Scholarly article on topic 'After Moral Error Theory, After Moral Realism'

After Moral Error Theory, After Moral Realism Academic research paper on "Philosophy, ethics and religion"

Share paper
Academic journal
The Southern Journal of Philosophy
OECD Field of science

Academic research paper on topic "After Moral Error Theory, After Moral Realism"

The Southern Journal of Philosophy Volume 53, Issue 2 June 2015



Stephen Ingram

abstract: Moral abolitionists recommend that we get rid of moral discourse and moral judgement. At first glance this seems repugnant, but abolitionists think that we have overestimated the practical value of our moral framework and that eliminating it would be in our interests. I argue that abolitionism has a surprising amount going for it. Traditionally, abolitionism has been treated as an option available to moral error theorists. Error theorists say that moral discourse and judgement are committed to the existence of moral properties, and that no such properties exist. After error theory is established, abolitionism is one potential way to proceed. However, many error theorists suggest that we retain moral discourse as a sort of fiction. I evaluate some attractions of both fictionalism and abolitionism, arguing that abolitionism is a plausible position. No one doubts that error theorists can be abolitionists. However, what has gone largely undiscussed is that it is open to others to be abolitionists as well. I argue that moral realists of a metaphysically robust sort can and perhaps should be abolitionists. 'Realist abolitionism' makes for a surprisingly neat theoretical package, and I conclude that it represents an interesting new option in the theoretical landscape.

If someone told you that we should get rid of moral discourse and moral judgement, you'd probably raise an eyebrow or two. These things are deeply ingrained in human culture and psychology, so the idea that we should eliminate them seems strange. Indeed, it seems repugnant. What sort of person

Stephen Ingram is a PhD student at the University of Sheffield. His interests lie primarily in metaethics, and his work is currently focused on the ontological commitments incurred by robust forms of moral realism.

1 I am grateful to two anonymous referees for this journal, to Jimmy Lenman, Miranda Fricker, Rich Healey, Pete Caven, Carl Fox, Lewis Brooks, Graham Bex Priestley, and Lizzy Kirkham. This paper was written during an AHRC doctoral scholarship.

The Southern Journal of Philosophy, Volume 53, Issue 2 (2015), 227-248. ISSN 0038-4283, online ISSN 2041-6962. DOI:

would want to eradicate the right and the good from their conceptual repertoire? Not a very nice one, you might think. But that is roughly what defenders of 'moral abolitionism' propose, and abolitionists don't regard this proposal as repugnant. Indeed, they think that we'd be better off if we got rid of our moral framework. Abolitionists think that we've overestimated the practical value of this framework, and I think that abolitionism has a surprising amount going for it.

Abolitionism has traditionally been treated as an option for moral error theorists. Roughly, error theorists about morality say that moral discourse and judgement are committed to the existence of moral properties, and that there are no such properties. If that's right, then our moral framework is systematically in error: all moral sentences and beliefs are false because they fail to refer to anything that actually exists. Error theory received its first major defence from J.L. Mackie (1977), and it finds contemporary advocates in Richard Garner (1990), Richard Joyce (2001; 2006; 2011), Bart Streumer (2008; 2011) and Jonas Olson (2010; 2014), among others. If error theory is true, an obvious question arises. What happens next? What should we do with moral discourse and moral judgement once we realise that they fail to refer to anything real?

Abolitionists like Hinckfuss (1987) and Garner (2007) say that we should get rid of the moral framework. They think that we'd be better off without it. However, some error theorists think that it is in our interests to retain moral discourse and judgement in some form. Perhaps we should conserve the moral framework as it stands despite its faultiness, as Olson (2011; 2014) suggests. Or perhaps Joyce (2001; 2005) is right and we should treat it as a useful fiction. Of course, the shoulds in all these claims cannot be moral. Error theorists hold that there is nothing that one morally ought to do, because there are no moral properties. But they accept that there are things that one pru-dentially ought to do. So when error theorists make claims about what we should do with moral discourse and judgement, these are to be interpreted as prudential claims about what is in our interests.

No one doubts that error theorists can be abolitionists. However, what has gone largely undiscussed is that it is open to others to be abolitionists as well.2 I am going to argue that moral realists of a metaphysically robust sort can and perhaps should be abolitionists. This metaphysically robust form of realism has become increasingly popular in recent years. It finds prominent defenders in Russ Shafer-Landau (2003), Michael Huemer (2005), Graham Oddie (2005), William Fitzpatrick (2008) and David Enoch (2011), among

2 The possibility that abolitionism could be accepted by those who reject error theory has been noted by Olson (2014, 179 n.5) and Campbell (2014).

others. I will show how it is open to these metaphysically robust realists to defend abolitionism. Not only that, I will also be arguing that 'realist abolitionism' is not just an abstract possibility. Realism and abolitionism make for a surprisingly neat package, and this package is an interesting new option in the theoretical landscape.

In §2 I sketch the commitments of error theory and metaphysically robust moral realism (henceforth just 'moral realism') in more detail. In §3 I argue that abolitionism must be taken seriously. Although there is not space here for a thorough cost-benefit analysis of the moral framework, abolitionism is surprisingly plausible. In §4 I show how abolitionism is compatible with moral realism, and I suggest that realist abolitionism is an interesting new package. I should emphasise that I am not going to argue for or against either error theory or realism. My topic is simply what happens after each of these theories is established. I want to know what happens after moral error theory, and after moral realism.

Realists and error theorists have much in common. Both accept the following commitments:

Discourse. Moral sentences are truth-apt.

Judgement. Moral judgements express beliefs.

Ascription. Moral sentences and beliefs ascribe irreducible and categorically authoritative moral properties to acts (and persons, institutions, and so on).

Discourse and Judgement are fairly straightforward. Together they constitute a commitment to cognitivism. (Some characterise cognitivism as the view that moral sentences are truth-apt. Others characterise it as the view that moral judgements express beliefs. I'm playing it safe by treating both as complementary elements of a broader cognitivist stance.) In accepting Discourse and Judgement, realists and error theorists are thus united as cognitivists. This puts them in competition with the non-cognitivism of A.J. Ayer (1936) and R.M. Hare (1952) and the expressivism of Simon Blackburn (1984; 1993a; 1998) and Allan Gibbard (1990; 2003).3 I will return to this competition in §3.3.1.

