David Estlund ‘What’s Unjust About Structural Injustice?’. Critical Precis by Peter de Marneffe.

29 Replies to “David Estlund ‘What’s Unjust About Structural Injustice?’. Critical Precis by Peter de Marneffe.”

1. Estlund’s paper is interesting, clear, and provocative and De Marneffe’s precis does a professional job of capturing the shape of his argument and explaining its significance.

   Estlund articulates an intuitive worry about structural injustice shared by moral philosophers and perplexed lay people alike: ‘How can this be wrong if there is no culprit at whom a grievance attitude should be directed?’ However, to those sympathetic to a structural approach this reactions seem odd. I think the key to the difference in reaction is a different understanding of injustice and its relationship with morality.

   Unlike a simple deontic account, structural approaches to justice don’t see justice as a subcategory of moral duties that should be enforced. Nor do they see justice as a set of goals to be achieved and define the right thing to do as to promote these goals as a telic view would. They have a different view of the relationship between morality and justice and hence a different view of the relationship between injustice and grievance attitudes.

   This alternative view of the relationship between morality and justice can be seen most clearly in discussion of interactional verses institutional approaches to human rights. In these debates supporters of an institutional view (early Pogge) identify human rights (a matter of justice) as primarily about social systems rather than individual wrongs. On this view, beating someone is not unjust it is morally wrong. What is unjust is if the social system positions some groups within society such that they can beat other groups with impunity. Thus, systematically turning a blind eye to domestic abuse or lynching’s is unjust but individual incidents of these moral wrongs that are difficult to pull off and are punished appropriately do not show that there is injustice. On a purely structural view only structural injustices should be considered injustices, other kinds of ‘injustice’ are in fact immoral acts.

   On this sort of view should injustices in systems are not just bad they are wrong. According to Young’s account social structure treat people unjustly when they advantaged some and disadvantaged others on an arbitrary bases. The inequalities that matter concern access to goods needed to develop and flourish and relations of domination. In both cases what matters is vulnerability to these occurrences.

   Should these injustices attract reactive attitudes towards the system itself? This is one way of replying to the problem Estlund identifies. I think whether this makes sense or not is something on which people have wildly divergent views. Does ‘Don’t hate the player hate the game’ make sense? I strongly suspect a plausible version of this idea could be developed but it will not be easy to persuade the sceptic.

   A second approach is to apply the grievance attitude to the aggregate that together reproduce the system. This might follow from seeing this aggregate as ‘sharing responsibility’ for the social structure in some sense. Then the difficulty is making sense of what the appropriate reactive attitude is in this case as the aggregate is not an agentwhose members are neither liable perpetrator nor an innocent bystanders. This, as Young and Estlund discuss, is tricky. Christopher Kutz work might help here in thinking through these issues.

   A third approach, that I believe looks promising, is to reject the idea that reactive attitudes should be aimed at injustice itself rather than at those agents who are related to it in various ways. Structural approaches can and do recognise moral agents as having responsibility or duties with regards to these systems. These relationships are different to liability or culpability (which in the cases Estlund is concerned with is absent hence the problem). However, as he notes, these relationships to injustice might properly attract reactive attitudes. Contra-Young our relative success or failure with regards to fulfilling our relationship to injustice grounded responsibilities or duties with regards to addressing it should attract reactive attitudes (as Nussbaum recognises in her preface to Responsibility for Justice). These might be understood as precautionary collectivization duties (as they are on my account), shared responsibility (as Young suggests), or duties of superintendence (as Estlund calls them). They are not simply positive duties due to their basis in the agent’s relationship to the unjust system. However, they are typically not negative duties because individuals usually are not liable for the shape of their structure. The appropriate reactive attitudes to agents who take up these responsibilities and/pr duties in different ways and different extents is a topic worthy of exploration.

   According to this third approach there is no reason to think the relationship between injustice and reactive attitudes should be direct as Estlund suggests it should. Instead, it should be indirect. This means rulings of structural injustice are a first step towards revising reactive attitudes towards agents variously situated with regards to that injustice. These judgements play an important normative role, however that role is more complex. They assist in helping us understand what we need to do and how we need to judge and react to thers.

   So far the structural injustice literature has focussed on the relationship between structural injustice, and responsibility or duty. Estlund’s paper challenges structural injustice theorists to be clearer on the relationship between structural injustice and reactive attitudes. Questions it provokes structural injustice theorist to address include: Does it make sense to have a reactive attitude to a system? Should those who share responsibility for the system also receive that attitude individually, aggregative, or collectively? Do these attitudes distribute down to participators, members, beneficiaries or contributors? How should judgements of structural injustice affect our reactive attitudes to moral agents standing in different relations to those structures? I look forward to responses taking up this challenge that can build on the excellent work on injustice and reactive attitudes already done by Srinivasan (2018) and Cherry (2021).

2. I want to start by thanking David Eslund for his intriguing paper and Peter de Marneffe for summarising the arguments so clearly.

   In his thought-provoking paper, Estlund puts pressure on the claim, implied by many structural injustice theorists, that structural injustice is wrong even when there is no individual culprit involved. Estlund posits what he calls the reach/grievance dilemma. As he rightly point out, structural injustice theorists (including Iris Marion Young and Sally Haslanger) tend to endorse an hybrid view that contends that ‘while injustice does not depend on individual wrongdoing (that widens its reach), it is nevertheless a wrong (that suggests that grievance attitudes are warranted)’ (p. 343). If I understood Estlund’s main argument correctly, what he finds problematic with the hybrid view is that (a) it does not explain why a structural injustice is wrong (and not just bad) if it is not because of individual wrongdoing and (b) why victims of injustice justifiably might hold feelings of grievance if there is no individual culprit to blame for their condition.
   I must say that there is much I admire in and agree with Estlund’s arguments. In particular, I share the intuition that the relation between structures and individuals’ (wrongful) actions is more complicated and in need of clarification than accounts of structural injustice tend to convey, and I have tried to hint at that in my own work on historical injustice. My sense is that the main reason why theorists like Haslanger and Young stress that structural injustice is wrong independently of individuals’ wrongdoing is that they aim to show that such an injustice is irreducible to the wrongs committed by individuals and that the focus of analysis (and target of change) should be structures and institutions, rather than individuals’ behaviours. However, I believe that, in reality, structural injustices rarely happen without any wrongful individual action, especially if we endorse a broader account of what counts as a wrongful behaviour/attitude. This observation leads me to a second point of agreement with Estlund: some structural injustice theorists tend to rely on a very (and too) narrow conception of individual culpability (see p. 349) – one that does not (at least not fully) theorise the culpability underlying, for instance, omissions, negligent conduct, attitude, and legacy wrongs (p. 349). Especially in the case of Young’s work on structural injustice, this narrow sense of culpability comes down to her preoccupation with the allegedly paralyzing and defensive reactions to be blamed: one of her worries seems to be that blame leads to inaction and, since structural injustices are conditions that need to be rectified, it’s better to find ways to make individuals feel responsible that do not involve blaming them. That being said, I agree that culpability is much more complex than such an account suggests and, indeed, more recent work on structural injustice precisely starts from similar types of unsatisfaction with the ‘orthodox’ approach to blame endorsed by theorists like Young.
   Nevertheless, there are a few issues that I would like to press Estlund on in the hope to reach more clarity about what makes structural injustice wrong and what place blame has in it.

