn’t it my account you’re supposed to be refuting?’ [Republic 349a]. Maybe Streumer softens this reply a bit when he reassures us that we can at least come close to believing error theory [2017: 177-8]: And to be in bad faith is to close one’s eyes to the truth not because one cannot believe it, but because one does not want to believe it. By coming close to believing the error theory in the ways I have described, I think I am as far from being in bad faith as it is possible to be. Would the fact that error theory cannot be sincerely endorsed, or cannot be fully believed, give us any reason for rejecting it? Perhaps not. As Cowie writes, ‘recall that moral error theorists have already bitten the bullet in accepting that that [sic] torturing the innocent for fun is not morally wrong’ [2016: 120]. For the error theorist to concede that he has no reasons to believe his own view is no more counterintuitive than many other of his commitments, such as the claim that the Trans-Atlantic slave trade was not morally wrong. Either that constitutes reductio ad absurdum argument against error theory on its own, rendering the Self-Defeat Argument superfluous, or else it does not, in which case it is unclear what the new objection is supposed to add. What is biting one more bullet to someone who has merrily bitten the rest of the arsenal? The assumption implicit in this reply is that the error theorist is immune to a certain kind of reductio ad absurdum argument. He need not worry about any counterintuitive commitment being exposed, provided that it is less counterintuitive than other claims he accepts. Consider, however, that some implications of error theory might be even more counterintuitive than the rejection of the Principle of Non-Contradiction. I can imagine being presented with an apparently sound argument that the only way for me to preserve my belief that there are reasons not to kill and eat my own children is to accept dialetheism, the view that there are some true contradictions (Graham Priest [2006] defends this view). In that case, accepting dialetheism might seem to be the smaller bullet to bite. Surely that would not show that an argument demonstrating that error theory commits us to rejecting Principle of Non-Contradiction would be without force. The error theorist could not credibly say: ‘Perhaps my view commits me to rejecting the Principle of Non-Contradiction, but that is ok, because I am already committed to biting bigger bullets than that.’ The lesson here is that even if the intuitive costs of error theory are already high, an argument that raises those costs further may nonetheless be worth making—especially if it establishes a different kind of intuitive cost. Biting the self-defeat bullet weakens the error theorist’s dialectical position in a way that other appeals to moral intuition do not. If the error theorist acknowledges that such things as sincerity or ‘faithful participation in the argumentative enterprise’ is irrelevant, then he has no grounds on which to object if his interlocutor argues in bad faith (by, for example, simply refusing to consider objections to his arguments). Streumer, or another hard-bitten error theorist who is willing to embrace self-defeat, might be happy to accept this. How should we adjudicate the dispute with him, then? Whenever we find an intuitive standoff like in philosophy, it is worth asking how things are likely to seem to a neutral arbiter—in this case, that would be someone who is neither committed to accepting error theory, nor committed to rejecting it. A neutral arbiter would presumably be a reasonable arbiter, someone who is committed to upholding basic norms of discourse, such as sincerity, that constitute philosophical best practice. Such a person could be expected to disfavor the error theorist in this exchange. We might also consider what would be the most that we could reasonably expect a critic of error theory to show. The strongest possible evidence we could have against any philosophical position is proof that it is logically inconsistent. Error theory appears to be consistent, but logical consistency is cheap. The next best thing, it seems to me, would be a demonstration that the position cannot be reasonably held by anyone. That this counts against a view’s likelihood of being true—to at least a very slight degree—seems plausible even if error theorists cannot be forced, on pain of contradiction, to accept it. 4.5 Does the Self-Defeat Argument prove too much? We might reasonably want to know whether there are any other philosophical positions beside error theory that are susceptible to this critique. Pyrrhonian skepticism has already been mentioned. We probably will not be alarmed or surprised if we were to discover that it is also self-defeating. If the Self-Defeat Argument overgeneralizes to impugn other apparently reasonable views, however, then we might have grounds to be suspicious of it. Plausibly, a similar concern applies to eliminative materialism about the mind of the sort already dicussed. Someone who thinks that there are no folk psychological states such as