Essays in Moral Skepticism brings together twelve essays by Richard Joyce. The collection is divided into three parts, each corresponding to the three philosophical themes for which Joyce is best known: moral error theory, the evolutionary debunking of morality, and moral fictionalism. The terrain covered is impressive, ranging from empirical issues in moral psychology to conceptual matters and technical questions in the philosophy of mind and language. But far from giving the appearance of a Monty Python sketch (‘And Now for Something Completely Different!’), the collection's essays hang together in a mutually informative and complementary way. Although only one of the essays is new (‘Evolution, Truth-Tracking, and Moral Skepticism’), the reader benefits from their combination. Joyce's purpose-made introduction is additionally helpful in this regard, providing the reader with something of a roadmap for the philosophical journey ahead.
I am unable to afford all of Joyce's essays their proper philosophical due here. In the interests of space, I'll devote my critical energies to one essay from each part, offering small glimpses into some of the others along the way.
Part I's essays showcase different angles of Joyce's argument for moral error theory. Error theorists are typically cognitivists; they take moral beliefs and assertions to be aiming at truth. But they deny that moral beliefs and assertions ever are true. (Or, to put matters more carefully, error theorists deny the truth of all moral claims within an important class; for example, positive, atomic, first-order, non-tautological ones.) In their view, the world is just not the right way for moral terms to refer, or for moral predicates to be literally satisfied. There are, then, no moral facts; no facts about what is morally right and wrong, or about what we morally ought or ought not to do.
‘Expressivism, Motivational Internalism, and Hume’ defends a hybrid variety of cognitivism, which carves out a place for a conative element in moral judgements. ‘The Accidental Error Theorist’ raises problems for some of the error theorist's metaethical rivals – most notably, dispositional analyses of moral properties – in an argumentative manner that wears its chutzpah on its sleeve. According to Joyce, dispositional theories often fail to build sufficient structure into their analysis of moral properties. The result, ‘moral goodness’ (or ‘rightness’, etc.), fails to refer, and dispositionalists turn out to be budding (though of course unwitting) moral error theorists.
‘Morality Schmorality’ homes in on the debate between error theorists and success theorists, who maintain that moral terms succeed in referring. Far from vindicating moral discourse, Joyce takes success theorists to merely vindicate a schmorality: ‘something bearing a resemblance to a morality … but which falls short of really being so’ (p. 56). These themes are expanded upon in ‘Metaethical Pluralism: How both Moral Naturalism and Moral Skepticism may be Permissible Positions’. Here, one finds a less virulent strain of moral error theory than one is used to seeing defended by Joyce. Error theorists usually construct their case by identifying what they take to be core conceptual commitments of moral discourse. These commitments help to build a kind of a job description; they tell us which kinds of objects, properties, or relations there must be in order for our moral terms to refer. Joyce is well-known for denying that anything satisfies this job description – for him, there is nothing in the world that answers to our ordinary concept of (e.g.) moral rightness. In ‘Metaethical Pluralism’, however, he is surprisingly hospitable to the suggestion that something may come sufficiently close. Here, Joyce affords serious consideration to Lewis's proposal that there are imperfect (though still deserving) deservers of our moral terms. When confronted with the choice between accepting that nothing is morally right (that is, moral error theory) and accepting that properties like moral rightness are not quite as we thought, Lewis (‘Dispositional Theories of Value II’, The Proceedings of the Aristotelian Society 63 (1989), pp. 113–37, at 137) favours the latter, ‘calm and conservative respons[e]’ – though he suggests that such judgement calls may ultimately come down to ‘a matter of temperament’. Joyce argues that there is no fact of the matter as to whether success theorists (e.g. moral naturalists like Lewis) or error theorists are correct; each has strong arguments on their side, and each arrives at their position via permissible albeit non-mandatory exercises in conceptual precisification. In borderline cases such as these, Joyce suggests that we follow Lewis in thinking that our philosophical judgement calls come down to a matter of temperament.
