Trauma and Compassionate Blame

  • Alycia W. LaGuardia-LoBianco (Grand Valley State University)


A common question in moral philosophy concerns how we should regard traumatic histories that have influenced wrongdoing. In this paper, I argue that one standard line of response—survivors are exempt from blame for a given wrongdoing because of a traumatic history that influenced this wrongdoing—is problematic. Instead of trying to determine how trauma categorically exempts survivors from blame, I argue that we should start from the fact that survivors are members of the moral community who can be as blameworthy as anyone, and then ask how considerations of trauma might impact a moral response to their wrongdoing. I consider and reject three attempts to exempt survivors on the grounds of traumatic histories, arguing that each risks divesting survivors of moral agency. I then develop an account of compassionate blame, a unique attitude in which we can blame the survivor for their wrong while also taking compassion towards the survivor in light of their traumatic history, and argue that this is the appropriate moral attitude to cultivate in these cases.

How to Cite:

LaGuardia-LoBianco, A., (2021) “Trauma and Compassionate Blame”, Ergo an Open Access Journal of Philosophy 7. doi:



Published on
23 Oct 2021
Peer Reviewed

The impact on me of realizing that our histories as agents are webs in which all that is a product of our wills is supported by things that are not is an ethical impact. It moves me towards humility and mercy, virtues that acknowledge the unfairness of life but also presuppose a morally structured context of interaction.

—Claudia Card, The Unnatural Lottery

1. Introduction

Those who love us can hurt us the most. And sometimes, those who hurt us—even when they love us—do so because they have been deeply hurt themselves. How we come to terms with this reality can test the limits of our relationships, our moral responses, and our compassion.

This problem is especially apparent when we look at the impacts of trauma. Trauma is not only devastating in its immediate effects. One characteristic feature of trauma is its ability to partly shape who a person is. And tragically, as a result of being shaped by trauma, some survivors may inflict certain harmful wrongs onto others. For instance, a veteran suffering from posttraumatic stress disorder may violently lash out at her family; a child who grows up in an abusive household may repeat that abuse in his adult relationships. This transmission of harm that can result from trauma poses a difficulty for those in close relationships with survivors, especially concerning everyday practices of blame. Though the harm at issue may be significant, the recognition of a survivor’s trauma may make blame seem inappropriate. In light of the influence of trauma, survivors may be exempted from blame on the grounds that they were unfairly shaped to commit those harms.

Philosophical discussions about agents who are “peculiarly unfortunate in . . . formative circumstances” (Strawson 1962/1974: 9), often try to determine whether certain histories (like trauma) mitigate responsibility or blame for behaviors that result from these histories.1 I argue that instead of trying to determine whether trauma is a mitigating condition, we have good reasons to start from the assumption that trauma survivors are responsible, if sometimes burdened, moral agents who can be blameworthy for the wrongful harms they have committed, and then ask how considerations of trauma may impact our blaming practices. Approaching the issue this way is not only truer to our interpersonal practices in which we do regard survivors as part of our moral communities who can be as blameworthy as anyone, but it also resists the dehumanizing view of survivors as exempted from some forms of moral regard. However, this does not mean that trauma makes no difference to our blaming practices. The recognition of trauma calls out for an amendment to our blaming practices, not the rescinding of them. And this, I argue, should involve cultivating a compassionate form of blame.

In this paper, I propose a notion of compassionate blame and argue that it is an appropriate response to survivors whose transgressions stem from trauma. Compassionate blame is not a composite of blame and compassion, but rather a unique attitude distinguished by a set of beliefs. Additionally, the intentional object of the emotional component of compassionate blame is broader than that of blame or compassion alone. Rather than focusing only on the survivor qua transgressor or the survivor qua sufferer, compassionate blame takes as its object the survivor’s broader history that connects these two aspects. This response acknowledges how the survivor’s trauma may have shaped them to be responsible and blameworthy for certain harmful wrongs, but does so in a way that is attuned to their suffering from that trauma. Because there is this connection, blaming someone for a transgression that results from trauma must be bound up with compassion; to recognize the source of the harm is not only to contextualize it, but also to realize the tragedy of a situation in which suffering begets suffering. Compassionate blame thus responds to survivors as fellow beings who have suffered while still recognizing them as responsible moral agents who have done wrong.

In order to motivate the notion of compassionate blame, I first briefly discuss psychological research on intergenerational transmission of trauma. I then discuss three reasons trauma might exempt survivors from blame: the survivor has suffered, and we ought not blame those who suffer; trauma is a form of constitutive moral luck, and one cannot be blameworthy for that which is outside of one’s control; and trauma damages agentive capacities so as to undermine responsibility. I argue that each reason is insufficient to justify exemption, and further, that efforts to exempt survivors problematically amount to a denial of their agency. Finally, drawing on Nancy Snow’s (1991) account of compassion, I motivate a view of compassionate blame as an attitude that situates blame of survivors within a broader view of their traumatic history.

One may wonder why I have centered my discussion on interpersonal relationships rather than the general conditions (for anyone) of blaming survivors, especially given the philosophical precedent for the latter perspective. The simple reason is that this concern originates in our actual lives before getting translated into philosophical puzzles. How do I come to terms with the fact that someone I love has hurt me (or others) because they have been traumatized? How do I negotiate the pain, anger, pity, compassion, and despair at knowing what their suffering has led them to do?—questions like these are an important part of many close relationships (especially given how common trauma is). Centering the perspective on loved ones of survivors is therefore true to a reality of interpersonal relationships, and it is worth working through the philosophical implications for the sake of our actual moral lives. Additionally, we can take philosophical lessons from real interpersonal experiences. Starting from the grounded conditions of interpersonal relationships can provide the considerations and guidance that abstract assessments of blaming conditions cannot access, and I’ll argue that these are worth paying attention to.

2. Trauma and Transgression

Before asking whether we should blame survivors who have transgressed, consider how trauma can shape survivors’ transgressions.2 Some insight into the characteristic effects of trauma will not only help yield a better understanding of these cases and their moral import, but may also help combat caricatures of survivors as inevitably morally ‘broken.’3 It is worth emphasizing upfront that trauma is a varied and personal phenomenon, and there is no single, predictive way that trauma (which can take many different forms) influences behavior.4 A variety of influences shape any human behavior so that trauma is always one influence among many. And there are individual differences in who may become traumatized by different events and to different degrees. While it is uncontroversial that trauma can change people, exactly how they are changed will depend on a number of variables.

Consider the effects of trauma in one survivor’s story. Psychiatrist Bessel van der Kolk (2014) describes the lingering effects of war trauma in one of his patients, Tom. Tom was a former Marine who had served in Vietnam. Van der Kolk recounts Tom’s difficulties interacting with his family ten years after his service ended. A backyard Fourth of July party was so intolerable that Tom had to avoid it: “When he got upset he was afraid to be around his family because he behaved like a monster with his wife and two young boys.5 The noise of his kids made him so agitated that he would storm out of the house to keep himself from hurting them” (van der Kolk 2014: 8).

Van der Kolk traces Tom’s posttraumatic stress symptoms to the loss of Tom’s closest friend, his fellow Marine Alex, who was killed in an ambush along with other members of the platoon Tom commanded. In addition to long-term symptoms of posttraumatic stress, Tom had a more immediate reaction to the ambush. He sought revenge: “The day after the ambush Tom went into a frenzy to a neighboring village, killing children, shooting an innocent farmer, and raping a Vietnamese woman” (van der Kolk 2014: 13).6 Van der Kolk writes, “After that it became truly impossible for him to go home again in any meaningful way. How can you face your sweetheart and tell her that you brutally raped a woman just like her, or watch your son take his first step when you are reminded of the child you murdered?” (2014: 13).