In accepting Ascription, realists and error theorists are further united on the essential nature of moral discourse and judgement. They agree that 'stealing is wrong' ascribes a certain property — wrongness — to stealing. And they

3 Complication: Blackburn and Gibbard endorse quasi-realism. Quasi-realists claim to capture Discourse by deflating the concept of truth, and they claim to capture Judgement by treating the concept of belief expansively enough to include certain conative states. I leave this complication aside here.

agree that this property is essentially irreducible and categorically authoritative. To say that wrongness is an irreducible property is to say that it cannot be identified with or analysed in terms of any more basic kind of property. And to say that it is a categorically authoritative property is to say that the normative force provided by wrongness is unconditional. Moral properties direct us to perform or refrain from performing certain courses of action. For instance, 'stealing is wrong' directs us not to steal. Wrongness is categorically authoritative insofar as the direction it provides is "independent of social or psychological contingencies" (Hampton 1998: 96).4

In accepting Ascription, then, realists and error theorists are united on the essential nature of moral discourse and moral judgement. As Joyce (2001: Ch. 2) puts it, they treat Ascription as a 'conceptually non-negotiable' feature of moral discourse and moral judgement. This means that a sentence or belief that ascribes anything less than an irreducible and categorically authoritative property is not a specifically moral sentence or belief. It is a sentence or belief of some other kind. Joyce's reason for thinking this is that moral direction does not feel 'escapable' in the way that, say, prudential direction does. Here is a piece of prudential direction: if you want your boss to like you, you should offer her a biscuit. The direction 'you should offer her a biscuit' is conditional on your wanting her to like you. If you don't care about that, this piece of direction will have no force for you. So prudential direction is escapable by being conditional on something that you might legitimately reject. Moral direction is not like that. It feels inescapable, and this is explained by the categorical authority of moral ascriptions.

We can now see how realism and error theory part ways. Realists accept the following:

Realist Metaphysics. There are irreducible and categorically authoritative moral properties.

Error theorists deny this, and so accept the following:

Anti-Realist Metaphysics. There are no irreducible and categorically authoritative moral properties.

Although realism and error theory have much in common, they have a major metaphysical disagreement. Error theorists find the notion of an

4 Note that this is not the same as the claim that moral properties are response-independent. Response-independence is a claim about the existential status of moral properties. Categorical authority is a claim about the nature of the direction provided by such properties. Natural properties can be response-independent, but error theorists and realists will deny that natural properties can be categorically authoritative.

irreducible and categorically authoritative moral property to be 'queer,' and they reject it.5 The upshot of the error theorist's commitments is that moral discourse and moral judgements are systematically in error. All moral sentences and beliefs are false. All ascriptions of moral properties fail to refer.

Error theorists clearly face the 'what happens next?' question alluded to in §1. What happens once it is shown that there are no moral properties? What should we do with moral discourse and judgement? As I will argue in §4, the same 'what happens next?' question also confronts moral realists. It will be easiest to see this, however, once we have discussed the question in its more traditional error theoretical context. This is the topic of §3.

We would need a thorough cost-benefit analysis of moral discourse and moral judgement in order to answer the 'what happens next?' question conclusively. I can't hope to be that thorough here. Moreover, my discussion will inevitably be speculative, as is every other discussion of this question. All I can do is be fair to each side of the debate.

There are three main options for how to proceed after error theory is established: conservationism, fictionalism, and abolitionism. Conservationists claim that we should continue to embrace moral discourse and moral judgement, despite their failure to refer to anything that actually exists. Fictional-ists claim that we should treat the moral framework as a useful fiction, so that we engage with it as part of a make-believe. Abolitionists claim that we should eliminate (fictive and non-fictive) moral ascriptions.6 The essence of the abolitionist position is a prohibition on uttering sentences and making judgements that ascribe moral properties to acts.7

Conservationists and fictionalists agree that moral discourse is prudentially worth keeping in some form, whereas abolitionists think that it is in our interests to eradicate it. Although conservationism is not undefended (see Olson 2011; 2014), I will set it aside. I am interested in the prudential value of

5 Mackie's argument from queerness has several aspects, and the queerness of categorical authority is just one of them. Olson (2014) identifies four queerness arguments (from motivation, supervenience, knowledge, and categorical authority). Olson follows Garner (1990) and Joyce (2001, Ch. 2) in taking the queerness of categorical authority as providing the strongest argument for error theory.

6 It is important for understanding the precise nature of the abolitionist position that fictive as well as non-fictive moral ascriptions are eliminated. I thank an anonymous referee for pointing this out. In the rest of the paper I will take it as given that fictive as well as non-fictive moral ascriptions are abolished.

7 Making judgements includes the process of deliberation that leads to the judgement.

retaining the moral framework in some form, and focusing on the debate between fictionalists and abolitionists will make it easier to expose the general costs and benefits of moral discourse and moral judgement. In §3.2 I explain fictionalism in more detail and evaluate its attractions, and in §3.3 I do the same for abolitionism. My conclusion will be modest: abolitionism is a plausible position that must be taken seriously.

According to fictionalists, we should see the moral framework as a useful fiction — a myth or pretence that it is in our interests to maintain. Fictionalism can be a descriptive or a prescriptive thesis. The descriptive thesis says that our moral practices are actually fictive, and the prescriptive thesis recommends that we reform our (non-fictive) moral practices so that they become fictive. I will concentrate on prescriptive fictionalism, as it seems quite unlikely that our moral practices are actually fictive.8

The guiding idea of prescriptive fictionalism (henceforth just 'fictionalism') is to treat moral discourse and judgement as part of a make-believe. Participating in make-believe involves thinking a proposition without believing that it is true. Here is Joyce:

When a child make-believes that the upturned table is a ship, she is thinking the proposition "The table is a ship" (with all its associated imagery), or perhaps simply "This is a ship," without believing that proposition. The proposition is, of course, false, but we could not on that account accuse the child of any error (2001: 197).

Joyce suggests that something similar can be recommended for moral judgements. If we construe moral judgements as fictive, judging that stealing is wrong involves thinking the proposition 'stealing is wrong' without actually believing it. This involves no error, for it doesn't purport to be anything other than make-believe. In short, fictionalists recommend reforming moral judgement so that we merely pretend to believe that moral ascriptions successfully refer.

