Can free will and moral responsibility skeptics recognize the existence of ‘basic desert’ or, more precisely, justify certain responses to moral behavior on the basis of considerations of basic desert?

Although I refer to those who are skeptical of moral responsibility, some involved in the free will debate have argued that there are different forms or senses of moral responsibility and that only some forms require free will. The idea that there might be multiple forms of responsibility seems to have originated with Gary Watson, who, in his essay ‘Two Faces of Responsibility’ (Reference Watson1996) distinguishes between what he calls ‘attributability’ and ‘accountability’. More recently, David Shoemaker has proposed a third form, which he calls ‘answerability’ (Reference Shoemaker2011, Reference Shoemaker2015). Shoemaker argues that agents with different impairments may lack some, but not all, of these forms, as they depend on different capacities of the agent (2015: 224); but he is silent on the issue of free will. After arguing for such moral responsibility ‘pluralism’, Derk Pereboom, a free will skeptic, contends that not all forms of moral responsibility are equally threatened by the absence of free will (Reference Pereboom and Shoemaker2017). Pereboom distinguishes different forms of moral responsibility on the basis of their different ‘aims and justifications’ (2017: 122). He endorses a notion of moral responsibility that has a ‘forward-looking’ justification, in terms of the goods that holding responsible (and in particular, blaming) may achieve—the goods of protection, reconciliation, and moral formation (2017: 125). Pereboom considers his developed view to be a form of responsibility-as-answerability, which depends on an agent's responsiveness to reasons; and he maintains that this capacity and the forward-looking justifications for blame he outlines are compatible with an absence of free will (2017: 127). What is not compatible with the absence of free will, Pereboom claims, is the ‘basic desert sense’ of moral responsibility (2017: 123).

Pereboom is not the only free will skeptic who distinguishes ‘basic desert’ moral responsibility from non-basic-desert senses. In fact, in his article on the subject for the Stanford Encyclopedia of Philosophy, Gregg Caruso (Reference Caruso and Zalta2018) defines moral responsibility skepticism as skepticism about the idea that humans are ever morally responsible in the basic desert sense, before going on to note, ‘Importantly, moral responsibility skepticism . . . is consistent with agents being [morally] responsible in others senses’. And while he is not a free will skeptic, Randolph Clarke has also toyed with the idea that there may be distinct forms of moral responsibility, only some of which are tied to the notion of basic desert.Footnote ¹ In a critique of John Martin Fischer's view of moral responsibility, Clarke considers a case of a hypothetical agent S whose personality and behavior have unwittingly been designed by another, and writes, ‘[Fischer] seems to take the question of such an agent's responsibility to be a “yes or no” matter. We might better take it to be a question of what sort of responsibility such an agent might bear. We might want to distinguish various aspects of responsibility, and we might wish to say that while some apply to S, others don't. It's particularly, I think, a certain type of moral desert that we might sensibly wish to deny application’ (Reference Clarke2010: 249).

Like Pereboom, Clarke seems to base his idea that there are different forms of moral responsibility on the premise that some judgments about and responses to the hypothetical agent might be appropriate, while others might not be. For instance, Clarke says it might be appropriate to assess the agent's character (as a charming creep), or to call on him to justify apparently untoward behavior and, if he cannot, to acknowledge his fault and resolve to correct it, or to feel certain reactive attitudes, such as disapprobation or disgust. It might even, says Clarke, be appropriate for the agent to feel guilty when he mistreats someone—but the agent might not deserve to suffer the unpleasantness of his self-directed attitude; and it might also be inappropriate for us to take up ‘retributive’ attitudes toward him (2010: 249). (I discuss the association of basic desert with retributivism below.)

Finally, there are those who seem to think that basic desert is a feature that one can ‘add on’—or not—to a favored view of moral responsibility. Michael McKenna, who has developed a ‘conversational’ model of moral responsibility (Reference McKenna2012), remarks in an essay on the concept of basically deserved blame, ‘The conversational theory of moral responsibility needs supplementing with normative resources that reach outside the conversational elements of the theory. My own view is that this can be done with resources that do not require commitment to basic desert. . . . Nevertheless . . . one can also supplement the conversational theory by way of a basic-desert thesis’. McKenna explains that this supplementation can be done via the claim that the harms involved in blaming on the conversational model are deserved by the blameworthy (Reference McKenna2019: 270).

