1. Introduction
Doing the right thing can be difficult. Doing the morally worthy thing can be even harder. What it means for an action to have moral worth, and what is required for an action to have moral worth, is the subject of continued controversy. Historically, two positions on moral worth have taken centre stage. According to one camp, an agent performs a morally worthy action if and only if they do it because the action is morally right (Herman 1981; Jeske 1998; Sliwa 2016). According to the other camp, a morally worthy action is that which an agent performs because of features that make the action right (Arpaly 2003; Arpaly & Schroeder 2014; Markovits 2010).1 These two views go by different names in the literature. The former view has been called the Kantian View (Isserow 2019; Johnson King 2020), the Conservative View (Howard 2021), and the Rightness Itself View (Singh 2020). The latter view has been called the New View (Johnson King 2020), the Right-Making Features View (Singh 2020), and the Responding View (Isserow 2019).
What sets these views apart is the kind of motivation each takes to be essential for an action’s moral worth. When an agent does the right thing because of the action’s moral rightness, she is motivated by rightness itself to perform that action. When an agent does the right thing because of a particular aspect of the action that makes it right, she is motivated by a right-making feature to perform that action. We refer to motivations from rightness itself as higher-order motivations and motivations from right-making features as first-order motivations, as right-making features are ground-level facts that make an action right, and rightness itself is a fact that rests on those ground-level facts.2 Higher-order theorists, like Paulina Sliwa (2016) and Zoë Johnson King (2020), argue that higher-order motivations are necessary and sufficient for moral worth, while first-order motivations are largely irrelevant. By contrast, first-order theorists, like Nomy Arpaly (2003) and Julia Markovits (2010), argue that first-order motivations are necessary and sufficient for moral worth, while higher-order motivations are irrelevant. In an important sense, higher-order and first-order views of moral worth are diametrically opposed to one another. The motivations that one camp argues are necessary and sufficient for moral worth are the very motivations that the other camp argues are irrelevant. Nevertheless, proponents of these opposing views share something important in common. With the exception of Arpaly (2003) and Arpaly and Schroeder (2014), they theorise about the nature of moral worth by focusing almost exclusively on the moral worth of, and praiseworthiness or creditworthiness for, right actions. But each of these properties has a negatively valenced counterpart that attaches to wrong actions. Just as agents can deserve praise or credit for doing the right thing, they can deserve blame or discredit for doing the wrong thing. While the former actions have moral worth, the latter actions have what we will call moral counterworth. Given this apparent symmetry, examining the moral counterworth of actions could shed light on the nature of moral worth.
We proceed as follows. In §2 we present two approaches to theorising about moral worth and reflect on the considerations in favour of each. In §3 we present cases that indicate that both higher-order and first-order motivations contribute to the moral counterworth of wrong actions. We then argue that, given the relevant similarities between moral worth and moral counterworth, these considerations undermine both higher-order and first-order views, and instead support Dual Pertinence: First-order motivations and higher-order motivations can each affect the moral worth of an agent’s action. In §4 and §5 we demonstrate the resilience of this view. First, we argue that higher-order and first-order theorists cannot easily explain away the results of our cases because, surprisingly, there are reasons internal to both positions that support Dual Pertinence. Consequently, higher-order and first-order theories have much more in common with one another than previously thought. Next, we defend the claim that reflecting on moral counterworth can shed light on the nature of moral worth. We do so by showing that there is a strong prima facie case for there being symmetry between the two, and that neither theoretical nor empirical work on the differences between praiseworthiness and blameworthiness undermines this claim. Lastly, in §6, we conclude by explaining how reflection on moral counterworth can serve to support recently developed accounts of moral worth that make room for the relevance of both higher-order and first-order motivations.
2. Two Approaches to Moral Worth
Accounts of moral worth aim to determine the kinds of motivations that elevate merely right actions—actions that happen to conform to the correct normative theory—to morally worthy actions—actions that merit praise or credit.3 Morally worthy actions aren’t simply those that elicit praise from others. Rather, agents perform morally worthy actions when they are praiseworthy for doing the right thing. And this, according to moral worth theorists, depends on agents’ motives for action. On one approach, morally worthy actions are those that are performed because they are morally right. On another, morally worthy actions are those that are performed for the reasons that make them morally right. In this section, we first present each approach and highlight the features that set them apart. Then, we turn our attention to what these approaches have in common.
2.1. Higher-Order Views
Several philosophers take higher-order motivations to be critical for morally worthy actions. Perhaps the most notable defender of a higher-order view is Kant, who argues that morally worthy actions must be performed “for the sake of the law” (1785/1998: 3) and “simply from duty” (1785/1998: 12). More recent defenders of higher-order views include Sliwa and Johnson King:
Sliwa: A morally right action has moral worth if and only if it is motivated by concern for doing what’s right (conative requirement) and by knowledge that it is the right thing to do (knowledge requirement). (2016: 394)
Johnson King: An act has moral worth just in case it is an instance of someone’s deliberately doing the right thing. (2020: 201)4
There are subtle but important differences between these views. For example, while Sliwa argues that one must know that one is doing the right thing in order for one’s action to have moral worth, Johnson King argues that one simply needs to deliberately do the right thing, which does not require knowledge that what one is doing is right. For Johnson King, an agent deliberately does the right thing so long as they try to perform the action that they believe has the property of being morally right, even if they don’t know that the action possesses this property (2020: 203). Nevertheless, these authors follow the Kantian tradition by arguing that a higher-order motivation to do the right thing is necessary, and, when paired with the relevant beliefs or knowledge, sufficient for moral worth.
One key impetus for higher-order theories comes from considerations of accidentality. An assumption shared by most everyone in the moral worth debate is that accidents are not morally worthy. Imagine a professor who gives her students generous extensions on assignments during a global pandemic, not because it is the right thing to do, but because she wants to receive positive student evaluations at the end of the term. Most would agree that the professor’s action does not possess moral worth. Though the professor does the right thing by giving her students extensions, she does the right thing by accident. Her selfish motivations would have led her to do the wrong thing (inflate grades, cancel class unnecessarily, etc.) in a wide range of circumstances, and it just so happens that in this case her selfish motivations result in doing the right thing. However, if the professor gives her students generous extensions during a global pandemic because she is motivated to do what’s right, and she knows that providing extensions is the right thing to do (or she deliberately tries to do the right thing by giving her students extensions), then there is no sense in which the professor does the right thing by accident. In this case, the professor is moved by a higher-order motivation to do what’s right, and this guarantees that the rightness of her action is not accidental.5 Thus, higher-order theorists argue that acting on a higher-order motivation to do the right thing, when paired with the relevant beliefs or knowledge, is sufficient for performing a right action with moral worth.
Higher-order theorists also take higher-order motivations to be necessary for performing morally worthy actions. Imagine a professor who gives her students generous extensions on assignments during a global pandemic because she cares about their well-being, and despite not knowing that her action is morally right. In fact, she may think that her action is morally wrong—perhaps because it violates certain university policies, which the professor (mistakenly) believes she ought never violate. Here too the professor’s motivations would have led her to do the wrong thing in a wide range of circumstances. As Barbara Herman (1981) argues, even caring for someone’s well-being can lead us to act wrongly, like when lending a helping hand to an art thief carrying a heavy painting. So, in this case, Sliwa and Johnson King would argue that the professor’s action does not possess moral worth. Since the professor does not think she is doing the right thing in offering extensions to her students, she is not motivated to perform this action because it is right. And, according to higher-order theorists, this renders her right action an accident, and one that cannot possess moral worth.