Fictionalists can allow that there are reflective contexts in which we acknowledge that there are no moral properties. As Joyce (2001: 196) puts it, the fictionalist can "enter the 'critical mode' should he care to." Admittedly, you might doubt that this is psychologically feasible. Whereas we easily disengage from other sorts of fiction by putting the book down or turning off the television, it will be harder to disengage from the moral fiction if we

8 For a defence of descriptive fictionalism, see Kalderon (2005). It is beyond the scope of this paper to explain what I think is wrong with descriptive fictionalism. See Lenman (2008) and Cuneo and Christy (2011, 88—92).

generally operate within it unreflectively. However, an obvious reply is that it will simply require some effort to periodically remind oneself of the fact that there are no moral properties. Fictionalism is a reforming doctrine. It tells us to revise the way we engage in moral discourse and judgement. It's not unreasonable for the implementation of such a reform to involve some psychological work. The more interesting question is whether this work is worth it. Although I cannot provide the thorough cost-benefit analysis required to answer this question definitively, I will outline and evaluate some potential attractions of fictionalism.

Daniel Nolan, Greg Restall, and Caroline West (2005: 310—314) suggest a number of putative advantages of fictionalism over abolitionism. I will discuss the three that seem most convincing. The first attraction of fictionalism over abolitionism identified by Nolan, Restall and West (henceforth 'NRW') is that fictionalism would be more psychologically convenient. For most people, abolishing thought and talk of phlogiston wasn't psychologically problematic because it was a specialised, technical concept. But moral ascriptions of rightness and goodness, for example, are pervasive and deeply ingrained, so their elimination from both discourse and judgement would involve a significant psychological upheaval. NRW think that fictionalism has the benefit of avoiding such an upheaval.

However, I doubt that fictionalism has much advantage over abolitionism here. Understood as a prescriptive thesis, fictionalism involves significant psychological inconveniences of its own. The move away from our ordinary moral framework into a fictional moral framework will involve a pretty inconvenient psychological upheaval, and this is a cost of fictionalism. Moreover, as I've noted, fictionalists must periodically enter the critical mode and recall that they are engaged in a fiction in order to avoid slipping back into the ordinary moral framework. This self-surveillance will require some effort, so will also be psychologically inconvenient. Any advantage the fictionalist has in terms of psychological convenience should therefore not be overstated, given the psychological inconveniences of undertaking the fictionalist's own recommended reforms.

A second advantage of fictionalism identified by NRW is that it allows us to avoid frequent forays into metaethics when considering what to do in normal life. In other words, rather than going through the inconvenience of prefacing practical recommendations with claims about how there's actually nothing that one morally ought to do, we can just proceed as usual. For instance, when someone asks a fictionalist whether she is morally obliged to keep her promise, the fictionalist need not say something like: 'actually, you're mistaken in thinking that there are any moral obligations at all because error theory is true, but keeping this promise would be in your long-

term interests because...' Fictionalism avoids such excursions into meta-ethics, which those who aren't professional metaethicists apparently find inconvenient.

The problem with this suggestion is that it only has any force if our group contains both fictionalists and realists. In that case, the fictionalists will have the convenience of not continually declaring their disbelief in moral properties, though it's worth noting that fictionalism has a certain cost here because realists may be understandably upset if they learn that they have been given moral advice by people who are only pretending to believe what they're saying. That aside, the ideal situation for almost any reforming theory is one in which belief in that theory is part of a background network of beliefs shared by the community. Abolitionists would like to see their position gain wide acceptance, as would fictionalists. Indeed, Joyce (2001: 177) specifically targets his fictionalism at a community of error theorists with broadly shared ends. If most people came to believe error theory, there would be no need to get into metaethics in ordinary discussions because error theory would already be common ground. So in our society of abolitionists, if you ask me 'Should I keep my promises?' I can straightforwardly reply 'Yes,' since both your question and my answer will be uttered in an implicitly non-moral conversational context. In that context, we will naturally interpret the 'should' as prudential rather than ethical.

Admittedly, creating a society of abolitionists will probably require a programme for educating people out of moral thinking. Even if everyone arrives at the conclusion that error theory is true, creating a society of abolitionists will be no mean feat. The moral framework is so ingrained that it will be hard to remove even after error theory is established. This is part of the psychological inconvenience of abolitionism. But creating a society of fictionalists will be no mean feat either. Fictionalists will also need an educational programme to spread the moral make-believe, that is, to turn ordinary moral ascriptions into pretend moral ascriptions. Of course, ordinary folk typically lack clear-cut metaethical opinions. So the idea of a community of error theorists is of course an idealisation. But it is an idealisation to which any reforming doctrine, including prescriptive fictionalism, is committed. So fic-tionalism has no significant advantage over abolitionism here.

The third and probably most obvious putative attraction of fictionalism identified by NRW is that it appears to have the

capacity to salvage the important role moral discourse is widely thought to play in co-ordinating attitudes and regulating interpersonal conflict in cases where people disagree about what they are to do, especially where collective action is needed or the proposed actions of different people interfere with each other (2005: 312).

We are used to discussing practical matters within a moral framework, and abolishing this framework would mean finding new ways to reach answers about the questions we face as a community. The framework within which we organise and police our social interaction would have to be replaced with something else, something that makes things go well for us without moral ascriptions. Fictionalism avoids the need to restructure the regulation of interpersonal conduct.

Of the putative attractions of fictionalism identified by NRW, this is probably the one that will be thought most effective. Actually, abolitionists think that moral discourse and judgement tend to hinder rather than help the resolution of interpersonal conflicts, and that replacing the moral framework is therefore in our interests. I will delay discussing this until §3.3.2, however, for it is on precisely this point that the main case for abolitionism rests. In general, it is worth noting that NRW's suggestions largely concern the convenience of retaining moral discourse and judgement. They think that abolitionism will require more effort (psychologically, linguistically, and interpersonally) than fictionalism. Part of the abolitionist's job is to show that these inconveniences are worth bearing.

