Moral Worth

Received: 6 September 2023 DOI: 10.1111/phc3.12982 ARTICLE - Revised: 24 April 2024 Accepted: 27 April 2024 Moral worth Euan K. Metz The American University in Cairo, New Cairo, Egypt Abstract The concept of moral worth, of being creditworthy for doing Correspondence the right thing, is often seen as essential feature of a moral Euan K. Metz. Email: euan.metz@aucegypt.edu theory. It forces us to provide a clear account of the relationship between moral motivation and moral action, raising important questions about the demands that morality makes of us. Work on moral worth has a long lineage, especially in Kantian scholarship. Recent years, however, have seen a more focused interest in the nature of moral worth outside of the Kantian tradition. Indeed, part of this interest stems from a rejection of an orthodox Kantian understanding of what moral worth is. In this article, I chart prominent reasons for rejecting the orthodoxy, and distinguish between two rival camps that have emerged: Right Reasons Accounts and Rightness Accounts. I delineate some of the demands that these accounts must meet, and end by discussing a potential way forward that has emerged via hybrid views and goal‐based views that attempt to utilise the most promising features of each. 1 | DEFINING MORAL WORTH It is a familiar thought within moral philosophy that a moral assessment of action is lacking if it pays sole attention to the resulting action, without regard for the way in which that action was done. Famously, Kant argued that there is a morally relevant difference between the shopkeeper who consistently charges their customers the same price for the same goods because he believes that his business will succeed only if he is perceived as honest, and the shopkeeper who performs the same action because he believes that doing so is right.1 The moral difference between these cases is reflected in the concept of moral worth. It is important to distinguish the concept of moral worth, which is standardly understood as a property of - action, from concepts of moral assessment which bear on the mental states and processes which rationalise or cause action. It is worth distinguishing in this regard between what determines an action, the features of that action Philosophy Compass. 2024;e12982. https://doi.org/10.1111/phc3.12982 wileyonlinelibrary.com/journal/phc3 © 2024 John Wiley & Sons Ltd. 1 of 14 2 of 14 - METZ (e.g. in light of its determinants), and the resulting action. We talk, for instance, of having “good” or “bad” intentions, motives, and desires, where such evaluations may have a moral dimension. Theories of moral worth often appeal to the connection between such states and processes and the action performed. For example, some think that an action Φ has moral worth if, and only, and because Φ‐ing is right, and A Φ’s on the basis of a good motive. In the same way, we should distinguish between moral worth as a property of an action, from moral assessment of the character of the agent who performs that action.2 It is generally understood that actions can have moral worth even if the agent who performs that action has a bad character.3 The shopkeeper may decide to do the right thing in one instance, but generally tends towards vicious behaviour. Still, his action, at least in this instance, is deserving of moral worth. Given these restrictions most theorists accept that moral worth, although related to positive moral assessments like praiseworthiness, is not identical to them.4 We say that an agent is praiseworthy, as opposed to an action. And it seems that there are ways that an agent can be praiseworthy that do not have anything to do with moral worth. Again, one can be praiseworthy for acting on the basis of a good intention, even if they fail to do the right thing, or even do the wrong thing. Thus, it would better to understand moral worth as a kind of value, an appeal to which could be one explanation of why we hold a positive moral assessment of an agent. This point leads to the last feature which identifies the most prominent modern conception of moral worth. Most theorists make a further assumption in identifying the concept of moral worth, that it is a property not merely of action, but of right action.5 Moral worth is thus “right action þ X”. It is for this reason that moral worth is taken to be a way of identifying whether or not an agent is creditworthy for performing the right action. To be creditworthy for performing a right action means that we can non‐accidentally attribute that action to her. She deserves credit for performing the right action in this sense.6 Restricting moral worth to right action leaves open further options with respect to whether or not the relevant sense of “right action” is an objective one or a subjective one. This choice will presumably make a difference to one's account of X. Suppose that a customer walks through the shopkeeper's door wanting to buy a knife, which the shopkeeper sells her at the appropriate price because they think that is the right thing to do. However, unbeknownst to the shopkeeper, the customer intends to use this knife to commit murder. On an objectivist conception, if right action is not determined relative to the information available to the agent at the time, the shopkeeper does the wrong thing, and his action fails to have moral worth. On a subjectivist conception, on which right action is determined relative to the information available to the agent at the time, the shopkeeper does the right thing, and his action may have moral worth. Here is a schematic analysis of moral worth that will provide us with a starting point and from which we can fill out the details: A’s action Φ has moral worth if, and only if, (i) Φ‐ing is [objectively/subjectively] right, and (ii) A Φ’s on the basis [intention/desire/motive] of X.7 Most of the literature on moral worth is focused on identifying X. I now turn to examine some recent proposals. 