Abstract
Dual-ranking act-consequentialism (DRAC) is a rather peculiar version of act-consequentialism. Unlike more traditional forms of act-consequentialism, DRAC doesn’t take the deontic status of an action to be a function of some evaluative ranking of outcomes. Rather, it takes the deontic status of an action to be a function of some non-evaluative ranking that is in turn a function of two auxiliary rankings that are evaluative. I argue that DRAC is promising in that it can accommodate certain features of commonsense morality that no single-ranking version of act-consequentialism can: supererogation, agent-centered options, and the self-other asymmetry. I also defend DRAC against three objections: (1) that its dual-ranking structure is ad hoc, (2) that it denies (putatively implausibly) that it is always permissible to make self-sacrifices that don’t make things worse for others, and (3) that it violates certain axioms of expected utility theory, viz., transitivity and independence.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Of course, since P must, at t, perform one of a1,..., a n , this set must include the act of not performing any voluntary bodily movement at t.
The symbol “\( \succ \)” means “outranks.”
See Sider (1993, pp. 126–127).
“\( {\text{o}}_i \, \succ _{\text{P}} \,{\text{o}}_j \)” stands for “o i outranks o j on the \( \succ _{\text{P}} \) ordering.”
See Sider (1993).
On why we should even bother looking for a version of maximizing act-consequentialism that accommodates such basic features of commonsense morality, see Portmore (2007). I’ll be restricting my discussion to versions of maximizing act-consequentialism. Although it’s true that satisficing act-consequentialism appears to be able to accommodate many of the same features of commonsense morality that DRAC can, satisficing act-consequentialism is, I believe, fatally flawed. Satisficing act-consequentialism implausibly permits agents to bring about a sub-optimal outcome (provided that sub-optimal outcome is good enough) merely for the sake of preventing a better outcome from obtaining. See Bradley (2006).
Here, I’m including the prevention of an evil as a benefit. So, on the self-other asymmetry, whereas it is permissible to do what will result in oneself suffering a greater harm for the sake of preventing others from suffering a lesser (net) harm, it is impermissible to do what will result in someone else suffering a greater harm for the sake of preventing others from suffering a lesser (net) harm. For more on the self-other asymmetry, see Slote (1985, chap. 1). On why consent is not at issue, see Slote (1984, pp. 190–191) and Gert (2004, p. 34).
The distinction between agent-sacrificing options and agent-favoring options comes from Slote (1985).
I call this “Schefflerian Utilitarianism,” because, like the theory that Scheffler (1994) argues for, it incorporates agent-centered options (or what Scheffler calls “agent-centered prerogatives”), but eschews agent-centered constraints (or what Scheffler calls “agent-centered restrictions”). SU does, however, differ from Scheffler’s Distributive Hybrid Theory in at least two important respects: (1) it doesn’t give priority to benefiting the worse off and (2) it incorporates not only agent-favoring options, but also agent-sacrificing options.
Unlike Ted Sider’s “Self/Other Utilitarianism,” SU permits agents not only to give their own wellbeing no weight but, alternatively, to give it extra weight.
I’ve emphasized the fourth and fifth columns with bold type since these are the two that are important in determining an act’s deontic status on SU.
U(o1) is 17, whereas U(o4) is 19.
The benefit to others is U−p (o3)−U−p (o1), which equals 5. The cost to herself is U p (o3)−U p (o1), which equals −6.
Whereas U p (o1) is 7, U p (o2) is only 2. And whereas U−p (o1) is 10, U−p (o2) is only 5.
The term ‘requiring reason’ comes from Gert (2004). Note, however, that I use the term differently. Whereas Gert uses the term to refer to reasons that are capable of generating rational requirements, I’ll use the term to refer to reasons that are capable of generating moral requirements. To avoid confusion, I might have instead used the phrase ‘moral-requirement-generating reason’, but I’ll stick with the more concise ‘requiring reason’. For more on this sort of account of supererogation, see Portmore (2007) as well as Portmore (manuscript).
To say that something is a moral reason to perform a i is to say that it is a consideration that, morally speaking, counts in favor of performing a i . See Portmore (manuscript) for a more precise definition.
A requiring reason generates a moral requirement when morally decisive. For a requiring reason to be morally decisive, it must, at a minimum, outweigh opposing requiring reasons. It may also have to meet other necessary conditions. For instance, it may have to outweigh all opposing reasons (requiring and non-requiring), which is what I argue in Sect. 3 below. For more on what it means for a reason to be morally decisive and for other possible necessary conditions for a reason’s being morally decisive, see Kagan (1989, pp. 65–70). Non-requiring reasons are reasons that are incapable of generating a moral requirement. They include, of course, all non-moral reasons, but they may also include any moral reasons that are incapable of generating a moral requirement. I want to allow for the possibility that some moral reasons are capable of making an act supererogatory even though they are incapable of making an act obligatory.
See Gert (2004, pp. 31–33).
This would explain why it is, on SU, always morally permissible to do what will be best for others overall, for, arguably, it is always morally permissible to do what one has best requiring reason to do. For more on this point, see Sect. 3 below.