In arguing that higher-order motivations are necessary and sufficient for moral worth, higher-order theorists take first-order motivations—i.e., motivations from the reasons that make an action right—to be largely irrelevant to moral worth. Since they consider higher-order motivations sufficient for moral worth, first-order motivations cannot be necessary. The professor who gives her students extensions because it is the right thing to do need not also be motivated to perform the action because it promotes her students’ well-being for her action to be praiseworthy.6 And, if higher-order motivations are necessary for moral worth, first-order motivations cannot be sufficient. The professor who thinks that it is wrong to give students extensions, but does so because she wishes to promote their well-being, does not perform a morally worthy action despite acting from the right first-order motivations.
In addition to arguing that first-order motivations are neither necessary nor sufficient for moral worth, many higher-order theorists go further and argue that first-order motivations can have no impact at all on the moral worth of actions. While they grant that being concerned with the right-making features of actions can be praiseworthy in some sense, they deny that acting from these motivations impacts the degree to which agents are praiseworthy for performing right actions. Johnson King argues:
It is true that an agent who acts rightly and has stronger moral concern will be more praiseworthy overall than one who acts rightly but has weaker moral concern (under otherwise identical circumstances). But the first agent is not more praiseworthy for acting rightly than the other. She is more praiseworthy for her stronger moral concern: it is her character that is more praiseworthy. (2020: 200–201)
So, while first-order motivations may be relevant to agents’ characters, which can themselves be praiseworthy, first-order motivations do not affect the degree to which agents are praiseworthy for acting rightly. Thus, on higher-order views, first-order motivations play no direct role in determining the moral worth of right actions.7
2.2. First-Order Views
In contrast, some philosophers take first-order motivations to be essential to morally worthy actions. Two notable philosophers who defend first-order views of moral worth are Arpaly and Markovits:
Arpaly: For an agent to be morally praiseworthy for doing the right thing is for her to have done the right thing for the relevant reasons—that is, in response to the features that make it right . . . (2003: 72).
Markovits: 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. (2010: 205)
Like Sliwa’s and Johnson King’s views, there are interesting differences that set these accounts apart. However, these views are alike in that they take first-order motivations to be necessary and sufficient for morally worthy actions.
First-order theorists tend to rely on what Arpaly calls cases of “inadvertent virtue” (2003: 10) to defend the sufficiency claim. These are cases in which individuals perform a morally right act because of its right-making features, but do not take themselves to be doing the right thing (and even take themselves to be doing the wrong thing). Take the professor who offers extensions to her students because she cares about their well-being, but believes that what she is doing is morally wrong. In stark contrast with higher-order theorists, first-order theorists would argue that the professor’s action possesses moral worth and she is praiseworthy for performing it. According to first-order theorists, performing the right action for the reasons that make it right ensures that the action is not an accident. As Markovits argues, “Actions motivated by right-making reasons . . . are not merely contingently or accidentally right. If I am motivated by right-making reasons, it is no coincidence that my motive issues in the right action” (2010: 211). In arguing that first-order motivations are sufficient for right actions to possess moral worth, first-order theorists reject the claim that higher-order motivations are necessary. The professor who takes herself to be doing the wrong thing by giving her students extensions during a pandemic is certainly not acting on a higher-order motivation to do the right thing. Nevertheless, first-order theorists like Arpaly and Markovits argue that such a right action is non-accidental and has moral worth.
First-order theorists also take first-order motivations to be necessary for moral worth. In doing so, they deny that higher-order motivations can be sufficient. Several first-order theorists follow in the footsteps of Michael Smith and argue that being motivated to perform the right action purely because it is the right thing to do reveals something defective about one’s moral character:
Good people care non-derivatively about honesty, the weal and woe of their children and friends, the well-being of their fellows, people getting what they deserve, justice, equality, and the like, not just one thing: doing what they believe to be right, where this is read de dicto and not de re. Indeed, commonsense tells us that being so motivated is a fetish or moral vice, not the one and only moral virtue. (1994: 75)
Brian Weatherson, a sympathiser with first-order views, distils the following claim from this passage: “The good, as such, should not be one’s only motivation” (2019: 46). Not only do first-order theorists take higher-motivations to be insufficient for moral worth, they take acting solely from these motivations to be morally criticisable.
In addition to arguing that higher-order motivations are neither necessary nor sufficient for moral worth, first-order theorists argue that they are inert when it comes to the moral worth of actions. This is because first-order theorists typically take the degree of moral worth to be determined exclusively by first-order motivations. Arpaly, for example, argues:
. . . an agent is more praiseworthy, other things being equal, the stronger the moral concern that has led to her action. . . . Moral concern is to be understood as concern for what is in fact morally relevant and not as concern for what the agent takes to be morality. (2003: 84)
Arpaly takes moral worth to be determined by agents’ concern for the right-making features of actions and not their concern for morality as such. In this way, Arpaly explicitly rules out the relevance of higher-order motivations to the moral worth of actions. Markovits also proposes an account of moral worth that comes in degrees, when arguing that “. . . right actions have moral worth to the degree that the noninstrumental motivations for their performance coincide with noninstrumental moral justifications for their performance” (2010: 238). Because the fact that an action is morally right is not a right-making feature of the action, acting on such a reason will not coincide with the moral justifications for the performance of the act. So, higher-order motivations are rendered irrelevant on Markovits’s account too.
In sum, first-order theorists take higher-order motivations to be irrelevant to moral worth—such motivations are neither necessary nor sufficient for moral worth, and they do not contribute to moral worth even in conjunction with first-order motivations.
2.3. Common Ground
Higher-order and first-order views of moral worth are very much at odds with one another. The motivations that higher-order views argue are necessary and sufficient for moral worth are the very motivations that first-order views argue are irrelevant, and vice versa. Given this sharp divide, it is perhaps no surprise that the proponents of these views have reached a dialectical impasse. Higher-order theorists are unmoved by the cases that first-order theorists rely on to support their views, just as first-order theorists are unmoved by higher-order theorists’ supporting cases.
Nevertheless, these different approaches to understanding the nature of moral worth have something important in common. The discussion from both camps has focused almost exclusively on right actions. A cursory examination of the literature reveals a multitude of examples, such as showing up to teach a valuable class (Henson 1979: 43), selling goods at a fair price (Herman 1981: 368), keeping a promise (Johnson King 2020: 190), saving drowning strangers (Markovits 2010: 210), treating a snake bite and telling a hard truth (Singh 2020: 162–63), giving a friend a ride to work and giving money to charity (Sliwa 2016: 401–2), helping an enslaved individual seek freedom (just about everyone) and more. This attention to praiseworthy actions is likely due to the fact that Kant introduced the term ‘moral worth,’ or “moralischen Werth,” in a discussion of the morally good:
For, in the case of what is to be morally good it is not enough that it [i.e., the action] conform with the moral law but it must also be done for the sake of the law; without this, that conformity is only very contingent and precarious, since a ground that is not moral will indeed now and then produce actions in conformity with the law, but it will also often produce actions contrary to law. (1785/1998: 3–4)
But notice that much of what is expressed here is also true of morally wrong actions. Sometimes, wrong actions possess what we call moral counterworth. Morally counterworthy actions are wrong actions for which agents deserve blame or discredit in virtue of doing the wrong thing.8 Not all actions that violate the moral law possess moral counterworth—that is, agents are not always blameworthy for doing the wrong thing.9 Some are coerced, some arise from non-culpable ignorance, and some are performed by individuals who lack the capacity to understand the moral law or its requirements. In such cases, those who perform actions that violate the moral law are not blameworthy for doing wrong and we do not take their actions to reflect poorly on them. These actions therefore fail to possess moral counterworth.