2 | WHAT IS X? REJECTING THE ORTHODOXY The orthodox answer to the ‘What is X?’ question identifies X roughly along the lines of the second shopkeeper in the example we began with. It is because the shopkeeper acted on the basis of the fact that it was the right thing to do which gave his action moral worth. This arguably aligns with Kant's own view8 that it is at least necessary for an action to have moral worth that the action be performed for the sake of duty.9 For the purposes of exegesis, I shall define what I will call Rightness Accounts more broadly, as those that identify X with rightness (where this could be further interpreted in terms of various deontic concepts, such as obligation, duty, permission, etc). - METZ 3 of 14 A plausible motivation for adopting a Rightness Account is that such accounts provide a straightforward answer as to how A's right action could be non‐accidental.10 As we have seen, it is possible to perform the right action, but do so for a reason other than the rightness of the action itself. The first shopkeeper acts on the basis of self‐interest, not because the action is right, and nevertheless hits on the right action. But suppose that he knows he can get away with overcharging a few credulous customers. Then his self‐interest would, in those cases, prompt him not to perform the right action. Thus, the shopkeeper's action fails to have moral worth since his motive (self‐ interest) and the circumstances he finds himself in happen to fortuitously align. He does the right thing, but only accidentally. The most obvious choice of X therefore would be rightness itself, since if A is always (or usually) moved to act on the basis of the rightness of the action, as opposed to self‐interest or other “inclinations” he will never fail, when he does the right thing, to do that thing accidentally. He is not always guaranteed to do the right thing, but when he does, in acting from the motive of rightness, his doing the right thing is no accident.11 However, much of the contemporary literature on moral worth begins from a common starting point: the rejection of such theories. Indeed, it is the rejection of the orthodoxy about moral worth that has set the terms for the contemporary debate. Thus, I will first outline some of the most prominent reasons for rejecting the orthodoxy, and then discuss alternative views. First, an appeal to cases. When we think about cases involving moral psychology more complex than that of the shopkeeper, we tend to find that acting on the basis of rightness fails to deliver the right result. A commonly discussed case in the literature is that of Huckleberry Finn. Here is the case: After having run away from home due to his violent father, Huckleberry Finn meets Jim, a runaway slave, and together they sail down the Mississippi river. After becoming close friends, Finn comes to realise that in helping Jim to escape he is thereby helping to steal from Jim’s “owner” Miss Watson. Finn eventually decides that he ought to turn Jim over to the authorities, and that it would be wrong not to do so. But at the crucial moment, Finn cannot go through with it, and ends up concocting a story that allows Jim to evade his captors. Despite acting against his best moral judgement, Finn ends up doing the right thing. Given Finn's belief that he is acting wrongly, he does not act on the basis of the rightness of the action. We can suppose that, if queried, Finn would respond that he had in fact done something wrong. Yet, plausibly, Finn does the right thing. According to Rightness Accounts, we ought to infer that despite the fact that Finn does the right thing at the crucial moment, his action fails to have moral worth, since he acts contrary to what he believes is the right thing to do. Indeed, this is how Finn conceives of his own action: as a coward who fails to be resolute in his moral convictions. And yet, the fact that Finn responds to (for example) the value of Jim's humanity in the face of what he believes is morally required strikes many as morally laudable. Second, Michael Smith's “moral fetishism” objection has been widely taken to present a serious challenge to Rightness Accounts of moral worth.12 Rightness Accounts of moral worth imply a distinctive view about the sort of motivation that the agent's action is done for. They imply that it is at least necessary that the motive of the agent must be that acting in that way is right, where this is understood in a de dicto sense. Describing an attitude under a de dicto reading indicates the way in which that attitude represents what it is about, while describing an attitude under a de re reading indicates that we are talking about the object of that attitude itself.13 For example, suppose that as a fan of horror fiction you desire to meet Stephen King, but are unaware of the fact that he has published work under the name Richard Bachman (indeed, suppose that you are familiar with Bachman's novels and dislike their pessimistic tone). In desiring to meet King, are you also desiring to meet Bachman? If your desire is interpreted as a de dicto reading, then no – you do not represent the object of your desire (King) under the description ‘Richard Bachman’ or ‘The author of Rage’. But if your desire is interpreted as a de re reading then yes, you do, since the object of your desire (King) is identical to Bachman. 4 of 14 - METZ The problem for Rightness Accounts is their commitment to the claim that an agent is moved by the rightness of the action, interpreted de dicto; that moral worth essentially depends on an agent being concerned with doing the right thing, whatever that turns out to be. But, as Smith claims ‘[g]ood people care non‐derivatively about [honesty, caring for their children and friends, people getting what they deserve, justice, equality, and the like]’14 and thus to treat rightness as the ‘one and only moral virtue’15 is to make a ‘fetish’ of morality, in the sense that it makes the legitimate moral concerns of good people conditional on a desire for the right as such. Suppose that you decide to donate a fifth of your salary to a charity to a that treats preventable diseases in impoverished places. Your motivation for doing so stems from a basic concern for the worth and value of human life. For Rightness Accounts, this motivation is flawed in the sense that it fails to include a motivation to do the right thing under that description. But, as Smith points out, this requirement ‘alienate[s] [you] from the ends at which morality properly aims’16 – in this case a direct concern for the value of human life. These objections: appealing to a more realistic moral psychology and the moral fetishism objection have prompted many to abandon Rightness Accounts of moral worth. How then should we account for moral worth? I now turn to examine some positive proposals. 3 | RIGHT REASONS ACCOUNTS The most popular response to the rejection of the orthodoxy is to look instead to the features that make an agent's action right,17 as opposed to rightness itself. Call such accounts Right Reasons Accounts. That should not be surprising, as an underlying motivation behind the objections above is to show that in failing to respond to the features of the action that make it the right thing to do, such as one's concern for the welfare of others, or honesty, or one's care for one's children is to fail to respond to the ‘ends at which morality properly aims’. The right‐making features of an agent's action are usually understood in terms of normative reasons.18 Distinguish between two concepts: (1) the reason for which the agent performed their action (a motivating reason) and (2) the reason which (in fact) counted in favour of the agent acting in that way (a normative reason). Following Arpaly19 and Markovits20 we can explain what it is for an action to have moral worth by “matching” an agent's motivating reason and the normative reason