   (1) One thought that I have is that the wrongness of structural injustice has to do with its evitability (and, in some cases, predictability). Here I take structural injustice generally to refer to conditions where (usually groups of) persons are in a position of vulnerability (that can lead to, for instance, exploitation and/or domination). What makes these conditions wrong, rather than just bad is that they can be avoidable. This is the case even when it is not so easy to find specific individuals committing wrongdoings.
   Consider the following example (which I also mention in my own work). In Italy, women obtained the right to a legal and safe abortion back in 1978 (with the passing of the law 194). However, because many (roughly seven out of ten) gynaecologists are conscientious objectors on religious grounds, women are very often prevented from exercising their reproductive right and, thus, they are arguably put in a very vulnerable situation. They are suffering a (gendered based) structural injustice. I believe that many would share with me the intuition that such a condition is not simply bad, but also wrong. However, it is difficult to trace that wrongness back to individual wrongdoings. If we accept the permissibility of doctors’ conscientious objection (which I grant for the sake of the argument here), it is not clear that the unjust structural condition Italian women seeking to have an abortion are likely to find themselves in is the individual gynaecologists’ fault. After all, they are doing nothing wrong in refusing to perform an abortion; quite the contrary, they are exercising their legal and (for the sake of the argument) moral right of freedom of conscience. What seems to be wrong in this situation is that the unjust position suffered by those Italian women could be well be avoided and prevented through, for instance, institutional measures (e.g., guaranteeing enough doctors willing to perform an abortion in public hospitals in each Italian regions). Of course, one might say that many doctors who are conscientious objectors also hold wrongful views about women and their sexual freedom, but, if we reduced the structural injustice in question to such individual attitudes, we would miss its main wrongness: the fact that Italian women are put in a vulnerable position when seeking an abortion, which can be institutionally and structurally prevented.
   (2) Although I agree with Estlund that many theorists of structural injustice rely on a too narrow understanding of culpability, I fear that Estlund, instead, endorses a too broad conception of it. For instance, should we be hold culpable (and blameworthy) for the unconscious biases we might fall prey of? Unconscious biases are certainly problematic, but I doubt that many would think that people should be blamed for holding them in the same way as, for instance, they are blameworthy if they intentionally commit a wrong. If this is the case, then, the claim that, sometimes, structural injustices can happen independently of individual wrongdoing has some bite. Consider the example of mandatory arrest policies in cases of intimate-partner abuse in societies like the US characterised by racial structural injustices. One very serious effect of such policies is that very often women racialised as non-white end up being arrested by police officers because they are perceived as particularly aggressive (even when they are allegedly the victims of abuse). Some police officers might hold racist attitudes and should be blameworthy for them. However, it does not seem implausible to suggest that some might simply display unconscious biases, which are difficult to eradicate even when one does not subscribe to a racist worldview. Is Estlund willing to include mechanisms such as unconscious biases into the reach of culpability?
   (3) A final observation. Estlund concludes by asking ‘[I}f there is no culprit, as well as no warrant for grievance attitudes, what does it mean to claim that it is wrong at all rather than only bad?’ (p. 342). I believe that many theorists of structural injustice would reply that the difference is that conceiving of structural injustice as wrong means that we have a responsibility to address it, rather than simply lamenting its misfortune.

3. I would also like to join the others in thanking both Peter de Marneffe and David Estlund for excellent and thought-provoking reflections.

   I think this paper is very timely and rightly pushes us to think further about what is wrongful about structural injustice at a time where the very concept of structural injustice is being questioned. I believe the paper captures an important internal tension within the literature on structural (in)justice that Elizabeth has also hinted at above. Young and many others who have worked on structural (in)justice seem to want to both have a broadly relational view of justice that insists on how individuals and groups are connected to one another such that some are made more vulnerable to relational wrongs (domination, exploitation, discrimination, etc.) and simultaneously acknowledge that many are non-culpably responsible for maintaining the structure that leads to these pockets of vulnerability. Yet, I think that Estlund is right to point out that it is then unclear in what sense structural injustices are unjust.

   I would like to explore two things in response to the text.

   First, I doubt that contemporary instances of structural injustice do not typically come with significant instances of wrongdoing, especially once we consider negligence and expressive wrongs. If we understand negligence broadly as unreasonably failing to act in a way that a person who would have taken the fundamental interests of others seriously would have acted – as, for instance, Moreau (2020) and Sangiovanni (2018) have both suggested – then it seems reasonable to argue that an important aspect of structural injustice is the generalized failure to treat the interests of some on a par as the interests of others. In a word, it amounts to a generalized failure to treat and regard others as equals, in the way relational egalitarians have suggested.

   In this way, it seems that those defending the relevance of structural injustice (beyond structural badness) could underline that grievances may be warranted precisely because individuals who reproduce and sustain social structures through their actions fail to consider them and treat them as equals (to perhaps varying degree and with varying responsibility of redress). Accordingly, by underlining that social structures – as they are typically understood – are reproduced and sustained through individual actions, there is a fundamental failure of treating others as equals, which may be excusable in some cases, but nonetheless warrants some grievances. Consequently, I am not sure I see how introducing negligence, understood as a failure to treat others as social equals, leads to a “mishmash of disparate ingredients, seemingly cobbled together ad hoc” (359).

   Second, the connection between deontic and telic understandings of equality are of fundamental importance, but nonetheless sometimes overblown. If one accepts – as I think it is reasonable to accept – that recognizing that a state of affairs is bad, then it at least gives us a pro tanto reason to change the state of affairs. Accordingly, recognizing that a situation where some are more vulnerable to relational wrongs provides a reason to change the state of affairs, especially for the actors who are in a good position to change it at relatively little costs to themselves. Consequently, underlining this connection between telic equality and deontic equality – in that recognizing that some things are valuable provides us with (some) reasons for actions – may be a way to deflate the dilemma. It may provide us with the reach necessary to establish that some state of affairs ought to be changed, and simultaneously explain why grievances against those who could change it at little costs to themselves but fail to do so may be justified.

4. Thanks to everyone involved in generating this conversation and contributing to it.

   I am sympathetic with the points made already about how political theory is related to moral theory. As Elizabeth Kahn suggests, “Unlike a simple deontic account, structural approaches to justice don’t see justice as a subcategory of moral duties that should be enforced. Nor do they see justice as a set of goals to be achieved and define the right thing to do as to promote these goals as a telic view would. They have a different view of the relationship between morality and justice and hence a different view of the relationship between injustice and grievance attitudes.” Very simply, on my view, the state has a basic responsibility to organize us in ways that are mutually beneficial (this includes protection of basic rights, fairness in distributions, etc.) – this is the justification of the state – and when it fails to do this, it is unjust and we can have a legitimate grievance against it. The state demands our compliance, and we demand it perform in certain ways. On this view, it is odd to even suggest that agents are just or unjust. (Though I don’t care much about grievance, per se. I still think there can be injustice, even if grievance is inapt. IMHO, the emphasis on reactive attitudes, in general, is misguided in moral philosophy and social philosophy.)

   A second concern I have is the use of the Hurricane Problem. As I understand it, Estlund uses the Hurricane Problem to suggest that there isn’t a morally significant difference between socially caused pain and suffering and naturally caused pain and suffering. So the defender of the hybrid view cannot say that socially caused pain and suffering is unjust if naturally caused pain and suffering isn’t.

   The Hurricane Problem: For any simple structural type of which an instance *could conceivably (and consistent with natural laws) be naturally produced*, as by a hurricane, being of that form does not by itself ever warrant grievance attitudes. But all simple structural types could conceivably have instances that were naturally and not socially caused. Therefore, no simple social structural form itself warrants grievance attitudes. (346, my italics)

   It seems that Estlund is saying that if a socially caused event *could have occurred naturally,* then the socially caused *type of event* doesn’t warrant grievance attitudes. But consider an analogy:

   The Bullet Problem: Suppose Smith aims a gun at Jones and shoots. The bullet enters Jones’ heart and Jones dies. There are conceivable instances when the same bullet enters Jones’ heart from the same gun and same trajectory (etc), but because a robot malfunctioned. But a malfunctioning robot is not a moral culprit. So murders, in general (not just by robots), don’t warrant grievance attitudes. [If grievance has to come from the individual wronged, not a bystander or loved one, then we can make it that Jones doesn’t die. And if the robot is “too social” and not natural enough, then imagine the wind caused the gun to fall which caused it to discharge. And if guns are too social, change the whole thing to a rock being dropped by a person or falling from the same height….]

   The Bullet Problem seems misguided; but how is it different from the Hurricane problem? Obviously, even on the deontological account, causes matter. And it is implausible that if the result *could have been caused naturally* it isn’t a moral issue. It does matter whether the pain and suffering was caused by a failed social infrastructure. This hurricane problem is not a compelling objection to the telic or hybrid approach.

5. Thanks for the excellent article David, and for the nice summary Peter. I wonder if you could say more about Slice and Patch that you introduce on p.351. You know this stuff much better than I do, but it seems these cases are helpful to escape the reach/grievance dilemma. Here’s a version of the case focused on racial injustice so that Soupers don’t need to find it:

   Racist Slice and Patch: Patient, who is black, needs an operation to prevent them suffering serious harm. Slice and Patch are racist doctors. Slice’s job is to slice. If Slice does that job without Patch patching, Patient will die. If Patch patches without Slice slicing, they will inflict harm to no benefit to Patient. Slice will not slice, whatever Patch does. Patch will not patch whatever Slice does.

