http://orcid.org/0000-0003-0957-5698
![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
ABSTRACT
The moral error theorist claims that moral discourse is irredeemably in error because it is committed to the existence of properties that do not exist. A common response has been to postulate ‘companions in guilt’—forms of discourse that seem safe from error despite sharing the putatively problematic features of moral discourse. The most developed instance of this pairs moral discourse with epistemic discourse. In this paper, I present a new, prudential, companions-in-guilt argument and argue for its superiority over the epistemic alternative.
Notes
1 I sometimes, for brevity, omit the ‘moral’ in ‘moral error theory’. I similarly suppress the ‘(or facts or relations)’ and formulate error theory (etc.) in terms of properties.
2 The locus classicus is Mackie [Citation1977: ch. 1]. See also Garner [Citation1990], Joyce [Citation2001], Bedke [Citation2010], Olson [Citation2014], and Streumer [Citation2017].
3 When deployed by opponents of the error theory, they are more accurately labelled ‘companions in innocence’, but I will follow the existing convention.
4 For assessment of a range of such arguments, see Lillehammer [Citation2007].
5 ‘Relevantly’ is crucial. There are many cases where your desire is part of the explanation of why you have a reason, without making this reason non-categorical. If you desire to hurt others, then you have a categorical reason to seek medical help. This reason is relevantly independent of your cares and concerns because it does not prescribe a means of realizing your aims.
6 For recent discussion, see Finlay [Citation2008] and Joyce [Citation2011]. See also Foot [Citation1972].
7 This is not to deny that one could combine error theory with the claim that all normative properties are identical, or reduce, to categorical reasons for action.
8 See, especially, Cuneo [Citation2007: ch. 4].
9 For discussion of a range of proposals about how the moral and epistemic might be interrelated, see Cuneo [Citation2007: ch. 2]. One possibility would involve there being pragmatic encroachment on the epistemic, because epistemic properties are appropriately related to moral reasons for action. On this, see Fantl and McGrath [Citation2007] and Hawthorne and Stanley [Citation2008].
10 There are other ways to reject epistemic error theory, such as by rejecting realism about epistemic discourse.
11 There are two versions of this response. The first is that epistemic discourse ascribes normative properties but that the discourse is in good standing because it ascribes reductively normative properties. The alternative is to hold that epistemic discourse is in good standing because the properties that it ascribes are non-normative. The difference between these two ways of replying does not matter for my purposes here, because either is sufficient to falsify the parity premise.
12 Perhaps Heathwood loads the dice by using ‘suffering’ rather than ‘pain’. As noted above, there are other possible objections to the epistemic argument. For example, one might be sceptical that a reductive account of probabilities is possible.
13 For discussion, see Cowie [Citation2016: sec. 4].
14 Someone attracted to (e.g.) a deontological conception of epistemic justification might deny that truth plays the kind of role that is built upon by the reductionist view that I have been considering.
15 However, as Cowie makes clear [Citation2016: 129; my emphasis], it's only a master argument against the epistemic argument: ‘I have argued that the companions-in-guilt argument—at least if it is understood as an argument by analogy with epistemic reasons—fails.’ See also Rowland [Citation2016] and Das [Citation2017].
16 At least some of these sentences can be used to make moral (and other non-prudential) claims as well.
17 We do not talk about prudential permissions and obligations. But we do talk of prudential good and bad, prudential reasons, prudential necessities, what someone prudentially should/must/ought to do. For discussion of the normativity of prudential discourse, see Fletcher [in preparation].
18 I am not taking a stand on whether all pains are prudentially bad. Perhaps pains that are relished are not. On this, see Bradford [in preparation].
19 There are plausible candidates for normative properties that prudential discourse might be reducible to. For example, see Darwall's [Citation2002] claim that prudential value is what it would be rational to want in so far as you care about someone.
20 These points are not decisive against the proposals. Rather, it is some evidence against the reduction, and provides the task for a reductionist account of explaining why it seems like there can be (e.g.) prudential discourse that does not ascribe the relevant properties.
21 See also Hare [Citation1952: sec. 9.4].
22 There is more to say about (i) how precisely we determine that these are prudential judgments, and (ii) what kinds of meta-prudential views we should seek to endorse on the basis of these considerations. These are interesting issues that I lack the space to pursue here.
23 I think that they are equally plausible, but I concede that they might only be of roughly equal plausibility. For readability, I will put the point in terms of equal plausibility.
24 Granted, the epistemic argument promises something that the prudential equivalent cannot—the possibility of showing that the moral error theorist cannot coherently argue for their view (although whether the moral error theorist would be so hamstrung if committed to the epistemic error theory is a vexed question). See Olson [Citation2011], Rowland [Citation2013, Citation2016], Cowie [Citation2014a, Citation2014b, Citation2016], and Das [Citation2017].
25 There are many other controversial entailments of each view: for example, on the epistemic side, that no one could know anything, that no belief-forming mechanism is epistemically superior to any other, and, on the prudential side, that no one could live well, that there are no rational, prudentially grounded, preferences.
26 This reduction, if possible, would insulate the moral error theorist from the charge of being unable to argue consistently for their view.
27 They would be free to combine this extended error theory with a reductive view of epistemic normativity of the sort discussed above.
28 For valuable comments and discussion, I am grateful to Debbie Roberts, Mike Ridge, Chris Cowie, and referees and editors of this journal.
