Introduction

In 1444, a pig named Verray “killed and murdered” a girl out in the fields of Saint‐Prix, France (MacGregor 2019, 6). It is unusual that the records name the pig. It is not unusual, however, that the legal records name the deed as murder, a term that implied a deliberate, premeditated deed of malice. Indeed, it was quite common across Europe in the medieval period to convict and execute domestic animals for murder (Evans 1906; Humphrey 1987). There is no indication that this was regarded as an extraordinary matter. Standard judicial practices were followed, notably imprisonment and public execution (typically at the town gallows). Verray was a member of the moral community, to be held responsible like any other.

Today, we are likely inclined to regard these medieval animal executions as forms of abuse. Even at that time, philosophical and theological thinkers had ample grounds for looking askance at such doings. According to the broad Aristotelian tradition of reflection, humans share appetite, emotion, and voluntary action with other animals, but deliberation and choice or decision (prohairesis) are confined to human beings; nonhuman animals do not choose one thing reflectively in preference to others (Deane‐Drummond 2009, 207). Yet the widespread European practice of criminalizing nonhuman animals offers powerful testimony of a social world within which there was no tidy human‐animal binary. “Placing the animals both physically and symbolically in a space usually occupied by humans imposed the normative boundaries of the human community on the nonhuman,” notes Lesley MacGregor (2019, 16). “By positioning an animal as a defendant, the very boundaries between humans and animals became less rigid.”

As in the present day, the boundaries between humans and animals are once again becoming less rigid, what are we to say about Verray and his executors? The actions for which human beings are properly held responsible are typically seen as our “rational,” “voluntary” actions, and as such they are our distinctively human actions. For instance, in an influential contemporary study of responsibility, R. Jay Wallace focuses on identifying “the conditions that make it morally fair for us to adopt the stance of holding people responsible” (1994, 5). He argues that it is fair to hold others responsible when they have “the ability to grasp and apply moral reasons, and to govern one's behavior by the light of such reasons” (1994, 1). To have this ability is to have the “normative competence” that is responsibility. Since grasping and applying moral reasons is a distinctively human activity, there would seem to be little to learn here from consideration of nonhuman animals, little here that is “animal.” It was not fair to hold Verray responsible for killing the child, not appropriate to label the killing a murder. The normative competence that is responsibility is a purely human affair.

Intention and Agency

Human bodies can, to be sure, be part of a causal chain of events without our being responsible for these events, in the sense of these being properly attributable to us as agents. Say that someone picks me up and throws me over a bridge onto a person walking on the footpath below. This is not something that I did, but something that happened to me, even if my falling body kills the pedestrian. Causal involvement (we could call this causal responsibility) is distinguishable from agent responsibility; in the latter case, we trace responsibility back to the thrower's own agency. There are also things that we humans do, in a broad sense, for which we are not ordinarily responsible—sneezing, for instance, or digesting. A bit of pollen finds its way into my nose, and I find myself sneezing. Here, it makes no sense to ask me, “why did you do that?” It was not something that I did on purpose or intentionally, and in this case, we can point to no other agent's intention. I might, however, seek to induce myself to sneeze, say, by sniffing pepper. When I do this, my sneezing becomes an act of which I have ownership, something for which I am properly responsible. I did it intentionally, and it is reasonable to ask me why.

Elizabeth Anscombe's hallmark study of intention, published in 1957, corrected the temptation to locate intentions through introspection (Anscombe 2000). Intentions are not a distinctive kind of mental content. Rather, we have to do with intentional action whenever why questions can meaningfully be posed to an agent, and an answer be given in terms of the agent's beliefs and purposes. Importantly, this means that habitual action can also be intentional action, for which one can properly be held responsible. You can legitimately ask me why I turned right at the corner, even if the answer is simply that this is the path of my usual evening commute, and I turned right without needing to think about it, without needing to deliberate or decide. I can also be responsible for omissions, say, for failing to stop at a red light. Here, I am responsible for my omission if I saw that the light was red, knew that drivers must stop at a red light, and the brakes of my car were properly functioning. I am also responsible even if the brakes of my car failed to work, if I knew that my brakes were in need of repair. In these cases, I am responsible for my omission because I am answerable for it; again, it makes sense to ask me why I failed to stop, in a way it would not make sense if I came from another planet and did not know the rules of the road.