Tom’s story is harrowing for a number of reasons: he both suffered significant damage and perpetuated significant damage onto others.7 This damage fell upon his close loved ones as well as strangers; it manifested as long-term symptoms of PTSD and as immediately atrocities committed in the wake of trauma. Tom’s story thus represents the range of ways in which trauma can potentially beget harmful wrongdoing.8

Intergenerational transmission of trauma—the hypothesis that the effects of trauma can impact a survivor’s family and subsequent generations—can help unpack some of the effects of trauma like Tom’s.9 Historically, research on this topic has focused on the effects of war captivity (Zerach & Solomon 2016) and the historical traumas of the Holocaust (Alford 2015) and colonialism (Prussing 2014) by tracing the downstream effects of these traumas onto others, often through symptoms of posttraumatic stress disorder (PTSD).10 For the purposes of this paper, the relevant effects of trauma are those that involve harmful treatment of others. Two such effects are heightened arousal and emotional detachment (APA 2013). Heightened arousal can involve aggression and anger (APA 2013: 265), unprovoked outbursts and heightened irritability when reminded of traumatic events, and violence against others (Herman 1997: 63). Emotional detachment represents the opposite emotional extreme and is characterized by detachment from others (APA 2013: 271–272), self-isolation and difficulties with intimacy (Herman 1997: 63), and “avoidance and emotional numbness” (Zerarch & Solomon 2016: 299).

Externalized aggression and emotional detachment, together and separately, can clearly harm those in close contact with survivors. A pattern of aggressive behavior towards others can cause significant damage even if it does not culminate in physical violence. Not only can it create a strained relationship with others, but it may also amount to conditions of psychological abuse (Herman 1997: 76–86). Similarly, prolonged or severe detachment can be harmful in relationships that require intimacy and emotional openness. For instance, one study of former prisoners of war who alternately lashed out at and withdrew from their children found that those children developed insecure attachment styles (Zerarch & Solomon 2016). These effects may even manifest as secondary traumatization, in which those close to the survivor experience posttraumatic symptoms (Zerach & Solomon 2016: 298).11 In line with this, another study found that among a sample of mothers in domestic violence shelters, risk factors for perpetuating child abuse were mediated by psychopathology, especially PTSD resulting from being a victim of intimate partner violence (Anderson, Edwards, Silver, & Johnson 2018: 86).12 Specifically, PTSD symptoms related to intimate partner violence “were a significant mediator of the relationship between distress from childhood sexual [but not physical] abuse and current potential for engaging in childhood physical abuse” (Anderson et al. 2018: 87). All this suggests that it is the long-term effects of trauma manifested as posttraumatic stress symptoms (rather than having experienced trauma itself) that are related to potential harmful behaviors.

How does all this bear on the question of blaming survivors for their harmful wrongs? Hopefully, this brief discussion yields some specific understanding of how trauma can, indeed, sometimes shape transgressive behavior. It shows that effects of trauma can lead to relational harms (though there is stronger evidence that these harms result from coping with posttraumatic effects rather than the direct repetition of traumatic victimization). But we should be careful not to draw too strong a conclusion from this: experience tells us that committing wrongful harms is not the inevitable (or even likely) result of surviving trauma. Abuse does not manufacture abusers (Herman 1997: 114); the reality is more nuanced.

Insofar as trauma is an influence on a survivor’s harmful behavior in a given case, questions of blaming the survivor for that wrongful harm become murky (whether the blame at issue is for wrongs done to the blamer personally, to others, or both). The difficulty arises when considering that trauma has relevantly contributed to the survivor’s behavior. Viewed in this light, trauma may seem to exempt survivors from blame for that behavior. I’ll consider this position, and ultimately reject it, in the next section.

3. Trauma as Exempting

For the sake of this discussion, I’ll assume an emotional account of blame, though I won’t argue for it here. That is, I understand blame to consist in the experience of a negative or hostile emotional response to another’s actions, whether this amounts to negative reactive attitudes like resentment and indignation (see Strawson 1962/1974), or some other cluster of emotions like anger, contempt, and disappointment. An emotional account of blame also involves relevant beliefs, such as ‘S did action X and X is wrong,’ but must involve the relevant emotional experience to count as blame. The reason for assuming an emotional account of blame (as opposed to, say, a cognitivist view, according to which blame is just a certain type of judgement) is grounded in the interpersonal nature of these cases, in which blaming loved ones who have survived trauma is often a deeply emotional experience. As will be discussed, some of the struggle with blaming survivors involves the tension between resenting them and feeling compassion for them. So, an emotional account of blame seems to comport with a common experience among loved ones of survivors.

On an emotional account of blame, the characteristic emotions of blame must be experienced by the blamer in order to count as blame but need not be expressed to the target of blame. In what follows, then, my concern will be primarily with the issue of the blameworthiness of the survivor, not with the question of whether blame ought to be expressed to the survivor (blameworthy or not). The issue of expressing blame is of course relevant when considering interpersonal relationships, and some of the considerations related to blameworthiness may also apply to expressing blame. Further, the extent to which one can successfully compartmentalize a blaming response without letting it impact the expressed treatment of the target of blame is not obvious. However, I take these two issues to obey separate considerations that may overlap at points. Again, this focus follows our actual practices: where intuitions suggest that trauma mitigates blame, it points towards exempting the agent, not only withholding the expression of (appropriate) blame from them. So, I’ll focus on whether blame as I understand it is an appropriate response to survivors—whether they are in fact blameworthy—independent of whether that blame ought to be expressed to them, though I’ll note where considerations of trauma may also bear on the issue of expressing blame.

How might traumatic influences on transgressions impact our assessment of blameworthiness, if at all?13 One intuitive view holds that traumatic influences exempt agents from blame. This view has received much philosophical attention, notably Gary Watson’s attempt to flesh out P. F. Strawson’s claim that being “peculiarly unfortunate in . . . formative circumstances” can disqualify agents from the participant stance (1962/1974: 9).14 It may do so as an exempting condition.15 In Watson’s recreation of Strawson, exempting conditions “show that the agent, temporarily or permanently, globally or locally, is appropriately exempted from the basic demand [for reasonable regard and goodwill towards others] in the first place” (1987: 206). The exempted agent is not the appropriate kind of target for the reactive attitudes since she is not subject to the basic demand for how to treat others that applies to the rest of us. She thus falls outside of the considerations that govern typical interpersonal relationships; we look on her not as a fellow responsible agent, but with the distancing “objective attitude” (Strawson 1962/1974: 9). Rather than participating in the moral community, the exempted agent exists at its margins.16

Traumatic pasts may exempt for at least three reasons. The first concerns the fact that trauma involves suffering; the second appeals to the source of the trauma as outside the survivor’s control (that is, as a result of moral luck); the third holds that the effects of trauma undermine responsible moral agency.

Consider the first reason that trauma may exempt. A survivor’s status as a victim of trauma, or more generally, the fact that they have undergone significant suffering, may be morally relevant.17 Put simply, the survivor may have endured enough. Blaming a survivor in spite of what they have suffered may seem unfair, callous, or even cruel.18 The reasoning is that if we correctly attend to a survivor’s suffering, we will realize that they are an inappropriate target of blame because of what they have been through. This concern is a version of the problem with blaming the victim since it holds that survivors should not bear the additional burden of blame given how they have suffered, and are therefore exempt from blame.19

While this concern about being sensitive to suffering has an admirable aim in responding to survivors with mercy, the fact that one is suffering does not actually bear on conditions of one’s blameworthiness. The motivation for this concern is to avoid further burdening sufferers with blame, and this consideration does rightfully influence how we deploy our blame towards others. If the survivor is still suffering terribly, expressing one’s blame to them may be unproductive or cruel. In this case, it may be appropriate to (perhaps temporarily) withhold expressions of blame. But this is a prudential concern about when and how to express blame which is independent of the survivor’s blameworthiness. And this is the problem: it does not seem that suffering or being a victim (from trauma or otherwise) itself has any bearing on one’s moral status with regard to blame. So merely appealing to the fact that a survivor suffers does not provide plausible ground for their exemption.20 If anything, suffering may indicate some other exempting condition, like an incapacitation of moral agency caused by trauma, but then suffering is just an indicator of the actual exemption and is not doing any work itself.