Yet it's not clear to me that we ought to take Lewis's talk of ‘temperament’ entirely at face value (especially given the characteristically whimsical nature of his prose). One suspects that what Lewis really had in mind here was philosophical temperament – that is to say, one's background philosophical methodology and principles. A general part of Lewis's own methodology is to proceed on the assumption that our repertoire of everyday (or ‘folk’) beliefs (e.g. about mental states or moral values) are more or less correct. Paired with this methodology is a ‘maxim of honesty’: ‘never put forward a … theory that you yourself cannot believe in your least philosophical and most commonsensical moments’ (On the Plurality of Worlds (Oxford, 1986), p. 135). Once we shift our attention to Lewis's more general philosophical framework, then, it is far less clear that the conservative response really is as optional for him as Joyce makes it out to be; for the conservative response seems to cohere far better with Lewis's framework than a response which denies a large swathe of the folk's first-order moral judgements. As far as this debate is concerned, then, there may very well be a fact of the matter regarding which side a particular philosopher with a particular background theory ought to take. This isn't a million miles away from the conclusion that Joyce himself reaches, but it has the virtue of emphasizing that even if these are ultimately judgement calls, they may be very principled ones.
Whereas Part I focuses upon the truth of our moral beliefs, Part II raises issues pertaining to their justification. ‘The Origins of Moral Judgment’ carves out an empirical hypothesis concerning the evolutionary origins of our moral beliefs. ‘Evolution, Truth-Tracking, and Moral Skepticism’ exploits that hypothesis in service of moral scepticism. (Arguments with this general structure often go under the name of ‘evolutionary debunking arguments’.) According to Joyce, there is an empirically respectable story to be told according to which our capacity to form moral beliefs is a biological adaptation. But the adaptive utility of our moral beliefs is one thing; their truth is quite another. Since we can supply a plausible evolutionary story of our moral belief-forming mechanisms which appeals only to their adaptive utility and nowhere to their truth, we should be less confident that these belief-forming mechanisms are truth-tracking, and thus, that our moral beliefs are indeed true (p. 152). Following exposure to this sceptical argument, we should take our moral beliefs to be unjustified – at least provisionally.
Joyce is keen to distinguish his argument from its close relatives. Some debunkers have fairly modest ambitions, aiming to show that our moral beliefs are unjustified if we assume a robustly realist understanding of moral facts – that is (roughly), if we assume that our moral claims have objective or mind-independent truth-conditions. This strain of debunking does not lead to wholesale moral scepticism; it is only if we assume realism that scepticism looms. Joyce, by contrast, takes his argument to have epistemological bite whether or not moral truths are construed realistically. He offers the example of Fred, who forms beliefs about the value of pieces of a foreign currency by consulting tea leaves. The truth-conditions for currency-value-claims are clearly intersubjective (being determined via human convention) rather than objective. Yet Fred's beliefs still seem unjustified in light of the non-truth-tracking processes that give rise to them. As is the case with objective facts, then, ‘subjectivist/constructivist facts’ are the sorts of things that our belief-forming mechanisms may be more or less adept at tracking (p. 146).
The foreign currency example serves as a nice analogue for intersubjectivist understandings of moral facts, according to which the truth of moral claims turns on the verdicts of some (actual or hypothetical) set of human reasoners. But it's not clear that the lesson applies in equal measure to subjectivist understandings. If moral truths are determined by each individual's subjective states (such that ‘φ-ing is wrong’ is true just in case subject S believes that φ-ing is wrong), then it's difficult to see why considerations of the kind that Joyce raises should lead us to think that we're grossly ineffective at tracking them – at least without any reason to think that we're ineffective at tracking our own beliefs more generally.