In this way, there is an important symmetry between moral worth and moral counterworth: morally counterworthy actions cannot be accidents, just as morally worthy actions cannot be accidents. Interestingly, agents can non-accidentally do the wrong thing even if they do not deliberately, intentionally, or even knowingly do the wrong thing. Take, for example, wrong actions performed out of culpable negligence or ignorance. Wrongs of this kind are not accidental. It is no accident that one hits a pedestrian if one knowingly drives around with faulty brakes, nor is it an accident that one votes for a dangerous political figure if one never bothers researching the candidate.10 The moral wrongness of these actions is a nonaccidental effect of the agents’ lack of moral concern, just as the rightness of morally worthy actions is “the nonaccidental effect of the agent’s concern” (Herman 1981: 366). And given these agents’ motivational profiles, they are very likely to perform relevantly similar wrongs in relevantly similar circumstances. Thus, like the moral worth of right actions, the moral counterworth of wrong actions is partially dependent on agents’ motivations. To assess the moral counterworth of an agent’s action, we need to know why they performed it.
Given these parallel features of the moral worth of right actions and the moral counterworth of wrong actions, reflecting on moral counterworth could shed light on the nature of moral worth and allow us to push the debate forward. In fact, judgments about moral counterworth may be more reliable than those about moral worth, as Susan Wolf suggests:
To blame someone undeservedly is, in any case, to do him an injustice. Whereas to praise someone undeservedly is apt to be just a harmless mistake. For this reason, I think, our intuitions about praise are weaker and less developed than our intuitions about blame. (1980: 155)
Perhaps Wolf’s observation helps explain why neither camp’s praiseworthiness-focused thought experiments have succeeded in moving their opposition. In any event, we believe that much can be gained by exploring the nature of moral counterworth.
3. Moral Counterworth
Higher-order and first-order theorists each argue that the motivations their opposition takes to be necessary and sufficient for moral worth are irrelevant. However, reflection on cases in which agents perform blameworthy actions suggests that both higher-order and first-order motivations can impact moral counterworth. And because moral counterworth is simply the negative analogue of moral worth, this indicates that both kinds of motivations can impact moral worth as well, contrary to the claims of first-order and higher-order theorists alike.
3.1. Cruel Intentions
Higher-order theorists are committed to higher-order motivations being necessary and sufficient for moral worth. In defending this position, they also argue that first-order motivations are irrelevant to moral worth. While this claim is defensible (though not beyond reproach) when it comes to the moral worth of right actions, it becomes extremely implausible if applied to the moral counterworth of wrong actions. Compare two agents who act on the same higher-order motivations, but very different first-order motivations, when doing the wrong thing:
Selfish Gossip: Cecile learns of a good friend’s embarrassing secret. She knows that it would be wrong to reveal the secret, and does not wish to do wrong. While at a party, an opportunity to be the centre of attention arises. Wanting to be popular, Cecile succumbs to temptation and reveals her friend’s secret.
Cruel Gossip: Sebastian learns of a good friend’s embarrassing secret. He knows that it would be wrong to reveal the secret, and does not wish to do wrong. While at a party, an opportunity arises to humiliate his friend by revealing the secret. Wanting to embarrass his friend, Sebastian succumbs to temptation and reveals his friend’s secret.
In this pair of cases, Cecile and Sebastian possess the same higher-order motivation to stay mum: both are motivated to keep their friend’s secret because they know that revealing it is wrong. However, Cecile’s and Sebastian’s higher-order motivations are outweighed by their respective first-order motivations to reveal the secret: Cecile’s desire to be the centre of attention, and Sebastian’s desire to humiliate his friend. We take it to be uncontroversial that Cecile and Sebastian are blameworthy for revealing the secret—their actions each possess moral counterworth. But it also seems clear that these agents are not equally blameworthy. Sebastian is (much) more blameworthy for revealing the secret than Cecile, and his action possesses more moral counterworth than Cecile’s action.
What could explain the difference in blameworthiness? Cecile and Sebastian share the same higher-order motivation. They also each decide to reveal a friend’s secret, and in so doing become the centre of attention and embarrass their friend. The only difference between Cecile and Sebastian lies in their first-order motivations. Cecile’s motivation to reveal her friend’s secret is morally neutral—a desire to be popular is not in itself morally objectionable. Yet, Cecile behaves selfishly when she ignores the moral reasons to keep the secret and acts on the neutral reason to be popular. In contrast, Sebastian’s motivation to tell the secret is itself morally objectionable. He desires to harm his friend by embarrassing them, which is cruel. We submit that Sebastian’s cruel first-order motivation renders him more blameworthy than Cecile. If this is right, then first-order motivations are not irrelevant to moral counterworth—for they can directly contribute to the degree to which an agent is blameworthy for doing the wrong thing.
The idea that agents’ first-order motivations are relevant to blameworthiness has not gone unnoticed in the literature. In fact, Arpaly explicitly argues that first-order motivations affect the degree to which agents are blameworthy. Arpaly is unique among moral worth theorists in that she develops an account of both moral worth and its negative parallel,11 which she takes to be synonymous with praiseworthiness and blameworthiness, respectively. She argues:
Other things being equal, a person is more blameworthy for a given wrong if she acts out of ill will (from sinister reasons) than she would be if she were to act out of a lack of good will (out of neutral reasons, while ignoring moral reasons to the contrary). (2003: 80)12
Notice that Arpaly’s account nicely captures the difference between Cecile and Sebastian in the cases above. Cecile reveals her friend’s secret in order to be the centre of attention. The desire for popularity is a neutral reason—it doesn’t necessarily conflict with morality (Arpaly 2003: 79). Yet, Cecile acts on this reason while discounting the moral reasons to keep the secret, namely the fact that sharing the secret will humiliate her friend. Thus, Cecile acts out of a lack of good will towards her friend and is blameworthy. However, Sebastian reveals his friend’s secret in order to humiliate them, which is a sinister reason and one that is in conflict with morality. This indicates that Sebastian acts out of ill will, which renders him (highly) blameworthy, and certainly more blameworthy than Cecile.
Cases of cruel intentions reveal that first-order motivations can directly affect moral counterworth. While the above considerations are compatible with first-order views of moral counterworth (particularly Arpaly’s account of the negative parallel to moral worth), further reflection on cases of wrong actions raises challenges for these accounts as well.
3.2. Evil Deeds
In addition to arguing that first-order motivations are necessary and sufficient for moral worth, first-order theorists also contend that higher-order motivations are irrelevant. There is some intuitive (though certainly defeasible) support for this claim when it comes to the moral worth of right actions. But when applied to the moral counterworth of wrong actions, the claim quickly loses its appeal. Compare Cecile to an agent who acts on the same first-order motivation, but a very different higher-order motivation, when doing the wrong thing:
Selfish Gossip: Cecile learns of a good friend’s embarrassing secret. She knows that it would be wrong to reveal the secret, and does not wish to do wrong. While at a party, an opportunity to be the centre of attention arises. Wanting to be popular, Cecile succumbs to temptation and reveals her friend’s secret.
Evil Gossip: Isabelle learns of a good friend’s embarrassing secret. She knows that it would be wrong to reveal the secret, and she wishes to do wrong. While at a party, an opportunity to be the centre of attention arises. Wanting both to be popular and to do wrong, Isabelle reveals her friend’s secret.
Cecile and Isabelle are similar in many respects. Both agents knowingly perform the wrong action of revealing a friend’s secret. The two are also motivated to reveal the secret in order to gain popularity. In this way, they possess the same first-order motivation, and act out of a lack of good will towards their friend (to use Arpaly’s terminology). Yet despite these similarities, we submit that these agents are not equally blameworthy. Isabelle is (much) more blameworthy than Cecile.
