Download The Concept of Amoral Actions and Its Implications for Moral Obligation and more Study notes Philosophy in PDF only on Docsity!
Amorality
ABSTRACT: Actions are usually grouped into one of several moral cat- egories. Familiar ones include the morally required, the morally permitted, and the morally prohibited. These categories have been expanded and/or refined to include the supererogatory and the “suberogatory”. Some eschew deontic categories such as the above, but nevertheless allow the existence of two comparative moral categories, i.e., the morally better or morally worse than some contextually specified baseline. At the risk of adding to the clut- ter, I want to explore the possibility of yet a further category, viz., amoral action, or actions that, strictly speaking, lack any moral status whatsoever. I’m interested in examining this category in part because I believe that it is plausible to say that such actions exist. In addition, I think that admitting the existence of amoral actions has additional consequences for our under- standing of the nature of moral obligation, in particular the semantics of the moral ‘ought’.
Actions are usually grouped into one of several moral categories. Familiar ones include the morally required (actions that, morally speaking, one must perform), the morally permitted (actions that, morally speaking, one can, but needn’t, perform), and the morally prohibited (i.e., immoral actions, actions that are not compatible with conformity to one’s moral require- ments). These categories have been expanded and/or refined to include the supererogatory (a subset of permissible actions that are morally superior to other permissible actions but that are not themselves required) and the “suberogatory” (a subset of the permissible that are morally “bad to do, but not forbidden”^1 ). Others have eschewed deontic categories such as the above, but nevertheless allow the existence of comparative moral categories, i.e., the morally better or morally worse than some contextually specified baseline.^2 At the risk of adding to the clutter, I want to explore the possibility of yet a further category, viz., amoral action, or actions that, strictly speaking, lack any moral status whatsoever. I’m interested in examining this category in part because I believe that it is plausible to say that such actions exist. In addition, I think that admitting the possibility of the existence of amoral actions has additional consequences for our understanding of the nature of
(^1) Julia Driver, “The Suberogatory” in Australasian Journal of Philosophy 70 (1992). (^2) Alastair Norcross, “Reasons without Demands: Rethinking Rightness” in Contempo- rary Debates in Moral Theory, ed. Dreier (Oxford: Blackwell, 2006).
moral obligation, in particular the semantics of the moral ‘ought’.^3 The plan of the paper runs as follows. In §1 I distinguish the sense of amorality with which I work from other uses of that term that sometimes crop up in moral theory. §2 offers a prima facie case for the existence of amoral actions. §3 poses an important problem for the possibility of amoral- ity. Given the semantics of moral requirement and its relationship to moral prohibition, it might appear that there can be no actions that entirely lack a moral status or valence. §4 discusses a variety of potential responses to this puzzle and offers my own preferred solution. §5 responds to an objection to that solution, and §6 concludes.
- Amorality: A Primer
In discussing the notion of amorality, it is important to make clear the precise concept I wish to explore. Importantly, “amoral” is used in many different ways, some of which are irrelevant for present purposes. Chief among these is the concept of an “amoral agent” or “amoralist”.^4 Familiar from discussions of moral motivational internalism, the amoralist is a char- acter who supposedly makes moral judgments but who feels no motivation to conform to the moral judgments she makes. I do not wish to discuss the amoralist here; the term “amoral” as I use it is intended to designate a category of actions rather than individuals. Furthermore, “amoral” might also be used to describe the choice between two actions that is, for moral purposes, trivial or neutral. For instance, if one is faced with the choice of saving one of two strangers from immediate drowning, each of these strangers is equally deserving of being saved, they will have equally good lives after having been saved, etc., one might describe the choice between saving one versus the other as “amoral”, in other words, of no moral consequence. Though certainly related to the concept I’m inter- ested in exploring, it is not the same issue. I’m interested in acts, not choices
(^3) Throughout this paper, I will treat the notion of moral requirement and the moral ‘ought’ as interchangeable—to say that one morally ought to φ just is to say that one is morally required to φ. Some have challenged this, holding that ‘ought’ instead suggests a form of recommendation, rather than requirement. See, most importantly, Paul McNa- mara, “Must I Do What I Ought (or Will the Least I Can Do)?” in Deontic Logic, Agency and Normative Systems (Springer, 1996). However, nothing rides on this for present pur- poses. If one prefers McNamara’s reading of the notion of the moral ‘ought’, he or she is welcome to simply substitute “requirement”; nothing in the paper or the significance of its argument will change. (^4) Cf. David Brink, Moral Realism and the Foundations of Ethics (Cambridge: Cam- bridge University Press, 1989), ch. 3.