Anscombe's intervention situated intention and agent responsibility within the social context of accountability, focusing intention on exchanges of reasons rather than on hidden brain states. Yet if responsible agency has its proper place within reason‐giving exchanges, this would seem to underscore its distinctively human character, or at the very least, its confinement to creatures who employ complex forms of symbolic language with generative, hierarchically structured syntax, of which humans are the only known instance. Yet matters turn out to be considerably more complex. The criteria for holding others responsible are murky and contested. Practices of responsibility are bound up with social expectations and emotional reactions, many of which are shared with other social animals. Practices of holding others responsible for their actions in various ways outrun both the capacity to give an account for one's actions and the capacity to reflect on the justifiability of holding another responsible. Here as elsewhere, it is salutary to attend to human animality in order to arrive at a more adequate grasp of just what about responsibility is distinctively human.

Plants sense the world, responding to changes in the environment. They also communicate with one another by way of hormones and ions. Making sense of a plant's activity involves reference to the cascade of changes that take place in plant cells and tissues in the presence of sunlight, water, and a host of specific chemical compounds. Animals, too, sense and respond to changes in the environment. In the case of many animals, however, these responses are not a direct result of environmental changes. They are, rather, mediated, not just by what the animal senses through sensory organs such as eyes, ears, nose, tongue, and skin, but also by what the animal believes, desires, and seeks. As Alasdair MacIntyre notes, many nonhuman animals are “guided by a kind of practical reasoning that is exhibited in their taking this to be a reason for doing that” (1999, 60; see also Deane‐Drummond 2019, 95–96; Berkman 2009, 26–27).

We cannot ask a mole or a vulture why it did what it did. However, we can meaningfully ask these questions about the mole or vulture, and the answers that we give go through the animals’ own beliefs, desires, and emotions, for example, “the turkey vulture swooped down onto the leak in the gas pipeline because it thought the leaking mercaptan came from a decomposing carcass” (e.g., Smith and Paselk 1986). That is, we regard many animals as intentional agents, doing things on purpose, and altering their behavior as their beliefs change: “disappointed at finding no carcass, it flew off and did not return to the site.” We tell the same sorts of sense‐making stories about fellow human agents. Such stories are important ethically, bound up with our affective‐cognitive evaluative construals of intentional action.

Reactive Attitudes and Holding Responsible

What, beyond intentional agency, is requisite for agent responsibility? Perhaps the most significant discussion of responsibility in the philosophical literature is P. F. Strawson's 1962 article “Freedom and Resentment,” which continues to serve as a key point of reference for contemporary discussions of the topic. Strawson's key contribution was to seek to make sense of responsibility by attending to acts of holding responsible, and to see holding responsible in turn as a matter of being disposed to adopt some reactive attitude toward another. Our reactive attitudes are emotions that we feel toward others insofar as we are bound up together with them in webs of interpersonal relationship. Our reactive attitudes reveal the expectations that we have of others and how they will act toward us. Strawson was particularly interested in the reactive attitude of resentment, noting that if someone steps on my hand while trying to help me out of a tight spot, the physical pain may be no less than if that person were to step on my hand out of spite. However, in the latter instance, I would be disposed to feel a resentment that would not be present in the former. When we feel resentment, we are disposed to blame someone, to hold them responsible for violating my legitimate expectations concerning how they will treat me.

Wallace, building on Strawson, noted that while we are subject to a wide array of emotions toward the persons with whom we are in relationship, there are special features of the reactive emotions of resentment, indignation, and guilt (1994, 12). These have a distinctive connection with expectations: “episodes of guilt, resentment, and indignation are caused by the belief that an expectation to which one holds a person has been breached” (12). If we cease to believe that an expectation has been breached, or cease to believe that the other can meaningfully be held to this expectation, we cease to feel the reactive emotion: “to be subject to the reactive emotions is to take this stance toward a person, and to adopt this stance is in turn to be subject to the reactive emotions” (24).

Strawson's contribution productively shifted the conversation about agent responsibility away from debates over indeterministic freedom. Our ordinary practices of responsibility‐holding do not assume or require strong freedom of the will of the sort incompatible with determinism (incompatibilist freedom) (Wallace 1994, 3). The key question is rather under what conditions it is appropriate to adopt particular reactive attitudes toward another agent (De Mesel 2017, 311). This is an inquiry, moreover, that goes on within the ethical. All sorts of facts may be relevant, but at the end of the day, the key question for this inquiry is whether or not it is fair to hold a given individual or class of individuals responsible—and this is an ethical, not merely an empirical, question.