The relevant difference between Cecile and Isabelle lies in their higher-order motivations. Cecile possesses a higher-order motivation not to reveal her friend’s secret—she knows that doing so is wrong and does not want to do the wrong thing. However, her first-order motivation to be the centre of attention overpowers this higher-order motivation and she chooses to reveal the secret. In contrast, Isabelle possesses a higher-order motivation to reveal the secret—she wants to reveal the secret because doing so is wrong.13 We submit that Isabelle’s motivation to do wrong renders her more blameworthy for telling her friend’s secret than Cecile. And if we are right that Isabelle’s motivation to do wrong enhances the degree to which she is blameworthy for doing wrong, then higher-order motivations are not irrelevant to moral counterworth.14
3.3. Dual Pertinence
Above, we argued that first-order and higher-order motivations are both relevant to moral counterworth. And moral counterworth is simply the negative analogue of moral worth. This provides good reason to think that the kinds of motivations that impact counterworth can impact moral worth as well, which leads to the following conclusion:
Dual Pertinence: First-order motivations and higher-order motivations can each affect the moral worth of an agent’s action.
Importantly, we do not claim that all higher-order and first-order motivations affect moral worth and counterworth. It may very well be that only some subset of these motivations is relevant. Similarly, we do not claim that these motivations always affect moral worth and counterworth. It’s possible that these motivations impact the moral worth or counterworth of only certain actions and/or that they are only efficacious in particular circumstances.15 We are also not making a claim about the degree to which these motivations affect moral worth and counterworth. While it’s plausible that some motivations can affect moral worth or counterworth to a greater extent than others,16 we remain neutral on this point. We are not, for example, claiming that evil higher-order motivations increase moral counterworth more than cruel first-order motivations do, or vice versa. Finally, we also remain neutral as to how motivations come to impact the moral worth or counterworth of actions. It’s possible that motivations have an additive or subtractive effect on moral worth or counterworth, but it could also be the case that these motivations have a multiplicative effect. Our claim is simply that both first-order and higher-order motivations can affect the moral worth and counterworth of actions.
As it stands, higher-order and first-order views of moral worth are incompatible with Dual Pertinence, since each view takes only one kind of motivation to be relevant to the moral worth of actions. However, there are various ways higher-order and first-order theorists could resist Dual Pertinence. Our defence of Dual Pertinence relies on two claims: (I) first-order and higher-order motivations can each affect moral counterworth, and (II) moral counterworth and moral worth are relevantly similar, such that the kinds of motivations that affect the former can also affect the latter. Our defence of (I) relies on what we consider to be the clear intuitive judgment that agents who act solely from selfish intentions can be less blameworthy than their cruel and evil counterparts. Our defence of (II) rests on the fact that moral counterworth is the negative analogue of moral worth, which gives us reason to think that the motivations that affect the former can affect the latter. Our opponents could resist (I) by simply denying the intuition that agents who act from selfish intentions can be less blameworthy than their cruel or evil counterparts. Alternatively, they could attempt to explain away the intuition by arguing that the fact that the purely selfish agent is less blameworthy does not indicate that their action possesses less moral counterworth. Additionally, our opponents could resist (II) by attempting to contain the intuition to moral counterworth and reject its relevance to moral worth. Although we will not directly address the outright denial of the core intuitions that our account rests on, we will address the attempts to explain the intuition away and to contain it. As we do so, we appeal to theoretical considerations that we believe undermine the blatant denial of the intuition.
4. Evil Deeds vs. Evil Agents
An agent who performs an action with moral worth is praiseworthy for doing the right thing. But agents can be praiseworthy for things beyond performing morally right acts. Indeed, higher-order and first-order theorists alike (e.g., Arpaly 2003; Johnson King 2020) distinguish between being praiseworthy for possessing certain character traits and being praiseworthy for performing morally right actions. In fact, these theorists often rely on this distinction to explain away the intuitive appeal of their opponents’ motivating cases. For example, here is how Johnson King accounts for the often-discussed case of Huckleberry Finn:
Huckleberry Finn has a praiseworthy character trait: he cares about Jim. And this leads him to perform an act of a good type: it is morally right. But it does not follow that Huck is praiseworthy for performing an act of this type. And, in fact, he is not praiseworthy for performing an act of this type. Huck accidentally does the right thing, and we are not praiseworthy for that which we do accidentally. (2020: 200)
Here, Johnson King grants that first-order motivations are relevant to a particular kind of praiseworthiness, namely praiseworthiness for character traits, while maintaining that such motivations are irrelevant to the kind of praiseworthiness that is related to moral worth.17
Likewise, one could distinguish between blameworthiness for character traits and blameworthiness for performing wrong actions. Higher-order and first-order theorists could respond to our cases of Selfish, Cruel, and Evil Gossip by claiming that Sebastian or Isabelle is more blameworthy than Cecile, but not because their actions possess more moral counterworth. Rather, Sebastian and/or Isabelle could be more blameworthy than Cecile because they possess worse character traits: Sebastian is cruel and Isabelle is evil, while Cecile is merely selfish. If this is right, then our cases do not illustrate that higher-order and first-order motivations are both relevant to the moral counterworth of actions, and cannot be used to support Dual Pertinence.
We agree that there is a valuable distinction to draw between blameworthiness for character traits and blameworthiness for performing wrong actions. However, we do not think it is plausible that our judgments about the relative blameworthiness of the agents featured in Selfish, Cruel, and Evil Gossip track blameworthiness for particular character traits. There is not enough information in any of these cases to make a judgment about what kinds of character traits the agents possess. And, even if it were stipulated that the agents possessed the same moral track records and characters at the moment of their decision to reveal the secret, our judgments about their relative blameworthiness would remain the same.
Still, one might argue that the agents’ motivations themselves manifest different character traits that are better or worse. Or, one could contend that certain motivations make us blameworthy all on their own, purely for having them. For example, a higher-order theorist could argue that Sebastian is more blameworthy than Cecile not because his action has more moral counterworth, but because his cruel first-order motivation expresses a worse character trait or is itself more blameworthy. Similarly, the first-order theorist could argue that Isabelle and Cecile perform actions with equal moral counterworth, but Isabelle is more blameworthy because her higher-order motivation to do wrong reflects a worse character trait or is itself more blameworthy.
While such responses are possible, we do not think they are particularly promising. Notice that each camp can utilise this strategy to address cases that challenge their respective views. The debate would then devolve into a standoff over which motivations manifest character traits, which motivations are blameworthy in and of themselves, and which motivations are relevant to the moral counterworth of actions. This would be an unfortunate development, since it would make for a purely lateral move in the dialectical gridlock.
Nevertheless, even if attempting to debunk the blameworthiness judgments about Selfish, Cruel, and Evil Gossip doesn’t give either the higher-order or first-order theorist an edge over the other, it puts the defender of Dual Pertinence in an uncomfortable position. This kind of debunking explanation is very difficult to rule out, so it would be unwise to rely solely on the judgments elicited by Selfish, Cruel, and Evil Gossip to defend Dual Pertinence. Luckily, there is independent reason to believe Dual Pertinence. In fact, we can derive reasons in support of this thesis from both first-order and higher-order theories. Therefore, by these theorists’ own lights, they ought to reject the character-based explanation of the Selfish, Cruel, and Evil Gossip cases and accept Dual Pertinence.