If the fact of suffering from trauma does not exempt survivors, perhaps the source of that trauma is morally relevant. A second reason trauma might exempt involves attending to the fact that who we become and what we do is often not up to us, but is rather a result of moral luck. Understood in this light, survivors who transgress are victims of constitutive moral luck: their characters are shaped in morally relevant ways by forces outside of their control (Williams 1981; Nagel 1979). And on some views of moral luck, the fact that we do not control those conditions that influence our choices undermines our responsibility for those choices—or at least disturbingly calls it into question (Nagel 1979).21 We may think this line of reasoning is apt towards survivors since trauma is a particularly tragic example of constitutive moral luck. Survivors are partly how they are because of terrible circumstances they in no way controlled, so cannot be blameworthy for what they do as a result.22 Alternatively, one may think that since a survivor’s transgressive behavior is a result of the bad moral luck of a traumatic upbringing, then I, the blamer, am as vulnerable to exhibiting that behavior as they were—I may have become the same way under those circumstances—I’ve just been lucky.23 And this means I am in no good position to blame a survivor. Recognition of one’s own moral vulnerability is thus similar to blaming hypocritically: though it doesn’t alter the blameworthiness of the target of blame, it may make my blaming them inappropriate.24 And since we are all equally vulnerable to moral luck, victims of traumatic histories are inappropriate targets of blame.

However, the first worry about moral luck—that one cannot be responsible for those actions that are shaped by forces outside of one’s control, and so that traumatic histories exempt survivors—meets with immediate problems when we consider how responsibility works in our actual lives. Following Margaret Urban Walker’s (1991) criticism of the conception of ‘pure’ agency, I want to suggest that constitutive moral luck is not an impediment to responsible agency, but rather a condition of it.25 Moral luck does not pose a problem for responsibility when we recognize that it is a normal feature of our moral agencies to take responsibility for who we are even though we have been shaped by forces outside of our control.26 The burdens that we face because of our pasts, the tools and limitations we are unfairly saddled with—these things make up the material of our agencies from which we make choices. To suppose otherwise would evoke an unrealistic portrait of agency according to which we are only responsible for those influences on ourselves that we chose (Walker 1991: 22). The texture of our moral lives is complex enough to account for the fact that while I don’t choose what I’ve been through, I may nonetheless find myself responsible for what I do given what I have been through. Thus, trauma may shape the contours of survivors’ moral agency and responsibility, but this needn’t, itself, exempt survivors from these burdens.

One may still worry about the second reason that moral luck may exempt survivors—that one can’t blame a survivor since, but for the grace of moral luck, one could have become the same way.27 But in response, notice that we need to take moral luck into account in order to make our actual moral relationships intelligible. Walker notes that we would have difficulty accepting an agent who appealed to moral luck to convince us to exempt her (1991: 19). I would add that we might think this agent is in bad faith by attempting to evade her responsibility in this way. If close loved ones of unlucky moral agents were to say of survivors, they’ve just been unfortunate, so I can’t blame them, they would be engaging in a form of projected bad faith in which they clear a path for others to evade their responsibility. And viewing another this way is in tension with the demands of close relationships. Holding others responsible—sometimes through blame—is one important part of close relationships that maintains moral expectations and boundaries. It is not clear how we could continue these moral interactions if we exempt others for their unlucky constitutions just because we, too, are vulnerable to moral luck.28 Rather, doing so might render our moral relationships anemic, for it would mean regarding survivors primarily as products of their trauma rather than as agents who have been influenced in many ways, trauma among them. And this seems to threaten the basis of a genuine relationship in which we take another’s history into account when we engage with them as a moral agent and sometimes assess them as blameworthy for their harms rather than patronizingly letting them off the hook.29

Perhaps, then, trauma exempts survivors because of what it can do to them. The third reason for exemption is that trauma may damage moral agency so that a survivor is not responsible for those actions influenced by trauma, which in turn exempts them from blame for those actions. We know that trauma can damage agency along a number of dimensions (Herman 1997), so it would not be surprising if this damage is sometimes moral. Many philosophers have appealed to this intuition or attempted to spell out the agential incapacitation at issue.30 For instance, trauma may undermine moral agency by making transgressive actions inevitable.31 If a survivor could not help but develop aggressive tendencies or emotional detachment in response to their trauma (and, perhaps as a result, cannot alter them now), she should not be blamed for them.

Alternatively, one may think that formative trauma can impair an agent’s normative capacities, either by hampering their recognition of moral reasons or their reaction to these reasons (or both).32 Consider the first type of impairment. Though not considering cases of trauma or PTSD, Wolf (1987) suggests that certain histories (such as those of “deprived childhood victims”) may render one unable to distinguish right from wrong (1987: 56). On Wolf’s view, since the capacity to recognize moral reality is necessary for responsibility, transgressors who lack this capacity are not responsible for their wrongs. If one were to apply a version of Wolf’s reasoning to cases of trauma, one might conclude that insofar as a traumatic upbringing results in a survivor’s distorted moral values or failure to recognize certain actions as wrong (if these values and actions were normalized under trauma), they cannot be responsible for the actions that result from these incapacitations.

A second impairment to normative capacities involves a failure to react to moral reasons. Even if survivors can recognize a moral reason for what it is, they may have difficulty choosing to act in accordance with that reason. Thus, their reactivity to reasons may be hampered. For instance, a survivor may recognize that they ought to respect another’s emotional boundaries or that their outbursts are hurting the people they love, yet have trouble actually choosing to act in a way that would address their behaviors. That is, these reasons may have no motivational pull despite a survivor’s recognition that they justify her action.33 This idea may be nascent in the minds of loved ones who recognize that trauma can result in certain limitations: he just doesn’t have the capacity to act any differently; she never got the tools she needed to do better. Here, a survivor is not insensitive to moral reasons, but lacks the ability to follow through on them.

While trauma can indeed damage moral agency through impairments to either of these normative capacities, the conclusion that this undermines responsibility is far too quick. As discussed in Section 2, while we should recognize the potential influences of trauma on moral behavior, we should be careful not to overstate their power. If transgressive behaviors are not the inevitable result of undergoing trauma, then trauma does not categorically undermine moral agency. Additionally, we can recognize that moral damage comes in degrees and may impact some aspects of responsible agency while leaving others intact. Moral agency isn’t all-or-nothing: the capacity for responsible agency can still exist overall despite damages to certain capacities.34 Perhaps in some extreme cases, moral agency is severely damaged enough to render the agent not responsible, but these will likely be rare cases that shouldn’t define the moral assessment of the average case (especially when we remember how common trauma is [Felliti et al 1998; Merrick et al. 2018], and that moral incapacitation that undermines responsibility does not seem to be a widespread phenomenon).35

A stronger worry is that assessing survivors as exempt because agentively incapacitated paints a problematic portrait of trauma survivors that effectively excludes them from the moral community. On a view that trauma exempts survivors because it incapacitates moral agency, wrongdoing is seen as a product of trauma rather than the responsible action of a traumatized agent. And while recognition of a survivor’s trauma does—and should—impact our response to survivors and their wrongdoing, it should not do so in a way that pushes these agents to the liminal spaces of the moral realm, as if they are too fragile to bear real responsibility. Exempting survivors on the grounds that trauma is wholly incapacitating effectively treats survivors as non-agents—as people so damaged that they have lost their status as agents. And this should be deeply worrying for anyone who is a survivor or cares for one, not only because it is not true to our actual experiences (including those of the demands of relationships discussed above), but because survivors ought not be regarded in this way. That is, I am suggesting that there are strong moral—and political—reasons to treat survivors as responsible, and that these are independent of the status of their normative capacities.