   These judgements seem nailed on:
   1) Slice does nothing wrong by not slicing, for doing so would kill Patient.
   2) Patch does nothing wrong by not patching, for patching would harm Patient for no benefit.
   3) Slice’s disposition, resulting from their racist views, causally contributes to Patient not being saved; same for Patch.
   4) Patient is warranted in having hostile attitudes or grievance to Slice and Patch given their failure to prevent Patient suffering serious harm; attitudes that would not be warranted towards a hurricane. And his hostile attitude is due because the harm Patient suffers is the product of the dispositions of Slice and Patch, based on their racist views. They are warranted in feeling grievance not just at the attitudes or dispositions of Slice and Patch, but because of their role in the harm Patient has suffered.

   There are difficult further questions about this case (who, if anyone, is responsible for the suffering? have Slice and Patch together acted wrongly if neither has acted wrongly individually?). But however we answer those questions, 1)-4) seem clearly right don’t they?

   Can’t we conclude, then, that hostile attitudes that would not be warranted in the case of the harmful effects of a hurricane are warranted in some cases where no one has acted wrongly – cases where problematic attitudes by several people together cause a bad outcome without anyone acting wrongly? So we have reach (injustice without wrongdoing) and we have warranted grievance. And it seems pretty plausible that many cases that seem like structural injustice cases are like this: cases where no one is in a position to correct a social structure because of the dispositions of others, but the dispositions of many result in problematic disadvantage.

6. First, let me thank the organizers for inviting me, and to Peter for such a clear and thorough precis.

   Thanks, Elizabeth. I think I agree with everything you say here, and I appreciate your list of some questions this should leave us with. You propose that in response to these challenges, we “reject the idea that reactive attitudes should be aimed at injustice itself rather than at those agents who are related to it in various ways. … rulings of structural injustice are a first step towards revising reactive attitudes towards agents variously situated with regards to that injustice.” That is certainly suggested (if not proven or assumed) by my arguments in the paper, so I lean the same way as you here. As you know, the general question I raise for that approach is whether it can explain a warrant for “grievance” reactive attitudes in all the cases we want it to, as we think about wrong associated with structural injustice. (“We” won’t all agree on which cases those are, so each of us asks this for ourselves.) What if no one has plausibly violated any duty, and yet the structural problem is in place? . Now, it’s always possible that once we feel clear about these matters some cases that we once responded to with resentment or anger no longer arouse those attitudes now that we recognize that we were too quickly supposing there must be wrongdoing. To avoid this we’d need an account that reaches those cases. As part of an attempt to meet this challenge of “reach” as I call it, it may be possible to find individual targets than we might think, and categories that I haven’t discussed, of grievance attitudes toward individuals that are warranted and relevant in the right way to the structural problems. As my paper leaves things, the question here is what might make a person a warranted target of such grievance other than wrongful contribution to the troubling structure, or violation of superintendence duties (of prevention, remedy, etc.) I don’t claim to have exhausted the possibilities, but nor am I aware of an adequate suggestion yet.

7. Thanks Hugo. I think you’re right to emphasize that in common cases we call structural injustice there are often enough culprits to go around—that is, some agents who are relevantly blameworthy to give warrant to some grievance attitudes. So, even if we insisted on there being guilty individuals for there to be a wrong of structural injustice, we could cover some of the central cases we’d want to cover.

   Now, Young and Haslanger have argued that structural injustice is a wrong in itself, not only bad. And we can recognize that intuition. But they each say that it not a wrong by any agent, so insisting that it’s not a wrong unless there are agentive wrongdoers would be to part ways with them.

   Suppose we do. Next, the question would be about cases where our sense that there is a wrong of injustice strikes us as irrespective of whether anyone is to blame for producing or failing to remedy it. Perhaps class hierarchy will puzzle us this way: its coming to exist, and even persisting (at least for some time) doesn’t seem to depend on any wrongdoing by anyone, even if in fact there is such wrongdoing. Will we rescind our judgment that it is a wrong in that case? Or should we hope and expect to find culprits of some kind we haven’t yet anticipated?

   You mention suggestions such as that many will have failed “to treat the interests of some on a par as the interests of others.” That sure does sound like an agentive case of wrongdoing, so we’d want to know if it’s bound to be present in the cases we’re hoping to cover. I think that this very existence of more powerful and less powerful classes doesn’t itself settle that there have been any such failures of respect either. If in a given deeply class-divided society many people are treating those in lower classes disrespectfully (which is surely common), that’s wrong by individual agents. And, maybe, it could even take a form that could only occur under class inequality—certain kinds of disrespect are thereby made possible. So, it’s got a close relation to the structure. But we should ask this: plausibly there could be that deep class inequality itself even if, in addition to no legacy or superintendence culprits, everyone fully meets their duties of interpersonal respect. In that case, do we admit that there is now no wrong of injustice (perhaps still telic injustice), and no warrant for grievance attitudes of resentment, indignation, righteous anger, etc. in that deeply class-divided society? If not, we haven’t yet been told what, in that society, warrants such attitudes. Even though, as you say, there will normally be such violations of duties, we still want to know what the wrong of structural injustice is if it is meant to be present even irrespective those violations.

   Finally, as you say at the end, if something is bad even if not wrong, as telic injustice is said to be, that might well give agents reason to help to prevent and remedy it. Yes, but as I argue, that doesn’t show that if it is present someone must have failed in those duties, as I argue in the paper. They might not know how to prevent or remedy it, and/or their part in such an undertaking, if it’s to be successful, could in principle by more than we think could be morally required of them, such as leaving no room for other important projects and activities (parenting, reflecting, helping people in need even where this wouldn’t advance that project, etc.). If there are no such violations of duty maybe we’ll decide that there could, in principle, be a society of deep class inequality without it being a wrong of injustice. But if that sense persists, we don’t yet have a warranting account.

8. Thinking more about some of the comments occurring before mine just above, I’d like to add that neither Young nor I think that structural injustice NEVER involves bad actors. Of course it does; and bad actions can have long term effects. The point is that bad actors are not necessary for structural injustice. The problem is that social phenomena are part of complex dynamic systems and it is very difficult to predict what will happen when we intervene in some part of such a system. We can often attempt to do the right thing with the best intentions, but things go badly wrong because of the system dynamics and interactions between sub-systems. These unintended effects are not just “natural” phenomena evolving, for human practices what sustain the system. Some, such as Jorge Garcia, have argued that we are then morally responsible because we are negligent in failing to stop the practices. But it is way too hard to understand the interactions between the systems and how the injustice is sustained to think that one is negligent for not knowing how to intervene effectively. I worry that a lot of moral and political philosophy treats social systems by analogy with games where the rules and outcomes are clear, and we just all need to obey the rules, or change the rules if they don’t work well; the people who don’t obey the rules and those who don’t try to change the problematic ones are morally responsible for doing wrong. But societies are not like this. We are parts of systems that we can’t control, but social systems are not things that just happen to us either. If our ordinary moral concepts can’t deal with this reality, then maybe we need new ones.

9. One might deny that Slice’s racist disposition contributes causally to Patient not being saved from the harm. If Slice were not racist, but everything else were constant, then Patient would still be harmed, because Patch would still be racist, so Patch wouldn’t patch, and so Slice would still be required not to slice. Thus it seems that Slice’s disposition has no causal effect. Have I got that right?

10. The situation looks symmetrical for Patch. If Patch were not a racist, but everything else were constant, Patient would still be harmed, because Slice would still be racist, and still not slice, so there would be nothing for Patch to patch. So Patch’s disposition doesn’t seem to have a causal effect on the outcome.

    Still, it does seem that Patch’s disposition makes a difference, not to what ends up happening but to what it’s permissible or obligatory for Slice to do. If Patch were not racist, then Slice’s refusal to slice would be wrong, and Patient would have a grievance against Slice for not slicing. Patch’s disposition doesn’t make a difference to the end result, but it makes a difference to how we get there. Patient could address the following complaint to Patch: “If you weren’t such a racist, I would have a legitimate complaint against Slice. It’s your racism that relieves Slice of his duty to help me.” Isn’t that a valid complaint?

11. Alasia, thanks for these, and sorry for getting to you out of order. I’m glad to see our points of agreement and that we’re working along some similar lines. Let me say something about each of the three issues you press toward the end.

    1.
    You describe the condition of women in Italy, where abortion is legal but a majority of doctors (permissibly, we’ll assume) conscientiously decline to perform the procedure. You write that “what seems to be wrong, rather than merely bad, is that this vulnerable position of women “could be well be avoided and prevented through, for instance, institutional measures.”