Another potential benefit of fictionalism is identified by Joyce. Joyce notes that merely being aware that an act is in our long-term interests doesn't ensure that we'll actually do it. Short-term gains have an appeal that "may subvert the agent's ability to deliberate properly so as to obtain a valuable delayed benefit, leading him to 'rationalize' a poor choice" (2005: 301). Joyce suggests that fictionalism lets us supplement and reinforce prudential direction with the categorical authority implied by moral direction. Retaining the impression of this categorical authority will help motivate us to act on our long-term interests, because ascribing a moral property to some act implies that this act is inescapably right. Saying 'you are morally obliged to keep your promises, and you'll be a morally bad person if you don't' gives more forceful (so more effective) direction than saying 'keeping promises tends to be in your long-term interests.'

Joyce's claim that fictionalism will help with weakness of will carries greater weight than NRW's suggestions. Joyce's is not merely a claim about the inconvenience of abolitionism. Rather, it is a claim about the mechanisms by which one can most effectively realise prudential value. Abolitionists may be on the back foot here because, once moral discourse and judgement have been abolished, practical recommendations will lack the appearance of categorical authority contained in by moral ascriptions. Individuals living in a society of abolitionists do not have a resource that fictionalists can appeal to here, so they must find other ways to avoid the lure of short-term over long-term gains.

Some of fictionalism's putative attractions have been overstated, then, but fictionalism does have certain things going for it. Not so much, however, as to make it obvious that moral ascriptions are worth keeping. In §3.3, I survey and assess some of the putative attractions of abolitionism.

3.3 3.3.1

I have said that abolitionism consists essentially in the recommendation that we eliminate moral ascriptions. That is, abolitionists advocate a prohibition on utterances and beliefs that ascribe moral properties to acts. Initially, abolitionism seems repugnant. However, abolitionists don't take the elimination of our moral framework to involve humanity regressing into nightmarish anarchy.9 For one thing, there would still be a legal framework to handle certain social issues. But abolitionists also say that we've overestimated the prudential value of moral discourse and judgement. They think that it is in our interests to eliminate the moral framework and replace it with something more effective.

Before outlining and assessing abolitionism's attractions, we should say something about what could replace moral discourse and judgement. Note that words like 'ought,' 'should,' and 'good' can still be used after the moral framework is abolished. As long as they are applied in a prudential rather than an ethical mode, we can still use such words to offer direction and express judgements about courses of action. In an abolitionist society the claim that one ought to U rather than W will be automatically interpreted as the claim that U-ing is more in one's interests than W-ing. It will be naturally interpreted as a prudential claim, stripped of the irreducible and categorically authoritative moral properties integral to moral ascriptions.

In essence, then, abolitionists propose that we replace the moral framework with a prudential one. As Garner (2007: 507) puts it, having abolished the moral framework "we must figure out what is in our short, middle, and long-term interest, and base our decisions on that." For abolitionists, the claim that you must base your decisions on what is in your interests is conditional on your caring about what happens to you. As it happens, humans do tend to care about what happens to them. Admittedly, we are sometimes weak-willed and thus fail to act in our own interests. A developed prudential framework would therefore benefit from the introduction of certain non-moral mechanisms to help combat weakness of will. For example, a system of rewards and penalties could be used to train children to have robust dispositions to act in their long-

9 Incidentally, nor do most fictionalists. For example, Joyce thinks that fictionalism would be prudentially better for us, not that abolitionism would be a disaster.

term interests. Such a prudential education could be as effective as the moral education we dole out to children now, and mechanisms like this would help provide an alternative to Joyce's proposal that we retain the moral framework in order to combat weakness ofwill.

It might be objected that abolitionism is grist to the expressivist mill. Blackburn (1993b) has famously argued along these lines. Blackburn's objection runs roughly as follows. Suppose you conclude that error theory is true, and you abandon moral discourse and judgement. You presumably won't stop evaluating things altogether, for at best that would be incredibly difficult. When you hear reports of murder, you'll experience strong disapproval and you'll hope that the murderers get locked up. When you see a Boy Scout helping an old man cross a busy road you'll approve, and you'll wish him well. Blackburn says that what you're doing here is a form of evaluation that can be called 'shmoralising.' Luckily, there's a ready-made vocabulary for articulating your shmoral judgements. You can say that murderers are vicious and that they ought to be locked up. You can say that the Boy Scout acted rightly and that he is a good person. Blackburn says that shmoralising looks and feels very much like moralising. So perhaps you were shmoralising all along. Perhaps shmoralising just is moralising. If that's right, an examination of abolitionism suggests there was actually never any error in our moral framework. All that we were ever doing when engaging in moral discourse and judgement was expressing attitudes like approval. Abolitionism collapses into expressivism.

It is beyond the scope of this paper to provide a detailed answer to Blackburn's objection, but what abolitionists should say here is obvious. They should reject the moral phenomenology on which Blackburn bases his objection. Realists and error theorists will both deny that shmoralising looks and feels significantly like moralising, because shmoralising does not include a commitment to irreducible and categorically authoritative moral properties. Moral ascriptions feel inescapable in a way that shmoral ascriptions don't. That inescapability is essential to moral ascriptions, so anything that lacks it is not a moral ascription. Expressivism therefore fails to capture what is essential to moral discourse and judgement.

Terence Cuneo (2006: 35) argues along these lines, claiming that "expressivism is false on account of its being unable to accommodate properly the illocutionary act intentions of agents who engage in ordinary moral discourse." And Joyce (2001: Ch. 2) says that the irreducibility and categorical authority of moral ascriptions is a conceptually non-negotiable feature of moral discourse and judgement. Even Blackburn (1993b: 157) acknowledges "a nagging feeling" that there aren't really any obligations on the expressivist picture he defends (though he doesn't think that this is a problem). It is

beyond the scope of this paper to show realists and error theorists have the phenomenology right here. But let's give them the benefit of the doubt for now, and assume that abolitionism does not collapse into expressivism. What, then, are the merits of eliminating the ascription of moral properties?

Hinckfuss provides two arguments for abolitionism. First, he suggests that moral societies like ours are inherently elitist, authoritarian, and inegalitarian because they involve some people being (morally) superior to others. Hinck-fuss imagines that these people will be elevated to a higher position because they are the authorities on what to do. As moral experts, what they say goes. Hinckfuss (1987: 28) says that in moral societies children are "morally propagandised by those whom they regard as their 'betters,' that is, those who they feel know more about what is right and what is wrong than they do." Abolishing the moral framework would therefore undermine these unappealing features of moral societies. Call this 'The Elitism Argument' for abolitionism.