Notice that this objection to the incapacitation argument—it does not regard survivors as moral agents—also applies to the suffering and moral luck arguments. All three attempts to exempt survivors try to show (in different ways) that because of unchosen suffering endured in the past that reverberates into the present, survivors are not proper bearers of one type of moral regard. Each argument presents an image of survivors as agentive victims to their trauma, released from responsibility because they have been too hurt, too unfortunate, or too damaged. And each therefore motivates the idea that survivors are less than full moral agents. Thus, the broader worry with any attempt to exempt survivors by virtue of their trauma is that these considerations fail to treat survivors as genuine agents who have done serious wrong.36

As the loved ones of survivors, we are in relationships with people, not pasts. And while another’s past is certainly relevant to our relationship with them, using trauma as the grounds for exemption risks reducing survivors to their pasts and erasing the whole person from view. Recognizing another as a full person with robust agency seems a basic necessity for genuine interpersonal relationships, and treating survivors as the damaged products of trauma can therefore undermine relationships with them in addition to being the wrong way to treat them. Instead, we can recognize the real influences of traumatic histories without ceasing to treat survivors as responsible moral agents who have hurt others. Rather than striving to explain why a wide swath of ‘unfortunate formative circumstances’ disqualifies agents from responsibility, I think we should take for granted that trauma survivors are genuine moral agents and members of the moral community, even if some may have been shaped in certain damaging ways that led to harmful behaviors.37 Additionally, since part of the function of blame is to acknowledge serious wrongs, viewing survivors as categorically exempt risks watering down the severity of the wrong at issue and denying the blamer their justified response. We need to keep the fact that others have suffered harm at survivors’ hands in moral focus in addition to recognizing survivors’ agencies, and we can do this by maintaining blame.

At this point, it may be tempting to think that in order to properly respect survivors as agents we should simply treat them as standardly blameworthy, just as we would any other transgressor. If trauma is not exempting of survivors, the thought goes, then it does not carry any special moral weight: these are standard cases of blameworthy wrongdoing. The problem with this conclusion is that it fails to attend to just those features that make trauma morally important, namely, that someone has suffered a great deal under unfair circumstances that have deeply changed them. So, just because trauma does not categorically exempt survivors from blame does not mean that it calls for no special considerations in our blaming practices. It does—just not those considerations that mitigate blame. Rather, we should recognize that another’s suffering has contributed to their transgression, even as we hold them responsible for it through blame. I suggest that this recognition calls for compassion. In particular, I argue that in these cases, we ought to have a particular stance of blame taken up in the mode of compassion, or compassionate blame. In compassionate blame, we can recognize that survivors have suffered from trauma in ways that have influenced their behavior while still treating them as moral agents who have done wrong.

4. Compassionate Blame

Drawing on Nancy Snow’s (1991) account of compassion, I argue that on a certain understanding of compassion that is tailored to the cases at hand, compassionate blame is a coherent attitude that we ought to cultivate towards survivors who have transgressed.38 On this view, blame towards a survivor is maintained, but is contextualized within a broader recognition of the survivor’s traumatic history and how it has shaped them to be an agent who is blameworthy for certain harmful wrongs. This means attending to the tragedy of these circumstances: not only that a survivor has endured trauma, but that as a result, they must now deal with the moral fallout at their own hands.39

In order to articulate the notion of compassionate blame, first consider standard definitions of compassion. Snow identifies compassion as an other-regarding emotion in which “the perception of the other’s negative condition evokes sorrow or suffering in the one who feels the emotion” (1991: 196). But in order to distinguish compassion from similar emotions like grief or pity, Snow argues that we need to regard compassion as a “[composite] of belief and feeling” (1991: 2) that is identified by characteristic beliefs associated with compassion. In particular, these are the beliefs that the other is suffering significantly, that this suffering is occasioned by vulnerability to misfortune, and that the feeler herself is similarly vulnerable to misfortune (even misfortune of a different kind) (Snow 1991: 198).40 While compassion is evoked by the recognition of another’s suffering, part of compassion is realizing that one is not insulated from the possibility of suffering: another’s suffering could as easily have been (or become) one’s own. This recognition of common vulnerability to suffering is sufficient to bridge the emotional gap in identifying with the suffering other (Snow 1991: 199).41

When considering survivors in light of the trauma they have suffered, as well as anyone who has suffered because of the survivor’s harmful wrongs, compassion is clearly appropriate. Yet we’ve seen that suffering is only one relevant moral feature of these cases, and is a complex feature at that. In order to attend to all the morally relevant features of these cases—the suffering, the bad moral luck of trauma and its constitutive influence on the survivor’s behaviors, the wrongness of the act at issue, the survivor’s responsibility for it, and a loved one’s relationship with the survivor—we need something more than standard compassion. Following Snow’s understanding of compassion as a composite of feeling and belief, I propose that compassionate blame involves both a characteristic emotional component of ‘suffering with’ the other (and the emotional components of blame, discussed below) as well as a set of distinguishing beliefs.42 It is this set of beliefs relevant to the features of these cases that will set compassionate blame apart from garden-variety compassion and garden-variety blame. Consider, first, the relevant set of beliefs that partly constitute compassionate blame.43

In addition to the beliefs that (1) the survivor has or continues to suffer as a result of trauma, and (2) the victim(s) of the survivor’s transgression(s) has or continues to suffer from that harmful wrong, compassionate blame also involves a recognition of the common vulnerability to suffering that Snow identifies. However, in order to address the contours of these cases, this recognition must also include an appreciation for moral luck and its role in both the suffering and the transgression at issue. It involves recognizing that the survivor’s suffering from trauma is a particularly unfair vulnerability to suffering as well as a form of constitutive moral luck that has helped shape later wrongs. And this means that when we have compassionate blame for a survivor, we are attending to their whole tragic context, from trauma to transgression: the fact that they have been shaped to be who they are and act as they do partly as a result of suffering they did not create and which has led them to hurt others. It means seeing the connections between a survivor’s suffering and their ultimately blameworthy action, while still holding them responsible through blame. The list of relevant beliefs for compassionate blame must therefore also include (3) the survivor’s suffering through trauma is a form of (constitutive) moral luck and has influenced their later transgression(s), and (4) they are blameworthy for that transgression(s).44

On a view of moral luck that is consistent with genuine responsibility—a view I’ve argued is appropriate in these cases—the recognition of vulnerability to moral luck as a component of compassionate blame does not preclude blame. It is perfectly consistent to believe that a survivor is blameworthy for some harmful wrong and that trauma helped constitute the person who performed that harmful wrong. So, we can coherently hold the requisite beliefs of compassionate blame since the conditions that make blame and compassion respectively appropriate are inextricably connected when viewed from the broader historical context.

Crucially, we must also recognize our common vulnerability to moral luck and to suffering when we have compassionate blame towards survivors, and we can do so without relinquishing our other moral assessments of them. Thus, another belief requisite for compassionate blame is (5) I am similarly vulnerable to suffering and to moral luck that may occasion suffering. To recognize that another has fallen victim to moral luck that has caused suffering is to recognize the same possibility for ourselves. But this recognition can motivate compassionate blame rather than make us ineligible to cast blame (as the objection considered above insisted). There is nothing inappropriate in my blaming another under these conditions; rather, it is an acknowledgment of an unfair world in which we all must bear moral burdens and deal with their fallout.