    You might mean either of two things. First, you mean that there is the relevant knowledge and ability in the population to install institutional measures that would remove this vulnerability. That’s very likely true, I think, even though not enough people are actually doing what is needed of them. In this case, though, I think we have wrongdoers, many of them blameworthy: the citizens who violate duties they have to do their part to remedy the situation (a duty of superintendence). So here, we have culprits, even though they’re not the doctors. As I point out, however, it’s doubtful that such a situation could only occur if not all individuals met their duties of superintendence. Suppose, for example, that they don’t know (or at least haven’t yet figured out) what institutional measures would fix this. Then we won’t have wrongdoing or at least blameworthiness, but the structure remains.

    Now the second thing you might mean is that even if there isn’t any agent who could and should contribute to the solution more than they do, it’s still clear that, in some sense, it “could well be avoided and prevented” by such measures. If that doesn’t indicate any agent who could and should do something they don’t do, though, then this wouldn’t seem to help us understand how wrongness is present and not only badness.

    2.
    You ask whether we are blameworthy for unconscious biases, and if culprits can be found there. Implicit bias and related phenomena are an important and vexing question in moral philosophy. And you’re right that it matters here, since much ostensible social injustice might be produced by, or even consist in, such attitudes. My stance in the paper was to let attitudes such as racial or gender bias (even unconscioius) count as culpable for the sake of argument. This is why I speak often about “actions or attitudes” instead of just actions. (For example, see my note 33 at p. 342, and P. 348, 353, and others.) I want to grant for the sake of argument (at least) that something along the following lines is correct: certain states of our motivations or character are in themselves to be morally disapproved of. Whether they are blameworthy is a further question, so that’s the complication. Sometimes a person has morally bad attitudes, to be disapproved, but this fact is not their fault, whereas for some others it might well be their fault. Maybe the latter, but not the former, could and should have taken steps to fix it. It might depend, for example, on how morally impoverished their (especially their formative) environment is.

    Even if a person is not to blame for being a bad person, they may yet be to blame for their wrong actions, which would give us the simpler kind of culpability. But imagine a society in which many people have and are known to have horrible attitudes toward people of a certain ethnicity. But since there are so few such people in their community, they rarely if ever act wrongfully on these attitudes (suppose merely telling someone about a wrongful attitude you have toward some other group isn’t necessarily itself wrong in itself). I allow that this might be enough to count a society as unjust, and that we’ll count those with the objectionable attitudes as culprits or their counterparts for our purposes. Allowing that for the sake of argument, I go on to press my challenges to cases where no culprits even in that extended sense are present, but there’s still, say, severe hierarchy. (Domination in Pettit’s sense is similar to this case in important ways.)

    3.
    Your third and final issue is your point that even if there aren’t culprits in the sense of wrongdoers who produced the bad structure, there will often yet be people who wrongly fall short of their responsibility to address it. Yes, absolutely. This is a version of violation of what I call duties of superintendence, so I agree with you that it’s a very important site of culpability. I just argue that even adding them in there are cases that are difficult to reach.

12. Hi Andrew,

    I’m no expert on causation, but here’s a quick analysis. In my earlier note I suggested that Slice and Patch each causally contribute to the harm Patient suffers. Perhaps more accurate is to say that they contribute to the cause. Compare: a plinth is on top of two pillars, and it needs both pillars to remain elevated. These two pillars are simultaneously destroyed. The destruction of the two pillars causes the plinth to fall. What is true of the destruction of pillar 1? If your analysis of Slice and Patch were right, the destruction of pillar 1 does not causally contribute to the plinth falling, because the plinth needed pillar 2 as well to stay up. But it can’t be that the destruction of pillar 1 makes no contribution to the cause. Because if that is true of pillar 1 it is true of pillar 2. And this can’t be true of both pillars. So the destruction of pillar 1 contributes to the cause of the plinth falling. Similarly, although it Patient would have been harmed even if Slice hadn’t been racist, Slice’s racism can contribute to the harm.

13. Thanks, Sally, for sharing those two concerns. (As for what you say in your second post at 6:56pm UK time, I agree with everything you say there, so I won’t reply to that further.)

    On your first concern: it sound as though you are not moved by the alleged puzzle in the first place, which I might put roughly in this way: how can it be a wrong (which you say it is) if no agent, by act or omission, commits it? You write,

    “Very simply, on my view, the state has a basic responsibility to organize us in ways that are mutually beneficial – this is the justification of the state – and when it fails to do this, it is unjust and we can have a legitimate grievance against it. The state demands our compliance, and we demand it perform in certain ways.”

    On one natural reading, you think the *state does* commit the relevant wrong: “the state has a basic responsibility.” “…we demand it perform in certain ways.” “…when it fails to do this, it is unjust and we can have a legitimate grievance against it.” It sounds as though the state has a duty, or an obligation, or at least is normatively required in some way, to do certain things—some ground or warrant for our “demand.” This doesn’t sound like a kind of structural wrong that is irrespective of wrongs by individual or collective agents. I address the group-agent way of finding culpability for structural injustice in the paper, but I’ll hold off on that here, since I’m not at all sure that’s what you do have in mind. That’s because you eschew that route elsewhere. For example, in “Oppressions,” p. 313 and the pages around that, you suggest (as I understand it) that structural oppression does, “not imply an oppressing agent (group or individual).” (Crucially, the parenthesis is in the original.)

    Your second concern is about my “Hurricane Problem,” and you quote my central claim, namely that what might conceivably have been naturally produced does not warrant grievance attitudes, and this applies to all social structure in the “simple” sense—that is, conceived apart from its causes. You say, “Obviously, even on the deontological account, causes matter.” Yes, that’s right, and it’s part of my point. So, social structure being social, so to speak, won’t yet explain how grievance attitudes are warranted. Such a warrant depends on something further, such as how it is caused. Similarly, in your bullet case, the victim’s death *considered apart from how it was caused,* does not warrant the grievance attitudes.

    So, on your Bullet Problem I…bite the bullet, but it’s perfectly comfortable: death as such, and social structure as such (“simple structure” as I dub it, so here add the idea of, “simple death”), don’t warrant grievance attitudes. So then we need to hear, presumably, about what kinds of causes must be introduced to warrant them. That’s all the hurricane argument shows, and I would expect you to agree…so far. You seem to agree with me that cause matters to whether attitudes like resentment and righteous indignation are warranted, in both cases.

    That is to set up the next step, which is this: there’s at least one kind of cause we know explains the warrant for grievance attitudes, namely blameworthy wrongdoing by some agent(s). So we should ask: Is there any other kind of cause that would do it? That is, if you deny that cause by agentive wrongdoing is necessary to warrant grievance attitudes (as I think you do), you must believe some other kinds of causes can warrant them. And then we would want some explanation of, a) what those causes are, and, b) how it warrants them. What is often offered as the warranting element is that the structure is not only social (which doesn’t do the trick as per the hurricane problem) but is *socially (rather than naturally) produced.*

    That offers an answer to (a) but without an answer to (b)–how it warrants the grievance attitudes–we aren’t given any reason to accept that answer to (a). As I say at that point in the paper, “…an account must be forthcoming of how something about a social structural condition’s being socially rather than naturally caused warrants such attitudes as resentment or righteous anger even without any culprit to be angry at or resented. It is far from clear, to me, how this would go, but I do not prejudge the matter.” (me at 346)

    Importantly, “grievance attitudes” is a term of art, just a handy name for what seems like a family of attitudes around resentment, indignation, and righteous anger. The fact that we might normally call something very different grievance, like being upset about the weather, or our crappy society (suppose), or grief (!) is not at all relevant here. That wouldn’t be taking seriously the question whether it is an attitude in an apparent family around resentment, etc. Obviously, these all seem to be attitudes that include some representation of blameworthiness (not necessarily believed, but as fear involves some representation of danger) but I proceed without assuming blameworthiness is necessary for warranted grievance attitudes. It might turn out that it is.

    What I expect is that many readers will recognize an inclination to attitudes akin to resentment, etc., in response to charges of structural injustice. They might wonder, as I do, if they can be vindicated even irrespective of culprits. Others might not feel it, or recognize it but now feel it disappear on reflection. In that case, it should be recognized that charges of structural injustice in the culprit-free sense ought to be clearly (and probably more explicitly to avoid misunderstanding) divorced from invitations or inclinations to be resentful, indignant, or righteously angry. Where there are relevant culprits, as there very often are, carry on.

14. Hi everyone, and thanks to David and Peter for a great paper and precis. Picking up on Victor’s comment, it seems like what we need to get out of the reach/grievance dilemma is a solution to the Slice and Patch puzzle: i.e. a view which lets us say that Patient is warranted in feeling aggrieved towards Slice and Patch for the harm he stands to suffer, despite it not being wrong for either surgeon to decline to do their parts of the operation.