It is easy to reject The Elitism Argument by pointing out that accusations of elitism, authoritarianism and inegalitarianism are themselves moral accusations. On this reading, the argument is self-incriminating. However, a more charitable interpretation is that retaining the moral framework has a certain cost because it produces things to which we are generally resistant. On this charitable interpretation, the question is whether moral societies actually include the sort of unappealing features that Hinckfuss imagines. We might deny that they do. Against what Hinckfuss says about children being morally propagandised by so-called moral experts, for example, we might maintain that people actually tend to avoid deferring to moral experts unless they are in a better position with respect to relevant non-moral facts (cf. McGrath 2009). If that's the case, the only moral 'elites' deserving of deference will be individuals in a position of epistemic (rather than moral) superiority. This does not seem so worrisome.

However, there is more to The Elitism Argument than is suggested by this. Even if moral discourse and judgement do not generate problematic social hierarchies, it may be that they can help to perpetuate them. If your group is in the business of subjugating some other group, one effective way to help sustain that subjugation is to convince everyone that your group is more competent at moral judgement. (This can be underpinned by the provision of better education for the oppressive group.) Plausibly, such methods have been used throughout history to help sustain oppressive social hierarchies. Not deliberately, one assumes, but with substantial influence

nonetheless.10 Take, for example, the subjugation of women. Opponents of women's suffrage commonly argued that married women did not need the vote because they would be best represented by their husbands (cf. Shanley 1986). Men were said to be more competent at making such judgements, so wives could defer to their husband's expertise. Admittedly, deciding who to vote for is not a purely moral judgement. But it includes moral judgements, judgements about the moral character of the candidates and the moral content of their policies, for example. This provides some evidence that social hierarchies have been perpetuated by means of moral hierarchies.

So one reason to think that it is in our interests to abolish the moral framework is that it can play a role in perpetuating problematic and oppressive social hierarchies. Whilst this shows that there is more to The Elitism Argument than is often supposed, the idea that this establishes the plausibility of abolitionism is questionable. After all, the moral hierarchy indicated above is based on the false belief that men are inherently better at making moral judgements than women. Other evaluative frameworks could also be used for nefarious ends if we operate with certain false beliefs. The real problem is not with the framework but with the beliefs. In short, whilst The Elitism Argument does identify a cost to the moral framework, this cost is not integral or unique to that framework.11 I therefore wouldn't want to rest the case for abolitionism on The Elitism Argument.

Hinckfuss' second argument for abolitionism is more convincing. Hinck-fuss (1987: Ch. 4) suggests that moral discourse and judgement often get in the way of conflict resolution. Recall from §3.2 that NRW's apparently most compelling reason for retaining a fictional moral framework was that this framework plays a useful role in regulating interpersonal conflict. The abolitionist reply is simply that this turns out to be incorrect. Perhaps surprisingly, moral discourse and judgement are often a hindrance to successful conflict resolution. This is a result of the categorical authority of moral ascriptions. Although this authority might help combat weakness of will, it can also lead to opposing parties becoming entrenched in their positions, making it harder to bring conflicts to a satisfactory conclusion. A benefit of abolitionism is that it avoids this problem by eliminating moral ascriptions. Call this 'The Conflict Argument' for abolitionism.

How does the categorical authority of moral ascriptions undermine conflict resolution? Well, conflicts are frequently resolved through compromise.

10 I am grateful to Katharine Jenkins for discussion of this point.

11 Perhaps the moral framework lends itself to nefarious uses more than other frameworks, in which case The Elitism Argument might be more forceful. However, I doubt that this is the case. I am grateful to Carl Fox for this suggestion.

Each party in the conflict gives up some of what they want in order to reach a middle-ground, thereby bringing the conflict to a close. If the conflict concerns what you judge to be morally right, then compromising will involve giving up some of what you judge to be morally right in order to reach a middle-ground. But moral judgements ascribe categorical authority. They issue unconditional, inescapable demands to agents by ascribing moral properties to acts. Compromising on your moral judgements in order to reach a middle-ground will therefore involve going against what you take to be unconditionally and inescapably demanded by moral reality. And if you judge that U-ing is unconditionally and inescapably demanded by moral reality, then compromise on U-ing does not appear to be a legitimate option. As David Enoch (2011: 23) puts it, in cases of moral conflict it seems appropriate to "stand your ground." Unless you can be convinced that you're mistaken, you might understandably regard the only legitimate way out of the conflict to be your way or the doorway.

This is how the categorical authority of moral ascriptions can give rise to entrenchment in certain cases of interpersonal conflict. Rather than such cases being resolved through compromise, the parties instead end up in a sort of deadlock or stalemate as each gets entrenched in their own moral view. Moreover, this happens not merely as a result of non-moral considerations, such as dogmatism and hubris (though no doubt these can hinder conflict resolution as well). Rather, it happens as a result of the essentially categorical nature of moral ascriptions. The very nature of moral discourse and judgement can undermine our ability to resolve conflicts. Even in cases where the conflict is in principle resolvable (that is, where the disagreement is not 'fundamental') the categorical authority of moral ascriptions presents a significant practical obstacle to the successful resolution of that conflict. Since it is generally in our interests to resolve conflicts, it is a cost of the moral framework that it creates this obstacle.

So, whilst the categorical authority of moral ascriptions has the benefit of combating weakness of will, it also has the cost of giving rise to an entrenchment problem in cases of conflict. I'm not suggesting that every conflict is affected by this problem. All that the abolitionist requires is that it happens often enough or with serious enough consequences to undermine the view that we should retain the moral framework. Abolitionism's attraction, then, is that removes a significant obstacle to conflict resolution, opening the door to more effective ways of dealing with interpersonal conflicts. As Garner (2007: 504) puts it, abolishing moral discourse and judgement would free us "from that well-established framework and those tacit understandings that lock us into interminable arguments, ultimate impasses, righteousness, rhetoric, and error." Once we have eliminated the ascription of categorically

authoritative moral properties, we can find more effective ways to deal with interpersonal conflicts.