So far, I’ve argued that the characteristic beliefs that constitute and distinguish compassionate blame are coherent. But one may worry that the emotional component of compassionate blame is not. This seems to be the heart of Watson’s worry that considering abusive histories causes “ambivalence” in our reactive attitudes by pulling us in opposite affective directions (1987: 275). When discussing Robert Harris, a victim of severe childhood neglect and abuse who committed horrible murders, Watson writes:

What appears to happen is that we are unable to command an overall view of his life that permits the reactive attitudes to be sustained without ambivalence. This is because the biography forces us to see him as a victim, and so seeing him does not sit well with the reactive attitudes that are so strongly elicited by Harris’s character and conduct. Seeing him as a victim does not totally dispel those attitudes. Rather, in light of the ‘whole’ story, conflicting response are evoked. The sympathy toward the boy he was is at odds with outrage toward the man he is. These responses conflict not in the way that fear dispels anger, but in the way that sympathy is opposed to antipathy. In fact, each of these responses is appropriate but taken together they do not enable us to respond overall in a coherent way. (1987: 275)

According to Watson, knowledge of Harris’s history evokes appropriate sympathy for his status as a victim, but this attitude is in tension with the (also appropriate) negative reactive attitudes towards his transgressions. We can have both emotional reactions and recognize that each is appropriate, but it is nonetheless difficult to “command an overall view of his life” that resolves the tension between these two perspectives (1987: 275).45

Watson hits on exactly the tension that makes cases of traumatic transgression so complex. While I grant that focusing on different aspects of the survivor can pull us in conflicting emotional directions—it can be difficult to attend to a survivor’s suffering when we are angry at them for what they’ve done—Watson’s claim that “taken together [each appropriate response does] not enable us to respond overall in a coherent way” (1987: 275) evinces a failure of imagination. Watson is worried about being pulled in opposite emotional directions as a result of focusing on one or the other aspect of the survivor. But the suggestion of compassionate blame holds that instead of narrowing our focus, we should expand it. While Watson may be right that these perspectives are in tension, he does not consider whether any broader perspective can in fact coherently accommodate both features and resolve the tension. Compassionate blame is just this response: since both the survivor’s history and their transgression are morally relevant and deeply intertwined with one another, the appropriate response to them must attend to both of these features—not by holding each view side by side, but by recognizing that each is connected in a broader historical and moral whole. In the broader perspective of compassionate blame, blaming emotions are nested within compassionate emotions—which, together with the distinguishing beliefs, gives compassionate blame a unique phenomenological character.46

Another way to put this is that the intentional object of compassionate blame just is the survivor in terms of a history of trauma that has made them blameworthy for wrongful harms, and thus necessarily includes blaming emotions as well as the emotions of compassion, inseparably felt towards the survivor. This is part of what makes compassionate blame a unique phenomenological experience that is not just reducible to blame plus compassion. The emotions associated with garden-variety blame and garden-variety compassion, respectively, have a narrower focus than those of compassionate blame. Blame is directed at a person by virtue of the wrongful action they’ve committed, while compassion takes the person-as-sufferer as its object. The emotional component of compassionate blame takes a broader object: when we feel compassionate blame for another, we view them as a blameworthy agent within the recognition of their history as a fellow suffering being.

Additionally, there is perhaps not as much tension between sympathy and antipathy as Watson suggests. After all, we hold multivalent, ostensibly conflicting assessments of and emotions towards people without much trouble (especially those we are close to). I can feel compassion for the person that I am also angry at, suffer with them while I blame them for what they’ve done.47 ­Anyone who has been a partner or parent knows this is a common psychological phenomenon. Think of being angry with a loved one for putting themselves in danger—wanting to hold them close and push them away, feeling both overwhelming love for them and anger at their folly. Complex emotional experiences like these are a common and coherent part of our interpersonal lives.

Further, we know that transgressors can suffer and victims can transgress, and none of this mitigates any of the pain they may feel. In her discussion of the social implications of compassion, Martha Nussbaum notes that “when we think about crime and criminals we need to get beyond the simple dichotomy between treating them as responsible and treating them with compassion. It is perfectly consistent to treat a criminal . . . as fully responsible for his crimes, and yet to acknowledge with compassion the fact that he suffered misfortunes that no child should have to bear” (Nussbaum 2001: 414). I want to add that it is not just that these responses are compatible, but that they call for a broader moral response that attends to their connection. We should not look at a traumatic past without considering its downstream effects, nor should we consider the transgression in isolation from its influences, but instead recognize how these things are connected and give rise to complex emotional experiences.

Compassionate blame shares certain similarities with Cheshire Calhoun’s (1992) discussion of contextualizing a person’s wrongdoing in terms of their history without thereby excusing or justifying their wrongdoing.48 Calhoun motivates an “aspirational” form of forgiveness, which “requires hunting for a deeper reading of unexcused and unrepented wrongdoing . . . that tries to understand how culpable wrongdoing fits into the larger pattern of a person’s life” (1992: 92). For Calhoun, a person’s history is relevant not because it is a new perch from which to make a more accurate moral assessment of excuse or justification. Rather, looking at a person’s history allows us to see that their wrongdoing is “intelligible” (1992: 77), though they are still blameworthy for it; that “although an agent’s wrongdoing fails to make moral sense, it does make biographical sense” (1992: 92). Calhoun notes that “It is possible to enter sympathetically into a life [like this]” (1992: 94). When aspirationally forgiving, our moral assessment does not change; the person is still blameworthy, and one is still entitled to resentment towards them (1992: 81). But this sort of forgiveness allows one to see the other as an imperfect agent with a meaningful history against which their actions make sense, and to forgive them without expecting them to change (1992: 95).

Calhoun’s foil is an account of forgiveness that effectively amounts to exempting the wrongdoer. Like Calhoun, I deny that history (specifically, traumatic history) exempts a person’s wrongdoing: trauma does not undermine a moral assessment of blame just as, for Calhoun, it does not excuse or justify the wrongdoing.49 Calhoun’s proposal of aspirational forgiveness allows for one to forgive while still feeling appropriate resentment towards the wrongdoer. Though I’m not concerned with forgiveness, I similarly hold that resentment and its ilk are compatible with the more sympathetic emotions we might have towards another. However, though history matters on Calhoun’s view, it does not matter morally since it does not undermine blame. By contrast, I hold that history does matter morally: traumatic histories are morally relevant not because they undermine blame, but because this history bears on our moral response by demanding compassionate blame. Excuse and justification are not the only ways to make moral sense of an action. Contextualizing blameworthy wrongdoing in trauma is a way of making moral sense of that wrongdoing since it helps us recognize how morally relevant suffering has shaped a person and their actions. And this matters morally by informing our response to survivors. Compassionate blame is a response that attends to these morally relevant features: it is sensitive to trauma without denying the agency of the survivor or the seriousness of their wrongs and thus avoids exactly those problems with exempting survivors (denying their moral agency) and treating them with standard blame (not being sensitive enough to their suffering), respectively.

Though I have centered this discussion on the loved ones of survivors, I think the morally relevant features of these cases mean compassionate blame is an obligatory attitude for anyone—strangers and loved ones alike—to have towards survivors.50 Compassionate blame ought to be cultivated towards agents whose blameworthy transgression is relevantly influenced by trauma, full stop.51 But the concern about how to relate to a survivor who has transgressed is often immediate in interpersonal relationships, and the meaning and experience of compassionate blame will take on a different tone for close loved ones than it will for strangers or the general population. Here, it is a question of how we continue to relate to a person or how we understand a person who was part of our lives. Insofar as compassionate blame offers a way forward with a survivor or an inroad to making sense of a relationship that has ended, it is much more than just a situated assessment of a person and their actions, as it is to a stranger. Loved ones are in a privileged position to see that the same reasons that make blame appropriate also call for compassion. And sometimes, the anger and resentment of blame is partly a function of the closeness with others that can make taking a compassionate stance towards them easier. We would not blame so strongly or anger so fiercely if we were indifferent to the other. Blame opens up a distance between people only because there is already a closeness—a closeness that might not be completely obliterated by blame and which, sometimes, may help motivate compassion.

As discussed in the previous section, in order to have a genuine relationship with another, we need to relate to them as a responsible agent. And having a close relationship with another does not always mean shielding them from all negative attitudes, but can rather involve holding them responsible for their wrongs through blame. Notice, too, that blame doesn’t always have to be the poison that some philosophers think it is. One can say to a survivor, I’m angry with you for your actions but I still love you; I resent you for what you did but I feel compassion for the person who did it. Blame can, but needn’t, completely undo a relationship. It can change the contours of a relationship by, for instance, setting certain boundaries or expectations, but in many cases, we can maintain ­functioning relationships with those we blame. Further, loved ones still have discretion over how blame is deployed. Compassionate blame as a recognition of another’s particular form of suffering can motivate patience, a gentler expression of that blame, a determination not to be cruel with one’s blame, while still treating the survivor as genuinely blameworthy.