    David, in your book Utopophobia, you offer a solution to this puzzle. Very briefly, you argue that Patient stands to be wronged because the surgeons are under a ‘plural requirement’ to operate on him, where this means:

    1. If Patch patches, then Slice is obligated to slice
    2. If Slice slices, then Patch is obligated to patch, and
    3. It ought to be that [Slice slices and Patch patches]

    This plural requirement is a hybrid of deontic (1 and 2) and telic (3) elements. If it is violated, then Patient is warranted in having grievance feelings. But Slice and Patch do nothing wrong by opting not to do their parts of the operation. Rather, a violation of a plural requirement is a ‘wrong without culprits’.

    I’m wondering why you don’t appeal to this solution to the the Slice and Patch puzzle in the paper? It seems like the ‘plural requirement view’ might give you the following response to the hurricane problem. If Patient is harmed as the result of a hurricane, rather an illness which could be prevent if both surgeons were to do their parts, then we lack the structure of conditional deontic obligations described in 1 and 2; it’s false that e.g. Slice is obligated to slice conditional on Patch patching, since here slicing and patching would not benefit Patient (he’d be harmed by the hurricane regardless). So, no plural requirement. This response looks like it scales up: when bad social arrangements have social causes, there’s a complex structure of conditional obligations akin to 1 and 2; but when these arrangements have natural causes, there’s no such structure of conditional obligations.

    Now, (as you know) I have my own reservations about the viability of the plural requirement view as a solution to the Slice and Patch puzzle. But I’m curious as to why you’ve chosen not to mention it in your paper, given how naturally it seems to fit into the argument. Are you sceptical about whether it avoids the hurricane problem?

15. In response to Victor, yes, I see that Slice and Patch’s dispositions are both part of a set that overdetermines the outcome. So maybe there is no problem; we just hold all participating causal factors responsible. The point I found interesting was that instead of appealing to a disposition’s effect on outcomes, we could appeal to its effect on what is permissible or obligatory for the other party to do, and then claim that this effect on what’s permissible / obligatory can give rise to a justified grievance (on Patient’s part). Does that make sense to you? It may be just another way of stating your points 3 and 4.

16. Thanks, Dave.

    I think one problem is that I don’t see the state as a “group agent”. The state is responsible for things and can “do” wrong – or “be” wrong, but not qua agent. You seem to have some ideas about group agents based on a social ontology that I reject, viz., that the social world is constituted by collective intentionality. I just think that isn’t in the least bit plausible when it comes to the state (or many other things). And I think that there are kinds of causation that aren’t billiard ball causation (see my paper on structural explanation). Do you not think that social structures can cause things? Why is that? So we have very different metaphysical starting points. Given that, I’m not surprised that we disagree.

    Moreover, I’d be willing to say that the deontologist’s account of wrong is at least as hard to comprehend (and to me, at least as incomprehensible) as you find the idea that the state can do/be wrong. I find myself in good company with Plato and Rawls, and others. People find different things plausible starting points. So we not only disagree about metaphysics, but about what is normatively plausible.

    I also mentioned, the whole issue of grievance doesn’t move me. I don’t think that we need to find *someone* to blame if we think there is wrong. I see wrong all over the place that no individual or group of individuals (intentionally) caused. And I can be angry about it, without thinking that I have to find an agent to be angry at. I have a hard time finding the assumptions you make about grievance plausible, and my best guess is that they derive from moral philosophy, not from the phenomenology of life under conditions of injustice. I believe that I have suffered injustice, and others have suffered much worse injustice, without there being someone who “did it to us.” I’m not looking to blame agents. To say that our beliefs that we have suffered injustice is misguided because in fact we were just struck by the equivalent of a hurricane comes a little too close (IMHO) to denying something deep about our capacity to make warranted moral judgments. Do you really want to go there?

    Also, I don’t think you understand the point of the Bullet Problem. You say that if the action/event is of a type *that can occur naturally*, then actions/events *of that type* aren’t morally significant. But a killing is a type of event that can occur naturally, so given your argument, a killing of any kind is not morally significant. But obviously murders are morally significant.

    Ultimately, I think we are talking past each other. But that’s OK. I’m glad you are thinking about these things, even if you are thinking about them with very different background assumptions.

17. Thanks, David, for your replies – they are really helpful. Just some extra thoughts:

    (1) Like Beth and Sally, I wonder whether you and structural injustice theorists do not simply hold very different assumptions about the relationship between morality and justice. Unlike you, structural injustice theorists see the two of them as different domains. This does not mean that, as Sally pointed out, structural injustice never involve actors behaving morally wrongly. The point is that what is unjust about structural injustice cannot (and should not) be reduced to wrongful individual actions. To go back to the case of Italian women seeking a legal abortion in Italy, I don’t think that the nature of the problem can be fully illuminated through the lenses of citizens’ failure to fulfil duties of superintendence. In a sense, I think that approaching the problem in this way would miss the workings of a complex net of structures (e.g., patriarchy and the relationship between respectability in the medical profession and acting as a ‘good’ Catholic) that must be changed to improve women’s condition.

    (2) One question that I have, which relates to my previous two last points, is what difference it makes for you to talk about blameworthiness and culpability in the context of structural injustice. You seem to be willing, even if just for the sake of argument, to stretch the concept of blameworthiness to include many unconscious attitudes. I think that structural injustice theorists would think that this is not necessary and might be counterproductive given that many would actually feel that being blamed for something that is partly behind their control is unfair (which means that they might be discouraged from taking up responsibility for changing unjust structures). This comes down again to your different starting point from structural injustice theorists. I read structural injustice theorists as being deeply political in their goals – the aim is to change unjust structures, not to find culprits. This does not mean, though, that structural injustices theories cannot be philosophically sophisticated – Sally’s work is a notable example of how rigorous philosophy can combine with social justice aims. But, it does implies that they start from different assumption (and probably have different goals).

18. I am very grateful to David’s article and the discussion here for bringing things into focus. Perhaps in some cases what we call “structural injustice” has been brought about non-wrongfully, and it cannot be remedied, or at least not in morally permissible ways (for example: when no one can be convinced to perform abortion for reasons of conscientious objections, and it’s impermissible to coerce anyone do do so.) I agree with David that, in those cases, deontic morality will not be able to diagnose an injustice. I have two questions, in response to both the article and the above exchanges.
    1. Does it matter how frequent such cases are, not on absolute terms but in the percentage they represent amongst all cases that we want to call “structural injustices”? If they are few and far between, we don’t really seem to be facing a dilemma. Rather, David draws our attention to the fact that structural injustice – where injustice is meant in a deontic sense – is more limited than some of us perhaps thought it was. Of course, David and others may think such cases are not rare, and even that they are frequent enough to throw into question the very existence of denontic structural injustice. I wonder how much of the disagreement between different participants in the present conversation hinges on this. If indeed this is where much of the disagreement lies, then the philosophical action is in identifying who has what duty to prevent/remedy/compensate; assuming – optimistically – that we could reach rough agreement on this issue, that would also yield a verdict re: the existence and scope of deontic structural injustice.
    2. The problem in (1), about the scope of justice in a deontic sense, is not specific to structural injustice. It is a pervasive problem for theories of justice in general: some people have much shorter, and worse, lives than other people, out of nobody’s fault and with nothing that anybody can do to prevent or compensate. Even those who have deontic sympathies are happy to call these cases “cosmic injustices”, to draw attention to the fact that there is injustice here in the telic, purely evaluative sense. This is in itself an indication that the telic, purely evaluative sense is primary and the deontic sense is downstream for it. (I think there are also proper arguments for claiming that much!) But even if this last observation – re what sense is primary – is incorrect, does anything speak against identifying different ways in which a situation can be unjust, and so against working with a (slightly!*) pluri-semantic sense of “justice”? Can’t we be content to call cases like the ones described in my opening para “structural injustices” in the merely evaluative sense? As long as we’re clear which sense of justice is at stake – and, most importantly, which sense of “justice” entails duties and therefore reactive attitudes – nothing of normative substance seems to ride on this slight pluri-semantism.

    *”slightly” because the telic/purely evaluative and the deontic sense of injustice are closely connected, at least in my view.

19. Victor, thanks for that suggestion of a way out of the Slice/Patch puzzle. You say this, and I wonder if it is it correct:

    “Slice’s disposition, resulting from their racist views, causally contributes to Patient not being saved; same for Patch.” …“And [Patient’s] hostile attitude is due because the harm Patient suffers is the product of the dispositions of Slice and Patch, based on their racist views.”