Although I was once attracted to the idea that there are different forms of moral responsibility (see Vicens Reference Vicens, Markham and Knight2022), I have since become somewhat skeptical. While there may be multiple conditions or capacities, such as certain forms of control and awareness, that are required for moral responsibility, and while these conditions may be present to greater or lesser degrees so that there may be both different degrees of responsibility and distinct reasons why one's responsibility is mitigated, I am inclined to think the concept of moral responsibility itself is univocal—and basic. What I mean by basic is that moral responsibility cannot be analyzed in terms of or reduced to something more fundamental, such as being the appropriate target of reactive attitudes (contra Strawson Reference Strawson1963). If anything, being morally responsible is the more fundamental thing that makes one the appropriate target of such attitudes—though even if one is morally responsible, it may be inappropriate for others to hold such attitudes toward one, for other reasons (see Smith Reference Smith2008: 379). And I think the different aims and justifications that Pereboom says correspond to different forms of moral responsibility are really just different reasons one might have for holding someone morally responsible—or engaging in a practice that is often (but possibly mistakenly) associated with the person being morally responsible (see Clarke Reference Clarke2016: 135).

So I do not distinguish ‘basic desert’ moral responsibility from other forms of moral responsibility in part because I am not sure that it makes any sense to talk of multiple forms of moral responsibility. But I also do not associate the one and only ‘form’ of moral responsibility that there is (or is not) with basic desert. For if there is such a thing as basic desert (or, more properly, basically deserved responses to one's behavior), one can have it without being morally responsible. (However, I also do not think of basic desert as a possible add-on to one's favored account of moral responsibility, as McKenna seems to. For while one might deserve to be treated in certain ways in virtue of how one has behaved, exactly what treatment is deserved may be in part determined by whether one is morally responsible for what one has done. I discuss this in greater detail below.) As I do not argue against moral responsibility pluralism in this essay,Footnote ² I offer two versions of my thesis, depending on whether the reader agrees that moral responsibility is a univocal concept:

Although both versions of my thesis can be summed up with the slogan ‘basic desert without free will’, I want to emphasize that free will skeptics who think that some form of moral responsibility is ruled out by their skepticism can still believe in basic desert. My thesis should not be taken as a form of semi-compatibilism—that is, the view that even if determinism (or some other condition) rules out free will, moral responsibility is untouched. In other words, my thesis is that even if, because we lack free will, we are not in some important sense responsible for who we are or what we do, certain responses to our behavior may be appropriate in virtue of the fact that we (‘basically’) deserve these responses. In what follows, I take moral responsibility to mean whatever form of moral responsibility is ruled out by the absence of free will.

But what is basic desert? The concept might be analyzed negatively (in terms of what it is not) or positively (in terms of what it is). I reject positive characterizations that tie the definition of (basic) desert to what responses or treatment someone may deserve, or that take certain sorts of responses or treatment to be entailed by the concept of desert. For instance, in commenting on the notion of basic desert at issue in free will debates, McKenna considers various interpretations of what might be meant in the case of deserved blame, including that it would be ‘a pro tanto good that the wronging party suffer’, or ‘prima facie right to cause the wronging party to suffer’, or ‘permissible that others do some harm to the wronging party’ (Reference McKenna2009: 12); he later defends an interpretation according to which the harms involved in deserved blame are noninstrumentally good (McKenna Reference McKenna2019). Joel Feinberg likewise proposes that ‘the various kinds of treatment that persons deserve from other persons . . . have at least one thing in common: they are generally ‘affective’ in character’ (Reference Feinberg1970: 61)—meaning that they are pleasant or unpleasant, favored or disfavored, and so on—while Dana Nelkin asserts that ‘what distinguishes desert from at least some other notions of fittingness is that what is deserved in any particular case is good or bad for the person’ (Reference Nelkin2013: 131n15). The problem with such characterizations of (basic) desert is that what is deserved may vary depending on the features of the wrongdoing; and even supposing someone is morally responsible ‘in the basic desert sense,’ if the wrongdoing is not serious the wrongdoer may not (basically) deserve suffering or harms of any sort. Moreover, if, as Nelkin suggests (Reference Nelkin2013: 124), being the target of negative reactive attitudes is not necessarily harmful to the target, this would seem to rule out the possibility that such attitudes could be basically deserved—a possibility that I do not want to rule out.Footnote ³ Interestingly, Thomas Scanlon once rejected the concept of (basic) desert, or desert-based justifications for treatment, precisely because he identified desert ‘with the idea . . . that it is good that people who have done wrong should suffer’—an idea that he regards as ‘morally repugnant’ (Reference Scanlon2013: 102). Again, we do not want to rule out the possibility that anyone could basically deserve any kind of responses to their behavior just because we are opposed to the idea that it could ever be, say, intrinsically good to make someone suffer.