which contributes to the rightness of the action. One influential rendering of this idea, due to Markovits, posits the following: Coincident Reasons Thesis (CRT) My action is morally worthy if and only if my motivating reasons for acting coincide with the reasons morally justifying the action – that is, if and only if I perform the action I morally ought to perform, for the (normative) reasons why it morally ought to be performed.21 Let us apply the CRT to a case of donating a portion of your salary in light of your concern for the value of human life. The fact that there is suffering is a reason for donating, and one that contributes to the rightness of your action. According to CRT, your action has moral worth where your motivation for acting is the fact that there is suffering. This provides a normative reason which justifies acting in that way, and it contributes to what you morally ought to do. This is the result that we want. CRT allows that your action has moral worth while allowing that you have a non‐ derivative concern for the value of others. The CRT can also explain Finn's case. Finn acts against what he believes to be right but ends up responding to a feature of Jim (e.g. his humanity) that in fact provides a normative reason and one that contributes to the rightness of his action. Further, proponents of CRT claim that it provides us with a way to meet the non‐accidentality criterion. Since it demands that the normative reasons which contribute to the rightness of the action must overlap with the agent's motive, it can explain why the charity‐giver's action has moral worth, while the shopkeeper's self‐ interested action lacks moral worth. This is because, while the shopkeeper's self‐interested reason to not - METZ 5 of 14 overcharge his customers is presumably a normative reason to do so, it is not a reason which contributes to the moral rightness of the action. Thus, there is no match or overlap between the shopkeeper's motivation and the moral features of the action which make it right. It thus avoids accidentality in one's action aligning with a non‐ moral motive.22 Defenders of the CRT23 argue that the thesis is necessary but insufficient to account for moral worth as it stands. In the rest of this section let me fill out some of the most influential additional conditions on CRT. 3.1 | Identifying the motive The first way in which CRT is insufficient as it stands is that it does not preclude agents from acting on the basis of a moral reason, but where they treat that moral reason in a way that undermines our judgement that their act has moral worth. For example, suppose that the shopkeeper acts on the basis of the fact that it would be dishonest, but only treats this reason as weighty in his deliberation because acting dishonestly causes him to feel revulsion, and acting honestly usually makes him feel better about himself. CRT implies that the shopkeeper's act has moral worth, despite the fact that he only accidentally did the right thing. We thus need to add to CRT a condition which specifies the way in which an agent treats the moral reason that they act upon. Markovits's response is to draw upon the distinction between instrumental and non‐instrumental moral reasons. In order for your act to have moral worth it must be that you treat the moral consideration which makes your act right as a non‐instrumental reason, which Markovits defines as … provided by the ends of our actions that are worth pursuing for their own sake (in the case of justifying reasons) or that we pursue for their own sake (in the case of motivating reasons).24 In treating the fact that it is the honest thing to do as a weighty reason only because it makes him feel better about himself, the shopkeeper is not treating this consideration as a non‐instrumental reason. Hence, his act fails to have moral worth. Others have argued that the appeal to instrumentality is insufficient. Some argue that we must attend to a broader category of reasons, of which instrumental reasons are a subset, that they call “conditional reasons”.25 A conditional reason is such that it is only a reason given a certain further condition. This includes instrumental reasons, such as the fact that you have reason to drink the water on the condition that it will quench your thirst. But not all conditional reasons are instrumental reasons. If you have reason to go camping because you enjoy it, but will only go if it is not raining, then whether or not it is raining conditions the reason you have to go camping.26 The issue for CRT is that there seem to be non‐instrumental conditional reasons that affect our judgements about moral worth. Coates gives the example of Dion, who helps the needy but does not condition this reason on whether those who need help do so out of moral need, or selfish ends.27 Thus, Dion would help a stranded motorist, but would also help an art thief steal a painting. The former act looks like it fails to have moral worth because it looks merely accidental that he did the right thing, for the right reason. Lockhart proposes that we amend CRT to accommodate this problem by requiring that the reasons for which an agent acts must ‘conditional only upon those reasons that the morally justifying reasons are conditional upon.’28 But is it too demanding to require that actions which have moral worth meet this criterion? Presumably there can potentially be many different sorts of conditions which in fact hold on one's moral reasons, and if one must have in mind all of these conditions, correspondingly fewer actions will have moral worth. Perhaps we need to further identify which conditions are appropriate to the case at hand, and which are not. CRT also does not specify the level of description of the non‐instrumental normative reasons which figure on its right‐hand side. Defenders of CRT‐style accounts tend to appeal to the sorts of considerations that would help to meet the objections of Smith, considerations that identify appropriate moral ends. But those considerations 6 of 14 - METZ could be the most fundamental right‐making feature of the action, such as, for example that the action will maximise net well‐being overall. That sort of reason may not be subject to the charge of “fetishism”.29 Finally, there is the issue of weighting. To what extent given CRT do the motivating reasons that you act upon need to overlap with the non‐instrumental normative reasons that morally justify that act? Markovits argues that we should think of moral worth as gradable and use the idea of overlap to capture this aspect. The more that your motivating reasons overlap with your normative reasons, the more moral worth that you act has.30 On the other hand, others31 do not think that moral worth can come in degrees, especially those who think that moral worth is primarily about non‐accidentally acting rightly, since as Portmore points out, you either did it by accident or not.32 3.2 | Modal