Of course, it would be foolish to imagine that abolishing moral discourse and judgement will create a paradise in which we all quickly converge on compelling answers to questions we face as a community. Entrenchment won't completely disappear just because we become abolitionists. The implication of The Conflict Argument is just that we will do better at resolving interpersonal conflicts if we operate within a prudential rather than a moral framework. In the strictly prudential framework, compromise will be easier to achieve. By removing the commitment to categorically authoritative moral ascriptions, we remove the reason to suppose that compromise is off the table. An important obstacle to conflict resolution is thus no longer present. Decisions will be based simply on prudential considerations, and if compromise is in one's interests (as will very often be the case) then compromise will clearly be the way to go.

How often does the moral framework actually hinder conflict resolution? And how problematic an effect does it really have? These are empirical questions, and we can only speculate as to their answers. Hinckfuss and Garner almost certainly overstate the extent to which moral discourse and judgement hinder conflict resolution. For example, although they stop short of claiming that there would be no war without the moral framework, Hinck-fuss (1987: 45) and Garner (2007: 507) are oddly confident about the role that the moral framework plays in the proliferation of war. I am more sceptical about this. Nevertheless, my suspicion is that the entrenchment produced by the moral framework has a more serious effect than fictionalists typically admit.

To take just one illustrative example, the United States congress nowadays frequently gets into deadlocks over important issues, including especially budgetary and fiscal issues that have economic implications at both a domestic and a global level.12 This is at least partly because many of those involved are entrenched in ideologies on government spending (among other issues) that they take to be not just economically correct, but morally authoritative. Abolitionists will suggest that this conflict would be easier to resolve within a purely prudential framework. In the abolitionist society, any fiscal disagreements that occur will be the result of disagreements over the economic facts and conflicts in economic interest. There will simply be no moral disagreement because the ascription of moral properties will have been eliminated.

12 A recent example is the so-called 'fiscal cliff.' Some consequences of this failure to reach budgetary agreement are discussed on the economics blog Econbrowser by Menzies Chinn (2012).

So the risk of ideological entrenchment is removed in the abolitionist society. Moreover, given that it is almost certainly in almost everyone's long-term interests for both sides of the US congress to negotiate a compromise, it will be obvious to abolitionists (who make decisions based on prudential considerations) that negotiating a compromise is the thing to do.

The Conflict Argument renders abolitionism plausible. This is a deliberately modest conclusion, for my discussion has been inevitably speculative and incomplete. We cannot know whether a thorough cost-benefit analysis of moral discourse and judgement would favour fictionalism or abolitionism. Nevertheless, I have shown that abolitionism has much more going for it than you might have expected. As we saw in §3.2, the inconveniences of abolishing the moral framework are not significantly worse than the inconveniences of becoming fictionalists. As we saw in §3.3.1, the abolitionist's replacement for moral discourse and judgement (namely, prudential discourse and judgement) can also supply mechanisms to combat the weakness of will traditionally dealt with by the categorical authority of moral ascriptions. Most significantly, we have seen in this section that abolitionism has the attraction of removing a significant obstacle to conflict resolution. Whilst there is no doubt that it would be very hard to abolish the moral framework, we must take seriously the view that it is in our interests to do so.

Suppose that you're not an error theorist. Suppose that you're a moral realist. You might think that this automatically licenses you to carry on with moral discourse and judgement. If Realist Metaphysics is true, moral discourse and judgement are not systematically erroneous. Business as usual, right? Not quite. As we've seen, we can identify notable costs of the moral framework. These costs don't just disappear because we accept Realist Metaphysics. In this section I will show how realists can be abolitionists, and I will argue that abolitionism complements realism surprisingly well.

The major insight of error theory is that the metaphysics of moral realism is separate from its semantic and psychological components (cf. Kahane 2013). Mackie showed that Realist Metaphysics could be false whilst Discourse, Judgement, and Ascription are nevertheless true. If that's right, we get an error theory. This separateness of the metaphysical component of realism highlights that we could eliminate the ascription of moral properties without undermining the existence of such properties. Moral properties are distinct from discourse and judgements that ascribe such properties, and moral properties would still exist even if we stopped ascribing them to acts. This opens the door to the possibility of a realist abolitionism. Indeed, if what I said in

§3.3.2 is correct it may well be that realists should be abolitionists. If The Conflict Argument renders abolitionism plausible then anyone who buys into Realist Metaphysics must take realist abolitionism seriously. It's not just a conceptual possibility, then, it's an interesting new package in the theoretical landscape.

What exactly is involved in being a realist abolitionist? Realists can recommend abolitionism without requiring that we believe or pretend to believe that there are no moral properties. The realist abolitionist says that we should stop ascribing moral properties. She doesn't say that we should adopt the false belief that there are no moral properties, or that we should engage in a fiction according to which there are no moral properties. Asking people to eliminate moral ascriptions from their discourse and judgement is not the same as asking them to believe that there are no moral properties. After all. there is a difference between aiming not to form beliefs about a domain, something that one can achieve by simply ignoring it, and aiming to form false beliefs about that domain, something that has a much more peculiar ring to it and may even be impossible.13 The realist abolitionist merely aims to ignore the moral domain.

Nor do realist abolitionists suggest that we have no epistemic access to moral properties. We can be capable of knowing certain truths without actually knowing them, and without using them in our discourse and judgement. Most of us are aware of some rudimentary scientific truths, but few of us engage seriously in scientific discourse or judgement. For instance, I get along fine without utilising the technical discourse of professional chemists, but I don't deny that this discourse refers to something real. If I were suitably inclined, I could learn something about what chemists say about chemical properties before merrily continuing to ignore this. Similarly, we can be broadly aware that there are moral properties but we need not engage with the practice of actually ascribing such properties to acts.14

13 Many think that belief has truth as a constitutive aim, though this idea is difficult to unpack. Nevertheless, the idea that there is no gap between deliberation about whether to believe that p and deliberation about whether p is true is a popular one. See, for example, Velleman (2000a; 2000b), Wedgwood (2002), Owens (2003), Shah (2003), Steglich-Peterson, (2009), and Whiting (2012).

14 Complication: motivational internalists think that moral judgements are intrinsically motivating. Realist abolitionists can defend motivational internalism, but only by requiring the complete removal of an agent's awareness of the actual moral properties. Without that, there would be a risk that these properties would motivate the agent, which would undermine the abolition of the moral framework. However, a realist abolitionist who defended motivational externalism — the denial of the internal connection between motivation and moral judgement — would not have to say this. Realist abolitionists who endorse externalism could allow some awareness of moral properties as long as the agent's motivational setup accords with prudential rather than moral considerations.