The reactive emotions give us a productive entry point into considering the involvement of nonhuman animals in responsibility relations. Reactive attitudes, as noted above, are emotions that we feel toward others with whom we are in relationship. Strawson is particularly interested in resentment, but he names also “gratitude, forgiveness, anger, or the sort of love which two adults can sometimes be said to feel reciprocally, for each other,” as well as “hurt feelings” (Strawson 2020, 111). Strawson considers these to be attitudes that belong specifically to “involvement or participation with others in inter‐personal human relationships” (116); he does not consider whether they are felt by, or properly felt toward any nonhuman animals. Our participant reactive attitudes reveal the expectations that we have of others and of how they will act toward us. They are to be contrasted with objective attitudes, which we do at times take up toward one another: “we look with an objective eye on the compulsive behaviour of the neurotic or the tiresome behaviour of a very young child, thinking in terms of treatment or training” (116). When we do this, we treat the other in a behaviorist fashion, as an object of manipulation, or (perhaps in the case of someone who is deranged, hostile, and very strong) a force to be “understood and controlled,” fought or fled (119). We suspend ordinary interpersonal involvement, ordinary expectations of others and ordinary reactions to other's good or ill will. The presence of the reactive attitudes reveals, Strawson thinks, that our ordinary expectation and demand of others is “for the manifestation of a certain degree of goodwill or regard on the part of other human beings towards ourselves; or at least … for, an absence of the manifestation of active ill will or indifferent disregard” (122). At times he describes them in ways that demand self‐conscious awareness and reflection, as when he says that those to whom one takes up participant reactive attitudes are those with whom one can reason (116). But when he describes them as “natural human reactions to the good or ill will or indifference of others towards us, as displayed in their attitudes and actions,” (117) those who work or live intimately with chimpanzees or dogs or other socially‐living animals, are likely to think the “human” can be removed from this description.

Reactive Attitudes in Chimpanzees and Other Social Animals

Social animals are highly dependent on one another. They have a host of expectations concerning how the other members of the group will behave toward them. When these expectations are violated, we observe not simply surprise, as one might if the ground suddenly gave way beneath one's feet and one tumbled into a hidden cave, but reactions like anger and resentment, reactions quite specifically to the violation of social expectations of other members of the group (Deane‐Drummond 2019, 148–53). The alpha male in a chimpanzee colony, for instance, expects to be “greeted” by other members of the group, who issue a sequence of short pant‐grunts while making a deep bow and looking up at the dominant individual (de Waal 1998, 78). He expects to be allowed to step or leap over the greeter in a so called “bluff‐over.” Greetings and bluff‐overs serve to confirm the dominance hierarchy, thereby keeping peace within the community. But the dominance of the alpha male can be challenged, a challenge that most often begins when another male begins to avoid greeting him, and begins to put on aggressive displays in his vicinity, pacing around with hair on end (de Waal 1998, 83–85). In such situations, the alpha male expects other subordinate male and female members of the group to come to his aid in hooting, screaming, and potentially physically attacking his challenger. He solicits this assistance by screaming and by embracing potential supporters. When a power takeover is underway, the outgoing alpha male may find himself bereft of supporters, who have now shifted their loyalties to the challenger. In this situation, the alpha male's despair is expressed by way of temper tantrums, which for chimpanzees are the typical behavior of children who are being weaned. Here is de Waal's description of one such situation in the Arnhem chimpanzee colony:

Yeroen began to have tantrums after the conflict had been raging for about a month. With an unerring sense of drama he would let himself drop out of a tree like a rotten apple and roll around on the ground screaming and kicking, while all the time Luit was displaying. These hysterical outbursts gave an impression of scarcely suppressed despair and abjectness. When he had regained some of his self‐composure, Yeroen would run yelping to the females, throw himself on the ground a few meters away, and stretch out both hands to them. This was not a begging gesture but a beseeching gesture, beseeching them for their support. If the females refused to help, or even went out of their way to avoid him, Yeroen would once again break down and have a tantrum. (1998, 98–99)

Yeroen, like chimpanzee toddlers whose mothers relent and allow them to nurse, would immediately snap out of his tantrum if he received the support he sought. Now, De Waal may anthropomorphize a bit exuberantly. Still, it is difficult to capture what is going on in this situation without reference to participant reactive attitudes. Yeroen has a host of social expectations of other members of his group. As these are serially violated, he experiences and expresses an array of emotional responses ranging from anger to despair. More than this; he does not merely express his emotions randomly; he displays these emotions to particular other individuals with whom he stands in specific forms of social relationship.