issues Many have pointed out that the issue of non‐accidentality is affected by modal considerations. Correctly attributing moral worth to an agent requires that they would have performed the right action non‐accidentally in nearby possible worlds. To see why this is important, consider the following case from Sliwa: Jean’s Case Jean’s friend missed her bus to work and frets over being late to an important meeting; coming late would be a great embarrassment to her. Wanting to spare her friend a major embarrassment, Jean gives her a ride. Let’s assume that giving her friend the ride is the right thing to do in these circumstances and the fact that it spares her friend a major embarrassment makes it right. Thus, Jean acts for a right‐making reason.33 Whether Jean's action has moral worth depends on its meeting the non‐accidentality condition. It must not be a mere fluke that Jean does the right thing. But it is possible that her doing the right thing is a fluke, since we can easily imagine circumstances in which the right‐making reason that Jean responds to (sparing her friend embarrassment) is outweighed or defeated by competing moral considerations. Suppose that Jean would be motivated to spare her friend embarrassment, even if it would involve shooting somebody. It appears that despite Jean acting on the basis of a moral reason which makes her action the right thing to do, it is an accident that she did the right thing. To what extent are counterfactual circumstances relevant to the CRT? Markovits argues that appealing to such counterfactuals can be misleading. She cites the example of a dog‐lover who saves strangers at sea at personal risk but would have saved their dog if they had been in harm's way. In this case she urges us not to ‘discount the worthiness of his actions… That he would have done so may be a sign of excessive concern for the dog, rather than of too little concern for the strangers – after all, the dog‐lover was willing to risk his own life to save theirs.’34 How are we to decide when and where counterfactuals are relevant to assessing moral worth via CRT? Let us consider the cases in more detail. In Jean's Case, Jean would have acted on the reason that she believes she has (that it would be embarrassing for her friend) even in the circumstances in which her action is not right (Howard calls this “sufficiency luck”35). In the dog‐lover case, the dog‐lover would have acted differently if the circumstances were different. Sliwa's thought is that if we vary the circumstances, but keep the agent's motives fixed, acting on the basis of right‐making reasons imports accidentality. One response to these cases is to argue that counterfactual considerations about what an agent would have done are irrelevant to the moral worth of the action, as this is to conflate moral worth as a property of the action performed, with moral virtue or disposition, which is a property of the character of the agent (as stressed in §1). It is important that we allow the moral worth of an action to be determined by the actual motives of the action at hand (what Isserow calls the “pertinence constraint”36). It is also relevant which circumstances we are considering. It is unhelpful to suggest that agents would keep fixed the same motives across very different counterfactual circumstances. It is plausibly too demanding to claim METZ - 7 of 14 that an agent lacks moral worth where they are generally responsive to right‐making reasons, but who fails, in marginal cases, to correctly identify the sufficiency of those reasons, acts for one of those reasons and unluckily does not to do the right thing.37 Defenders of CRT‐style accounts tend to emphasise the connections between agency and action here.38 Although moral worth is not moral virtue, it is not implausible to think that there are internal connections between the two concepts. A morally virtuous person is the sort of person who is good at responding to their moral reasons, which would be a starting place to secure the right sort of counterfactual safety. There are at least two difficulties for this move. First, these defenders need to justify and explain the connection between virtuous agency and good action. Unsurprisingly, those who defend Rightness Account views tend to emphasise the difference between these two concepts. For example, Johnson King argues that ‘… defenders of [Right Reasons Accounts] take there to be a much closer connection between praise for character traits and praise for act‐types than actually obtains.’39 She argues that in the Finn case, we ought to praise an element of Finn's character, but not his action (though she emphasises a slightly different sense of accidentality here – that Finn has “no idea” that he is doing the right thing). Second, defenders of Right Reasons Accounts need to explain exactly how it is that agents generally respond correctly to their moral reasons, without invoking the condition that they be motivated by rightness itself to guarantee non‐accidentality. Work has been done on this score, especially due to Lord40 and Cunningham41 who defend the idea that correctly responding to one's reasons in the required way is a matter of know how. Howard also pushes this line, stressing the continuity between acting with moral worth, and achieving well in other domains such as football and music.42,43 3.3 | Subjective or objective? A final aspect of further specifying CRT is whether or not the moral ought or moral reasons in CRT should be understood along objectivist or subjectivist lines. A subjectivist reading of moral ought is that it is determined by the considerations which are within the epistemic ken of the agent. Thus, if a doctor has reasonable evidence to believe that by administering a drug you would be harmed, then they ought not to do so. That is true even if, unbeknownst to the doctor, the drug would in fact prevent your disease from worsening and has no other bad side‐ effects. Markovits understands moral ought claims to be determined only by reasons which depend on what an agent has evidence to believe about their situation, that ‘we are morally required to do only what we have sufficient epistemic reason to believe it would be best to do, not what it would (in fact) be best to do.’44 Others understand moral worth as attaching only to right actions understood objectively.45 Sliwa objects that accepting that allowing for a subjectivist interpretation of right‐making reasons leads to implausible results, such as where an agent who due to her sheltered upbringing is morally ignorant about the morality of homosexuality and has (by her lights) morally justifying reason to sign a petition against same‐sex marriage.46 The choice over an objectivist or subjectivist reading of CRT is, without further argument, optional. However, adopting a subjectivist interpretation on which the evidence available to the agent determines what it is right for them to do arguably affects whether or not they do that thing by accident. Think back to Jean's Case. Jean acts on the basis of saving her friend from embarrassment. Since the same evidence that is available to Jean both constitutes a reason for her to act and makes that act right, the sort of accidental cases that Sliwa considers would not arise.47 If the fact that it would save her from embarrassment does not give her a reason (from Jean's evidence) then it fails to be right for her to act on that basis. Thus, for example, Howard argues that the only way to defend the idea that Jean's reason could be a fluke is to appeal to objectivism about rightness.48,49 If this is right, taking a stand on subjectivism/ objectivism about moral concepts may have significant implications for the debate between the orthodox view about moral worth and Right Reasons Accounts. It is also worth noting that there is comparatively much less discussion of this point in the recent literature on moral worth, so there is more work to be done on this score. 