Perhaps this awareness of moral properties will be restricted to certain reflective contexts — the critical mode — with the prohibition on moral ascriptions being operative in everyday life.

The realist abolitionist might be accused of encouraging a problematic practical inconsistency. She is asking people to believe that there are things that they ought to do, whilst telling them that actually they ought not to act on those oughts. This might seem odd at first blush, but there is really nothing problematic here. This is because some of those oughts are moral and some are prudential. There are things that we morally ought to do, but there are also things that we prudentially ought to do. The realist abolitionist simply maintains that, in the long run, the prudential oughts have greater weight overall than the moral oughts. In cases of conflict, for example, it remains the case that there are morally right and wrong ways to proceed. But if engaging with that moral framework tends to lead to more costs than benefits, we have prudential reasons to eliminate moral ascriptions and proceed with a prudential framework instead.

You might object by saying that moral considerations are always overriding. If that's right, it might be psychologically feasible to ignore moral properties but in doing so we would always be irrational because moral oughts would always take precedence over prudential oughts. If that's right, perhaps realists cannot legitimately be abolitionists after all. One might try to respond to this objection by rejecting the claim that moral considerations are always overriding. The idea is a controversial one, after all. Many are sceptical of it, especially given that it seems to have the odd implication that even very weak moral considerations will always override very strong prudential considerations. However, that there are exceptions to the overridingness of morality is not enough to answer the objection. After all, if moral considerations were typically overriding then it will typically be irrational to ignore moral oughts in favour of the prudential oughts with which they conflict. So to make this response work one would have to say that categorically authoritative moral considerations are typically not overriding. That would be a difficult thing to defend, so this first response does not seem very compelling.

The second, more appealing response is to say that, even if moral considerations are overriding, there are moral reasons to abolish moral ascriptions. Plausibly, many moral properties that apply to humans will be concerned with realising what is in our interests. The moral realist claims only that moral properties do not depend on our interests for their existence and authority. They do not say that our human psychology is utterly irrelevant to judgements of what is right and wrong in particular cases. For instance, the moral badness of pain may be a property that pain has unconditionally. Yet the particular things that humans happen to find painful will be highly

relevant to the instantiation of this moral property. So if categorically authoritative moral properties are linked with what is in our interests, and if abandoning the moral framework is in our interests, then there are plausibly moral reasons to be abolitionists.15 Thus, realist abolitionism is intelligible and could be rationally defended even if moral considerations are always overriding.

Indeed, this would allow the moral realist to embrace abolitionism. Abolishing the moral framework would be the morally right course of action. This highlights one way in which realist abolitionism offers a surprisingly neat theoretical package. If realism is true, and if a thorough cost-benefit analysis of the moral framework sides with abolitionism, the most effective way to realise moral value will be to eliminate moral ascriptions. There is not space here to argue that realism is plausible, but I have already argued (in §3.3.2) that abolitionism is plausible. So we can at least say that abolitionism deserves to be taken seriously by those already committed to realism.

I'd also add that abolitionism complements realism in another way. A common objection to realism appeals to the persistence of moral disagreement. One of Mackie's (1977: 36—38) arguments for error theory draws on this phenomenon, for example. Roughly, the idea is that the persistence of moral disagreement provides evidence against the existence of moral properties. In matters about which there are genuinely objective truths, we tend to expect progress towards convergence on those truths (at least among expert investigators). This is what appears to happen in science, for example. The ongoing lack of moral convergence provides evidence against the realist claim that there are genuine moral truths waiting to be discovered. In short, the idea is that persistent moral disagreement is most compellingly explained by the non-existence of response-independent moral properties.

Note that this evidence against realism is highly defeasible, for realists can deal with many cases of moral disagreement by appealing to what John Doris and Alexandra Plakias (2008) call 'defusing explanations.' As Doris and Pla-kias use this term, these are explanations of moral disagreement that appeal to epistemic disadvantages like incomplete information and imperfect rationality. More broadly, we can understand a defusing explanation as one that explains the existence or the persistence of moral disagreement whilst being consistent with the existence of moral properties. It is incumbent upon realists to provide such defusing explanations, in order to show that there are no 'fundamental' moral disagreements, that is, disagreements that are

15 The appeal to the moral value of becoming abolitionists does not make realist abolitionism self-undermining. It does, however, make it self-effacing in the same way that some versions of consequentialism are self-effacing (cf. Parfit 1984).

irresolvable in principle. By successfully deploying defusing explanations, we need not reject Realist Metaphysics in order to account for the persistence of moral disagreement.

My suggestion is that attaching abolitionism to realism helps the realist in this matter, and that abolitionism therefore complements realism quite nicely. Abolitionists contend that lots of moral disagreements persist as a result of the very nature of moral discourse and judgement. The categorical authority involved in moral ascriptions hinders the resolution of moral conflicts. This doesn't explain why cases of moral disagreement arise, but it might explain why some of them are so persistent.16 After all, the point of The Conflict Argument is that moral discourse and judgement can produce entrenchment and deadlock in cases of moral conflict. This way of explaining the persistence of moral disagreement is consistent with the existence of moral properties. I am not suggesting that this covers all cases, but it strengthens the realist's ability to deal with a major objection to her view. It thus seems that The Conflict Argument not only shows that abolitionism must be taken seriously, it also provides moral realists with a way of defusing the persistence of certain cases of moral disagreement. So this is another way in which realism and abolitionism make for a surprisingly neat theoretical package.

This section has shown how realists can be abolitionists. Not only that, it has shown that abolitionism complements realism in at least two interesting ways. First, abolitionism may provide the most effective way to realise moral value. Second, the central argument for abolitionism reinforces the realist's ability to respond to one of the major objections raised against the existence of moral properties. So realism and abolitionism combine quite nicely. And since abolitionism is a position that must be taken seriously, realist abolitionism must be taken seriously too.17 It is not just a conceptual possibility that we can note and then quickly dismiss. It is an interesting new package in the theoretical landscape, and deserves more attention than it has so far received.