here without further specification; the cases I offer should fit this category straightforwardly.
- Are There Amoral Actions?
At this point I’ve clarified the concept I wish to explore. It seems per- fectly respectable qua concept. But why should we care about it? Why is it worth exploration? Why isn’t an investigation into the concept of amorality like an investigation into chmess, or some similarly cockamamie category? Indeed, given the conceptual analysis of amorality, it may seem extraor- dinarily implausible to think that such actions exist. How could it be that there is some particular action open for an agent at a time, but which lacks moral valence? This action must not simply fail to be morally required or morally wrong, it must fail to be morally permissible and supererogatory, be morally better or morally worse. How could a real action (however one wishes to understand this notion), not simply a reflex or simple bodily move- ment, that we can perform fail to be morally evaluable or fall under one of the recognized moral categories? However, this possibility should not be entirely mysterious. For some normative domains, evaluation is not pervasive. Take etiquette. Etiquette simply has nothing to say about some actions. Consider my decision while at a coffee shop to order coffee with cream. This action is neither rude nor polite; it neither conforms to protocol nor is a violation. One might say that doing so is, perhaps, more polite than some baseline. Imagine the action of flipping my barista the bird. Certainly ordering my coffee with cream is more polite than flashing my server an obscene gesture. But this example seems to mistake two separable actions, both of which I perform when I order my coffee with cream: the action of ordering my coffee with cream, and the action of refraining from flipping the bird.^7 Certainly the
(^7) Note: this relies on a fine-grained conception of action-individuation. See Goldman, chs. 1 and 2. Though I prefer it for ease of exposition, one needn’t accept this account to make my point. Even if one accepts a coarse-grained theory, normative valence will apply to actions under particular descriptions. So even if the action of refraining from flipping the bird and ordering coffee with cream is the same, it maintains an etiquette valence under the former description, but certainly lacks one under the latter description. Etiquette, therefore, is not pervasive. I will accept a fine-grained view here, but my arguments can be translated mutatis mutandis. One further point here. You might think that a “fine-grained” theory of action- individuation (or the claim that moral valence applies to actions only under descriptions) renders the argument for amoral actions too easy: if, for instance, I perform a myriad of actions at any time at which I perform some action (i.e., if, whenever I steal Susie’s candy,
latter action is more polite than the alternative (i.e., flipping the bird). But there’s nothing about the action of ordering my coffee with cream that, in itself, is more or less polite than anything else I might do. Etiquette doesn’t say anything about it. Seen in this light, however, it is much less clear that amoral actions cannot exist. Everything hinges on whether or not morality’s evaluation in- cludes all potential actions in a way that etiquette’s does not. Some believe it does. Scheffler, for instance, writes: “all human conduct is in principle morally assessable.”^8 If this is correct, there is no category of amorality. But I think there are reasons to second-guess Scheffler’s confidence. Two arguments follow, focusing on roughly the same problem.