Strawson writes at times as though human beings have a set of natural and universal expectations concerning how other human beings will treat us, that is, with good will, or at least the absence of malice. Indeed, he characterizes the reactive attitudes generically as expressing a “demand for inter‐personal regard” (123). But in fact our expectations vary tremendously, depending on the particular social relationships in which we stand. I expect thoughtful love and care from my spouse, and may be angry if he gives me a gift, however extravagant, that shows that he does not know my tastes. I may, in contrast, be quite grateful for a useless or ugly hostess gift, regarding it as a generous sign of good will from someone from whom I expect little. And while generally resenting the enemy with which my country is at war, I may nevertheless be grateful to those caring for me as a prisoner‐of‐war, if they treat me with some modicum of kindness. The reactive attitudes are thus highly flexible and keyed to particularities of relationship. We see much the same in the participant reactive attitudes evident in the scene I have just described from the Arnhem chimpanzee colony. What one chimpanzee expects from another is dependent on gender, age, social position, family relationships, and friendships. A subordinate individual does not expect to be greeted by the alpha male, and does not resent not being greeted. A challenger does not expect nonfriends to support his challenge and does not resent their lack of support. Rough play is tolerated in a youngster, not in an adult. Wild chimpanzees anticipate hostility, not good‐will, from other chimpanzee colonies.

Now, we can of course debate precisely which reactive attitudes can be attributed to chimpanzees. This will depend a good deal on how much cognitive complexity we build into the concepts. Resentment is anger at having been offended, injured, or treated badly by another. Is it possible to feel resentment without having the concepts of offense, injury, and fairness, and second‐order reflection on their reasonableness? If having these concepts requires having these words, and the ability to debate their application, then young human children are as excluded as nonhuman animals (Deane‐Drummond 2019, 55). But if we are willing to attribute resentment to preverbal children who differentiate unintentional harms from intentional injuries, and whose reactive emotions are moderated accordingly, then we should be willing to do the same for chimpanzees and other animals whose reactions display similar forms of differentiation. And we do see these sorts of differentiated reactions in nonhuman animals. Capuchin monkeys who see that the human experimenter has both cucumber and strawberry available become enraged if the experimenter gives them cucumber rather than strawberry rewards for completing a task—even though they are perfectly content to receive cucumber if no strawberry is available (Brosnan et al. 2010, Tomasello 2016, 33–34). Capuchin monkeys and chimpanzees also compare outcomes with other subjects, showing obvious signs of displeasure and anger and refusing to continue to participate if the partner receives better rewards. “Chimpanzees were sensitive to harmful behavior and/or intent, but not to simple disparity over which the partner had no control” (Jensen et al. 2007, 13048). I conclude that it makes little sense to regard the reactive emotion of resentment as unique to humans. We are not the only animals that distinguish bad luck from malice, intentional from unintentional harm, and respond through the expression of resentment. Like our own reactive attitudes, in other animals, too, these are backward‐looking evaluative responses. Even if they potentially function prospectively, to mold the future behavior of the target, this is not their proximate cause, any more than our own resentment of others is ordinarily a tactic to manipulate or behavioristically shape their future activity.

What of indignation, anger at a norm's having been violated by another? If this requires a capacity for explicit articulation of the norm and its violation, then it will be confined to those who have the use of conventional language. But if a norm in this context is a social expectation, then we do witness anger at the violation of such social expectations. Third parties intervene in response to aggression on the part of freeloaders in experiments on chimpanzee cooperation (Suchak et al. 2016). Moreover, the alpha male rushes to the defense of females and subordinate males under attack by a rival. Indignation at violations of close social associates (kin, friends, and allies) appears to be quite common. In humans, too, indignation is most commonly felt on behalf of those with whom we feel some sort of identification.