Further, compassionate blame is an appropriate attitude towards survivors given how they may regard their own responsibility. Survivors are often extremely ashamed and highly punitive of themselves for their transgressions. As van der Kolk reminds us, “deep down many traumatized people are even more haunted by the shame they feel about what they themselves did or did not do under the circumstances. They despise themselves for how terrified, dependent, excited, or enraged they felt” (2014: 13). A survivor may be well versed in self-blame, and therefore not only deserve compassionate blame as someone who suffers, but may be especially in need of it because they lack self-compassion. At the same time, survivors should be held to account for their actions and the seriousness of the harm should be recognized. Herman writes that it is incumbent on those close to the survivor to help rebuild their sense of selves by being tolerant to their oscillating needs, however this “does not require that others tolerate uncontrolled outbursts of aggression; such tolerance is in fact counterproductive, since it ultimately increases the survivor’s burden of guilt and shame” (1997: 63). Extending this point, having compassionate blame towards the survivor is a way of reestablishing the bounds of a healthy relationship, not a coddling one in which all violence is immediately excused. The thought is that compassionate blame may strike a balance between appropriately blaming a survivor without needlessly piling on to their own self-blame.

Compassionate blame may change how a blamer relates to the survivor moving forward. Continuing a relationship with a survivor may require a deeper understanding of why a survivor acts the way they do and a recognition of what they have been through. However, it is crucial to note that not all relationships with survivors ought to be maintained. A survivor’s behavior may make the relationship unhealthy or harmful, especially if one has been a victim of the survivor’s harm, and one may have to consider whether it is a safe or worthwhile relationship to continue. And in some cases, it won’t be. However, even if a loved one ends their relationship with a survivor, or if it cannot continue for other reasons, compassionate blame is still an appropriate stance to take given the nature of the wrong and the suffering that helped create it. This is because it recognizes the reality of trauma, and therefore can situate the harmful wrong at issue in a way that can be explanatorily satisfying without denying the loved one their feelings.

Of course, there will be a lot of variety in close relationships with ­survivors, and though compassionate blame is appropriate in these cases, the extent to which it is practically achievable by a given person will vary.52 Some may be so angry at the survivor that they can only feel compassionate blame in occasional short bursts; others may have to remind themselves that the other is in fact blameworthy, and work to conjure those appropriate emotions; the severity of the transgression may make compassion blame incredibly difficult for others (though this doesn’t change its appropriateness). And I think we should recognize that some cases are like this, and that cultivating compassionate blame can be challenging. But adjacent to these cases are the rest, in which we continue to have relationships with people who have hurt and are hurting, and for this, compassionate blame may offer an appropriate and achievable perspective. The complexity of our close relationships seems to necessitate more options in responding to survivors than garden-variety blame and exemption.53 I hope to have offered a realistic alternative with compassionate blame.

5. Conclusion

Though I’ve centered the discussion on the effects of trauma on transgressions, if the idea of compassionate blame is at all compelling, we might think it applies to a wider scope of cases. Since we are all vulnerable to moral luck, and since many forms of suffering unrelated to trauma can influence our moral behaviors, compassionate blame may be an appropriate attitude whenever formative suffering (that is or is not traumatizing) influences transgressions. Perhaps, then, we should endeavor to cultivate compassionate blame whenever we recognize the moral impact of an unfair and tragic world in which we all must still act.


I’d like to thank the participants at the 2019 Blame, Punishment, and Health Workshop at Rutgers University for helpful comments and discussion that contributed to the substantial development of this paper and two anonymous referees for Ergo for extremely thoughtful and encouraging comments.