2.1. Moral Requirability^9
I also perform the separate action of moving my arms in a particular way, of omitting to buy milk at the store, etc.), surely some of them (i.e., moving my limbs in such-and-such a way) will not be morally evaluable. And hence, or so it would seem, it is utterly trivial that there should be amoral actions. However, even if my argument here is committed to a fine-grained theory of action-individuation, this does not entail that the existence of amoral actions is uninteresting or obvious. Take, for instance, my stealing of Susie’s candy. Surely that action is morally evaluable. But it seems unwarranted to me to say that moving my limbs in such-and-such a way is not morally evaluable without much more information. Indeed, depending on the circumstances, it seems perfectly acceptable to de- scribe that action—considered in and of itself—as permissible. However, if circumstances are otherwise, it could also have other moral valences: moving my limbs in such-and-such a way, after all, is the sort of thing that could generate good or bad consequences (espe- cially in this case), to name but one significant moral factor. It thus seems to me not so implausible to believe that doing so could even be required. Furthermore, it certainly does not lack any of the relevant disqualifying factors (discussed below) that paradigmatically un-requirable actions lack. Of course, this is just one example, but the general point holds: the existence of amoral actions is not simply guaranteed or a trivial matter given my as- sumptions here. However, even if I’m wrong about this, even if the existence of amoral actions is trivial, their existence has important upshots when it comes to the semantics of moral requirement, upshots I explore beginning in §3. (^8) Samuel Scheffler, Human Morality (Oxford: Oxford University Press, 1992), 25. (^9) Classically, many have thought that the concept of an excuse removes the action so excused from moral consideration. Excuses, it is said, are different than justifications: while justifications serve to show that the action in question, though it may have had some bad qualities, is nevertheless justified, an excuse does not serve to justify the action in question. (See, classically, J. L. Austin, “A Plea for Excuses” Proceedings of the Aris- totelian Society 57 (1956-7), 2-3.) As Austin writes, in offering justifications for an action, “we accept responsibility but deny that it was bad”; in offering excuses, “we admit that it was bad but don’t accept full, or even any, responsibility, ” (Austin, 2). However, though I’m compelled to believe that some actions for which excuses are available are those that lack moral valence, further argument is required to show that actions for which there is a legitimate excuse possess no moral valence at all. The most
know what the right characters are. But it would seem sensible to believe that if one of the actions eligible to form the object of a moral requirement for me includes curing cancer, a moral theory would have to be crazy not to select that action as the one that, in fact, is required. But of course this is absurd. There are plenty of actions that we can perform but that we almost certainly won’t perform for reasons that render a moral requirement of that action implausible. For instance, it could be that opening a safe to which one does not know the combination produces an astoundingly good set of results. Or it could be that one has the power to rescue one’s friends from calamity if only one can beat their captor—who happens to be a Grand Master—at chess. Surely one can beat the Grand Master; all one has to do is make the right moves in the right order. But while one certainly can beat the Grand Master, it’s just overwhelmingly unlikely that one won’t.^13 Virtually everyone agrees, then, that actions that one can perform merely in this maximally expansive sense of ‘can’ cannot be eligible for moral re- quirement. So much, it seems to me, is simply platitudinous. What is required, however, is an account of what is missing in such actions that renders them ineligible for moral requirement. Different theorists have di- agnosed this problem in different ways, and have searched for alternative accounts of actions that are eligible to be morally required or the object of a moral ‘ought’. Ecumenically, we can at least say that any such action will maintain a set of “disqualifying factors”. These factors will typically include relations borne between a particular agent and the action in question, and might include, for instance, the fact that the relevant agent has positively no clue how to perform the action, or is unable to perform the action as a result of reasoned deliberation, etc.^14 The project of identifying precisely the disqualifying factors is, of course, controversial. But whatever these factors are, they must explain the relevant datum, viz., that most people cannot be morally required to cure cancer, despite the fact that they can do so. But however this dispute is resolved, the central point (for my purposes) remains. For virtually any person at virtually any time, there are genuine action options (actions that he or she can perform) that are not eligible to be morally required of that person at that time. This problem also extends to the category of immorality. There will be (^13) For similar cases, see Frances Howard-Snyder, “The Rejection of Objective Conse- quentialism” in Utilitas 9 (1997); Dale Miller, “Actual Consequence Act-Utilitarianism and the Best Possible Humans” in Ratio 16 (2003). (^14) I have argued elsewhere that these disqualifying factors generally center around the fact that an action cannot be performed as a result of deliberative agency, but I’ll leave this aside here. See [Author’s Forthcoming Paper ], op. cit.