Can any positive characterization of basic desert be given that does not involve or entail claims about what responses or treatments one might deserve? We may begin with a rather thin characterization. Following Feinberg (Reference Feinberg1970), it may be said that the responses or treatment one (basically) deserves are ones that are fitting. But more will need to be said, for certain responses may be fitting in virtue of features of a situation that have little to do with facts about basic desert. For instance, it might be fitting to punish a child because he broke a school rule and the school's policy prescribes a certain punishment for the violation of that rule. Thus we must try to rule out reasons why a response might seem fitting that do not ultimately seem to have to do with basic desert. For this, we next turn to a negative analysis of the concept, beginning with a passage from Pereboom:

The first sentence of this quotation is defining a ‘sense’ of moral responsibility—the sense that Pereboom thinks we lack—in terms of basic desert: to be responsible in this sense is to be basically deserving of certain responses to one's behavior. The second sentence is indicating what is meant by basic desert, in terms of what it rules out. One would not count as basically deserving of imprisonment, say, if the only reason for imprisonment was to keep others safe, for this justification for imprisonment is merely consequentialist. Scanlon, commenting on an earlier work of Pereboom's, offers a nearly identical characterization of (basic) desert: ‘What is distinctive about a desert-based justification for treating a person in a certain way is that it claims that the treatment is made appropriate simply by certain facts about that person or what he or she has done. The ‘simply’ in this formulation is intended to exclude justifications that appeal to the effects of treating the person in that way and justifications that appeal to the fact that this treatment is called for by some institution (that is itself justified on some basis other than an appeal to desert)’ (Reference Scanlon2013: 101).

Similarly, Caruso and Stephen Morris understand the notion of basic desert to be ‘backward-looking and non-consequentialist’; on their view, responses are justified in terms of basic desert when they are ‘justified on purely backward-looking grounds’ that do not ‘appeal to consequentialist or forward-looking considerations’ (Reference Caruso and Morris2017: 838).

So far, I have characterized basic desert in terms of the appropriateness of responses to or treatment of a person in light of her behavior when consequentialist and contractualist justifications of that treatment are ruled out. Before adding a further condition to basic desert below, I show that a view of moral protest developed by Thomas Hill (Reference Hill1979) and Robert Adams (Reference Adams1999) fits well with this characterization so far—and contrasts with Pereboom's view of the purpose and justification for protest. Pereboom (Reference Pereboom and Shoemaker2017) largely adopts Angela Smith's view of blame as a form of moral protest; and Smith writes, ‘To blame another is . . . to modify one's own attitudes, intentions, and expectations toward that person as a way of protesting (i.e. registering and challenging) the moral claim implicit in her conduct, where such protest implicitly seeks some kind of moral acknowledgement on the part of the blameworthy agent and/or on the part of others in the moral community’ (Reference Smith, Coates and Tognazzini2013: 43).