  1. See, for instance, Strawson (1962/1974), Wolf (1987), Watson (1987), Buss (1997), and Ebels-Duggan (2013).
  2. I am deliberately using words like ‘shape’ and ‘influence’ rather than ‘cause’ because I think it is a fairer assessment of the effects of trauma and because we do not need to have causal certainty to know that trauma played a part in a behavior.
  3. A traumatic event is defined as “Exposure to actual or threatened death, serious injury, or sexual violence” (American Psychiatric Association 2013: 271). This exposure can be through direct experience, witnessing the event(s) in person, “learning that the traumatic event(s) occurred to a close family member or close friend,” or through “repeated or extreme exposure to aversive details of the traumatic event(s)” (APA 2013: 271). According to Judith Herman, a central figure in trauma research, traumatic events “confront human beings with the extremities of helplessness and terror, and evoke the responses of catastrophe” (1997: 33). Traumatic events consist in overwhelming horror that cannot be fought or escaped, and this powerlessness in the face of danger triggers the reactions to trauma that can lead to posttraumatic stress disorder (PTSD).
  4. I mean my discussion to apply to a wide variety of traumas, including both trauma that is characterized by the survivor’s victimization from another agent (like sexual assault and domestic violence) and that which is not (like surviving a natural disaster), since both can have similar influences on behavior and evoke the intuitions about responsibility that I discuss below (though our intuitions about the blameworthiness of traumatic victimization may be stronger than those of trauma that does not involve victimization).
  5. I find the term ‘monster’ problematic, but include it here because survivors may think of themselves in terms like these, warranted or not.
  6. Psychiatrist Jonathan Shay calls this reaction to combat trauma the berserk state, and argues that it is a profound moral injury in veterans (1994: 77–99).
  7. Transgressions committed against others, such as Tom’s atrocities towards strangers, are clearly morally problematic in their own right regardless of what impact they may have on survivors’ loved ones. However, knowledge of these transgressions can impact loved ones’ assessment of and relationship with survivors in ethically relevant ways. This focus is not meant to imply that the impact on the survivor’s relationships is the primary ethical issue around these transgressions—far from it—but just to acknowledge that one may blame a loved one for the harmful wrongs they’ve committed against others. Thanks to Craig Agule for highlighting this point.
  8. I’ll use ‘transgression’ and ‘harmful wrong’ interchangeably in what follows.
  9. Research on the intergenerational transmissions of violence is also relevant here. However, this research is notoriously inconsistent, and where there is evidence that growing up in a violent family increases the likelihood of perpetuating or becoming a victim of violence in adulthood, the effects are small (Smit-Marek et al. 2015). A safe assessment seems to be that while those who abuse others are more likely to have been abused themselves (Oka 2017: 912; see also Bijleveld et al. 2016), the predictive conclusion does not hold; it is still the case that, for example, “the great majority of survivors neither abuse nor neglect their children” (Herman 1997: 114). Since there is better evidence for the intergenerational transmission of trauma, and since harmful wrongs can be part of these effects, I’ve elected to focus on this research instead.
  10. While many of the studies discussed frame effects in terms of PTSD, I don’t assume that the behaviors I’m concerned with need be mediated by a diagnosis of PTSD; traumatic fallout may fall short of this diagnosis yet still lead to transgressive behavior. I cite this research because it is a useful way to think through what may be going on in some of these cases.
  11. These harms may be a direct result of secondary traumatization, but need not be. That is, a close loved one need not experience secondary traumatization in order to have been harmed by the trauma survivor; nor do the harms at issue just amount to causing secondary traumatization in others.
  12. The authors of the study note that these findings are not generalizable since women in domestic violence shelters are subject to a significant number of burdens and additional risk factors for perpetuating child abuse.
  13. In these cases, I’m assuming other facts independently support the survivor’s ­blameworthiness—that she caused the wrong, that it was actually wrong, etc.—and that other considerations that may excuse or justify the agent do not obtain in order to isolate trauma as a potentially mitigating factor. Hereafter, when I refer to a transgression, I mean it as shorthand for a transgression that was relevantly influenced by the effects of trauma, as discussed in the previous section.
  14. I take it that abusive and other traumatic conditions are the main case of concern in ‘particularly unfortunate circumstances.’ Watson, at least, focuses on abusive conditions with the case of Robert Harris, a man who suffered horrible child abuse and grew up to commit cold-hearted murders; a man whose upbringing left him, Watson claims, uninterested in participating in the moral community though able to understand its demands. Watson’s question is whether this historical explanation is itself exempting, and not just evidence of some independent agentive incapacitation. I think it is worth noting that ‘formative’ experiences need not happen when one is young; one can be ‘formed’ by trauma as an adult, and so cases of later trauma, if severe enough to prompt the question of exemption, should be considered here too.
  15. Though Strawson doesn’t discuss blame as a reactive attitude, I take my broad understanding of ‘blame’ as an interpersonal emotional experience to mean that what Strawson and Watson have to say about the reactive attitudes can be applied to blame as discussed here.
  16. Exempting conditions are contrasted with excusing conditions, in which an agent’s action on a particular occasion does not meet “the internal criteria of the negative reactive attitudes” (Watson 1987: 260). Accidental hand steppers meant no ill will, so while the offender is generally eligible to be the subject of the reactive attitudes, on that occasion they are excused.
  17. Trauma that does not directly involve victimization from another agent (such as trauma from surviving a natural disaster) can evoke this response as well as traumatic victimization like sexual assault; in the former case, one can be viewed as a ‘victim of trauma’ in a metaphorical sense.
  18. Kyla Ebels-Duggan considers a version of this objection (2013: 154).
  19. To be clear, in these cases the blame is directed at an act done by the victim, not blame for the conditions or events that made them a victim; trauma survivors are not being blamed for the trauma that shaped them, but for their actions that are influenced by that trauma. Compare with Superson (1993).
  20. Notice that if suffering did exempt agents on the grounds that we should not further burden sufferers with blame, then sufferers would be exempt whether or not trauma or suffering meaningfully contributed to the wrong at issue. If, for instance, one suffers trauma after they commit some transgression, or if a transgressor is clinically depressed (and this is independent of their committing a transgression), then we should not blame them for fear of adding to their suffering. And this seems to miss the point of questions of wrongdoing influenced by trauma. Thanks to David Boonin for suggesting these cases.
  21. This conclusion follows if we also hold the control condition: we are only responsible for what is under our control (Nagel 1979).
  22. I’m assuming that our characters are in some way predictive of our actions. So, constitutive moral luck is a matter of how people are shaped to be and what this leads them to do.
  23. This is one of the arguments Watson considers for why a traumatic history may exempt. He argues that the bad moral luck of Harris’s upbringing does not change our judgment that he is “vile,” but it does make “one feel less in a position to cast blame . . . The awareness that, in this respect . . . others are or may be like oneself clashes with the distancing effect of enmity” (Watson 1987: 276).
  24. This objection evokes standing to blame considerations, in which whether or not one ought to be blamed has to do with features of the blamer. See for instance Wallace (2010).
  25. Walker focuses on resultant and circumstantial moral luck, so I am expanding her argument here.
  26. Claudia Card (1996) makes a similar point about constitutive moral luck. She first notes that there are two perspectives of responsibility: a backward-looking perspective, focused on attributing responsibility to past actions, and a forward-looking perspective, focused on taking responsibility for future actions (Card 1996: 25–27). It is the forward-looking perspective that can accommodate responsibility in the face of constitutive moral luck, for it is concerned with taking responsibility given the conditions one finds oneself in.
  27. The preceding considerations also speak to this concern. If moral luck is a condition for responsibility for all of us, then the fact that another had especially bad luck does not seem to bear on whether we can blame them.
  28. Walker also considers the practical implications of a moral world of ‘pure’ agents and finds them similarly lacking (1991: 22–26).
  29. I am not suggesting that conditions of moral luck and formative trauma in particular can never be exempting. Perhaps in extreme cases they can be. But whether they are exempting will depend on the details of a given case, and not simply be by virtue of a survivor’s exposure to moral luck or trauma. That is, I’m contesting the notion that formative trauma, and by extension, constitutive moral luck, are categorically exempting conditions.
  30. See for instance Wolf (1987; cf. Wolf 2015), Bell (2008: 683), Watson (1987), Strawson (1962/1974), Buss (1997). Smith (2005) argues that particular histories may mitigate blame for certain attitudes without undermining responsibility for them.
  31. Watson (1987: 277–278) considers this reason for exemption and ultimately rejects it. Though not discussing traumatic histories in particular, Wolf alludes to the inevitability of certain attitudes and abilities given one’s formative circumstances (1987: 54, 57).
  32. For an account of this distinction, see Fischer and Ravizza (1998).
  33. See Fischer and Ravizza (1998: 41–46).
  34. This applies to those capacities that some think are necessary for responsible agency.
  35. Though, again, not discussing trauma or PTSD, Wolf evokes an othering of agents with damaging formative histories, contrasting their lack of responsibility with ‘our’ normal responsibility (1987: 56–57). This sort of thinking may have encouraged a strict dichotomy between the ‘healthy,’ complete moral agency ‘we’ undamaged folks have, and the morally depraved, non-responsible Other. In fact, we are all probably morally damaged in more or less extreme ways, yet still largely manage to be responsible agents.
  36. Martha Nussbaum articulates this point in her discussion of the social applications of compassion: “We know . . . that disaster can strike earlier and harder . . . affecting people’s very ability to form plans and aspirations, affecting their ability to be good . . . We may have sympathy for misfortunes that are utterly undeserved, but when people commit crimes, and do so with hostile intent, it is condescending not to blame them and hold them fully responsible. To treat them as if they could not help it negates their human dignity and treats them like ‘children’ or ‘animals without a soul’ ” (2001: 410).
  37. That is, the presumption should be towards treating survivors as full agents unless particular details of the case suggest they are exempted because of their trauma—not the converse.
  38. It is worth pausing to explain why I am focusing on compassion instead of some other other-regarding emotion like pity or empathy. As Nussbaum classifies these other emotions, “ ‘Pity’ has recently come to have nuances of condescension and superiority to the sufferer . . . ‘Empathy’ is often used . . . to designate an imaginative reconstruction of another’s experience, without any particular evaluation of that experience” (2001: 301–302). However, recall the morally relevant features that led us to consider compassionate blame: we need to recognize the impact of another’s suffering without denying their agency. While the empathetic person can imagine another’s suffering, this does not involve a judgment about that tragedy of that suffering (Nussbaum suggests that empathy “is quite different from and insufficient for compassion; it may not even be necessary for it”, 2001: 302), and while pity does make a judgment about suffering, this judgment looks down upon the piteous sufferer and thus recognizes the sufferer as inferior. Compassion offers a way to recognize and evaluate another’s suffering while still treating them as an equal.
  39. When I use ‘tragedy,’ I don’t mean to evoke the classical associations with pity. Rather, it is just a recognition of the misfortune and sadness of the situation.
  40. Similarly, Nussbaum identifies three necessary cognitive components in the Aristotelian understanding of compassion: I must believe another’s suffering is serious, did not come about through their own fault, and that I may encounter similar suffering (2001: 306–322). Nussbaum argues that this last condition is not in fact necessary for compassion, but is rather an epistemic aid to the cognitive component that is necessary for compassion: the eudaimonistic judgment that the sufferer is important to my goals and plans (2001: 321).
  41. Snow denies that one needs to be able to imagine another’s suffering in order to have compassion for them: we can feel compassion for others because we love them and recognize their suffering and our common vulnerability to it, even if we are unable to imagine their experience (1991: 197). This means Snow does not think empathy is necessary for compassion (a point Nussbaum, 2001: 328, agrees with). Additionally, Snow argues that compassion characteristically involves “an altruistic concern for the other’s good . . . [which] can be manifested in a variety of other-regarding thoughts and desires, such as the wish that misfortune had not occurred and the desires to help the other in various ways” (1991: 196–197). This concern can give rise to altruistic actions that attempt to relieve the other’s suffering, but need not; for Snow, the desire for this relief is enough for compassion.
  42. The suggestion that suffering, and suffering from formative trauma, are relevant for compassion is not at odds with my argument above that suffering is not sufficient for exemption; suffering is morally relevant for a different reason.
  43. That is, what fundamentally distinguishes compassionate blame from garden-variety blame or garden-variety compassion is the requisite set of beliefs, though in addition, compassionate blame has a unique phenomenology through its emotional component that is not just that of blame plus compassion.
  44. I have focused my discussion on trauma since I think it poses particular challenges for these cases. However, I want to leave open the possibility of a broader view in which we ought to have compassionate blame whenever extreme or prolonged suffering (including that which was not traumatic) relevantly shapes an agent’s transgressions. Such a view could be spelled out in the relevant beliefs of compassionate blame by augmenting each instance of “trauma” with “formative suffering” (e.g., (1) the survivor has or continues to suffer as a result of trauma or formative suffering).
  45. Though Watson’s claim doesn’t imply this, moral intuitions about the tension of our responses may proceed from a false dichotomy between victims and transgressors: either you suffer or you cause suffering; either you’re evil or you couldn’t help what you did. The assumption is that trauma survivors who transgress can fit in one box or the other, but not both (though again, Watson’s claim is not this strong).
  46. Relying on the metaphorical imagery of ‘nesting’ is perhaps a shortcoming here, but I mean it to capture the experience of continuing to blame someone while also feeling compassion for them and their broader history (including those features that contributed to their blameworthiness), and that holding these emotional responses together and tracking their connection is itself a unique phenomenological response of compassionate blame.
  47. Though again, I’m not thinking of compassionate blame as just compassion plus blame.
  48. Calhoun does not discuss traumatic formative history in particular, but her discussion of history is still relevant to these cases.
  49. Kyla Ebels-Duggan takes up the role of history in a person’s blameworthiness and argues that it does not make any moral difference to responsibility or blameworthiness (2013: 142). Rather, history plays a purely epistemic role in discerning those normative attitudes that do make a moral difference (2013: 158). I disagree with both Calhoun and Ebels-Duggan on this point: I hold that (traumatic) history does make a moral difference, though I agree with both that this difference is not that it mitigates blame or responsibility. The moral difference that traumatic history makes is that it calls for compassionate blame as a moral response to survivors, which both holds them responsible through blame and attends to the fact of their suffering.
  50. Of course, strangers may not know a survivor’s relevant history, or may not know it the way a loved one does. So, insofar as strangers recognize the relevant influence of trauma, they ought to have compassionate blame towards survivors. Insofar as this history is unknown, strangers may be epistemically justified in garden-variety blaming survivors (though not because this type of blame is now morally appropriate).
  51. Of course, this will not be the case for all blameworthy agents. An agent may deserve compassion independently, for suffering that is unrelated to their blameworthy action, but this is not compassionate blame. Thus, compassionate blame is inappropriate when there is no relevant traumatic influence on a blameworthy action; here, standard blame is appropriate. Notice this implies that compassionate blame is appropriate whenever trauma and transgression are relevantly historically linked, and so may be appropriate quite frequently.
  52. That is, I think we all ought to try to cultivate compassionate blame in the relevant cases, but this may be difficult or even impossible for some people (particularly loved ones) given the nature of the wrong. The issue of achievability is separate from considerations of appropriateness, however.
  53. I thank Larisa Svirsky for this nice point.