What of forgiveness? Chimpanzee societies are full of conflict and just as replete with reconciliation. It is common to find adversaries, not long after a fight, grooming one another in order to restore peaceful relationship (Samuni et al.2021). Again, if we require conceptual explicitness here, we are unlikely to recognize this as forgiveness. If, on the other hand, we are focused on the character of reactive attitudes as social emotions, we find in chimpanzees an intimately familiar range of emotional reactions to the doings of socially significant others.

But do chimpanzees feel guilt? About as little, apparently, as young human children. A female does not want to get caught having sexual relations with a subordinate male, as the alpha male will punish her, and she goes to some lengths to perform the act discreetly, out of view of the alpha male. But we do not see her hang her head and mope after her secret dates. Guilt, a specific negative feeling associated with the consciousness of having violated a norm, or, better, with having done something one knows to be wrong, requires both awareness of the norm and identification with it. In young children, chimpanzees, dogs, and other highly social animals, we see awareness of social expectations/norms, and anticipation of the negative reactions of others in the face of violation of these expectations/norms, but we do not clearly see emotional identification with these norms (Horowitz 2009; Boehm 2012).

Current research is enabling us to provide ever more fine‐grained accounts of some of the underlying cognitive differences that are at play here. For instance, chimpanzees are aware of what others know and do not know, and act accordingly. However, they do not represent what others believe (Martin and Santos 2016, Horschler et al. 2020). Adult humans grasp that others believe things that they themselves know to be false. This involves “representational relations,” which allow “an organism to predict how an agent will behave when that agent's representation of the world conflicts with the organism's own idea of reality,” and it is a capacity that develops within the first two years of human life (Martin and Santos 2016, 374).

Chimpanzees, in contrast, appear instead to use “awareness relations” simply to track whether another individual is aware of the same things as the self. If not, the chimpanzee adopts an agnostic attitude concerning what the other individual knows (Martin and Santos 2016, 379). The ability to attribute false belief is very important for our own ethical assessments, because it has important implications for determining intentionality. For instance, if Tondra has managed to get the baby to take a nap by riding around town in the car, and Bill believes that the baby is napping in her crib, I assess Bill's action of leaving the baby in the hot car while he shops at the grocery store very differently than if he did not have this false belief. Without an ability to represent and attribute false belief, patterns of responsibility‐holding would go off the rails in key respects—we would be forced to remain agnostic, as chimpanzees are, when it comes to a host of matters relevant to assessing others’ actions.

I conclude that at least some of our fellow social animals target one another (that is, fellow members of their social groups) with participant reactive attitudes. These attitudes are not, then, as Strawson would have it, simply natural human responses. They are natural responses of social animals, animals whose relationships with one another are of great significance, and whose responses to the world and one another are mediated by emotions. They require the recognition of individuals, the association of individuals with their actions, memory of actions, and the appreciation of how actions affect others. They also require that certain social expectations or norms be in place, which can be either affirmed or violated by individual actions. Emotions, as we have noted, have cognitive dimensions; they are complex evaluative construals of an animal's social worlds. Whether these are interpersonal depends of course on how “person” is to be defined, but certainly they are interagential, and to be contrasted with objective attitudes, as for instance a response to a branch that falls onto one's head.

Social animals typically take up objective attitudes toward prey, regarding them as objects of analysis and manipulation, even if also as subjects that see and hear certain things and not others. Prey are not objects of reactive attitudes because predator and prey do not participate in a common social world of shared norms and expectations (Bekoff and Pierce 2009, 16). This is not to insist that reactive emotions are necessarily confined to the intra‐specific domain. Where symbiosis or mutualism occurs, it is perfectly reasonable to expect that reactive attitudes may be expressed across species. Certainly, we see this to be the case where human beings live in close relationship with members of other social species, including dogs, elephants, and chimpanzees.

From Resentment to Responsibility

Insofar as there are other social animals that target one another with participant reactive attitudes like resentment and indignation, there are other social animals that hold one another responsible, albeit only in the sense that their emotional responses embody assessments of others in accordance with shared implicit social norms and expectations. It is critical to acknowledge that all of this is up and running long before there is conscious awareness of shared norms, and long before the question can arise of whether these attitudes are or are not fair or deserved or the norms justified. Symbolic language makes it possible for reactions to be articulated, made explicit. This in turn makes it possible for us to differentiate between our reactions and the judgments that they embody, and thus also to ask whether the judgments are accurate and the reactions justified. We can then recognize that it is proper to hold another responsible only if they are worthy of some reactive attitude (Zimmerman 2010, 110).

