![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
Losing an arm to rescue a child from a burning building is supererogatory. But is losing an arm to save two children more supererogatory than losing two arms to save a single child? What factors make one act more supererogatory than another? I provide an innovative account of how to compare which of two acts is more supererogatory, and show the superiority of this account to its chief rival.
Acknowledgements
This paper was originally presented as part of the Jerome S. Simon Lectures at the University of Toronto, March 17–19, 2021. I am grateful for comments and suggestions on earlier versions of this material from James Goodrich, Sidra Goldman-Mellor, Johann Frick, D Black, members of the audience at the Simon Lectures, including Thomas Hurka and Evangeline Tsagarakis, and helpful referees for this journal.
Disclosure Statement
No potential conflict of interest was reported by the author(s).
Notes
1 Rothstein [Citation2017: 69]. My phrasing. This account suggests a widespread view that some acts are more heroic than others.
2 See McNamara [Citation2011: 216], Benn [Citation2018: 2408], Harman [Citation2016: 376, Hurka and Tsagarakis [Citationforthcoming], and Little and MacNamara [Citation2017]
3 Benn [Citation2018: 2403] is one of the few who reject this requirement. Unlike Benn, I adhere to a fine-grained view of act individuation according to which no act token would be both obligatory and supererogatory. See Smith [Citation2020].
4 Conditions 1 and 2 imply that this is a conception of objective supererogation, not subjective supererogation.
5 Others who invoke non-demanding value for this reason include Archer [Citation2016], Horgan and Timmons [Citation2010], and Little and MacNamara [Citation2017].
6 The Compact Edition of the Oxford English Dictionary does not list ‘erogatory’, but lists ‘erogation’ as an obsolete term meaning ‘the bestowal of gifts’ [274]. There is a sufficiently close connection between this meaning and the kind of value I am designating ‘erogatory’ to warrant use of the term. On deontic value, see Smith and Black [2019].
7 In a similar vein McNamara [Citation2011: 212] stipulates that an action ‘beyond the call’ must be morally better than the alternative which is ‘the least she could have done’.
This definition assumes that any supererogatory action has some permissible alternative that has the least erogatory value. Similarly, the subsequent definition of ‘personal cost’ assumes that any supererogatory action has some least costly permissible alternative. However, there may be cases involving an infinite number of alternatives in which there is no permissible alternative having the least erogatory value or the least cost. See Muñoz [Citation2020] for discussion. In other work I will explore this issue. The accounts in this paper are only intended to apply to finite-alternative cases.
8 This is a variant of an example given by Benn [Citation2018: 2404]. Benn [Citation2018] and Muñoz [Citation2021] both argue, for different reasons than mine, that an agent’s prerogatives (which are grounded in the agent’s cost) must be comparative.
9 For simplicity I do not include, among the costly acts, those that risk some comparative loss to the agent, even though she never experiences it. One could plausibly include these (as well as acts that might but do not benefit the intended beneficiary), and certainly must include them in any account of subjective supererogation. See Lazar [Citation2017] for a detailed discussion of such risks.
10 Horgan and Timmons [Citation2010: 4748] and Driver [Citation1992: 290] also argue that supererogatory acts can be performed in situations in which the agent has no duty.
11 This point was originally emphasized to me by James Goodrich and D Black (in separate personal correspondence). Hurka and Tsagarakis [Citationforthcoming] also accept it.
12 On some moral systems that include a prima facie duty to preserve one’s own well-being, Act A might be deemed wrong, and so not permissible. I shall continue to construe it as permissible rather than wrong. This will not affect this or any subsequent analyses, since an act’s violating Condition 5 and its being wrong equally disqualify it from counting as supererogatory.
13 During the presentation of this material at the Simon Lectures Thomas Hurka pointed out that Conditions 5 and 6 may be too strong. Condition 6, for example, entails that someone who makes a great sacrifice to rescue a child would not thereby perform a supererogatory act if she also had available, at no greater sacrifice, another very slightly more valuable alternative, such as rescuing the child plus saving his hat. He suggests Condition 6 should be revised to read ‘There is no permissible alternative to act X that would have non-trivially greater excess erogatory value but impose an equal or smaller opportunity cost on the agent’. Similarly a revision of Condition 5 could utilize ‘a non-trivially smaller cost’ to avoid a parallel counter-example. Such revisions may be plausible. For our purposes I will retain Conditions 5 and 6 as stated in the main text.
14 Theorists who hold praiseworthiness is a necessary feature of supererogatory acts can think of this account as identifying the features an action must have such that an agent performing that action in full knowledge of its features, and for the reason it has relevant ones of those features, would thereby perform a supererogatory and praiseworthy act.
15 I am grateful to Sidra Goldman-Mellor for pointing out the availability of this technique and assistance with this material. For an example of such work, see Song, Lin, Ward, and Fine [Citation2013]. The welfare-and-environment example is from MacAskill, Cotton-Barratt, and Ord [Citation2020: 63, n. 6].
16 MacAskill, Cotton-Barratt, and Ord [Citation2020] argue that what they call the ‘variance’ technique is superior to others for making inter-theoretic comparisons and interpersonal comparisons of utility. They note that under certain standard assumptions, their variance technique is equivalent to comparing variables by their z-scores [83], as I do here. Interestingly, they speculate that using their technique will also enable us to aggregate incommensurable goods [63, n. 6].
17 When the measures do not exhibit these features, it will be necessary to utilize Formula II (§2.3) to compare supererogatory acts whenever possible. Not all multiple-agent cases will have comparable acts.
18 In this case the sample is too small to satisfy the requirement that each variable is continuous and has a bell-shaped distribution. For the sake of illustration, I here set this issue aside.
19 More accurately, if the cost of Act X is , then X’s opportunity cost is
.
20 Note this rationale only makes sense when the moral value in question is deontic value, not erogatory value.
21 In cases such as , there is no need to actually calculate the two acts’ z-scores, since evaluating which act is more supererogatory only involves comparing opportunity costs and comparing excess values. When data points i and ii from the same population are both (for example) opportunity costs, and i is greater than ii, then necessarily i’s z-score is larger than ii’s z-score.
22 H&T’s assumptions about the sizes of the gaps differ from mine, but their assumptions are inconsistent with their assumptions about the gaps in the first two cases [H&T Citationforthcoming: Section 1].
23 Except in cases of extremely rare absolutist moral theories, which deem certain acts wrong whatever their alternatives.
24 In , the absolute cost of Act D is greater than zero. We could also construct a counterexample in which the absolute value of Act D is greater than zero.
25 James Goodrich has argued in private correspondence that the Cost+Value account has a bizarre feature, namely that on this account an act having a fixed cost for the agent would become more and more supererogatory as its value increased, until a point is reached at which the act would suddenly stop being supererogatory and become obligatory instead. If the increasing value were deontic value, which underlies obligation, then this might seem puzzling. (Although, if what makes an act obligatory is not solely its deontic value, but rather some balance between its cost and its deontic value, then it is not puzzling at all, since that balance changes as the act’s value increases but its cost does not). But the relevant increasing value in the Cost+Value account is erogatory value, which has no effect on obligation whatsoever.