Abolitionism is more plausible than most people assume. The categorical authority integral to moral discourse and judgement can undermine our ability to resolve moral conflicts, so it may well be in our interests to eliminate

16 I am grateful to Pete Caven for discussion of this point.

17 At least to the extent that we take realism seriously, something that there is not space to go into here.

the ascription of moral properties to acts, persons, institutions, and so on. Moreover, abolitionism is independent of the moral error theory with which it is typically associated. Abolitionism can combine with metaphysically robust forms of realism, for example, and this combination shows some promise. The realist abolitionist certainly cannot deny that abolishing moral ascriptions will be practically challenging. But if this challenge is accepted, the rewards could be significant.


Ayer, A.J. 1936. Language, Truth and Logic. London: Penguin Books, 1990. Blackburn, Simon. 1984. Spreading the Word. Oxford: Oxford University Press.

- 1993a. How to Be an Ethical Anti-Realist. In Essays in Quasi-Realism. New

York: Oxford University Press, 166-181

- 1993b. Errors and the Phenomenology of Value. In Essays in Quasi-Realism.

New York: Oxford University Press, 149-165. - 1998. Ruling Passions. Oxford: Oxford University Press.

Campbell, Eric. 2014. Breakdown of Moral Judgement. Ethics 124/3, 447-480. Chinn, Menzies. 2012. The Fiscal Cliff: International Implications. Accessed 12/06/

2013. []. Cuneo, Terence. 2006. Saying what we Mean: An Argument Against Expressivism. In R. Shafer-Landau (ed.). Oxford Studies in Metaethics: Volume I. Oxford: Oxford University Press, 35-71.

Cuneo, Terence and Christy, Sean. 2011. The Myth of Moral Fictionalism. In M.

Brady (ed.). New Waves in Metaethics. Basingstoke: Palgrave MacMillan, 85-102. Doris, John M. and Plakias, Alexandra. 2008. How to Argue about Disagreement: Evaluative Diversity and Moral Realism. In W. Sinnott-Armstrong (ed.). Moral Psychology Volume 2: The Cognitive Science of Morality: Intuition and Diversity. Cambridge, MA: The MIT Press, 303-331. Enoch, David. 2011. Taking Morality Seriously: A Defence of Robust Realism. Oxford: Oxford University Press.

Fitzpatrick, William. 2008. Robust Ethical Realism, Non-Naturalism, and Normativity. In R. Shafer-Landau (ed.). Oxford Studies in Metaethics: Volume III. Oxford: Oxford University Press, 159-206. Garner, Richard T. 1990. On the Genuine Queerness of Moral Properties and Facts.

Australasian Journal of Philosophy 68/2, 137-146. - 2007. Abolishing Morality. Ethical Theory and Moral Practice 10/5, 499-513.

Gibbard, Allan. 1990. Wise Choices, Apt Feelings: A Theory of Normative Judgement. Oxford: Oxford University Press.

- 2003. Thinking How to Live. Cambridge, Massachusetts: Harvard University


Hampton, Jean. 1998. The Authority of Norms. Cambridge: Cambridge University Press. Hare, R.M. 1952. The Language of Morals. Oxford: Oxford University Press.

Hinckfuss, Ian. 1987. The Moral .Society: Its .Structure and Effects, Discussion Papers in Environmental Philosophy. Canberra: Australian National University.

Huemer, Michael. 2005. Ethical Intuitionism. Hampshire: Palgrave MacMillan.

Joyce, Richard. 2001. The Myth of Morality. Cambridge: Cambridge University Press.

- 2005. Moral Fictionalism. In M.E. Kalderon (ed.) Fictionalism in Metaphysics,

Oxford: Oxford University Press, 287-313.

- 2006. The Evolution of Morality. Cambridge, Massachusetts: MIT Press.

-2011. The Accidental Error theorist. In R. Shafer-Landau (ed). Oxford Studies in

Metaethics: Volume VI, Oxford: Oxford University Press, 153-180.

Kahane, Guy. 2013. Must Metaethical Realism make a Semantic Claim? The Journal of Moral Philosophy 10/2, 148-178.

Kalderon, Mark Eli. 2005. Moral Fictionalism. Oxford: Oxford University Press.

Lenman, James. 2008. Against Moral Fictionalism. Philosophical Books 49/1, 23-32.

Mackie, J.L. 1977. Ethics: Inventing Right and Wrong. London: Penguin.

McGrath, Sarah. 2009. The Puzzle of Pure Moral Deference. Philosophical Perspectives 23/1, 321-344.

Oddie, Graham. 2005. Value, Reality and Desire. Oxford: Oxford University Press.

Olson, Jonas. 2010. In Defence of Moral Error Theory. In M. Brady (ed.) New Waves in Metaethics, Basingstoke: Palgrave MacMillan, 62-84.

- 2011. Getting Real about Moral Fictionalism. In R. Shafer-Landau (ed.).

Oxford .Studies in Metaethics 6, Oxford: Oxford University Press, 181-204.

- 2014. Moral Error Theory: History, Critique, Defence. Oxford: Oxford University


Owens, David. 2003. Does belief have an aim? Philosophical Studies 115/3, 283-305.

Parfit, Derek. 1984. Reasons and Persons. Oxford: Oxford University Press.

Shafer-Landau, Russ. 2003. Moral Realism: A Defence. Oxford: Oxford University Press.

Shah, Nishi. 2003. How Truth Regulates Belief. The Philosophical Review 142/4, 447482.

Shanley, Mary Lyndon. 1986. Suffrage, Protective Labor Legislation, and Married Women's Property Laws in England. Signs 12/1, 62-77.

Steglich-Peterson, Asbjorn. 2009. Weighing the aim of belief. Philosophical Studies 145/3, 395-405.

Streumer, Bart. 2008. Are There Irreducibly Normative Properties? Australasian Journal of Philosophy 86/4, 537-561.

- 2011. Are Normative Properties Descriptive Properties? Philosophical Studies

154/3, 325-348.

Velleman, David. 2000a. The Possibility of Practical Reason. In The Possibility of Practical Reason, Michigan: The Scholarly Publishing Office, University of Michigan Library, 170-199.

- 2000b. On the Aim of Belief. In The Possibility of Practical Reason, Michigan:

The Scholarly Publishing Office, University of Michigan Library, 244-281.

Wedgwood, Ralph. 2002. The Aim of Belief. Philosophical Perspectives 16, 267-297.

Whiting, Daniel. 2012. Does Belief Aim (Only) at the Truth? Pacific Philosophical Quarterly 93/2, 279-300.