8 of 14 - METZ 4 | BACK TO KANT? As we have seen, Right Reasons Accounts encounter difficulties meeting the non‐accidentality criterion if (a) we insist on drawing a sharp distinction between moral character and moral worth, and (b) we are objectivists about rightness. Recently therefore, some theorists have urged that we should re‐examine the merits of Rightness Accounts. Sliwa's own reaction to Jean's Case (§3.2) is that in the end Right Reasons Accounts are unable to avoid accidentality insofar as they allow for variance of rightness on the basis of various kinds of ignorance. Sliwa argues on this basis that we should look back to the orthodoxy and reconsider the idea that moral worth must be captured by a desire to do the right thing (read de dicto). However, in order to secure non‐accidentality (what Sliwa calls “counterfactual robustness”50) we require not only that an agent has a desire to do the right thing (read de dicto) but that knowledge of what right thing is to do must be part of this motivation.51 Appealing to knowledge helps to secure non‐accidentality since if ‘an agent knows that an action is the right thing to do, then it is the right thing to do and she could not have easily been mistaken about it’s being the right thing to do.’52 Some have argued that the knowledge condition fixes the accidentality issue, but at the cost of making reaching moral worth too demanding. For example, Isserow puts forward the case of John, who is unsure whether saving a drowning child in the ocean would violate duties he has to himself, and despite not knowing whether doing so is right, saves the child anyway.53 If knowledge is a constraint on moral worth then John's action by hypothesis has none, but this seems a hasty judgement. Again, this is a good example of how the sharp distinction between moral character and moral worth that Rightness Accounts insist upon can lead to implausible results. Further, although the knowledge condition may secure non‐accidentality, it is not clear that a Right Reasons Account could not simply co‐opt this condition into their own view. For instance, moral worth could be determined by acting on the basis of a non‐instrumental sufficient moral reason which you know is a sufficient right‐maker. This would have the advantage of securing non‐accidentality in just the way Sliwa wants as well as avoiding the worry about de dicto moral motivational fetishism.54 Another defender of the Rightness Account, Johnson King, rejects the stringent requirement for knowledge, and instead adds the condition that an agent's act must be an ‘instance of someone's deliberately doing the right thing’.55 This secures non‐accidentality (trivially) since, as she puts it ‘deliberate’ and ‘accidental’ are antonyms. If you do something deliberately, ipso facto, you do not do it accidentally. Moreover, to deliberately do the right thing is a distinctive sort of achievement, of ‘trying to act rightly and succeeding’.56 However, as noted at the end of §3.2. defenders of Right Reasons Accounts can point out here that successfully trying to act rightly may involve a form of skill, rather than an explicit concern (de dicto) to do what is right. Given the availability of this option, it is not clear that Johnson King's account provides decisive reason to opt for a Rightness Account. What of Smith's moral fetishism objection? There is a fairly large side literature on responses to Smith's worry, so here let me briefly mention some prominent criticisms. Some argue that the charge of fetishism applies only to certain cases, and that the lesson does not generalise. For example, a standing desire to do what is right (read de dicto) may provide a corrective where an agent lacks a desire (read de re) to do what is right. Lillehammer gives the example of a wife who is tired of their husband and is tempted to cheat on him.57 Although she thinks it would be wrong to have an affair, she finds herself ambivalent about how her husband would feel. In this case, it seems less plausible to suggest that her standing desire to do what is right (read de dicto) undermines our intuition that her action has moral worth. Others have argued that we should think more carefully about the sort of motivation a concern for rightness can be. As Svavarsdóttir points out,58 we may too easily dismiss moral motivation de dicto if we have primarily in mind somebody ‘with a rigorous obsession with morality or a resistance to examine hard reflective questions about morality.’59 Finally, another common objection to Right Reasons Accounts is that in arguing that moral motivation (de dicto) is problematic, they find it difficult to handle cases in which the agent is acting under conditions of uncertainty. For example, suppose that you are unsure about what the right thing to do is, and you decide (rationally) to put effort into determining what the right thing to do is. It seems plausible that you may not want this moral knowledge for its own sake, but rather because you want to do what is right (whatever that turns out to be).60 A related issue is that METZ - 9 of 14 Right Reasons Accounts may struggle to adequately deal with cases of moral testimony. Suppose for example, that you are blamelessly out of touch with what moral concerns are relevant to your action, but you want to avoid wrongdoing. To do so, suppose that you may permissibly appeal to a morally competent adviser to tell you what the wrong thing would be to do. Since Right Reasons Accounts tend to insist on moral motivation de re moral worth in cases of this sort seem to be ruled out, but it is not clear why they should be.61 The debate continues. One potential way forward is to recognise that although de dicto moral motivation understood as a non‐derivative concern may face the fetishism charge, a derivative concern to do the right thing (de dicto) may be less problematic. And it seems plausible to read