We see no evidence that other social animals consider whether their reactive attitudes are or are not justified, or whether their targets are or are not deserving of them (see Zimmerman 2010, 110). It is difficult to conceive of how such reflection could get underway in the absence of language of a high degree of complexity, allowing for the formation of recursive judgments, judgments about judgments (MacIntyre 1999, 54). Questions about the justifiability of reactive attitudes are questions that human beings routinely ask of themselves and one another, and also, at times, of other animals. While such reflection is the stuff of daily life, it is also an arena for considerable debate. Philosophical reflection here characteristically seeks to describe actual human practices of holding responsible, seeking to articulate the implicit principles that make sense of these practices, in order then to apply these principles in ways that identify inconsistencies, clarify confusions and recommend changes that would lead to more coherent, more justified practices. We ask, first, when do we tend to adopt particular reactive attitudes? And when do we reflectively judge that it is appropriate, fair, and justified to do so?

Conditions for Attribution of Agent Responsibility

Wallace notes that while we typically consider adult human beings as appropriate targets of blame, we are reluctant to blame “even those who have committed quite horrible crimes when we learn that they were subject to unusual deprivation in their youth” (231). He argues that we can justify this reluctance by showing that it is unreasonable to expect such individuals—who have suffered “physical and verbal abuse, emotional neglect and inattention, withdrawal of love and concern, extreme arbitrariness and hypocrisy in the application of punishments and rewards, and an atmosphere of violence, insecurity, and hopelessness”—to possess the powers of reflective self‐control, the capacity to grasp appropriate reasons for acting and to control their behavior in light of these reasons (226, 231). For blame is not merely a disposition to sanction another for something they have done (51). It is an evaluative construal that incorporates a judgment that the other ought not to have done this thing. It is not merely a reaction to the violation of certain expectations, but a way of holding the other to these expectations. It is not about retribution, in the sense of thinking that it would be good for the other to suffer harm (60). And while it is an emotion, expressing blame and other reactive emotions is not about venting or desiring to inflict harm, but about “demonstrating our commitment to certain moral standards, as regulative of social life” (69). When we judge that it would not be fair to hold the other to these expectations, our evaluative construal shifts, and with it, our emotions.

But do we in fact treat the powers of reflective self‐control as the necessary condition for agent responsibility? Others have argued that this is too strong a demand, noting that we hold ourselves and others accountable for a host of things over which we lack active control, including involuntary reactions, feelings, and beliefs. We commonly consider persons responsible for things like careless inattention, racist attitudes, and the like, which are not under their voluntary control and which they did not bring about through voluntary action: “emphasizing control has distracted us from noticing that we have very little control … over a wide range of beliefs, cares, inattentions, and even volitions that we commonly consider persons responsible for and struggle over whether to punish or to forgive” (Couenhoven 2013, 127).

Blame, Respect, and Repentance

Is there good reason for regarding ourselves as justified in holding one another responsible for things over which we lack active control, but which meet the ownership and proper function requirements? Say I realize that I have been wholly ungrateful to someone who has been very generous toward me, and in fact have responded with deep resentment.¹ Having come to recognize this, it is appropriate that I repent, reproach myself for my ingratitude, and take responsibility for it. In repenting, I need not see my ingratitude as something voluntary, or something over which I had control. In order to repent I must, though, see it as my own ingratitude, not simply as something that happened to me (1985, 15). Robert Merrihew Adams, developing this example, argues that

The deepest reason for accepting this responsibility, if I am right, is that it is rightly ours. It is important for a correct ethical appreciation of one's own life. To refuse to take responsibility for one's emotions and motives is to be inappropriately alienated from one's own emotional and appetitive faculties. (15)

This is a slippery matter, since repentance involves its own kind of alienation, one that Adams does not regard as inappropriate. Repentance “enlists the desire to satisfy one's responsibilities in support of the desire to change; whereas if one says, ‘I'm not to blame for my ingratitude because I can't help it,’ one takes some of the pressure off oneself by seeking refuge in an excuse” (15). We are appropriately alienated from our involuntary faults when we recognize them as ours, as woven into the selfhood with which we reflect and act, while also taking up a stand against them, regretting our ownership of these faults and desiring to be rid of them. When we do this, we regard ourselves as active rather than as passive, as mere sites for faults that play themselves out, as it were, in us.