Pereboom says that he endorses this account, but he does not think that blame requires protesting a moral claim implicit in the blameworthy conduct (2017: 129). In any case, both Smith and Pereboom take the nature of blame to be communicative, since as Smith puts it, its ‘constitutive aims’ are registering a certain fact with, and seeking acknowledgment of that fact from, others (2013: 44). As mentioned already, Pereboom thinks that blame, understood in this way, can be justified in terms of its positive consequences: the goods of protecting oneself and others from further wrongdoing, reconciliation with the wrongdoer, and the wrongdoer's moral formation. Smith likewise suggests that blame-as-protest has a consequentialist justification when she confronts the objection that, if blame is aimed at getting the wrongdoer to acknowledge his wrongdoing, then it would not make sense to blame a dead wrongdoer, as in the case of blaming antebellum slaveholders in the United States (2013: 44). Her response is that blame can still have a ‘point’ since its secondary aim is ‘moral recognition on the part of the wider community’ (2013: 44). However, she notes that, on her view, it would not make sense for ‘a present-day Norwegian to blame southern slaveholders’ (2013: 45). Though she does not spell out her reasoning here, her point seems to be that, while the Norwegian's expressed blame would still be a form of communication with those around her, it would do no good. In contrast, Smith writes, ‘We in the United States have a particular reason to continue to blame southern slaveholders (rather than merely judging them blameworthy), which has to do with our current relationships with members of our community who were deeply affected by this disgraceful chapter in our history. By continuing to blame, we continue to protest the ‘outrageous falsehood’ that the practice of slavery embodied’ (2013: 45). Because Black Americans alive today are negatively affected by the behavior of slaveholders of the past, our protesting their past behavior has the potential to do some good for them.

In contrast, Hill (Reference Hill1979) and Adams (Reference Adams1999) argue that the main purpose or value of protest lies not in the positive consequences it may have. Adams considers a case in which Dietrich Bonhoeffer saluted the Nazis, apparently reasoning that refraining from saluting would have undermined his conspiratorial involvement in the Abwehr; Adams then asks us to imagine that instead Bonhoeffer had engaged in a kind of silent protest by refusing to salute—an act of protest that would have been ‘costly’ and ‘ineffective’, but, Adams thinks, admirable nonetheless (1999: 216). Similarly, Hill considers cases of protest against injustice done to others that ‘cannot reasonably be expected to end the injustice, to prevent its recurrence, or to rectify it in any way’ and which may involve some harm to the protestor (1979: 84–85); he likewise thinks there may still be reason to engage in such protest. Both Adams and Hill argue that protest which does not have hope of realizing the sort of ends Pereboom identifies (such as the reformation of the wrongdoer) may be justified by appeal to protest's symbolic value; Adams emphasizes that the value of protest lies not in ‘what it causes’ but ‘what it symbolizes or stands for’ (1999: 217; see also Bennett Reference Bennett, Coates and Tognazzini2013: 67). Adams identifies protest as a way of being ‘for what one loves and against what one hates’ (1999: 219), while Hill describes it as a way to ‘honor the persons, groups, and causes that, from a moral point of view, most deserve it’ (1979: 99).

Hill and Adams seem right that one may engage in protest for non-consequentialist reasons, but more needs to be said supporting the claim that the justification of protest may be basic-desert based. For unlike Caruso and Morris, I do not think that ruling out contractualist, consequentialist, or ‘forward-looking’ justifications for some treatment is enough to show that the treatment is basic-desert based, since there are other reasons we might judge it fitting (or unfitting) to treat someone a certain way which have little to do with basic desert. Consider, for instance, a case of hypocritical blame: imagine a wife blaming a husband for making them late for something, when she perpetually makes them late herself. However we might construe blame—whether a kind of reactive attitude, sanction, or protest—it would seem inappropriate for the wife to engage in the blaming activity; her blaming him would not seem fitting. The inappropriateness or unfittingness does not stem from contractualist or consequentialist considerations; in fact, it seems to be ‘backward-looking’; and yet it has nothing to do with the question of whether the husband basically deserves blame. So in addition to ruling out consequentialist and contractualist considerations, we need to rule out relational considerations that might affect our judgments of the appropriateness of blame, in order to home in on only intrinsic features of the individual and his behavior. I take this to be what Feinberg means when he insists that ‘the facts which constitute the basis of a subject's desert must be facts about that subject’ (1970: 59); what Pereboom means when he says that for an agent to be morally responsible for an action in the basic desert sense is ‘for it to be hers’ (2014: 2) or for it to ‘belong to’ her (Reference Pereboom1995: 22); and why Scanlon emphasizes that (basic) desert-based justifications are made appropriate ‘simply by certain facts about that person or what he or she has done’ (2013: 101).