4.1. Quality of Will and Higher-Order Motivations
Interestingly, Arpaly’s first-order account can offer an explanation of how certain higher-order motivations can contribute to moral counterworth. Recall that Arpaly analyses moral worth in terms of good will, and the negative parallel to moral worth in terms of deficiency of good will and the presence of ill will:
If good will—the motive from which praiseworthy actions stem—is responsiveness to moral reasons, deficiency in good will is insufficient responsiveness to moral reasons, obliviousness or indifference to morally relevant factors, and ill will is responsiveness to sinister reasons—reasons for which it is never moral to act, reasons that, in their essence, conflict with morality. (2003: 79)
Given this characterisation of ill will, certain higher-order motivations would seem to count as being responsive to sinister reasons. Take the case of Evil Gossip in which Isabelle reveals her friend’s secret because it is wrong to do so. The fact that an action is wrong, while not a wrong-making feature of the action, is surely a sinister reason to perform it. Such a reason is necessarily at odds with morality—an action’s being wrong could never serve as a moral reason to perform it.18 So, one could argue that Isabelle’s decision to reveal her friend’s secret because it is the wrong thing to do demonstrates ill will, which renders her blameworthy by Arpaly’s own lights.
The fact that we can find an explanation for the relevance of higher-order motivations to moral counterworth within a first-order account is surely surprising. And, in the sentence that follows the passage quoted above, Arpaly argues:
. . . the person who is deficient in good will acts without regard for the wrong-making features of his action, while the person who has ill will performs his action exactly because of its wrong-making features. (2003: 79)
Here Arpaly equates sinister reasons with wrong-making features. On this characterisation, an action’s being wrong could never count as a sinister reason. This is because an action’s being wrong is not a wrong-making feature of the action, just as an action’s being right is not a right-making feature of the action. On this statement of Arpaly’s view, higher-order motivations are once again rendered irrelevant to an agent’s quality of will and to the moral counterworth of her actions.
What’s gone wrong? How could it be that one characterisation of a view makes room for higher-order motivations while the very next sentence renders them irrelevant? Arpaly assumes that the only reasons at odds with morality are those that make actions wrong. But as the case of Evil Gossip illustrates, this is not correct. The fact that an action is wrong, though it is not a wrong-making feature of the act, is nevertheless a sinister reason to perform it. Acting on such a reason is clearly an expression of ill will. So, given the very machinery of Arpaly’s account, certain higher-order motivations can be shown to be relevant to the moral counterworth of actions. Thus, reasons internal to this first-order view support the claim that both first-order and higher-order motivations can affect moral counterworth. Given the relevant similarities between moral counterworth and moral worth, these same reasons also support Dual Pertinence.
4.2. Deliberateness and First-Order Motivations
Johnson King’s higher-order account of moral worth also ends up providing reasons to accept Dual Pertinence. On Johnson King’s view, an agent performs an act with moral worth if and only if she deliberately does the right thing (2020: 202). While Johnson King is clear that deliberately doing the right thing need not involve knowledge that one is performing the right action (2020: 203), she is less clear on what precisely it means to deliberately do right. And while she takes only higher-order motivations to be relevant to deliberateness, there is good reason to think that first-order motivations impact how deliberately an agent acts.19 Johnson King suggests the following understanding of deliberate action:
If someone As deliberately, she wants to A and succeeds in A-ing by exercising effort and skill; this differentiates A-ing deliberately from A-ing accidentally, with mere foresight, or as part of a deviant causal chain. (2020: 203, fn. 9)
On this view, if an agent acts deliberately and non-accidentally, she acts with sufficient effort and skill.20 Indeed, as Johnson King notes, this helps ensure the kind of counterfactual success characteristic of non-accidentality. However, this approach to deliberateness points exactly in the direction of Dual Pertinence. Considerations of counterfactual success, along with considerations of effort and skill, appear to bolster the relevance of first-order motivations to an action’s moral worth. This is because the extent to which agents are sensitive and responsive to the right-making and wrong-making features of actions impacts the effort and skill with which they do the right thing.
We see this phenomenon in many domains outside morality. Compare a professional pâtissier to a novice baker, for instance. Though both wish to produce delicious baked goods, the pâtissier puts more effort and skill into their baking than the novice. But what sets the pâtissier apart from her novice counterpart? One important difference is that the pâtissier is much more sensitive to, and more likely to be motivated by, a wide-range of deliciousness-making features of baked goods: texture, complexity of flavour, richness, and so on. A pâtissier could be motivated to add cinnamon to her banana bread, for example, because it enhances the complexity of the bread’s flavours. In contrast, a novice is not responsive to these first-order deliciousness considerations, though she could be motivated to add cinnamon to her banana bread because a recipe calls for it or because an expert has told her to do so. In this case, the pâtissier acts more skilfully than the novice, and this is because the pâtissier acts on what may be called “first-order deliciousness motivations” while the novice does not.21
Acting with effort and skill, be it when doing the right thing or when producing delicious baked goods, involves sensitivity and responsiveness to a wider range of properties than Johnson King’s account admits. While higher-order motivations may be relevant to deliberateness and non-accidentality characterized by effort, skill and counterfactual success, so are first-order motivations. And because Johnson King analyses moral worth in terms of deliberateness in this sense, by her own lights, first-order motivations must also be relevant to the moral worth of actions. Thus, reasons internal to this higher-order view support Dual Pertinence.
5. What Does Cruelty Have to Do with Kindness?
We have argued that because higher-order and first-order motivations can each impact the moral counterworth of actions, and because moral counterworth is relevantly similar to moral worth, Dual Pertinence is true—both kinds of motivations can impact moral worth as well. But higher-order and first-order theorists could reject Dual Pertinence by denying that reflections on moral counterworth tell us much about the nature of moral worth. They could grant that first-order and higher-order motivations are relevant to moral counterworth but maintain that only one kind of motivation impacts moral worth. And, since most of those who work on moral worth have not been theorising about moral worth’s negative counterpart, they would have to change nothing about their views to accommodate the considerations raised so far.22
Furthermore, there is independent theoretical and empirical work to suggest that praiseworthiness and blameworthiness are relevantly different, which could render moral worth and moral counterworth importantly asymmetrical (Wolf 1990; Nelkin 2011; Knobe 2003). We now explore key ways in which praiseworthiness and blameworthiness come apart, and argue that none of these differences can undercut our inference to Dual Pertinence.
5.1. Theoretical Arguments for Asymmetry
Two prominent defenders of the asymmetry between praiseworthiness and blameworthiness are Wolf (1990) and Dana Nelkin (2011). Both argue that the ability to do otherwise is not necessary to be praiseworthy but it is necessary to be blameworthy. This is surely a significant difference, from which one might infer that reflections on the nature of blameworthiness could not shed light on the nature of praiseworthiness. If this is right, then the fact that first-order and higher-order motivations are relevant to blameworthiness and moral counterworth tells us nothing about the kinds of motivations that are relevant to praiseworthiness and moral worth.
However, a move from an asymmetry in the ability to do otherwise to the claim that reflections on blameworthiness can tell us nothing about praiseworthiness would be hasty. Indeed, Wolf and Nelkin come to the view that the ability to do otherwise is unnecessary for praiseworthiness but necessary for blameworthiness by defending a much deeper symmetry between these two forms of normative evaluation. On their views, what is required (and on Nelkin’s view sufficient) to be praiseworthy and blameworthy are one in the same thing—the ability to recognise and act for the right reasons.23 But exercising the ability to recognise and act for the right reasons requires different things in different circumstances. When an agent does the right thing for the right reasons, she exercises her ability to recognise and act for the right reasons. It needn’t be possible for the agent to do otherwise in order for her to act with this ability. In contrast, when an agent performs a wrong action, she fails to recognise and act for the right reasons. So, in order for the agent to have exercised the ability to recognise and act on the right reasons, it really must be possible for her to have done other than she did. This is why the ability to do otherwise is necessary for blameworthiness but not for praiseworthiness. It is not because blameworthiness and praiseworthiness bear no relationship to one another. Rather, these forms of normative evaluation require that agents act with precisely the same ability, and this simply calls for different things in different circumstances. Thus, one cannot rely on Wolf and Nelkin’s asymmetry thesis to support the claim that blameworthiness has little to do with praiseworthiness, or that moral counterworth has little to do with moral worth.