    I want to reconstruct your reasoning as I understand it in order to expose more of its structure before replying to it. I grant that each doctor’s omission is a causal contributor to Patient’s death. But I see why you go to the motives rather than the omissions. The omissions are morally right, but the motives are morally bad, thus a possible ground for Patient’s grievance attitudes. If the motive causes the omission then the motive too is a causal contributor to the death. On this basis, you appear to argue, Patient has a legitimate grievance against each doctor for what happens to him at noon. Voila?

    I’m not convinced (though maybe I haven’t gotten the argument right). First, I wouldn’t myself say that either Slice or Patch is to blame for anything in this case, and then, as I’ll suggest, it’s unclear how things are supposed to go.

    As I mentioned earlier (in part 2 of my reply to Alasia) as I see it, sometimes motives are not only morally bad but also the agent’s own fault, and blameworthy because of the agent’s own choices. But there’s no suggestion of that in the example, so I assume that the motives are morally bad but themselves also blameworthy. Now, morally bad motives (even if not themselves blameworthy) can ground blame for any *wrong* actions from those motives. But we don’t have any wrong actions in this example.

    And, next, I don’t believe right actions can be blameworthy for the following reason: The question of blameworthiness, I think, is whether the agent is (to whatever degree) to blame rather than (to whatever degree) excused for it. But if the act in question is not wrong, then: excused for what? No such blame/excuse question arises. Right action, it seems to me, doesn’t admit of blame because it doesn’t admit of excuse.

    So, what would this tell us about Slice and Patch? We want to know if Patient has a legitimate moral complaint against each of them for what happens to him at noon. Without the applicability of blame, I don’t see how. Of course, as I say, each doctor’s racist motives are to be morally disapproved. But (and this seems to me to be the key) I fail to see how this warranted disapprovability (so to speak) of the motives can find its way into their rightful actions—and those actions are what especially happened to Patient. (The actions, while right, do lack moral worth, in Kant’s phrase, but probably so did their making coffee that morning. That absence is not a ground of complaint in either case.)

    When someone makes a financial donation to a good cause, but only to burnish his reputation, shall we say that the benefitted charity has a reasonable complaint against him for this? That act for that motive, we can say, manifests moral deficiency, but it’s just the motive that is morally deficient. The action is wholly right, and the charity is warranted in being wholly glad about it. Their judgments of the donor’s character is a fair but separate matter. Likewise, Patient, like any of us, is warranted in disapproving of the doctors’ motives. He even has a complaint that not everyone has on their own behalf: the doctors’ bigotry about Blacks like himself. But all that is no more true on the day on the day of the needed surgery, than on the day or week before. It isn’t available to explain what, if anything, grounds complaint by him especially regarding what happens to him at noon on the crucial day: Slice’s rightly, and fortunately, not doing the surgery which would have led to a much more painful death.

    There’s obviously more that would be needed on this tough issue, but I’ll leave it here and see whether we want to delve more deeply.

20. Thanks for an excellent paper, Dave, and thanks all for a helpful precis/discussion!

    I tend to think that Dave is right both that the reach/grievance dilemma is important and that, at this point at least, the right response to it is not to become skeptical about whether many putative cases of structural injustice really are that.

    One response to the dilemma, which Dave describes, is to extend the reaches of culpability and, thus, ultimately the reach of deontic injustice. Dave mentions attitudes, but, as far as I can see, does not say much about this extension possibility (unlike the legacy wrong extension) (p. 349). Hence, here is a suggestion (perhaps) along the attitudes extension lines:

    In some cases, people have grievance attitudes towards others on account of the other people’s attitudes even when these people at not at fault for having the outrageous attitudes they have. As in: “I realize that given the environment you are in, you cannot be faulted for your outrageous ableist attitudes. Still, these are outrageous attitudes, and I resent you for having them”.

    Suppose this makes sense. (At least, it strikes me psychologically plausible even if, ultimately, we should reject that such grievance attitudes are justified.)

    If so, could one way in which the second disjunct of the Disjunctive Criterion of Wrongness (“or someone with warranted grievance on some other basis ” could be true be this: its being the case that the relevant putatively structurally unjust distribution of goods results from a pattern of actions (none of which, individually, might amount to wrongdoing, though, in real-life cases (as Dave acknowledges), plausibly a lot of them will) that (lots of) people have the wrong attitudes towards. “Wrong views” here could amount to favoring? Or weaker: condoning? Or even much weaker, e.g., not strongly oppose or the like?

    Perhaps this suggestion cast some light on the hurricane challenge. Unlike ableist social structures resulting from or being constituted by patterns of action, which people favor, condone, or simply do not disidentify with, people typically strongly dislike hurricanes. Perhaps that is why social structures that result in bad distributive outcomes are unjust in a different sense from hurricanes that result in the same outcomes?

    Admittedly, this suggestion implies that if, say, people in general had welcomed Hurricane Katrina, victims of it would have a grievance.

    But I am not sure if this is a problem. (Again: that many people would in that case, rightly or not, harbor grievances against hardships caused in this way does not seem psychological implausible.) Hence, I put this one table, not as a claim, but as a possible way of pursuing of extending culpability (or, more to the point, grievance).

21. Thanks, Sally, for that rejoinder. It doesn’t sound like we’ll be coming into agreement about all this today, but I do want to clear up several things that come up in your post. Let me say, though, that I very value any insight into where and why we are in disagreement (and where not), so even though it’s frustrating in a way, I’m glad even to know where your reaction is one of incredulity!

    1.
    You write,

    “I think one problem is that I don’t see the state as a “group agent”. The state is responsible for things and can “do” wrong – or “be” wrong, but not qua agent. You seem to have some ideas about group agents based on a social ontology that I reject, viz., that the social world is constituted by collective intentionality. I just think that isn’t in the least bit plausible when it comes to the state (or many other things).”

    I’m puzzled by this, because I don’t think the state is likely to be a group agent either, and didn’t mean to suggest anything else. I wanted to know if you do, since then I would rehearse my arguments against that kind of solution. So, no different starting points here; I’m with you on that. And knowing this at least helps a bit interpretively, since that language of yours that I quoted sounded a lot like it treated the state as a group agent. OK, so that’s not what you mean by its “responsibilities,” to “organize us,” to “perform in certain ways,” our being warranted (apparently) in making “demands” of it, etc.

    As for your saying just now that even though the state is not an agent it can, notably in your own scare-quotes, “do” wrong, this suggests that maybe the other agential-sounding language could just as well be read as in scare quotes too, no? If so, fine, but then it seems to me only to highlight the very questions we’re trying to unpack (and maybe you intended no more than that). One could, mischievously, play with the idea that the sun “does” wrong sometimes too (it’s actually quite “admirable” this morning as I write), but you clearly mean to claim more than that about the state, still without ascribing genuine rational or intentional agency to it. But I don’t yet know what that more is. As you say, that’s fine for this brief exchange. We can only make as much progress as we can make. As I say, at least I now know you don’t mean group agency.

    2.
    You write,

    Do you not think that social structures can cause things? Why is that? So we have very different metaphysical starting points. Given that, I’m not surprised that we disagree.

    But, of course I agree that social structures can cause things. Hierarchy can cause instability, etc., etc. I’m not sure how I might have indicated otherwise, but I hope this clears it up. Here too, I don’t see that we differ on that (metaphysical) starting point at all.

    3.
    You say,

    Moreover, I’d be willing to say that the deontologist’s account of wrong is at least as hard to comprehend (and to me, at least as incomprehensible) as you find the idea that the state can do/be wrong. I find myself in good company with Plato and Rawls, and others. People find different things plausible starting points. So we not only disagree about metaphysics, but about what is normatively plausible.

    Well, just to clarify what I’ve said about Rawls in this paper: I don’t think you’re wholly in company with Rawls, who certainly thinks there is such a thing as wrong in the deontological sense. The right is crucially different from the good for him (its having “priority,” etc.) Rather, what I do argue about Rawls is that for lots of reasons we’d expect social injustice, for him, to be a matter of the right and not of the good. But it’s not clear that the view comes out that way. And then it’s not clear whether, as we might expect, he even thinks it does. You are suspicious of deontological right itself, but he’s definitely not.