Joyce is unlikely to be moved by this suggestion. He is at pains to emphasize that it won't do merely to point towards some conception of moral facts that facilitates an easy escape from scepticism. That conception must also be plausible in its own right (p. 155). Subjectivism might turn out to be true, and moral facts, so construed, might find a place in the debunker's evolutionary story. But reinstating ‘epistemic justification requires more than a vague “might” ’ (p. 158). One must explain how these facts earn a place in the evolutionary story, and motivate the metaethical plausibility of subjectivism. Joyce is likely to find himself in good company if he insists that the latter is a tall order. But what of naturalist varieties of moral realism, which take moral facts to be identical to (objective) natural facts that it would be in one's reproductive interests to track – facts about what enhances co-operation, say? Such views carry the promise of establishing a link between fitness-tracking and truth-tracking in the moral domain, and they seem more promising than subjectivism. In response, Joyce seems content to appeal to the ‘general metaethical arguments against moral naturalism’ (p. 154). Yet one can't help but wonder just how strong such arguments are likely to be – especially given the philosophical ambivalence that Joyce espouses in ‘Metaethical Pluralism’. If the choice between moral error theory and moral naturalism really does come down to a matter of temperament (as Joyce appears to think), then it seems brash to dismiss the latter swiftly as metaethically implausible here.
The organizing themes of Part III are projectivism and fictionalism. ‘Patterns of Objectification’ explores the relationship between Mackie's moral error theory and his ‘objectification thesis’, according to which the apparent objectivity of moral facts arises from our tendency to objectify the moral attitudes (e.g. anger, sympathy) that we project onto the world. ‘Moral Fictionalism’ shifts the focus to life after moral error theory. Here, Joyce argues that moral error theorists ought to preserve moral discourse in the spirit of a useful fiction. More specifically, he recommends that we adopt attitudes of make-believe (rather than belief) towards moral propositions, and pretend to assert (rather than assert) them. The motivations for fictionalism are twofold. The first is epistemic; unlike beliefs, fictionalist attitudes aren't ontologically committing, and so, being a moral fictionalist need not involve believing any error-ridden moral propositions. The second motivation is practical; moral thought is said to provide us with a number of desirable practical goods. In particular, Joyce argues that moral thinking serves as a useful bulwark against weakness of will; we are more likely to resist temptation and act in accordance with our (long-term) practical interests when we enlist moral concepts in our deliberations.
Joyce distinguishes his brand of moral fictionalism from the ‘prefix’ variety. The latter recommends adopting attitudes of belief towards moral propositions that are prefixed by a(n explicit or implicit) story operator. (For example, ‘according to the moral fiction, it is wrong to steal’.) Joyce offers a number of considerations against prefix moral fictionalism. Among these is the complaint that prefix fictionalism is unable to accommodate the validity of moral arguments such as:
P1. If my cousin is an infant, then it is wrong to kill my cousin.
P2. My cousin is an infant.
C. Therefore, it is wrong to kill my cousin.
(Joyce frames this part of his discussion in terms of colour fictionalism. But since he clearly takes the problem to generalize, I reframe things in terms of moral fictionalism here.) According to Joyce, the prefix fictionalist faces a dilemma. On the one hand, she might take only sentences containing moral terminology (P1 and C) to be prefixed by a fictional operator (‘according to the moral fiction …’). The problem with this move is that the argument is no longer valid. Alternatively, she might prefix all three claims with a fictional operator. This maintains validity, but Joyce complains that it is utterly unmotivated; why think that the moral fiction will include any claims about the age of someone's cousin? It seems to me that the prefix moral fictionalist should simply take the second horn. It is, after all, commonplace that some things that are true in fictions are true in real life as well. We very often import truths about the actual world into fictional stories. We typically hold fixed, for example, logical and mathematical truths, the laws of physics, and perhaps even generalizations about human psychology. (Fictions are also governed by rules of export, of course; people often learn facts about the past from great works of literature.) It doesn't seem to be particularly problematic for the prefix fictionalist, then, if some things that are true in the moral fiction – the age of someone's cousin, say – are true in real life as well.
As was noted at the outset, the terrain covered in Essays in Moral Skepticism is impressive. There really is something for just about everyone with an interest in metaethics. The collection would be of particular interest to those seeking to reject a sceptical metaethical package. Here, they will find a systematic and well-supported one to reckon with.