Still, there might be other important differences between praiseworthiness and blameworthiness. One noted difference between praiseworthiness and blameworthiness is that being blameworthy opens an agent up to being the target of blame, which is characteristically sanction-like and harmful (Feinberg 1986; Hieronymi 2004), but this is not the case with praise. As Wolf argues:
Acts of moral blame are more connected with punishment than acts of moral praise are connected with reward. So . . . we have stronger reasons for wanting acts of blame to be justified. If we blame someone or punish him, we are likely to be causing him some pain. But if we praise someone or reward him, we will probably only add to his pleasures. (1980: 155)
We do not deny this particular asymmetry between blameworthiness and praiseworthiness. And we grant that this asymmetry should be reflected in how we understand these two forms of normative evaluation. For example, perhaps it should be harder, all things considered, to be blameworthy than praiseworthy. Or perhaps the limit on the amount of blame agents could possibly deserve should be lower than the limit on deserved praise. But Dual Pertinence is entirely compatible with these commitments. Dual Pertinence makes no claim about the relative difficulty or degree of being blameworthy versus praiseworthy, nor are such claims even suggested by the thesis. So, this asymmetry cannot be used to undermine Dual Pertinence.
In sum, we can grant these theoretical arguments for an asymmetry between praiseworthiness and blameworthiness without placing Dual Pertinence in jeopardy. While there are surely differences between praiseworthiness and blameworthiness, these differences do not support the general claim that we don’t stand to learn about the nature of praiseworthiness by reflecting on the nature of blameworthiness, nor do they undermine our argument for Dual Pertinence.
5.2. Empirical Work on Asymmetry
In addition to the theoretical arguments discussed above, there is also a growing body of empirical work that indicates that there is an asymmetry between praiseworthiness and blameworthiness. Take, for example, Joshua Knobe’s (2003; 2006) work on the Side-Effect Effect, which is the tendency to judge that foreseen side effects are brought about intentionally when they are negative but not when they are positive. Some have argued that this asymmetry arises because we take agents who engage in foreseen but unintended wrongdoing to be blameworthy, but do not take agents who knowingly but unintentionally act rightly to be praiseworthy.24 This explanation indicates that the motivations that are relevant to being blameworthy are not the same motivations that are relevant to being praiseworthy, which could undermine our argument for Dual Pertinence. In fact, Johnson King argues that the research on the Side-Effect Effect indicates that deliberateness and higher-order motivations are necessary for praise but not for blame:
[the participants’] reactions highlight an asymmetry between praise and blame: we can be blamed for performing an act of a bad type as long as we are aware that our act is of that type, but we merit praise for performing an act of a good type only if we do so deliberately. (2020: 202)
However, we do not think that the Side-Effect Effect can be used to support a deliberateness condition on moral worth or adjudicate between first-order and higher-order views. First, the results from studies on the Side-Effect Effect are entirely compatible with first-order accounts of moral worth according to which deliberately doing the right thing is not a necessary condition for moral worth. To see this, consider the most commonly used cases to illustrate the Side-Effect Effect:
Harm: The vice-president of a company went to the chairman of the board and said, “We are thinking of starting a new program. It will help us increase profits, but it will also harm the environment.” The chairman of the board answered, “I don’t care at all about harming the environment. I just want to make as much profit as I can. Let’s start the new program.” They started the new program. Sure enough, the environment was harmed.
Help: The vice-president of a company went to the chairman of the board and said, “We are thinking of starting a new program. It will help us increase profits, and it will also help the environment.” The chairman of the board answered, “I don’t care at all about helping the environment. I just want to make as much profit as I can. Let’s start the new program.” They started the new program. Sure enough, the environment was helped. (Knobe 2003: 191)
In the Help vignette, it’s true that the chairman does not act on a higher-order motivation to do the right thing when they decide to start the new program. But they are also clearly not acting on the reasons that morally justify the action (as Markovits requires) nor are they responsive to the features of the action that make it right (as Arpaly requires). After all, the chairman doesn’t care that the program will help the environment; they are only motivated by considerations of profit. Consequently, the chairman fails to satisfy the criteria that first-order theorists consider necessary for moral worth just as much as they fail to satisfy the criteria of higher-order theories.
Second, no deep asymmetries between the necessary conditions for praiseworthiness and blameworthiness arise if one adopts a first-order approach to these cases. On Arpaly’s view, the chairman in the Harm vignette is blameworthy because they act out of a lack of good will—they act on the neutral reason of profit maximisation while ignoring the moral reasons not to start the program. And the chairman in the Help vignette is not praiseworthy for the same reason—they act on the profit reason and ignore the moral reason to enact the program. In other words, we can explain why the chairman is blameworthy in the Harm vignette and not praiseworthy in the Help vignette by examining their first-order motivations alone. Not only does the Side-Effect Effect fail to establish any particular necessary conditions for the moral worth of actions, it also doesn’t indicate that there is an asymmetry between moral worth and counterworth.
Moreover, even if the empirical work on the Side-Effect Effect did indicate that particular higher-order or first-order motivations are necessary for moral worth but not moral counterworth, this would not undermine Dual Pertinence. Dual Pertinence is a claim about the relevance of higher-order and first-order motivations to moral worth—and not a claim about the motivations that are necessary or sufficient for moral worth. In order to reject Dual Pertinence, one must show that only one kind of motivation, either first-order or higher-order, affects moral worth. But the work on the Side-Effect Effect does not illustrate this. Whatever asymmetry may be revealed by the Side-Effect Effect, it would not threaten Dual Pertinence.
6. Conclusion
In this paper, we reflected on the nature of moral counterworth with the hope of illuminating the nature of moral worth. We argued that because moral counterworth is sensitive to both higher-order and first-order motivations, this is likely also true of moral worth. So, we proposed Dual Pertinence—the view that first-order and higher-order motivations can each affect the moral worth of an agent’s action. And while traditional higher-order and first-order views of moral worth are unable to accommodate Dual Pertinence, other approaches may be more successful. In fact, some have recently offered hybrid views of moral worth that cut across the traditional divide between first-order and higher-order accounts. For example, Keshav Singh’s (2020) Guise of Moral Reasons account takes both first-order and higher-order motivations to be required for moral worth. And Jessica Isserow (2019; 2020) and Douglas Portmore (2022) defend disjunctive views of moral worth, according to which higher-order and first-order motivations can each make an action morally worthy without the presence of the other kind of motivation. These theorists, without reflecting on the nature of moral counterworth, also come to the conclusion that both first-order and higher-order motivations play an important role in determining the moral worth of actions.25 Thus, our arguments can serve as a novel source of support for hybrid models of moral worth over the traditional first-order and higher-order approaches. Given the lack of attention paid to moral counterworth, it’s likely that further reflection on morally counterworthy actions could shed light on ways to develop and expand these hybrid approaches of moral worth as well.
Acknowledgments
We are grateful to Christopher Howard, Luara Ferracioli, Douglas Portmore, Connie Rosati, and two anonymous reviewers for providing comments on earlier drafts of this paper. We would also like to thank the participants of the 2020 Slippery Slope Normativity Summit for their very helpful feedback on this project.