When repentance is appropriate, moreover, so is reproach and blame from others, the appropriate purpose of which “is not to crush us but to lead us to repentance, and to acknowledge moral realities” (23). When we blame another, we reproach them for some fault, which might include faulty attitudes and desires over which they have no direct control. This attitude expresses respect for the other, as one we take to be capable of taking responsibility for herself, repenting of her faults, and throwing her agency behind the good. It is justified, not merely in a forward‐looking way, because it has the potential to trigger this process of transformation, but primarily because it is appropriate to respect the other in this way, by issuing this kind of summons to the other person as responsible agent.

What, then, would Adams have to say in response to Wallace's reluctance to blame the person subject to trauma, abuse, and deprivation? Adams's view, I believe, is that even in these circumstances, blame appropriately expresses respect for the other as capable of taking responsibility for herself and throwing her agency behind the good. At the same time, he would add that blame does not itself always justify punishment. Blame can go hand in hand with mercy and recognition of just how challenging it may be for this individual to change for the good. Wallace's attitude inappropriately objectifies the traumatized subject, encouraging the individual to be inappropriately alienated from her desires and attitudes, and thereby both disrespecting her and likely further weakening her capacity for responsible agency.

This is a powerful line of argument, which offers an important corrective to approaches, like that of Wallace, which limit attributions of responsibility to the sphere of reflective self‐control. Adams is, I believe, right to hold not simply that it is the case that we do often praise and blame others for attitudes and desires which they have not chosen and over which they have no direct control. It is also the case that these responses are often appropriate and justified, rather than a set of reactions and associated social practices that we should set about reforming.

Partial Self‐Ownership

As we have seen it may seem unjustifiable, even irrational, to hold persons responsible for actions flowing from things about themselves that result from constitutive luck, and over which they have no direct control. Until, that is, we reflect on how it is that we express respect toward others by holding them responsible for their desires and attitudes, by treating them as having the potential to take responsibility for themselves and either to stand behind who they are or to distance themselves from who they are by standing instead for the good and thus creating the conditions of the possibility of positive change. Targeting and expressing the reactive attitudes may thus not only be permissible, but may in fact be something that we owe to another, even when she lacks direct control over who she is.

We respect those we target with reactive attitudes by taking them to be capable of taking responsibility for themselves and throwing their agency behind the good. In some instances, we are uncertain of whether this is the case, but we rightly resist exempting another too quickly from responsibility‐holding, since in exempting another, we take up toward them the objective attitude. Someone judged to be not guilty of a crime by reason of insanity is judged not to have known the nature or quality of what they were doing or not to have known that it was wrong. This is the paradigm of a condition that exempts from blame and justifies the adoption of the objective attitude. Yet persons can slip in and out of insanity, and it is appropriate in the face of uncertainty to keep the door open to readoption of participant reactive attitudes. Here, respect is demonstrated through taking it to be an open question whether someone once judged insane is now still insane. Participant reactive attitudes might be expressed in this context in a kind of tentative and probing way, to invite the other to take responsibility for herself.

We are, I think, quite familiar with this sort of softened, probationary form of responsibility‐holding when it comes to young children. Here, we express the reactive attitudes, but in a somewhat tentative or moderated fashion, sensitive to the range of reasons to which this youthful agent can be expected to be responsive. Having witnessed Spot, the family dog's, yelp of pain at having been struck on the nose, toddler Ace can be expected to grasp that hitting the dog on the nose is not to be done. Early on, Ace may only grasp that hitting Spot reliably elicits Mommy's displeasure, but the expression of that displeasure invites Ace to repent of his Spot‐hitting, thereby distancing himself from Spot‐hitting Ace and identifying himself with a potential non‐Spot‐hitting Ace, integrated into his emerging moral identity (Blasi 1980, 2005). Mommy respects that potential in Ace by expressing the participant reactive attitudes toward him, albeit in this moderated fashion. Being held responsible necessarily outpaces holding oneself responsible, as it is only by way of the former that we learn the latter—even as we are justified in holding responsible only those who are capable of identifying their agency with the good, and thus rightly moderate our reactive attitudes as the scaffolding of responsible agency is built up.