some actions that an agent can perform, but which seem ineligible for the category “morally prohibited”, in virtue of the relationship to which an agent stands to the potential not to perform the action. Imagine, for instance, that Phil stands before a safe, and that opening the safe will prevent a very serious calamity. Phil, however, does not know the combination of the safe, and hence does not know how to open it. It would seem inappropriate to say that opening the safe is morally required, given that he doesn’t know the combination (though, strictly speaking, he can open the safe, by entering the right numbers in the right order). But it would also seem inappropriate to say that failing to open the safe is immoral. Given the presence of relevant disqualifying factors, the actions of opening the safe or not are not eligible to be morally required or immoral. I think the above set of claims are reflective of moral experience; some actions are just ineligible to be morally required despite the fact that one can (if one is, e.g., really lucky) do so, and hence refraining from such actions is ineligible to be morally prohibited. But this isn’t yet enough to describe such actions as amoral. Note that failing to be required does not entail failing to be evaluated. A number of potential evaluations could remain; these actions could be morally permissible, as well as morally better or worse than potential alternatives. However, if we allow a potential action φ cannot be morally required on the basis of various disqualifying factors, this entails that φ-ing must be amoral. The key is the relationship between moral requirement and moral reasons. As I understand the latter concept, a moral reason is the evaluative currency of morality: it is a fact that helps to determine whether a particular action is morally better, morally worse, or maintains some other moral deontic status, etc. So to allow that φ-ing is morally evaluable is to allow that moral reasons apply to it, and will count in favor of, or against, various deontic valences possessed by the action. But if we allow that moral reasons will apply to opening the safe, it seems hard to say how doing so could fail to be morally required. Any plausible theory, other things being equal, will require Phil to open the safe if doing so prevents a very serious calamity at basically no cost—on the assumption that moral reasons apply to so doing. Thus, if moral reasons tell in favor of opening the safe to any degree (which they must if this action is to be evaluated at all), it is hard to see how these moral reasons could fail to generate a moral requirement. Hence to claim that Phil’s opening the safe is not morally required—which, it seems to me, it surely isn’t—we must say that reasons don’t apply to it, or that any potential reasons are silent given its failure to be morally requirable. Parallel reasoning applies to Phil’s potential failure to open the
Independence of Evaluable Alternatives (IEA): whether φ is morally evaluable for a person A in circumstances c is independent of the availability and moral valence of alternative non-φ actions in c.^17
But if we accept IEA, it would appear that any action that maintains the rel- evant disqualifying factors cannot be evaluated under any circumstances— whether or not, were it to be evaluated, it would be required. This is because (a) it cannot be evaluable if to evaluate it would be to require it and (b) whether or not to evaluate it is to require it depends on the alternative ac- tions available in those circumstances. Thus to say that the relevant action can be evaluated so long as to evaluate it would be to not render it morally required violates IEA: it holds that an action can be morally evaluated or not depending on the alternative actions available in a particular set of cir- cumstances. Hence, or so it would seem, the only way to guarantee that actions maintaining the relevant disqualifying factors will never be morally required is to outlaw moral evaluation of such actions in any circumstances. Moral evaluation is, therefore, not pervasive: it does not evaluate actions that, for particular agents, maintain the relevant disqualifying factors.