I think protest can be appropriate or fitting in virtue of such intrinsic features, even in the absence of free-will-requiring moral responsibility—and so one can basically deserve protest without being morally responsible or having the sort of moral responsibility that requires free will. To see this, consider the focus of attributionist accounts of moral responsibility: attributionists hold that a person is morally responsible for what she does if it is attributable to her, in the sense that it reflects something ‘deep’ or significant about her. Different attributionists identify the significant in different ways, such as one's ‘evaluative judgments or appraisals’ (Smith Reference Smith2005: 237) or one's ‘cares’ in the sense of what matters to one, or what one is ‘emotionally tethered’ to (Brownstein Reference Brownstein2016: 773). One's actions (or omissions) can reveal (intentionally or not) such judgments or cares as a self-centered orientation, sexism, or racism or other forms of bias, or even callousness or outright cruelty. Attributionists reject what Smith (Reference Smith2005: 237) calls ‘voluntariness as a precondition for moral responsibility’, that is, the requirement that morally responsible action be under (or traceable to an instance of) the agent's voluntary control, as well as what might be called consciousness conditions, that is, that consciousness of certain facts about one's action is required for moral responsibility. But suppose, with some detractors, we maintain these requirements and further reason—whether because of concerns about general metaphysical issues like determinism, ‘ultimate’ control, or luck (Pereboom Reference Pereboom2014; Strawson Reference Strawson1994; Levy Reference Levy2011) or more specific empirical concerns about human cognitive or motivational structures or processes such as automaticity or the role of consciousness (see Caruso Reference Caruso2012; Levy Reference Levy2014)—that moral responsibility is extremely limited or nonexistent. Still, as free will and moral responsibility skeptics, we should acknowledge the appropriateness of certain forms of protest. Imagine, for instance, that a person's actions (or inactions) reveal a deep-seated prejudice against people of a certain ethnicity: the individual regards them as inferior and unworthy of respect or decent treatment. Surely it would be appropriate for us (barring certain relational facts, such as that we ourselves are similarly prejudiced) to take a stand against the person's actions and underlying attitudes. And what grounds the appropriateness of such protest is simply the intrinsic features of the prejudiced person and her action itself.Footnote ⁴

Caruso and Morris maintain that ‘the key question with regard to basic desert moral responsibility is whether it would ever be appropriate for a divine all-knowing judge (who didn't necessarily create the agents in question) to administer differing kinds of treatments (i.e., greater or lesser rewards or punishments) to human agents in a hypothetical afterlife on the basis of actions and/or decisions that these agents performed during their lifetimes’ (2016: 842). Caruso and Morris understand basic desert and moral responsibility to be conceptually interconnected; and they conceive of basically deserved treatments in terms of rewards and punishments. As indicated already, I question both of these moves: one might be basically deserving of certain responses to wrongdoing without being morally responsible; and one might basically deserve a response to one's behavior that is not a reward or punishment. But Caruso and Morris's thought experiment (inspired by Galen Strawson's Reference Strawson1994 conception of ‘heaven and hell’ responsibility) is helpful in allowing us to test our intuitions about what responses count as basically deserved, since (we are to imagine) the all-knowing divine judge would not mete out rewards and punishments on the basis of consequentialist or contractualist considerations, and (we might suppose), such divine treatment would not be affected by relational features such as a lack of standing to blame. (I take it this is why Caruso and Morris say that the divine judge did not necessarily create the human agents—to avoid the issue of a creator god's complicity in their wrongdoing.) Thus, we may use their thought experiment to support the conclusion that protest might be basically deserved even in the absence of moral responsibility; for we can coherently imagine a divine judge taking a symbolic stand against the evil actions that humans could not avoid committing due to a lack of free will. Such protest would serve in part to reveal to those who committed these acts the evil that is attributable to them, as well as to cast judgment upon it. If there is no further punishment (we might imagine that after the judgment has taken place, God reforms humans, who then live harmoniously ever after), such divine protest would seem appropriate, and thus deserved.

One remaining question about my position is whether the protest that I think people can basically deserve, even without free-will-requiring moral responsibility, counts as a form of blame. Suppose blame is deserved only by the blameworthy, and to be blameworthy is to be morally responsible (Clarke Reference Clarke2016). If, additionally, as I suggested earlier, there is only one form of moral responsibility (version 1 of my thesis), then the protest at issue here is not a form of blame. However, if we consider the question under the assumption (version 2 of my thesis) that there are multiple forms of moral responsibility, only some of which are ruled out by the absence of free will, then the sort of protest that one can basically deserve even without free will may still count as a form of blame. Perhaps this is a reason to favor moral responsibility pluralism after all—if it seems to one that protest is indeed a form of blame. Yet, importantly, this form of moral responsibility is every bit as much a ‘basic desert’ form as that which requires free will.