    4.
    You write,

    … I see wrong all over the place that no individual or group of individuals (intentionally) caused. And I can be angry about it, without thinking that I have to find an agent to be angry at. I have a hard time finding the assumptions you make about grievance plausible, and my best guess is that they derive from moral philosophy, not from the phenomenology of life under conditions of injustice. I believe that I have suffered injustice, and others have suffered much worse injustice, without there being someone who “did it to us.” I’m not looking to blame agents. To say that our beliefs that we have suffered injustice is misguided because in fact we were just struck by the equivalent of a hurricane comes a little too close (IMHO) to denying something deep about our capacity to make warranted moral judgments. Do you really want to go there?

    We don’t disagree about whether there is severe wrongness all around us, so we can put that aside. I can agree on that even if we disagree about whether wrongness depends on wrongdoing, because…there is so much wrongdoing in any case. Again, we haven’t turned up different starting points here.

    Beyond that, as I read you here (is this accurate?) you are indicating that even if it’s granted in some case that it is not a matter of wrongdoing, its being not only bad but wrong, and even anger-warranting can be so obvious that it’s obtuse to wonder (as I confess I do wonder) what “wrong’ is meant to add in that case. Certainly, my arguments are addressed to those that find some force in the question how there can be a wrong if no agent commits it. By asking, I’m not denying that there can be. I’m asking. But I hear you indicating that you just don’t find that question to point to anything needing explanation. Here, we finally seem to be at different “starting points” as you say. Where each of us finds it difficult to even understand the other’s starting point here (but, so I’ve argued, on not as much else as you seem to have thought), we might wish for some philosophical route to get underneath this dispute and diminish the incredulity. But I don’t have one ready for today.

    5.
    You write,

    Also, I don’t think you understand the point of the Bullet Problem. You say that if the action/event is of a type *that can occur naturally*, then actions/events *of that type* aren’t morally significant. But a killing is a type of event that can occur naturally, so given your argument, a killing of any kind is not morally significant. But obviously murders are morally significant.

    My view would indeed be a very weak if it were that view, so we agree on that! But you won’t find the phrases “can occur naturally,” or “morally significant” in my paper, for good reason. Also, as I explain in the paper, I stipulate that by “produced naturally” I will mean “not by actions of people, alone or together.” (346) If anyone dislikes that terminology, they can just substitute the longer phrase. Obviously, then, in my stipulated sense many killings are not produced naturally—like murder. So, I’m not making any argument that would imply, absurdly, that murder is not “morally significant” or (what is the more precise issue at stake at that point in the paper) that murder wouldn’t warrant grievance attitudes. (And the issue just here is not whether one is very interested in whether it would warrant them or not. My argument is taking one step at a time.)

22. Rowan, you’re right, it might be helpful to address why I don’t rest more on what I call Plural Requirement. I actually say, in Utopophobia that Plural Requirement, “is an account of one important form of structural injustice.” (232-233) So, wouldn’t that be the thing whose violation would warrant grievance attitudes in cases like Slice/Patch, and by extension, some large set of similar scaled up cases?

    In a word, no. That is: my account of Plural Requirement in Utopophobia has limited aspirations. It’s easy (from what I have observed) to think it’s meant as a way to explain what warrants grievance attitudes about the Slice/Patch case, or in what way Patient is wronged. It is not. After pressing the difficulty of finding deontic wrong or blameworthiness in Slice/Patch cases, “I do not [in the account of Plural Requirement] propose to pull a rabbit out of a hat.” (232) “Plural Requirement is a structured set of facts partly about agential moral obligation, though not an agential obligation in its own right.” (252) Violation of Plural Requirement, “does not … imply any acts or attitudes by any agent that entail any criticism of any agent, not even those constituting the required plurality.” (p. 233). (These three quotations are from Utopophobia.)

    Also, as part of this broad qualification, I don’t ever say (as far as I can find) that Patient is “wronged” by violation. In your really valuable recent article (“Joint Ought,” Philosophy and Public Affairs (2024) 52 (1):42-68.) about Slice/Patch-type cases (which, despite this being my own example, have been discussed by others before and after me), you are careful to acknowledge most of this, although you do say, erroneously I think, that I propose, “that when Slice and Patch go off golfing, Patient is wronged even though no agent/s contravene/s any moral requirement.” (your p. 52) I don’t think I do. That’s fine, it’s still important for you to point out that Plural Requirement wouldn’t explain that thing we might want explained, even if I admit that it doesn’t.

    These important qualifications on my argument seem often to be missed, and for a couple of reasons. One is that Slice/Patch cases (again from observation) seem to get their hooks into many of us, partially eclipsing the fact that Plural Requirement is not meant to offer a fully satisfactory solution. A natural response is, then, what good is Plural Requirement? The answer as I see it has two main parts. First, even though it doesn’t “solve” the Slice/Patch puzzle, it gives an analysis (in those three conditions you quote) of a distinctive, and rather large I agree, set of cases that prompt us to have, but also to be puzzled about, an inclination toward such attitudes as resentment, indignation, righteous anger, and so on. It’s an interpretation of otherwise undefined statements such as: Required(S does x & T does Y). There is no agent under that supposed requirement, so what kind of requirement is it? We seem to have many thoughts of that form, but there is no interpretation of “required” or “ought” according to which it is well-formed (as far as I know). Plural Requirement allows it to be well-formed, but only by lacking some of the things we might have wished from it. If the most we can say about such cases is this 3-part thing, then even if it doesn’t solve the puzzle, it at least marks out that important family of cases. (It also gives me something to plug in for the “ought” or “required” in what I call the requirement of “prime justice” but I leave details about that aside here.)

    But also, second, while (again) it does not find any deontic wrong by any agent (nor, then, any blameworthy wrong) and so doesn’t explain the warrant we might want for the grievance attitudes, the conjunction of the three conditions is notably morally salient in intriguing ways, not unconnected to why it (perhaps illegitimately) tempts grievance attitudes. For one interesting thing: It’s true that when the conditions apply, that isn’t enough to imply any deontic requirement on any agent. But, when both parties do their part as stated in the conditions, each is indeed required, in the standard deontic moral sense, to do so. Each has acted as deontically morally required. It’s easy to see how we might, unless we look very closely at the matter, infer that there are those deontic requirements whether or not they do their parts. That, I emphasize, wouldn’t follow.

    There is another reason the qualification are not always noticed that is my own fault I suppose. Even though I explain all those qualifications, I propose to regard violations of Plural Requirements as, in a certain way, “wrongs without culprits” (p. 237) I think it was asking too much of the reader to hear that in the complicated key I meant it in.

    So, no rabbit out of a hat here: violation of plural requirement is not a violation by any agent of any requirement. When we ask what, if anything, about structural injustice without agentive wrongs might warrant grievance attitudes, it offers no answer. That is why I don’t rely on it for this specific purpose.

23. Andrew, that idea regarding Slice/Patch is interesting: “instead of appealing to a disposition’s effect on outcomes, we could appeal to its effect on what is permissible or obligatory for the other party to do, and then claim that this effect on what’s permissible / obligatory can give rise to a justified grievance (on Patient’s part).”

    First, it’s helpful to see that your suggestion has something of the flavor of Victor’s: Both of you observe that each doctor’s inaction permits the other’s inaction. But so far, since the actions are right, we don’t find a basis for grievance. The motives themselves are morally bad, but they aren’t special to noon on the important day. So, you both suggest in your different ways: What if we focus on the bad motives, and then also on their consequences at that crucial time? Victor tries physical consequences for Patient, you try moral consequences of interest to Patient: changes to Patch’s obligations with respect to Patient (though these don’t affect anything physically for Patient).

    Without more discussion, though, I find it a little difficult to understand your proposal, namely that Patient might have warranted grievance attitudes against Slice because Slice’s objectionable motives make it permissible for Patch not to stitch. Notice that Slice’s bad motives do not make it permissible for Patch to refrain from stitching even if Slice would cut. Patching would be required to stitch in that case, motives aside. So you couldn’t say that Slice’s bad motives make it permissible for Patch not to “help” Patient. Rather, they only make it permissible (and indeed required) for Patch to refrain from gratuitously harming Patient: stitching even though this would be painful and pointless. We already know Slice’s motives are objectionable, as they were on the previous day too. But what’s objectionable about Slice’s motives specifically owing to their having that deontic consequence on the crucial day? Again, the allegedly troubling consequence is this: because Slice is racist and so won’t cut, this makes it permissible (and required) for Patch not to stitch, which would only make things worse.

    It’s a complicated situation of motives, actions, and requirements, so I might not have gotten your idea right.