2.2. Regret^18
The problem of moral requirability succeeds in defending the existence of amoral actions for views that accept the existence of moral requirements. But some views deny this. For instance, according to a scalar view, there are no moral requirements: actions are simply rank-ordered from best to worst.^19 Because this view doesn’t admit of requirements, it need not as- sign a valence of amoral to avoid having actions that display the relevant
(^17) One question is worth discussing here. IEA can also be stated as follows: take two sets of circumstances, c and d. According to IEA, any action in c that is evaluable in c must also be evaluable in d if c and d differ only in the potential available actions. But, or so it would seem, this requires me to say that a particular action φ in c is identical to another action ψ, in d. But surely these actions are not identical. After all, they bear different relations (including being present in different circumstances). However, IEA can be easily restated as follows: according to IEA, the evaluability of any iteration of a particular act-type cannot vary simply on the basis of the alternatives available to this iteration. In other words, any iteration of the act-type “killing Caesar” cannot be evaluable simply as a result of the alternative actions available in the set of circumstances in which that iteration is available (though it can be evaluable or not depending on the presence or absence of, e.g., relevant disqualifying factors in such circumstances). (Goldman, 12.) Thanks to [... ] for urging me to discuss this issue. (^18) Thanks to an anonymous reviewer for prompting me to include this argument. (^19) Norcross, op. cit.
disqualifying factors classified as morally required. This proposal (and others like it^20 ) would still allow the open alternative of curing cancer to be “morally best”, “morally optimal”, or “supported by strongest moral reasons”, even if not strictly speaking required. But while this proposal is not as implausible as the claim that these actions are morally required, it is bad enough. Here’s an argument for this claim. Consider the following principle con- cerning the appropriateness of a particular attitude—regret—in response to not having performed specific actions. It runs like this: if an action is morally best (or is supported by strongest moral reasons, or is morally op- timal, etc.), it is, other things being equal, appropriate to feel regret for not having performed the action in question. This is a very modest principle. By “appropriate” here, I do not mean that the morally upright agent will always feel regret for not having performed the morally best action, or that someone is somehow a morally imperfect agent for not feeling such regret. I mean only that regret would not be considered “out of place”. I also do not mean to adopt a position that accepts an analytic link between the ap- propriateness of such an attitude and failing to perform the morally optimal action. Nor do I mean to say that it is appropriate to blame those who fail to perform the morally optimal action or the action supported by strongest moral reason (indeed, such a claim would be manifestly implausible espe- cially if we accept the existence of the supererogatory). I simply mean that whenever an action is morally best, morally optimal, or supported by the strongest balance of moral reasons, it would not be out of place to regret failing to perform it even if it’s not actually the subject of a moral obligation. One might argue that the principle I propose is too strong. Instead, one might hold that regret is appropriate only if one is also required to perform the action for which one has strongest moral reason. (In other words, moral regret is appropriate only if one has behaved immorally.) But, first, this is implausible especially for a scalar view. If it is appropriate to regret failing to perform an action only if it that action is required, and if no actions are ever required, it is never appropriate to regret failing to perform an action, no matter how morally favored. This is an absurd result. But, second, this claim is implausible on any view. Take, for instance, failure to perform a supererogatory action, e.g., my refusal to go out of my way to
(^20) This includes a potential view according to which moral requirements exist, but ac- cording to which the relevant disqualifying factors block moral reasons (which remain applicable to such actions) from generating a moral requirement. Thanks to an anony- mous reviewer for this proposal.
option to push a button to ensure high-quality lives for all humanity until the end of time. Instead, one simply sits down at one’s computer, punches a bunch of random keys, and by sheer lucky guess, cures cancer. Wouldn’t we say that curing cancer, in this case, is immoral (or at least morally worse, or morally negative), given the alternative? In response, I think the answer is no. Again, I think this case mixes up two separable actions, one evaluable, the other not. It is plausible to say that we could morally criticize someone for omitting to ensure high quality lives for all when they had the chance. This action does not maintain the relevant disqualifying factors. But this is different than saying that the action of curing cancer, which is distinct, is itself evaluable.^22 And given the reasoning on display above, it seems right to say that it cannot be evaluable, given that it cannot be requirable. This is not to say that there is nothing morally criticizable in what someone does in that case. One natural reaction to this case might be that despite the fact that one didn’t press the welfare button, which was wrong, at least typing random keys wasn’t a waste: one cured cancer by doing so; but this just illustrates that we will generally not morally evaluate the curing of cancer itself, but rather will treat the curing of cancer as a morally salient effect or result of an action that is, genuinely, evaluable, viz., punching random keys rather than pressing the welfare button. Curing cancer will not be the focus of moral criticism. A similar objection arises in this context.^23 Consider the following plau- sible principle of reasons:
Incompatible: if r is a reason for x to φ at t, and if x ’s ψ-ing at t is incompatible with x ’s φ-ing at t, r is a reason against x ’s ψ-ing at t.