cases of acting under conditions of uncertainty in this way. There is clearly a difference between an agent who cares solely for doing whatever the right thing turns out to be, and an agent who cares about doing the right thing (under that description) insofar as adopting that motive helps to achieve the ends to which morality aims. Drawing that distinction may help to assuage Smith's worry that de dicto moral motivation necessarily alienates oneself from those aims. 5 | MOVING FORWARD: HYBRID VIEWS AND GOAL‐ORIENTED VIEWS How then to capture non‐accidentality without being too demanding and facing “fetishistic” worries (Rightness Accounts) or becoming embroiled in the knotty issue of identifying how one tracks good reasons without undermining the pertinence constraint (Right Reasons Accounts)? I end this article by looking forward to more recent proposals that have tried to look beyond both the Right Reasons/Rightness dichotomy. One strategy has been to reject the standard assumption lying behind the Right Reason Account/Rightness Account dichotomy, that either rightness or right‐making features are a necessary condition for moral worth, but not both. We might challenge this assumption, building accounts on the premise that both rightness and right‐ making can be accommodated. Singh, for example, argues that a concern for rightness is a ‘precondition’ of the ‘proper functioning’ of motivation by right‐making features.62 His idea is that we can build into CRT the idea that (i) the agent's moral reasons are sufficient, and (ii) they act for those reasons ‘in virtue of taking it to contribute to the overall moral status of the action.’63 Thus a concern to do the right thing can guide the agent's motivating reasons, as opposed to being those motivating reasons. This allows us to say in Jean's Case (§3.2) that she fails to have moral worth insofar as a concern to do the right thing does not guide her action. Compare with Isserow, who argues that it is necessary and sufficient for moral worth that one be motivated by either what makes one's act right or the right‐making facts.64 Neither motive is, in itself, morally speaking, better than the other. However, clearly while advocates of hybrid approaches can make use of the most promising features of each of the accounts, they will also need to demonstrate that they avoid the problems that beset these accounts.65 Singh's account needs to respond to criticisms related to the idea that taking a moral reason to contribute to the overall moral status of the action is not necessary for moral worth66 while Isserow's account needs to respond to the criticism that responding to the right‐making facts is not sufficient for moral worth as it fails to secure against accidentality.67 Portmore provides an alternative approach.68 In cases which involve accidentality, like Jean's Case, rather than appealing to counterfactuals about what Jean would have done under other circumstances (which violates the pertinence constraint) or to the specific concern that is typically identified as motivating the agent to act (e.g. that Jean wishes to spare her friend from embarrassment) we should appeal to the wider nexus of concerns, both in terms of the features which favour that action, and in terms of the goals that performing that action serves.69 Portmore understands concerns in a wide sense (importantly including lack of concern – e.g. Jean's willingness to shoot somebody to save her friend from embarrassment shows a lack of concern with human life), and as context‐ sensitive. He argues that differing intuitions about attributing moral worth can be partly explained by paying closer attention to context, and as grounded in what he calls “ultimate moral concerns”.70 10 of 14 - METZ On the context‐sensitivity of concerns, consider the differing interpretations of the Huck Finn case. Arpaly71 and Markovits72 (Right Reasons Accounts) interpret Finn's action as morally worthy even though not from a motive of rightness, while Johnson King73 and Sliwa74 (Rightness Accounts) interpret Finn's action as not morally worthy as not performed from the motive of rightness (it was an accident, in some important sense, that Finn got it right). If the relevant context is whether Finn will do the wrong thing from a desire to help his friend, even if he believes that it is wrong, then his doing the right thing does indeed look lucky. But if the relevant context is whether Finn's justified concern for Jim's humanity would lead him to do wrong, Finn's action looks less lucky, in the sense that this concern is plausibly closely bound up with right action. The context‐sensitivity of concerns thus goes some way to diagnosing what drives each account to opposing conclusions about the same moral act.75 Second, Portmore's notion of an “ultimate moral concern”, or goal provides us with a tool to advance upon the focus on identifying the motive “X” through integrating within the theory of moral worth both the immediate motive of the agent and the moral goals which inform the presence of that motive. This allows Portmore to say, for example, that moral worth does not demand a concern for rightness as an ultimate goal, though it may in certain circumstances provide an appropriate (immediate) motive. This aspect of Portmore's theory is especially helpful in answering concerns around modal issues around the non‐accidentality condition, since acting from the ultimate concerns that an agent should have helps to ensure that an agent's acting from the right reason is stable across nearby possible worlds. For example, the view helps to explain why Jean's action fails to have moral worth, despite acting for a right‐ making reason, yet allows us to say that Markovits's dog‐lover's action has moral worth in responding to the right reason. Also, as we have seen, a number of theorists have pointed to the role that having at least a derivative concern to act rightly can play in assessing moral worth. This view appears to resolve this issue insofar as having a derivative concern to act rightly is a ‘means to furthering the end for which they should have an ultimate moral concern.’76,77 6 | CONCLUSION Recent work on moral worth has made strides in developing a clearer account than has been previously available. As we have seen, defenders of the two dominant views on the nature of moral worth sharply disagree over how far moral worth and moral character are distinct, and over the right way to secure the “counterfactual robustness” of right actions, without violating the pertinence constraint. A significant recent strand of the debate has thus tried to move beyond the two dominant views, instead attempting to incorporate the insights of both sides either into hybrid‐style views, or views that incorporate a telic element. A CK N O W LE DG M EN T S My thanks to two anonymous referees for Philosophy Compass for their helpful and constructive feedback on this paper. O R C ID Euan K. Metz https://orcid.org/0009-0001-4759-8842 E N D NO T E S 1 See Kant (1998/1785), 4:397. 