Alford C. F. (2015). Subjectivity and the Intergenerational Transmission of Historical Trauma: Holocaust Survivors and Their Children. Subjectivity, 8(3), 261–282.

American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). American Psychiatric Association.

Anderson R. E., Edwards L., Silver K. E., and Johnson D. M. (2018). Intergenerational Transmission of Child Abuse: Predictors of Child Abuse Potential among Racially Diverse Women Residing in Domestic Violence Shelters. Child Abuse and Neglect, 85, 80–90.

Bell M. (2008). Forgiving Someone for Who They Are (and Not Just What They’ve Done). Philosophy and Phenomenological Research, 72(3), 625–658.

Bijleveld C., Hill J., and Hendriks J. (2016). Sexual Abuse within the Family: The Intergenerational Transmission of Victimhood and Offending. In Kury H., Redo S., and Shea E. (Eds.), Women and Children as Victims and Offenders: Background, Prevention, Reintegration. Vol. 2. (905–921). Springer International.

Buss S. (1997). Justified Wrongdoing. Noûs, 31(3), 337–369.

Calhoun C. (1992). Changing One’s Heart. Ethics, 103(1), 76–96.

Card C. (1996). The Unnatural Lottery: Character and Moral Luck. Temple University Press.

Ebels-Duggan K. (2013). Dealing With the Past: Responsibility and Personal History. Philosophical Studies, 164(1), 141–161.

Felitti V. J., Anda R. F., Nordenberg D., Williamson D. F., Spitz A. M., Edwards V., . . . Marks J. S. (1998). Relationships of Childhood Abuse and Household Dysfunction to Many of the Leading Causes of Death in Adults: The Adverse Childhood Experiences (ACE) Study. American Journal of Preventative Medicine, 14(4), 245–258.

Fischer J. M. and Ravizza M. (1998). Responsibility and Control: A Theory of Moral Responsibility. Cambridge University Press.

Herman J. L. (1997). Trauma and Recovery. Basic Books.

Merrick M. T., Ford D. C., Ports K. A., and Guinn A. S. (2018). Prevalence of Adverse Childhood Experiences From the 2011–2014 Behavioral Risk Factor Surveillance ­System in 23 States. Jama Pediatrics, 172(11), 1038–1044.

Nagel T. (1979). Moral Luck. In Mortal Questions (24–38). Cambridge University Press.

Nussbaum M. C. (2001). Upheavals of Thought: The Intelligence of Emotions. Cambridge University Press.

Oka M. (2017) Interpersonal Violence. In Carlson J. and Dermer S. B. (Eds.), The SAGE Encyclopedia of Marriage, Family, and Couples Counseling (908–915). SAGE.

Prussing E. (2014). Historical Trauma: Politics of a Conceptual Framework. Transcultural Psychiatry, 51(3), 436–458.

Shay J. (1994). Achilles in Vietnam: Combat Trauma and the Undoing of Character. Scribner.

Smith A. M. (2005) Responsibility for Attitudes: Activity and Passivity in Mental Life. Ethics, 115(2), 236–271.

Smith-Marek E. N., Cafferky B., Dharnidharka P., Mallory A. B., Dominguez M., High J., . . . Mendez M. (2015). Effects of Childhood Experiences of Family Violence on Adult Partner Violence: A Meta-Analytic Review. Journal of Family Theory and Review, 7(4), 498–519.

Snow N. E. (1991). Compassion. American Philosophical Quarterly, 28(3), 195–205.

Strawson P. F. (1974). Freedom and Resentment. In Freedom and Resentment and Other Essays (1–25). Harper & Row. (Reprinted from Proceedings of the British Academy (1962), 48, 187–211.)

Superson A. M. (1993). Right-Wing Women: Causes, Choices, and Blaming the Victim. Journal of Social Philosophy, 24(3), 40–61.

van der Kolk B. (2014). The Body Keeps the Score: Brain, Mind, and Body in the Healing of Trauma. Viking.

Walker M. U. (1991). Moral Luck and the Virtues of Impure Agency. Metaphilosophy, 22(1 & 2), 14–27.

Wallace R. J. (2010). Hypocrisy, Moral Address, and the Equal Standing of Persons. Philosophy and Public Affairs, 38(4): 307–341.

Watson G. (1987). Responsibility and the Limits of Evil: Variations on a Strawsonian Theme. In Schoeman F. (Ed.), Responsibility, Character, and the Emotions: New Essays in Moral Psychology (256–286). Cambridge University Press.

Williams B. (1981). Moral Luck. In Moral Luck (20–39). Cambridge University Press.

Wolf S. (1987). Sanity and the Metaphysics of Responsibility. In Schoeman F. (Ed.). Responsibility, Character, and the Emotions: New Essays in Moral Psychology (46–62). Cambridge University Press.

Wolf S. (2015). Character and Responsibility. The Journal of Philosophy, 112(7), 356–372.

Zerach G. and Solomon Z. (2016). A Relational Model for the Intergenerational Transmission of Captivity Trauma: A 23-Year Longitudinal Study. Psychiatry, 79, 297–316.