Incompatible seems plausible: if I can’t have a hamburger and a hot dog, any reason in favor of eating the hot dog would also seem to be a reason against having a hamburger, and so forth. With this in mind, consider the argument of the last paragraph. Ex hypothesi, pressing the high-quality-lives button is incompatible with curing cancer. But if there are a number of moral reasons in favor of pressing the button, as would surely seem to be the case, then Incompatible implies that these self-same reasons are also reasons against curing cancer, and hence it would appear, at least in some circumstances, that curing cancer can be evaluable (and potentially immoral).
(^22) Cf. Goldman, ch. 2. (^23) Thanks to an anonymous reviewer.
Despite its plausibility, I resist Incompatible. Indeed, there is indepen- dent reason to reject it. This is because it doesn’t restrict the actions to which moral reasons could apply at all. Take, for instance, the action of my reversing the direction of time for the purposes of stopping the holocaust. Doing so, let’s say, is incompatible with my eating a hamburger in at least this straightforward sense: were I to reverse time to stop the holocaust, I would not eat my hamburger. But this does not entail that any reason in favor of eating a hamburger is also a reason against going back in time to stop the holocaust. This is because, plausibly, we don’t evaluate actions one cannot perform. To accept Incompatible one would have to revise it to avoid this result. Instead, one would have to accept something along the lines of:
Incompatible*: if r is a reason for x to φ at t, and if x ’s ψ-ing at t is incompatible with x ’s φ-ing at t, and if x can ψ at t, r is a reason against x ’s ψ-ing at t.
This would avoid the problem just cited. But notice that to restrict the application of Incompatible* to actions that one can perform just is a choice on our part: it is a choice made given the plausibility or implausibility of ap- plying moral status to actions one cannot perform. Notice that one needn’t accept my intuitive judgment that all acts one cannot perform are not evalu- able; the general point stands if one more restricts Incompatible in a different way. And the general point is: Incompatible must be restricted, and it is re- stricted simply on the basis of our first-order judgments concerning the class of actions to which moral reasons (or reasons of any other domain, for that matter) legitimately apply. But I have already given strong reason to believe that applying moral status to actions that possess the relevant disqualifying factors is itself implausible. And hence, if that argument succeeds, there is little reason to accept Incompatible* (or some other restriction) and not
Incompatible**: if r is a reason for x to φ at t, and if x ’s ψ-ing at t is incompatible with x ’s φ-ing at t, and if x can ψ at t, and if φ does not maintain any of the relevant disqualifying factors for x at t, r is a reason against x ’s ψ-ing at t.
But obviously Incompatible** causes no problem at all for the existence of amoral actions, at least as I defend them here.