Just as being held responsible runs out ahead of holding oneself responsible, ontogenetically, so also, as we have seen, the reactive emotions, and their expression, run out ahead of the possibility of questioning whether these are justified. Among our fellow social animals, we see the unhemmed expression of various participant reactive attitudes, free of reflection concerning whether these attitudes are or are not justified, and their targets deserving or not. Holding responsible was present long before there were responsible agents. In the absence of self‐conscious awareness and the capacity to reflect on one's reactive emotions or their justifiability, there are differential expectations of animate as opposed to inanimate things, of conspecifics as opposed to members of other species, infants as opposed to adults, individuals who conform to social norms as opposed to those that do not, and so forth. The reactive emotions are expressed in ways finely sensitive to such distinctions. In human beings, these distinctions have become open to reflective assessment, but we begin with dispositions for interagential assessment that are broadly shared with our nonhuman kin.

Conclusion

I have argued that since other animals experience and express a range of reactive emotions, they also participate in ways of holding one another responsible, even if these are unreflective and unselfconscious. Grasping this casts our own capacities for responsibility in a new light. Couenhoven aptly notes that we are “agents who, though they have a certain unity, also find that they are conduits of influences they do not fully understand or even perceive and who cannot pull themselves together” (2013, 133). As Adams says, “who we are morally depends on a complex and incompletely integrated fabric that includes desires and feelings as well as deliberations and choices” (1985, 10). This does not render our practices of accountability illegitimate, but it does mean that responsibility is multilayered and comes in degrees: “the more strongly, broadly, or deeply one owns a state of affairs, the higher a degree of responsibility one bears for it” (Couenhoven 2013, 128). We may be justified in regarding someone as the owner of an action without being justified in blaming them for that action, or sanctioning them for it.

What of Verray the pig? We know too little of the story here, too little about what provoked Verray's killing of the child. Most likely the child did something that was perceived by Verray as a threat. Pigs are highly social and employ aggression to establish strong dominance hierarchies. Early socialization, via the mixing of siblings and of piglets from different litters, facilitates the formation of dominance hierarchies without significance violence or injury. Social deprivation leads to heightened aggression (Camerlink et al. 2018; Weller et al. 2019). The child may have been perceived as challenging Verray's dominance, and as failing to display appropriate signs of submission. One way or another, it is likely that the child had not yet learned (as doubtless many medieval humans did learn, living as they did in intimacy with domesticated animals) how to navigate effectively amidst implicit swinish norms and social expectations. A different kind of respect for Verray was expressed by executing him for murder, holding him accountable as a member of the community. Yet nothing that we know currently about pig cognition suggests that Verray could have known the nature or quality of what he was doing or that it was wrong, could potentially have thrown his agency behind the good by refraining from killing. The expression of respect thus misfired, as it does when we fail properly to apply the insanity defense in the trial of a human being. Yet this is not to deny that certain forms of responsibility‐holding can be appropriate in human‐nonhuman relationships.

Practices of accountability, starting with the reactive emotions, work to knit us together in communities of shared expectation and response, and there is a competence here that humans typically share with other social animals and that enables interspecific as well as intraspecific social relationships. Along with other social animals, we regard one another as agents, who act intentionally in ways expressive of who they are and what they are for. None of us is ever ultimately responsible for ourselves and our doings, yet we rightly express respect for one another through our practices of accountability, inviting one another to take ever‐fuller responsibility for who we are and what we do in answerability to the goods we are capable of grasping and loving. This is a lifelong process. In order to become the kind of agents capable of reflectively being for the good, we—like other social animals—first need secure attachment and stable social bonds—we need to be embraced by communities of care and nurture, in which we are cherished even as the messy bundles of impulses that we are. Trauma leaves chaos in its wake; gentle care affirms us as potential objects of the reactive attitudes, potential participants in societies of shared responsibility to one another and to the good.

Acknowledgements

This project/publication was made possible by a grant from the John Templeton Foundation, Collaborative Inquiries in Christian Theological Anthropology. I am grateful to the Foundation for its support and to the leaders and fellow members of the project, who offered helpful feedback on a draft of this essay. I would also like to express my gratitude to Celia Deane‐Drummond, Mary Ann Myers, and Eric Priest for their invitation to contribute to this Symposium in Zygon: Journal of Religion and Science and for their thoughtful input along the way.