U−p (o3) and U−p (o4) are 15 and 20, respectively, whereas U−p (o1) is only 10.
On commonsense morality, there is often a range of supererogatory acts, where some supererogatory acts go further beyond what’s required than others. For instance, it seems that not only is my choosing to spend my Saturdays helping the poor supererogatory, but so is my choosing to spend both my Saturdays and my Sundays helping the poor. Although both choices would seem to be supererogatory, the latter seems to go further beyond what’s required of me than the former does.
Agent-centered constraints impose limits on an agent’s freedom to permissibly pursue the best consequences, impersonally construed. They include both restrictions and special obligations. Restrictions prohibit the performance of certain act-types (e.g., murder) even to prevent numerous others from each performing that very same act-type. Special obligations arise due to past promises or through institutionally defined roles, and it is wrong to violate such obligations even if doing so will produce the best consequences, impersonally construed. See Kagan (1989, esp. p. 4).
On how maximizing act-consequentialism can incorporate agent-centered constraints, see Portmore (1998, 2001, 2003, and 2005). For now, though, here’s how we could fairly easily, although rather crudely, modify SU’s ranking principle so as to accommodate agent-centered constraints:
Ranking Principle: \( {\text{o}}_i \, \succ _{\text{P}} \,{\text{o}}_j \) if and only if both (1) U−p*(o i ) > U−p*(o j ) and (2) U+p*(o i ) ≥ U+p*(o j ), where Uadj(o i ) denotes what the total adjusted aggregate wellbeing would be were o i to obtain—adjusted both by multiplying any disutility due to harms that P intentional caused to innocents by M1 and by multiplying any disutility due to P’s non-fulfillment of a special obligation by M2, where U−p*(o i ) = Uadj(o i )−U p (o i ), where U+p*(o i ) = U−p*(o i ) + [M3 × U p (o i )], and where M1, M2, and M3 stand for distinct multipliers—to illustrate, let M1 = 1,000, M2 = 500, and M3 = 10.
An evaluative ranking is a ranking that is generated by a principle that ranks outcomes in terms of their desirability, such that one outcome, o i , outranks another, o j , if and only if there are better object-given reasons to prefer o i to o j than to prefer o j to o i . One’s reason to prefer o i to o j is an object-given reason (as opposed to a state-given reason) if and only if it is provided by facts about o i and o j (the objects of one’s preference) as opposed to facts about one’s state of preferring o i to o j . See Portmore (2007) for more on this.
All-things-considered value isn’t a kind of value in the sense that moral value and non-moral value are two kinds of value. So I’ll be using the phrase ‘all-things-considered value’ as short for ‘value, all things considered (i.e., considering both moral and non-moral value)’. And ‘moral value’ and ‘non-moral value’ are being used as technical phrases to denote, respectively, the kind of value that does, and the kind of value that doesn’t, give rise to requiring reasons. A ranking of outcomes in terms of their moral value is a ranking of outcomes in terms of their moral desirability, such that one outcome, o i , outranks another, o j , in terms of their moral value if and only if there are better (object-given) moral reasons to prefer o i to o j than to prefer o j to o i . Likewise, a ranking of outcomes in terms of their all-things-considered value is a ranking of outcomes in terms of their all-things-considered desirability.
This accounts for why I call theories like SOU “dual-ranking act-consequentialism.” Although it’s true that SOU takes the deontic status of an action to be a function of a single ranking (viz., the \( \succ _{\text{P}} \) ranking), act-consequentialism must hold that the deontic status of an action is a function of some evaluative ranking or, if not a function of some single evaluative ranking, then a function of some set of evaluative rankings. So SOU is a dual-ranking theory in that it makes the deontic status of an action a function of not just a single evaluative ranking but of dual evaluative rankings.
This is not to deny that agents are sometimes morally required to do what promotes their own wellbeing. Nor is this to deny that there are duties to self. The claim here denies only that agents have such duties in virtue of the effects that abiding by them will have on their own wellbeing. For more on this issue, see Portmore (2003).
See Portmore (2003, manuscript).
If it weren’t for the self-sacrifice involved in performing a a , the agent would be morally required to perform a a .
See Unger (1996).
The phrase ‘purely justificatory reason’ also comes from Gert (2004). But, again, I use the phrase differently. Whereas Gert uses the phrase to denote reasons that have the power to rationally justify but not the power to rationally require, I use the phrase to denote reasons that have the power to morally justify but not the power to morally require.
For a less compressed version of this argument, see Portmore (manuscript).
See Howard-Snyder (1994).
I use the phrase ‘no worse reason’ as opposed to ‘at least as good a reason’ to allow for the possibility that the relevant reasons may be incommensurable such that there is no fact as to whether the reasons to perform a i are better than, worse than, or at least as good as the reasons to perform some other available alternative.