One might think that I have just given a reason to favor semi-compatibilism, since we would seem to have the ingredients required for a robust form of moral responsibility: one can be basically deserving of blame in the absence of free will. But I think this is a mistake: for there are certain ways of holding responsible that would not be appropriate, if we lacked free will for our actions. If retributivism involves the intent to cause suffering or harm as payback for wrongdoing, then I think retributivist responses to people's behavior are ruled out in the absence of free will. (But protest need not be retributivist, since it need not aim at or involve the intent to cause suffering or harm.) In saying this, I come close to Scanlon's position. In a discussion of punishment, Scanlon distinguishes between the ‘condemnatory aspect’ of punishment and the ‘hard treatment aspect’, and he maintains that the ‘liberty of indifference’ is not required for one to deserve the former: ‘Purely condemnatory responses (as distinct from ‘hard treatment’) are justified simply by what an agent is in fact like, psychologically, as revealed in his or her actions. . . . The cost of simply being (properly) condemned can be justified simply on the ground that this condemnation is deserved’ (2013: 104–5). However, Scanlon does not think hard treatment can be justified on the basis of desert, even if people have the liberty of indifference; he maintains that hard treatment can only be justified on other grounds, such as consequentialist considerations. I would think that this is precisely the difference that free will would make: it would make people deserving of more severe or substantial forms of treatment.Footnote ⁵

We might pause here to consider an objection to the position I have defended. One might argue that my position is not significantly different from that of Pereboom or other free will skeptics who recognize other forms of moral responsibility not threatened by the absence of free will and who acknowledge that protest and other kinds of blame can be appropriate in response to the immoral behavior of non-free agents. Of course, such skeptics do not say that protest is basically deserved in the absence of free will—they say that it is a fitting way of holding responsible. But is the insistence that it is in fact basically deserved a significant difference or, in the words of one anonymous reviewer, ‘does it amount to a minor clarification that no one would mind conceding?’ After all, Pereboom and I seem to be in agreement over a seemingly more important point, that in the absence of free will, no one deserves treatment intended to cause suffering or harm as payback for wrongdoing. Is that not the main point at issue in the free will debate—whether we have the sort of freedom that could justify retributive responses to our behavior?

In response, I take there to be substantive disagreement between myself and the other parties of the free will debate, in that I am offering distinct grounds on the basis of which someone might deserve something and treatment of them might be justified. Pereboom and other free will skeptics favor ‘forward-looking’ or consequentialist justifications for the treatment of non-free agents: a criminal may be isolated in order that society be protected, while a friend may be forgiven so that friendship can be restored. Those who believe in free will, on the other hand, think that treatment may also be justified on the grounds that an agent had an important form of control over her action, making her the author of her deed: she did it on purpose and it reflected her intention and will, they might say. I am proposing that a response to one's immoral actions might be justified in the absence of forward-looking justifications and in the absence of free will, on the grounds that a person has a ‘deep self’—one that is revealed and implicated in her immoral action, even if she has not acted consciously or freely (a point I develop further in ‘Can the Unfree Person Have a Deep Self?’ [Vicens Reference Vicens, Visala and Vainioforthcoming]). Consider, for instance, a person who makes a callous or cruel comment—and suppose that free will skeptics rule that this comment was, for whatever reason, not made freely. I propose that the commenter might deserve protest, even if such protest cannot be justified by any consequentialist (or contractualist) considerations, simply in virtue of the intrinsic features of the agent and her action—in this example, the callousness or cruelty revealed in her comment. Callousness or cruelty are moral qualities that might serve as the target of protest and which arguably do not require a strong form of (current or precedent) control for their manifestation.

In summary, I have argued that if we assume with the free will skeptics that people lack moral responsibility, or at least a certain important form of it, we may still maintain that people may be ‘basically’ deserving of certain treatment in response to their behavior. In particular, protest is a response that may be basically deserved, on the grounds that it is a fitting response to the intrinsic features of a person and her actions. Questions remain, however (perhaps to explore another day): Is such protest in fact a form of blame? What other responses to behavior might be basically deserved besides protest?