Notes
- Recently, some have offered views that cut across these two camps, like Isserow’s accounts of non-accidentality and moral worth (2019; 2020), Singh’s Guise of Moral Reasons account (2020), and Douglas Portmore’s Concerns View (2022). We discuss these views in more detail in §6. ⮭
- Often, these motivations are distinguished using the de re/de dicto distinction: To be motivated by an action’s rightness itself is to be motivated by rightness de dicto, and to be motivated by an action’s right-making features is to be motivated by rightness de re (Johnson King 2020; Smith 1994; Weatherson 2019). However, this way of discussing motivations has recently been criticized for failing to track what’s at stake in the debate (Singh 2020). We agree that there is need for a clearer way to refer to the two kinds of motivations, and opt for a first-order/higher-order distinction. This has the benefit of being valence-neutral, and is also used in other philosophical debates. For example, the literature on disagreement features a distinction between first-order evidence, evidence concerning p, and higher-order evidence, evidence concerning the rationality of believing p (Christensen 2010; Kelly 2010). Importantly, the first-order/higher-order distinction is different from the Kantian distinction between primary and secondary motives, which regards an agent’s foreground and background motivations (Baron 1995; Herman 1996; Isserow 2021), and also different from the distinction between first-order and second-order desires, which regards an agent’s desire to X and their desire concerning their desire (Frankfurt 1971). ⮭
- While we largely write in terms of praiseworthiness in this paper, we intend to remain neutral between those who think moral worth is best conceived of in terms of creditworthiness (e.g., Portmore 2022; Singh 2020) and those who analyse moral worth in terms of praiseworthiness (e.g., Arpaly 2003; Johnson King 2020; Markovits 2010). ⮭
- Mason (2019) defends a view of praiseworthiness that is similar to Johnson King’s view of moral worth. On Mason’s view, agents are morally praiseworthy when they try to do well by the correct normative theory (2019: 75). However, Mason is not concerned with moral worth. And while theorists of moral worth typically distinguish between right actions (those that are required by the correct normative theory) and morally worthy actions (those that merit praise), Mason does not operate with such a distinction and takes rightness to correlate with praiseworthiness. ⮭
- See Singh (2020) for criticism of this claim. ⮭
- It’s possible that the professor has to rationally believe or know that giving her students extensions will promote their well-being in order to have the right kind of higher-order motivation for her action to possess moral worth. Higher-order theorists typically do not take a mere belief that one is doing the right thing to be the kind of motivation sufficient for moral worth. Sliwa (2016), for example, argues that one must be motivated by the knowledge that one is doing the right thing, which could require that one also believe or know that one’s actions have the relevant right-making features. Importantly, however, the higher-order theorist argues that one need not be motivated by these first-order beliefs in order for one’s action to possess moral worth. ⮭
- This isn’t to say that first-order motivations couldn’t play an indirect or causal role in an agent’s coming to have the higher-order motivations needed for moral worth. An agent who has a first-order motivation to tell the truth because she cares about honesty, and also believes that telling the truth is the right thing to do because it is honest, could come to develop the higher-order motivation to tell the truth because it is the right thing to do. However, such a first-order motivation is neither necessary nor sufficient for developing the higher-order motivation. ⮭
- While we write mainly in terms of blameworthiness in this paper, we intend to remain neutral between understanding moral counterworth in terms of blameworthiness and deserved discredit. ⮭
- Similarly, not all agents who are blameworthy (or deserving of discredit) have performed a morally counterworthy action. On our view, moral counterworth is a term of art that picks out a particular kind of blameworthiness: blameworthiness for doing the wrong thing. But agents can be blameworthy for other things as well, like character traits, motivations, and beliefs. These forms of blameworthiness are distinct, and likely come with a different set of necessary and sufficient conditions than blameworthiness for doing the wrong thing. ⮭
- It’s noteworthy that non-deliberate wrong actions can be morally counterworthy, since it is typically thought that only deliberately performed right actions can be morally worthy. But there is a unifying explanation for this difference. When non-deliberate wrongs possess counterworth, as in cases of culpable ignorance or recklessness, it is because they are non-accidental—it is no accident that agents do the wrong thing when they are insufficiently attentive to the moral reasons to refrain from acting wrongly. But non-deliberate right actions are very often accidents. If an agent is insufficiently attentive to the moral reasons to perform a right action, but does so anyway by chance or for some other non-moral reason, we would not expect such an agent to act similarly in relevantly similar circumstances. And because non-accidentality is of primary importance when theorising about moral worth, we take the symmetry between morally worthy and counterworthy actions to hold so long as both kinds of actions must be non-accidental. We explore other challenges to the symmetry between moral worth and counterworth in §5. ⮭
- Arpaly refers to the negative analogue of moral worth as “negative moral worth” (2003: 69). ⮭
- There is a potential asymmetry between Arpaly’s view of moral worth and its negative parallel, since only good will impacts moral worth, while ill will and a lack of good will impacts that negative parallel. However, Arpaly and Schroeder (2014) provide updated versions of these views according to which both good will and reverse moral indifference (a lack of ill will) affect praiseworthiness, just as ill will and moral indifference impact blameworthiness. They argue: “a person is praiseworthy for right action A to the extent that A manifests good will (or reverse moral indifference) through being rationalized by it” and “a person is blameworthy for a wrong action A to the extent that A manifests ill will (or moral indifference) through being rationalized by it” (2014: 170). ⮭
- Though Isabelle’s motivational profile is admittedly unusual, it is compatible with a fair number of theses about motivation. For example, one needn’t reject the guise of the good thesis (Velleman 1992) to make sense of Isabelle’s motivational profile. On this thesis, agents are only motivated to perform actions that they take to be good simpliciter. And Isabelle, though she is motivated to do the morally wrong thing, can take doing the morally wrong thing to be good simpliciter. Isabelle’s motivational profile could very well conflict with certain versions of motivational judgment internalism, however. On these views, there is a necessary connection between judging that something is wrong and being motivated (at least to some degree) not to perform that action. Thus, some motivational judgment internalists may reject the possibility of an agent like Isabelle. However, it is unlikely that first-order theorists would follow suit. In fact, Arpaly (2003: 36–46) explicitly rejects Smith’s (1994) motivational internalist position. ⮭
- The idea that the motivation to do wrong is normatively relevant has not gone unnoticed in the literature on evil. Perrett (2002), for example, argues that in order for an agent to be evil, they must be motivated to engage in wrongdoing because it is wrong. And recently, Mason focuses on the motivation to do wrong in her discussion of Milton’s Satan. Satan, a fallen angel, purposefully commits himself to a life of evil, averring: “So farewell Hope, and with Hope, farewell Fear, Farewell Remorse: all good to me is lost, Evil, be thou my good” (1667/2000: bk. IV, 109–11). According to Mason, Satan is blameworthy precisely because “he pursues evil for its own sake” (2019: 165). ⮭