2.4. Objection: Objective/Subjective
For an appeal to the problem of moral requirability to be useful in de- fending the existence of amoral actions, one must share the judgment that
cancer. On no plausible account of the nature of the subjective ‘ought’ am I not required to cure cancer, given that I believe, rightly (or so we are as- suming for the sake of argument) that its moral value trumps anything else I might do, and that I believe (rightly) that I can do so. I suspect that the suggestion that I am not subjectively required to cure cancer rests on the error of confusing the moral value of an action with the moral value of the means to that action. To say that I am not sub- jectively required to cure cancer is to treat the subjective moral value of curing cancer as deriving from the subjective moral value of sitting down at a computer and typing random characters, which is the only means by which I could cure cancer (and which, obviously, has a pretty low subjective moral value). But notice that these are distinct acts. Even if typing random characters does not have a particularly high subjective moral value, cur- ing cancer sure does, and hence we can still be required to cure cancer (by whatever means).^27 Thus there is reason to believe that we should exclude some actions from the category of subjective moral requirability. Hence, given that non-requirability entails amorality, there are subjectively amoral actions. But it is hard to see why we should accept subjectively amoral actions without also accepting objectively amoral actions—especially given that it remains implausible to hold that we objectively ought to perform actions that we can perform only as a matter of sheer luck. And hence there are reasons to accept amoral actions at both the subjective and objective level. Given this, I hereby abstract from this distinction for the remainder.
2.5. Objection: Justifying Reasons
Here’s a restatement of the first argument for amoral actions. Moral evalu- ation of any action in any context depends on there being reasons in favor or against performing the relevant actions. But if moral reasons apply to a particular action, it would appear that, in at least some contexts, these actions (which nevertheless maintain the relevant disqualifying factors) can themselves be morally required (or immoral). One could say that they are not morally evaluable only in cases in which were they to be evaluated, they would be morally required. But this response has two problems: it continues to allow the existence of amoral actions, and (more importantly) violates IEA. And hence actions that maintain the relevant disqualifying
(^27) Note that even if we adopt a coarse-grained theory of action-individuation, we cannot escape subjective moral obligations to cure cancer (and to act in other disqualified ways): given its extremely high subjective moral value, I am surely morally required to perform the single act of typing keys in the order necessary to cure cancer.
factors cannot be morally evaluated. As should be clear from the above argument, whether actions that main- tain the relevant disqualifying factors, and are hence not requirable, can be morally evaluable turns on the nature and function of moral reasons. So far I’ve been writing as if all moral reasons have the power to require actions. In other words, once moral reasons in favor of a certain action φ are strong enough, or strong enough in comparison to the moral reasons in favor of an alternative action ψ, φ-ing is required. But, as some have argued,^28 moral reasons can have different functions. It could be, for instance, that some moral reasons have, at most, the power to justify actions without also main- taining the power to require actions. And if this is correct, we might say that reasons that count in favor of actions that maintain the relevant dis- qualifying factors maintain only justifying strength. No matter how strong the reasons in favor of curing cancer are, on this proposal, they could never require doing so, given that they fail to have the power required to do so. However, I think this suggestion is unattractive. Note that justifying rea- sons still maintain at least some, indirect, relationship to moral requirement. If an action is supported by justifying reasons, but justifying reasons that are not sufficient to justify (given the circumstances in which it is placed), this action is thereby morally unjustified, and hence one is morally required not to perform the action. In addition, if a single action passes the relevant threshold of sufficient justification, this action necessarily will be required: it is the only morally justified action. But if we allow that justifying reasons count in favor of actions that maintain the relevant disqualifying factors, it becomes inevitable that some actions that maintain the relevant factors will not be favored by justifying reasons to a sufficient degree, given po- tential alternatives, and hence will be morally unjustified. But this itself is problematic. Take again Phil’s case. It certainly seems wrong to say that his failure to open the safe is immoral. But, if evaluable, it seems wrong to say that failing to open the safe could possibly be justified by sufficient justifying reason. (After all, this failure results in some horrible calamity.) But if this is right, given that actions for which one lacks sufficient justifying reason are not justified, and hence immoral, one must say that Phil’s failure to open the safe, if evaluable, is immoral.^29 But this is no better than saying
(^28) Most notably, Joshua Gert in Brute Rationality (Cambridge: Cambridge University Press, 2004), ch. 2. See also Douglas Portmore, “Are Moral Reasons Morally Overriding?” in Ethical Theory and Moral Practice 11 (2007). (^29) Even if we reject deontic valences for actions that possess disqualifying factors, a roughly similar problem remains; if this action is not supported by strong enough justifying reasons, even if not immoral, it would certainly appear eligible for appropriate regret;
the “simple ability” to) perform, but that also are morally evaluable, as R. Now assume that at time t, agent x faces the possibility of performing an amoral action. At this time, then, there will be a distinction between the members of S and the members of R for x. But, for the members of R, there will be a subset—call this Rp—of R the members of which pass the thresh- old of moral permissibility. And hence, given that one performs a morally permissible act if and only if one conforms to an action one is morally re- quired to perform, one will be required to perform some member of Rp. But given that one behaves immorally if and only if one fails to conform to one’s moral obligations, if one fails to perform a member of Rp, one will behave immorally. But any non-R member of S will, obviously, not be a member of Rp. And hence, in performing a non-R member of S, one behaves in an immoral way, and hence not an amoral way. Schematically, this puzzle runs like this:
- Assume that, in any choice circumstance, a proper subset of R (Rp) is constituted by those members of R that are morally permissible.