In line with what I’ve said about supererogation in the previous section, I could add the following sufficient condition for an act’s being supererogatory:
$$ \begin{aligned}{} {\bf MS}\quad & {\text{P's performing a}}_i {\text{ is }}morally{\text{ }}supererogatory{\text{ if there exists an act a}}_j {\text{ such that (1) P is both }} \\ & {\text{morally permitted to perform a}}_i {\text{ and morally permitted to perform a}}_j {\text{, (2) P has better }} \\ & {\text{requiring reason, on balance, to perform ai than to perform a}}_j {\text{, and (3) P's performing a}}_i {\text{ is }} \\ & {\text{more morally praiseworthy than P's performing a}}_j {\text{ in that P's performing a}}_i {\text{ involves }} \\ & {\text{greater self-sacrifice}} \\ \end{aligned} $$See Stroud (1998).
Strictly speaking, there could be more than two options if there were more than one act that was tied for first place either in terms of what there is best requiring reason to do or in terms of what there is best reason to do, all things considered. In Portmore 2003, I defended a view where these were the only two options. I now think that that was a mistake, as I explain below.
This is what SOU implies.
This example is inspired by one of Kagan’s (1989, pp. 16, 240)
To say that M p (o i ) > M p (o j ) is to say that P has better (object-given) moral reasons to prefer o i to o j than to prefer o j to o i . Such a claim is consistent with someone else, say, Q, having better (object-given) moral reasons to prefer o j to o i than to prefer o i to o j . Thus it could be that M q (o j ) > M q (o i ). This is why we have to talk about moral value relative to P.
Remember that I picked the number “10” only for the purposes of illustration. The actual multiplier might be much greater.
See Portmore (2007).
Let me just stipulate that, as I’ll use the phrase ‘higher normative status’, a permissible act always has a higher normative status than an impermissible one.
I thank Ted Sider for pointing out to me the need to address the transitivity and independence axioms.
As Peter Vallentyne points out, “There is an important difference between (1) directing agents to maximize some ranking relation, that the theory happens to hold to be the basis of goodness, and (2) directing agents to maximize goodness, however that may be determined. A theory must be of the latter type in order to be teleological [i.e., consequentialist in my parlance]. Being teleological is like directing one’s son to marry the richest woman in town—whoever—that may be. For any particular woman, one’s judgments about whether he should marry her depend on one’s judgment about how rich she is. Being teleological is not, on the other hand, like directing one’s son to marry Ms. Jones, whom one happens to believe to be the richest woman in town. One’s judgments about whether he should marry any particular woman do not in that case depend on one’s judgment about how rich she is” (1988, pp. 253–263).
Along with other consequentialist theories, DRAC will also have the advantage of embodying the compelling idea that that it always morally permissible for an agent to act so as to effect the outcome that, of all those available to her, she has best reason to want to be actual. See Portmore (2007).
For a defense of this claim, see Portmore (2007).
I thank Ted Sider and an anonymous referee from this journal for their helpful comments on an earlier draft of this paper.
References
Bradley, B. (2006). Against satisficing consequentialism. Utilitas, 18, 97–108.
Gert, J. (2004). Brute rationality. Cambridge: Cambridge University Press.
Howard-Snyder, F. (1994). The heart of consequentialism. Philosophical Studies, 76, 107–129.
Kagan, S. (1989). The limits of morality. Oxford: Clarendon Press.
Portmore, D. W. (1998). Can consequentialism be reconciled with our common-sense moral intuitions? Philosophical Studies, 91, 1–19.
Portmore, D. W. (2001). Can an act-consequentialist theory be agent-relative? American Philosophical Quarterly, 38, 363–377.
Portmore, D. W. (2003). Position-relative consequentialism, agent-centered options, and supererogation. Ethics, 113, 303–332.
Portmore, D. W. (2005). Combining teleological ethics with evaluator relativism: a promising result. The Pacific Philosophical Quarterly, 86, 95–113.
Portmore, D.W. (2007). Consequentializing moral theories. The Pacific Philosophical Quarterly, 88, 39–73.
Portmore, D.W. (manuscript). Are moral reasons morally overriding? available at <http://www.public.asu.edu/∼dportmor/Are_Moral_Reasons_Morally_Overriding.pdf>.
Scheffler, S. (1994). The rejection of consequentialism (Revised ed.). Oxford: Oxford University Press.
Sider, T. (1993) Asymmetry and Self-Sacrifice. Philosophical Studies, 70, 117–132.
Sinnott-Armstrong, W. (2003). Consequentialism. In E. N. Zalta (Ed.), The stanford encyclopedia of philosophy (Summer 2003 ed.). URL = <http://plato.stanford.edu/archives/sum2003/entries/consequentialism/>.
Slote, M. (1984). Satisficing Consequentialism. Proceedings of the Aristotelian Society 58, 139–163.
Slote, M. (1985). Common-sense morality and consequentialism. London: Routledge and Kegan Paul.
Splawn, C. (2001). The self-other asymmetry and act utilitarianism. Utilitas, 13, 323–333.
Stroud, S. (1998). Moral overridingness and moral theory. The Pacific Philosophical Quarterly, 79, 170–189.
Unger, P. (1996). Living high and letting die. Oxford: Oxford University Press.
Vallentyne, P. (1988). Gimmicky representation of moral theories. Metaphilosophy, 19, 253–263.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Portmore, D.W. Dual-ranking act-consequentialism. Philos Stud 138, 409–427 (2008). https://doi.org/10.1007/s11098-006-9059-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11098-006-9059-7