2 See Portmore (2022), Johnson King (2020), §3. 3 See Hurka (2006). 4 An exception to this is Arpaly, who identifies moral worth with ‘… the extent to which the agent deserves moral praise or blame for performing the action.’ See Arpaly (2002), p.224, Arpaly (2003), p.69. METZ - 11 of 14 5 Not everybody shares this view. Markovits (2010) conceives of moral worth as ‘what else must be true of right actions’ but modifies her stance by allowing that her ‘Coincident Reasons Thesis’ can ‘explain when wrong actions have (at least some) moral worth.’ (p.240). Herman (1993) advances the view that moral worth is ‘ubiquitous – governing all our actions.’ (p.17). 6 Some writers thus explicitly identify the concept of moral worth with the concept of (moral) creditworthiness, for example Lord (2017), Portmore (2022), Singh (2020), Way (2017). 7 Note that this schematic analysis leaves room for accounts that hold that (ii) is explained by (i) since X can be that Φ‐ing is [objectively/subjectively] right. 8 Though compare Markovits (2010), §1. 9 Cf. Baron (1984) but compare Korsgaard (2009), ch.1 who argues that Kant's view is more nuanced here as his concept of duty itself contains the purpose for which an act is done. 10 Though compare Singh (2020), pp.165‐7. 11 For an influential interpretation of Kant along these lines, see Herman (1993), ch.1. 12 Smith (1994), §3.5. 13 Howard (2021b) argues (on the basis of work by Janet Fodor) that this bipartite distinction is not exhaustive, and identifies a third reading, namely one that is “nonspecific” in the sense that an agent may desire X without either wanting a particular X, or wanting X via a particular description of X. 14 Smith (1994), p.75. My emphasis. 15 Ibid. 16 Smith (1994), p.76. 17 In the following I make the simplifying assumption without argument that right‐making features are normative reasons. 18 As Sharadin (2019) points out however, the view that right reasons are identical to right‐making reasons on such accounts is optional. Moreover, he argues that this ‘assumes that a particular kind of first‐order normative theory is correct (p.126). 19 Arpaly (2002, 2003). 20 Markovits (2010). 21 Markovits (2010), p.205. 22 Understanding overlap between motivating reason and moral normative reason looks relatively unproblematic in ideal cases. But there are more complex cases in which an agent may act on the right normative reason but is also motivated by other reason which are morally irrelevant, or worse. Right Reasons Accounts therefore need to unpack the concept of ‘overlap’ in greater detail. I discuss this further in §3.1. 23 For those who take something close to CRT as a necessary condition on moral worth, see Arpaly (2002, 2003), Arpaly and Schroeder (2013), Cunningham (2021), Fearnley (2022), Lockhart (2017), Markovits (2010, 2012), Massoud (2016), Way (2017). 24 Markovits (2010), p.229. 25 Such as Coates (2021), Lockhart (2017). 26 The example is from Coates (2021), p.393. 27 Ibid. 28 Lockhart (2017), p.617. 29 Arpaly and Schroeder (2013), p.164 argue that a constraint on acting for the relevant reason is that it be performed from a ‘good will’ from which ‘[t]he concepts deployed in grasping the correct normative moral theory are the concepts through which one must intrinsically desire the right or good in order to have a good will.’ Markovits argues for a more permissive account of the constraints on the content of one's motivating reason. See Markovits (2010), p.228, (2012) p.292. Johnson King (2022) argues that if someone who cares directly about rightness treats it as a fetish, then someone who cares directly about other, downstream “thick” moral features such as well‐being, should also be treated a fetishist (in this case, about well‐being). See Johnson King (2022), §3. 12 of 14 - METZ 30 See Markovits (2012), p.293. For further discussion of the gradability of moral worth see Fearnley (2022), §3, Massoud (2016), Nelkin (2016). 31 Portmore (2022), pp.277‐8. See also Singh (2020) pp.163‐4. 32 Portmore (2022), pp.277‐8. 33 Sliwa (2015), p.398. 34 Markovits (2010), p.210. 35 See Howard (2021a), p.306. An action fails to have moral worth due to sufficiency luck where the motivating reason of the agent “could easily” or in nearby possible worlds fail to be a sufficient moral reason. 36 See Isserow (2019). 37 See esp. Howard (2021a), §5, Isserow (2019), §§3‐4. 38 See Coates (2021), p.391. The idea behind appealing to agency is to think ‘not what your actual motives would have you do in counterfactual situations, but what motives you would have in those situations.’ 39 Johnson King (2020), p.201. 40 Lord (2017). 41 Cunningham (2021). 42 See Howard (2021a), §4. 43 A different tactic (adopted by Fearnley (2022)) is to identify those counterfactuals which are relevant to an assessment of moral worth, and those which are not, by providing a criterion for relevancy itself. She defines relevancy (or ‘normalcy’ in her terminology) in terms of those possible worlds that adhere to ‘statistical and prescriptive norms’ tied to the actual world. 44 Markovits (2010), p.219. 45 See Herman (1993), p.16, Sliwa (2015), p.395. 46 Sliwa (2015), p.404. 47 See Howard (2021a), pp.312‐3. 48 Another issue that has been much discussed in the literature is over whether and to what extent moral worth comes in degrees. However, I do not see this issue as central to the debate between Rightness Accounts and Right Reasons Accounts. See Fearnley (2022), Massoud (2016), Nelkin (2016) for discussion. 49 See Herman (1993), p.16 for defence of objective rightness in the context of a discussion of moral worth. See also Benson (1987), p.367. 50 See Sliwa (2015), §4. 51 Sliwa (2015), pp.400‐1. 52 Ibid. 53 Isserow (2019), p.258. 54 Singh makes this point in Singh (2020) p.167. 55 Johnson King (2020), p.201. 56 Ibid. 57 Lillehammer (1997), p.192. 58 Svavarsdóttir (1999), p.200. 59 Similar points are made by Carbonell (2013), §1, and Isserow (2020), p.543ff. 60 See Lerner (2018) for discussion of this case. For further arguments that de dicto moral motivation is unproblematic, or even rationally necessary, see Aboodi (2017), Carbonell (2013), Copp (1997), Lillehammer (1997), Olson (2002), Svavarsdóttir (1999). 61 Though compare Markovits (2010), p.218‐9. See Hills (2009) on how what she calls moral understanding can generate practical instrumental reasons not to defer to moral experts. See also Field (2022), Grant (2023). METZ - 13 of 14 62 Singh (2020), p.170. 63 Singh (2020), p.171. 64 Isserow (2020), p.570. 65 Portmore (2022) argues along these lines. 66 In particular, the objection from Lord (2017) and Cunningham (2021) mentioned towards the end of §3.2. 67 See the discussion in §4. 68 Portmore (2022). 69 Portmore (2022), pp.278‐9. See also Howard (2023) who stresses the telic aspect of moral worth. 70 Portmore (2022), p.290. 