- One is morally required to perform a morally permissible action.
- Because amoral actions are, ex hypothesi, not members of R, no amoral action is a member of Rp.
- Hence to perform an amoral action φ is to fail to perform a morally permissible action.
- Hence to perform an amoral action φ is to fail to conform to a moral requirement.
- Hence, φ-ing is immoral.
- Hence, φ-ing is not amoral.
- Responses
To defend the existence of amoral actions, it must be the case that the argument framed above goes wrong somewhere. I’ll explore three potential responses in this section.
4.1. Amorality as Conformity to Moral Obligation
First, one might deny (4). One might claim that simply because one per- forms an action that is not morally evaluable (i.e., not a member of R), this does not entail that one does not also perform a member of Rp. Take, for instance, the act of not physically assaulting the person sitting next to me.
This action is surely morally required. But imagine that instead of physi- cally assaulting the person next to me I whip out my laptop, type a bunch of random keys, mail the result to a top medical journal and, as a result, widely disseminate the cure for cancer. Here I have performed an amoral action. I have performed an action that is not morally requirable given my cognitive limitations. However, it’s also the case that I have behaved in a morally permissible way: I haven’t physically assault the person next to me. Hence simply because one can be morally required to perform a particu- lar action or set of actions does not mean that one cannot also perform an amoral action in the effort.^30 Sometimes performing an action that lacks moral valence entails also performing an action that is morally permissible. Of course, this is perfectly correct. (4) does not hold, and hence (5) also does not hold, in cases in which one can perform an amoral action coincidentally with a morally permissible action. But to deny (4) on the current grounds simply opens a new route to the conclusion. To see this, consider that there may be cases in which it is impossible to fulfill my moral obligations by performing an amoral action. Recall the original case that introduced the problem. At t, R includes: {ψ, π}. All one would have to do is to interpret ψ-ing, which produces far more moral good than any other members of this set and so, plausibly, is the sole member of Rp, as unable to be performed as a result of φ-ing, which is a non-R member of S. In this case it would appear that an ex hypothesi amoral action is incompatible with performing an action in Rp, and hence doing so is immoral, and hence doing so is not amoral. One might respond by holding that this proposal still leaves the possi- bility that there could be some amoral actions, i.e., just in those cases in which one’s alternatives are such that one can perform a morally permissible action by performing an amoral action. But this would seem to render cer- tain actions amoral only in cases in which one’s alternatives are such that to perform an amoral action is also to act in a morally permissible man- ner. This result violates IEA. Whether an action is morally evaluable for a person at a time should be a feature of the properties of that action—in particular, whether it displays the relevant disqualifying factors—not the extent to which that action is compatible with the performance of some other action. And if this is right, then if we have reason to believe that performance of some particular action φ can be immoral, and hence can be evaluable, for a person A in one case with one set of alternatives (such as the case of curing cancer when one is required to donate one dollar instead),
(^30) Or, if one prefers a coarse-grained scheme, an amoral action description in the effort.