beliefs seems committed to saying ‘Eliminativism is true, but I do not believe that it is true,’ which is straightforwardly Moore-paradoxical. Lynne Ruder Baker advances this criticism [1987: 134]. I think it is relevant how radical a revision to our folk psychology eliminativism requires. If the eliminativist thinks that a completed neuroscience will turn up nothing at all that remotely resembles belief, or an affirmation of a proposition, then there is indeed no way to avoid Moore-paradoxicality. I think that this would count against eliminative materialism. The eliminativist might think that although there are no beliefs per se, she is open to the idea that neuroscience will eventually uncover something similar to it. So an eliminativist could say something like ‘Eliminativism is true, and although I do not believe it is true, there is some sense, a sense that neuroscience will eventually uncover, in which I affirm that it is true’. Perhaps that’s enough to make eliminativism somewhat less vulnerable to this kind of objection than error theory. One version of epistemic foundationalism, the view that there are some epistemically justified beliefs that are not epistemically justified by anything further, might also be subject to a self-defeat objection. Suppose that A is one of those beliefs. The foundationalist would then be committed to saying ‘A, but I have no epistemic reason to believe that A’ (see Reid [1997: 168-9]; Sampson and Hyun [2014: 634]. Not every version of foundationalism would be threatened by self-defeat. For example, some foundationalists maintain only that some beliefs are not justified by any further belief, though they may be justified by something else, such as appearances [Huemer 2001: 93-118]. But even the first kind of foundationalist might have strategies available for avoiding the problem. First, she might draw attention to the fact that the Self-Defeat Argument concerns philosophical positions, not foundational beliefs. Relying heavily on this distinction might seem ad hoc, but a foundationalist can be expected to take the position that we should treat philosophical positions differently than foundational beliefs, since the former arise from reflection on foundational beliefs. A second strategy is to note that foundationalism concerns only epistemic reasons, not all normative reasons. A foundationalist can concede that we might still have practical reasons to believe foundational beliefs: ‘I believe that my senses are reliable not because I can give any argument to support that, but because it would be bad for me to adopt a policy of doubting that’. The error theorist, by contrast, necessarily rejects both epistemic and practical reasons. So there is room to accept the Self-Defeat Argument against error theory and also accept this form of foundationalism. This second line of reply also applies to a different position that also seems to be threatened by the self-defeat argument. Some philosophers such as Charles Côté-Bouchard [2017] have expressed skepticism about epistemic normativity, or about a normative domain of “rationality” that governs consistency in belief and action, without committing themselves to skepticism about normativity generally. Kolodny [2005] for example is skeptical that there are specific reasons to be rational in this way. For a comprehensive discussion of the normativity of rationality in this sense, see Way [2010]. But this sense of ‘rationality’ is distinct from all-things-considered ought, or Just Plain Ought. Philosophers who reject epistemic reasons are, no less than error theorists, committed to ‘There are no epistemic reasons, and I have no epistemic reason to believe that.’ This also has a paradoxical ring to it, though to my ear it is not quite so pronounced as ‘Error theory is true, but I have no good reason to believe that.’ The reason is that the foe of epistemic reasons can fill out the sentence in a way that mollifies our intuitions somewhat, for example: ‘There are no epistemic reasons, and although I have no epistemic reasons to believe that per se, I nonetheless acknowledge that I have reasons of some sort to believe that.’ I do not think a similar move is available to the error theorist. He might try something like: ‘Error theory is true, and although I have no normative reasons to believe that, there are nonetheless compelling indications that error theory is true.’ That sounds o.k., but I think it only forestalls the trouble. We can ask whether those “compelling indications” are reasons to believe or not. A negative answer would be baffling (‘compelling considerations to believe X that are not reasons to believe X’?), an affirmative answer is inconsistent with error theory. Attempts to have it both ways will return us to the difficulties discussed in section 3. Maybe further interrogation of the view that there are no epistemic reasons would reveal similar difficulties, though this is beyond the scope of my present investigation. Conclusion I introduced the Self-Defeat Argument against normative error theory and considered how an error theorist might resist it. One strategy is to cash out epistemic reasons in descriptive terms to avoid paradoxical-sounding commitments. I claim that this fails because if ‘epistemic reasons’ are understood to lack normative force, then the paradox remains. The error theorist might also embrace self-defeat. Although this proposal is logically consistent, it is extremely counterintuitive and compromises the error theorist’s dialectical position. I conclude that the Self-Defeat Argument provides us both with good reasons to reject error theory and with good reasons for thinking that it is false. I would like to thank, in no particular order, Alastair Norcross, Michael Huemer, Chris Heathwood, Terence Cuneo, Graham Oddie, Jim Skidmore, Derek Baker, Tyler Paytas, Rishi Joshi, Oliver Traldi, Joshua Blanchard and Stephen Kershnar for comments on previous drafts of this paper. I also owe thanks for the three anonymous reviewers at this journal. References Baker, Derek. 2018. Skepticism about Ought Simpliciter, in Oxford Studies in Metaethics Vol. 13, ed. Russ Shafer Landau, Oxford: Oxford University Press: 230-252. Baker, Lynne Rudder. 1987. Saving Belief: A Critique of Physicalism. Princeton: Princeton University Press. Booth, Anthony Robert. 2014. Epistemic Ought is a Commensurable Ought, European Journal Of Philosophy, 22/4: 529-539. Boyd, Richard. 1988. How to Be a Moral Realist, in Essays on Moral Realism, ed. Geoffrey Sayre-McCord. Ithica, NY: Cornell University Press: 181-228. Churchland, Paul S. and Patricia M. 1998. On The Contrary: Critical Essays 1987-1997. Cambridge, Mass: MIT Press. Case, Spencer. 2016. Normative Pluralism Worthy of the Name is False, Journal of Ethics and Social Philosophy 11/1: 1-20. Copp, David. 1995. Morality, Normativity, and Society. New York: Oxford University Press. Côté-Bouchard, Charles. 2017. Belief’s Own Metaethics? A Case Against Epistemic Normativity. Dissertation, King’s College London. Cowie, Christopher. 2016. Good News for the Error Theorist: A Master Argument Against Companion in Guilt Strategies, Australasian Journal of Philosophy 94/1: 115-130. Cuneo, Terence. 2007. The Normative Web. New York: Oxford University Press. Foot, Philippa. 1978. Are Moral Considerations Overriding? in Virtues and Vices and other essays in Moral Philosophy. Berkeley: University of California Press: 181-188. Hintikka, Jaakko. 1962. Knowledge and Belief: An Introduction to the Logic of the Two Notions. Ithaca: Cornell University Press: 64-71. Huemer, Michael. 2001. Skepticism and the Veil of Perception. Lanham, Md.: Rowman & Littlefield Publishers. Huemer, Michael. 2005. Ethical Intuitionism. Houndmills, Basingstoke, Hampshire; New York: Palgrave MacMillan. Huemer, Michael. 2007. Moore’s Paradox and the Norm of Belief in Themes From G. E. Moore: New Essays in Epistemology and Ethics ed. Susana Nuccetelli and Gary Seay, Oxford: Clarendon Press. Husi, Stan. 2013. Why Reasons Skepticism Is Not Self-Defeating, European Journal of Philosophy 21/3: 424-49. Hyun, Alexander and Eric Sampson. 2014. On Believing the Error Theory, Journal of Philosophy 111/11: 631-40. Joyce, Richard. 2001. The Myth of Morality. Cambridge: Cambridge University Press. Kolodny, Niko. 2005. Why Be Rational? Mind 114/455: 509-563. Lillehammer, Hallvard and Niklas Möller. 2015. We Can Believe the Error Theory, Ethical Theory and Moral Practice 18/3: 453-9. Mackie, J.L. 1977. Ethics: Inventing Right and Wrong. New York: Penguin Books. Moore, G.E. 2013a. Moore’s Paradox, in G.E. Moore: Selected Writings, ed. Thomas Baldwin. New York: Routledge: 207-12. Olson, Jonas. 2014. Moral Error Theory: History, Critique, Defense. New York: Oxford University Press. Parfit, Derek. 2011. On What Matters: Volume One, New York: Oxford University Press. Plato. 1997. Republic. Translated by G.M.A. Grube, rev. C.D.C. Reeve in Plato: Complete Works, ed. by John Cooper. Indianapolis/Cambridge: Hackett Publishing Company. Priest, Graham. 2006. In Contradiction: A Study of the Transconsistent. Oxford: Oxford University Press. Reid, Thomas. 1997. Inquiry into the Human Mind on the Principles of Common Sense, ed. Derek R. Brookes. University Park, PA: Pennsylvania State University Press. Shafer-Landau, Russ. 2003. Moral Realism: A Defence. New York: Oxford University Press. Slote, Michael. 1983. Admirable Immorality, in Goods and Virtues. Oxford: Clarendon Press: 77-107. Streumer, Bart. 2013. Can We Believe the Error Theory? Journal of Philosophy 110/4: 194-212. Streumer, Bart. 2017. Unbelievable Errors. New York: Oxford University Press. Stroud, Sarah. 1998. Moral Overridingness and Moral Theory, Pacific Philosophical Quarterly 79/2: 170-89. Tiffany, Evan. 2007. Deflationary Normative Pluralism, Canadian Journal of Philosophy, 37/5: 231-62. Way, Jonathan. 2010. The Normativity of Rationality, Philosophy Compass 5/12 :1057-1068. Williams, John N. 2015a. Moore’s Paradox in Speech: A Critical Survey, Philosophy Compass 10/1: 10-23. Williams, John N. 2015b Moore’s Paradox in Thought: A Critical Survey, Philosophy Compass 10: 24-37. Williamson, Timothy. 1996. Knowing and Asserting, Philosophical Review 105/4: 489-523. 24