- This latter possibility is compatible with Aboodi’s claim that higher-order motivations are necessary for moral worth only in situations that feature “uncertainty concerning underived moral beliefs” (2017: 227). ⮭
- This leaves room for sympathisers of Williams’s (1981) “one thought too many” criticism, like Sorensen (2004) and Isserow (2021), who argue that acting on a higher-order motivation to do the right thing can detract from the moral worth of right actions without eliminating their worth. ⮭
- Johnson King (2019) also argues that both kinds of motivations can themselves be praiseworthy. ⮭
- Notice that one can accept this claim even if one also accepts a buck-passing account of wrongness (e.g., Scanlon 1998), according to which the wrongness of an action isn’t itself a reason to not perform the action. This is because an action’s being wrong, while neither a wrong-making feature of the action nor a reason to not perform it, can still be the kind of thing that shouldn’t motivate an agent to perform an action. Thanks to Connie Rosati for discussion on this point. ⮭
- In fact, Singh specifically criticises Johnson King’s view on the grounds that deliberately doing the right thing needn’t involve being motivated to perform an action because it is right (2020: 175). Furthermore, Mason (2019) and Singh (2020) can be read as providing accounts of deliberately doing the right thing that rely on higher-order and first-order motivations. ⮭
- While we’ll refrain from objecting to Johnson King’s understanding of deliberateness in this paper, it is worth noting that that the link between effort, skill, and deliberateness is not clearly a strong one. We can deliberately do things effortlessly, or that we’re demonstrably unskilled at. For example, we can deliberately eat a whole bag of cookies, or watch yet another episode of a show, with hardly any effort. And we can deliberately haggle for a better price despite clearly lacking the skill to do so. And, if one were to insist that such deliberate acts are skilful or effortful, then we may begin to wonder what kinds of actions fail to have these features. ⮭
- This discussion is related to an objection Howard (2021) raises to Sliwa’s claim that knowledge of an action’s rightness is necessary for an action to have moral worth. According to Howard, forms of expertise that are relevantly similar to moral worth do not require agents to have knowledge about their own expertise or skilfulness. For example, an agent can skilfully play football or write a song without having the belief that one is playing or writing skilfully (2021: 307–8). ⮭
- This is not true of Arpaly. Arpaly defends views of both moral worth and its negative parallel, and argues that only first-order motivations are relevant to each. So, to accommodate the considerations raised here, Arpaly would have to both argue that moral worth and its negative parallel are relevantly different and revise her account of the negative parallel of moral worth to make room for the relevance of higher-order motivations. ⮭
- According to Wolf, “the freedom necessary for responsibility consists in the ability (or freedom) to do the right thing for the right reasons . . . to choose and to act in accordance with the True and the Good” (1990: 94). And on Nelkin’s view, “one is responsible if and only if one acts with the ability to recognize and act for good reasons” (2011: 3). Wolf and Nelkin understand moral responsibility in terms of blameworthiness and praiseworthiness. ⮭
- While we’ll grant this explanation for the sake of argument, it’s important to note that there is considerable disagreement about how to account for the Side-Effect Effect, and many of the candidate explanations do not rely on an asymmetry between praiseworthiness and blameworthiness. See Knobe (2006) and Feltz (2007) for overviews of these accounts. ⮭
- For example, Isserow notes that it’s typically thought that wrong actions cannot possess moral worth: “Traditionally, only right actions are taken to be candidates for moral worth. . . . Although we may be less blameworthy (if at all) when noble motives lead us to act wrongfully, having one’s heart in the right place does not suffice. The road to hell has no moral worth, regardless of whether it is paved with good intentions” (2019: 252, fn. 3). ⮭
References
1 Aboodi, Ron (2017). One Thought Too Few: Where De Dicto Moral Motivation is Necessary. Ethical Theory and Moral Practice, 20(2), 223–37.
2 Arpaly, Nomy (2003). Unprincipled Virtue: An Inquiry into Moral Agency. Oxford University Press.
3 Arpaly, Nomy and Timothy Schroeder (2014). In Praise of Desire. Oxford University Press.
4 Baron, Marcia (1995). Kantian Ethics Almost without Apology. Cornell University Press.
5 Christensen, David (2010). Higher-Order Evidence. Philosophy and Phenomenological Research, 81(1), 185–215.
6 Feinberg, Joel (1986). Harm to Others. Oxford University Press.
7 Feltz, Adam (2007). The Knobe Effect: A Brief Overview. The Journal of Mind and Behavior, 28(3/4), 265–77.
8 Frankfurt, Harry G. (1971). Freedom of the Will and the Concept of a Person. Journal of Philosophy, 68(1), 5–20.
9 Henson, Richard G. (1979). What Kant Might Have Said: Moral Worth and the Overdetermination of Dutiful Action. Philosophical Review, 88(1), 39–54.
10 Herman, Barbara (1981). On the Value of Acting from the Motive of Duty. Philosophical Review, 90(3), 359–82.
11 Herman, Barbara (1996). The Practice of Moral Judgment. Harvard University Press.
12 Hieronymi, Pamela (2004). The Force and the Fairness of Blame. Philosophical Perspectives, 18, 115–48.
13 Howard, Nathan Robert (2021). One Desire Too Many. Philosophy and Phenomenological Research, 102(2), 302–17.
14 Isserow, Jessica (2019). Moral Worth and Doing the Right Thing by Accident. Australasian Journal of Philosophy, 97(2), 251–64.
15 Isserow, Jessica (2020). Moral Worth: Having It Both Ways. The Journal of Philosophy, 117(10), 529–56.
16 Isserow, Jessica (2021). Doubts about Duty as a Secondary Motive. Philosophy and Phenomenological Review. Advance online publication. http://doi.org/10.1111/phpr.12821
17 Jeske, Diane (1998). A Defense of Acting from Duty. The Journal of Value Inquiry, 32(1), 61–74.
18 Johnson King, Zoë (2019). Praiseworthy Motivations. Noûs, 54(2), 408–30.
19 Johnson King, Zoë (2020). Accidentally Doing the Right Thing. Philosophy and Phenomenological Research, 100(1), 186–206.
20 Kant, Immanuel (1998). Groundwork of the Metaphysics of Morals (Mary Gregor, Trans.). Cambridge University Press. (Original work published 1785)
21 Kelly, Tom (2010). Peer Disagreement and Higher-Order Evidence. In T. Warfield and R. Feldman (Eds.), Disagreement (183–217). Oxford University Press.
22 Knobe, Joshua (2003). Intentional Action and Side-Effects in Ordinary Language. Analysis, 63(279), 190–94.
23 Knobe, Joshua (2006). The Concept of Intentional Action: A Case Study in the Uses of Folk Psychology. Philosophical Studies, 130(2), 203–31.
24 Markovits, Julia (2010). Acting for the Right Reasons. The Philosophical Review, 119(2), 201–42.
25 Mason, Elinor (2019). Ways to Be Blameworthy: Rightness, Wrongness, and Responsibility. Oxford University Press.
26 Milton, John (2000). Paradise Lost. Penguin Books. (Original work published 1667)
27 Nelkin, Dana (2011). Making Sense of Freedom and Responsibility. Oxford University Press.
28 Perrett, Roy W. (2002). Evil and Human Nature. Monist, 85(2), 304–19.
29 Portmore, Douglas (2022) Moral Worth and Our Ultimate Moral Concerns. Oxford Studies in Normative Ethics (Vol. 11, 276–298). Oxford University Press.
30 Scanlon, T. M. (1998). What We Owe to Each Other. Harvard University Press.
31 Singh, Keshav (2020). Moral Worth, Credit, and Non-Accidentality. Oxford Studies in Normative Ethics (Vol. 10, 156–81). Oxford University Press.
32 Sliwa, Paulina (2016). Moral Worth and Moral Knowledge. Philosophy and Phenomenological Research, 93(2), 393–418.
33 Smith, Michael (1994). The Moral Problem. Blackwell.
34 Sorensen, Kelly (2004). The Paradox of Moral Worth. The Journal of Philosophy, 101(9), 465–83.
35 Velleman, J. David (1992). The Guise of the Good. Noûs, 26(1), 3–26.
36 Weatherson, Brian (2019). Normative Externalism. Oxford University Press.
37 Williams, Bernard (1981). Moral Luck: Philosophical Papers 1973–1980. Cambridge University Press
38 Wolf, Susan (1980). Asymmetrical Freedom. The Journal of Philosophy, 77(3), 151–66.
39 Wolf, Susan (1990). Freedom within Reason. Oxford University Press.