71 Arpaly (2002). 72 Markovits (2010). 73 Johnson King (2020). 74 Sliwa (2015). 75 Howard agrees, that ‘[t]his position is striking, given the past 50 years of debate on desires for rightness as such. That a certain desire could be thought by some as inconsistent with moral worth and by others as necessary for it is a powerful testament to how poorly that desire is understood.’ See Howard (2021a), p.303. 76 Portmore (2022), p.294. 77 In a similar vein, Howard (2023) argues that the very same fact can favour an action in different ways (morally, prudentially, etc). For example, the fact that donating money to charity X will improve the lives of needy children can be both a moral reason, and (given that you feel pleasure in giving to charity) a prudential reason. But, plausibly, only an act based on the former could have moral worth. In order to distinguish the contents of these reasons (moral, prudential, …) Howard appeals to goals. We can distinguish between a moral and a prudential reason by appealing to a distinctively moral or prudential goal. REFERENCES Aboodi, R. (2017). One Thought Too Few: Where De Dicto Moral Motivation is Necessary. Ethical Theory & Moral Practice, 20(2), 223–237. Arpaly, N. (2002). Moral Worth. Journal of Philosophy, 99(5), 223. https://doi.org/10.2307/3655647 Arpaly, N. (2003). Unprincipled virtue: an inquiry into moral agency. Oxford University Press. Arpaly, N., & Schroeder, T. (2013). In Praise of Desire. Oxford University Press. Baron, M. (1984). The Alleged Moral Repugnance of Acting from Duty. Journal of Philosophy, 81(4), 197–220. https://doi. org/10.2307/2026121 Benson, P. (1987). Moral Worth. Philosophical Studies, 51(3), 365–382. https://doi.org/10.1007/bf00354044 Carbonell, V. (2013). De Dicto Desires and Morality as Fetish. Philosophical Studies, 163(2), 459–477. https://doi.org/10. 1007/s11098‐011‐9825‐z Coates, A. (2021). Moral Worth and Accidentally Right Actions. Analysis, 81(3), 389–396. https://doi.org/10.1093/analys/ anab011 Copp, D. (1997). Belief, Reason, and Motivation: Michael Smith's “The Moral Problem”. Ethics, 108(1), 33–54. https://doi. org/10.1086/233787 Cunningham, J. J. (2021). Moral Worth and Knowing How to Respond to Reasons. Philosophy and Phenomenological Research, 105(2), 385–405. https://doi.org/10.1111/phpr.12825 Fearnley, L. (2022). Moral Worth, Right Reasons and Counterfactual Motives. Philosophical Studies, 179(9), 2869–2890. https://doi.org/10.1007/s11098‐022‐01805‐6 Field, C. (2022). At Least You Tried: The Value of De Dicto Concern to Do the Right Thing. Philosophical Studies, 179(9), 2707–2730. https://doi.org/10.1007/s11098‐022‐01791‐9 Grant, J. (2023). Moral Worth and Moral Belief. Ethics, 133(2), 216–230. https://doi.org/10.1086/722120 Herman, B. (1993). The Practice of Moral Judgement. Harvard University Press. Hills, A. (2009). Moral Testimony and Moral Epistemology. Ethics, 120(1), 94–127. https://doi.org/10.1086/648610 Howard, N. R. (2021a). One Desire Too Many. Philosophy and Phenomenological Research, 102(2), 302–317. https://doi.org/ 10.1111/phpr.12642 14 of 14 - METZ Howard, N. R. (2021b). Moral Fetishism and a Third Desire for What’s Right. Journal of Ethics and Social Philosophy, 20(3), 370–881. https://doi.org/10.26556/jesp.v20i3.1249 Howard, N. R. (2023). The Goals of Moral Worth. In R. Shafer‐Landau (Ed.), Oxford Studies in Metaethics (Vol. 16, pp. 157–182). Oxford University Press. Hurka, T. (2006). Virtuous act, virtuous dispositions. Analysis, 66(1), 69–76. https://doi.org/10.1093/analys/66.1.69 Isserow, J. (2019). Moral Worth and Doing the Right Thing by Accident. Australasian Journal of Philosophy, 97(2), 251–264. https://doi.org/10.1080/00048402.2018.1463270 Isserow, J. (2020). Moral Worth: Having It Both Ways. Journal of Philosophy, 117(10), 529–556. https://doi.org/10.5840/ jphil20201171035 Johnson King, Z. (2020). Accidentally Doing the Right Thing. Philosophy and Phenomenological Research, 100(1), 186–206. https://doi.org/10.1111/phpr.12535 Johnson King, Z. (2022). Deliberation and Moral Motivation. Oxford Studies in Metaethics, 17, 254–279. Kant, I. (1998/1785). Groundwork of the Metaphysics of Morals. Cambridge University Press. Korsgaard, C. M. (2009). Self‐Constitution: Agency, Identity, and Integrity. Oxford University Press. Lerner, A. (2018). The Puzzle of Pure Moral Motivation. Oxford Studies in Metaethics, 13, 123–144. https://doi.org/10.1093/ oso/9780198823841.003.0006 Lillehammer, H. (1997). Smith on Moral Fetishism. Analysis, 57(3), 187–195. https://doi.org/10.1093/analys/57.3.187 Lockhart, J. R. (2017). Moral Worth and Moral Hobbies. Ergo: An Open Access Journal of Philosophy, 4(20201214). https://doi. org/10.3998/ergo.12405314.0004.021 Lord, E. (2017). On The Intellectual Conditions for Responsibility: Acting for the Right Reasons, Conceptualization, and Credit. Philosophy and Phenomenological Research, 95(2), 436–464. https://doi.org/10.1111/phpr.12306 Markovits, J. (2010). Acting for the Right Reasons. Philosophical Review, 119(2), 201–242. https://doi.org/10.1215/ 00318108‐2009‐037 Markovits, J. (2012). Saints, Heroes, Sages, and Villains. Philosophical Studies, 158(2), 289–311. https://doi.org/10.1007/ s11098‐012‐9883‐x Massoud, A. (2016). Moral Worth and Supererogation. Ethics, 126(3), 690–710. https://doi.org/10.1086/684713 Nelkin, D. K. (2016). Difficulty and Degrees of Moral Praiseworthiness and Blameworthiness. Noûs, 50(2), 356–378. https:// doi.org/10.1111/nous.12079 Olson, J. (2002). Are Desires De Dicto Fetishistic? Inquiry: An Interdisciplinary Journal of Philosophy, 45(1), 89–96. https://doi. org/10.1080/002017402753556634 Portmore, D. W. (2022). Moral Worth and Our Ultimate Moral Concerns. In M. Timmons (Ed.), Oxford Studies in Normative Ethics (Vol. 12, pp. 276–298). Oxford University Press. Sharadin, N. (2019). Consequentialism and Moral Worth. Utilitas, 31(2), 117–136. https://doi.org/10.1017/s0953 820818000146 Singh, K. (2020). Moral Worth, Credit, and Non‐Accidentality. In M. Timmons (Ed.), Oxford Studies in Normative Ethics (Vol. 10, pp. 156–181). Oxford University Press. Sliwa, P. (2015). Moral Worth and Moral Knowledge. Philosophy and Phenomenological Research, 93(2), 393–418. https://doi. org/10.1111/phpr.12195 Smith, M. (1994). The Moral Problem. Blackwell. Svavarsdóttir, S. (1999). Moral Cognitivism and Motivation. Philosophical Review, 108(2), 161–219. https://doi.org/10.2307/ 2998300 Way, J. (2017). Creditworthiness and Matching Principles. In M. Timmons (Ed.), Oxford Studies in Normative Ethics (Vol. 7, pp. 207–228). Oxford University Press. A U TH O R BI O GR A PH Y Euan K. Metz works mainly on questions concerning the nature of fundamental normative concepts, their structure, interrelations, and deployment. He is currently an Assistant Professor at the American University in Cairo. How to cite this article: Metz, E. K. (2024). Moral worth. Philosophy Compass, e12982. https://doi.org/10. 1111/phc3.12982