24. Hi Dave – maybe we should stop here, but just a couple of quick points just because I don’t think I understand:

    1) You say, “But you won’t find the phrases “can occur naturally,” or “morally significant” in my paper, for good reason. Also, as I explain in the paper, I stipulate that by “produced naturally” I will mean “not by actions of people, alone or together.” (346)

    I still don’t understand your argument. Here is your hurricane text:
    For any simple structural type of which an instance *could conceivably (and consistent with natural laws) be naturally produced*, as by a hurricane, being of that form does not by itself ever warrant grievance attitudes. But all simple structural types could conceivably have instances that were naturally and not socially caused. Therefore, no simple social structural form itself warrants grievance attitudes. (346, my italics)

    Yes, I mistakenly replaced “grievance attitudes” with “is morally significant.” Apologies. But killing is a type that can be naturally produced. So am I right that you misspoke? You don’t mean that if I push a rock off a ledge onto your head and kill you, then the killing could also occur by the rock falling off the ledge? Why are you excluding actions from the types of event where the condition is met (that it could be produced naturally)? Why impose a stringent condition on structures that you don’t impose on agents?

    2) You say, “As for your saying just now that even though the state is not an agent it can, notably in your own scare-quotes, “do” wrong, this suggests that maybe the other agential-sounding language could just as well be read as in scare quotes too, no? If so, fine, but then it seems to me only to highlight the very questions we’re trying to unpack (and maybe you intended no more than that).”
    But you also acknowledge that structures can cause things. Isn’t causing things doing something? You seem to suggest that doing wrong can only be done by agents, but as you point out, that just begs the question the other way.

    3) I’m not anti rights talk. I just don’t understand the ontological basis for rights talk. Mostly I see Kantian explanations. Those are the ones I find obscure. If you are asking for the ontological basis for structural injustice, I think I can ask for the ontological basis for rights talk. And I’ve made significant efforts to explain the ontology of structures and their effects. What I was suggesting is that we can just accept that there is some explanation or other and go on, or we can demand that others accept our ontology. I do think that it is possible to accept certain moral claims without demanding a full, comprehensible, ontology. I do that all the time.

25. Anca, I’m glad these exchanges are, at least to some extent, “bringing things into focus” for you. Same here. Your two questions are helpful.

    1.
    You agree with me that (paraphrasing) in some cases that we might want to regard as structural injustice they may not involve deontic wrongs of any kind, but at most telic bads. Then you ask, “Does it matter how frequent such cases are, not on absolute terms but in the percentage they represent amongst all cases that we want to call ‘structural injustices’?” Maybe, as you say, such cases are a big fraction of the “target” cases of “structural injustice.” But suppose, as you consider but don’t assume, they are “few and far between” as a ratio to cases that can deliver deontic wrong? Does the point matter much in that case?

    Now, you’re granting (and not just for the sake of argument) that the point “matters” philosophically. That is, you aren’t saying that it really comes to nothing philosophically, as if it turned out to seem more significant through some play on words or something. You agree with the philosophical substance, if I understand you. So, it seems that what you are asking is how much this matters in some other sense. One possibility you might mean is mattering about what else to think, and another is mattering about what to do.

    If there are very few real world cases that seem structurally unjust but which also involve no deontic wrong, does this matter much for what else we should think? I believe it still does. Once we see how that is even a possibility, we can now extend the point in this way: Even though in the majority of cases there is, in fact, relevant deontic wrong, it might, as an empirical matter, be that the structural conditions we seem troubled about don’t depend on that wrongdoing. Alternatively, it might be that they always do seem to so depend. If they don’t so depend, the question arises whether that structural condition without those wrongs would still be, in some way, wrong. If instead they do always depend on relevant deontic wrong, I think many would then want to consider whether the wrong is, after all, nothing but the deontic wrongs. These are philosophically very different conclusions about the nature of right, wrong, justice, and injustice. They have very different implications for yet other things we should think, and what inferences they license, and so on. Which of these is correct isn’t much affected by whether, in actual cases we usually find relevant deontic wrong.

    As should be clear, I myself take it to be far from clear that all important target cases do depend on relevant deontic wrong (I’m in the Haslanger/Young camp on this). One case, for example, might be certain kinds of class hierarchy. If, as I think, some such cases could be present even if there isn’t and hadn’t been relevant deontic wrong, there is a big philosophical question about whether they still count as wrong rather than bad, what are the various things that might mean, and so on.

    Turning then to whether it matters for what to do: First, just to mention it and put it aside, I think the idea, sometimes suggested, that all that matters is what matters for what to do, is deeply and darkly implausible. But it’s one very important category of what matters, so let’s turn to it in this context: If it turned out that the vast majority of cases we’re interested in counting as structural injustice were in fact accompanied by relevant deontic wrong, since this might be a fine vindication of the grievance attitudes we experience, would it matter for what to do if it were also the case that in principle there could in principle, though not yet in reality, be those very same troubling structural types even without relevant deontic wrong?

    It seems to me that it very well might. One possibility is that it matters practically which of those is true for this reason: if instead it had turned out that the only kind of wrong is deontic wrong, that might, depending on circumstances, have been some guide to how to engage in reform—focusing on reforming individual acts and attitudes. That’s not guaranteed, but that possibility is often noted in treatments of structural injustice: it’s said to be important whether there is more than deontic agentive wrong since that might bear on whether to pursue reform by targeting individual acts and attitudes, or by aiming for broad structure reform instead. Now, it might yet be that directing your energy directly at structural reform, or at personal reform, would be best in either case. That would depend on a million things. But which kinds of bad and wrong there are could very easily be relevant to these practical questions.

    Another kind of practical relevance would be this: even if there is only a small fraction of cases that are structurally wrong or bad even without deontic wrongdoing, it would be practically important to appreciate this and to guard against unjustified blame and anger in those cases. Indeed, even if there is always some relevant wrongdoing, but not always very much, that would be important too in not exaggerating the degree and scope of anger, etc., that is justified. This might sound like a small matter about being polite or civil. What I mean is the consequences of these stances for unjustified social division. Social division is sometimes valuable as a step toward something better. But, as every day’s headlines now remind us, it can also be profoundly dangerous, and usually most dangerous for the most vulnerable.

    2.
    I can be brief(er) on your second question. Yes. That is, I have little or no problem about several very different kinds of “structural injustice.” I have “no” problem with saying that structural racism is structural injustice, even if its wrongness were wholly the wrongdoing of the racists. It can still be profoundly structural too. Then there might also be, as many think, a very different kind of “structural injustice” that is not at bottom wrongs of individuals. I have problem with pluralism there.

    I have “little” problem (but not “no”) problem with speaking about structural (and more generally social) injustice in a telic sense. I obviously accept it in the paper. There’s a sense in which an author can use any word to mean anything they want as long as (if it’s possible!) they can succeed in keeping that meaning clear. But words come to have very robust conventional meanings, and departing from them risks leading to equivocation between several different things.

    In an important paper that I cite (“The Significance of Injustice,” in Rawls’s “A Theory of Justice” at 50, ed. Paul Weithman (Cambridge: Cambridge University Press, 2023), 95–106.) Peter de Marneffe lays out a strong argument that prior to Rawls, in the tradition of thought about what was called “justice” and “injustice,” those terms overwhelmingly referred to one person wronging or not wronging another person. Still, neither Peter nor I think there’s any point in trying to turn back the linguistic clock; who cares as long as we can keep our meanings clear? But that change does risk serious misunderstanding, and part of what I’m doing is trying to guard against some of it. For example, if one is told their society is deeply unjust on the ground of significant economic inequality, they could be forgiven for taking this to mean that some people are taking advantage of, and benefitting at the expense of, others—injustice in the traditional sense. But, if what the speaker meant is not that, but only the fact of inequality itself as a certain kind of (telic) bad but not wrong (as in “luck egalitarianism”), many important implications for thought, and probably also action, can change dramatically. But sure, there are a number of different things “structural injustice” can helpfully be used to mean if we keep the meaning clear

26. Thanks for that suggestion, Kasper. As it happens, I endorse almost exactly what you say about grievance attitudes toward objectionable attitudes, even if it’s not properly blame for them. See part 2 of my reply to Alasia’s first post, toward the beginning of this whole thread. I’m glad to see you find a similar view plausible. (I’m at about 8k words now, so no worries about not reading or recalling every little bit!) As I say there, I allow in the paper, for the sake of argument at least, that objectionable attitudes can count for my purposes as warranting relevant grievance attitudes even though there not strictly about blameworthiness.

    So, here, I think is how you propose to put it to use. Socially produced structural conditions, unlike hurricane-caused ones, will often indicated some range of attitudes including some objectionable attitudes. In that case, and in that indirect way, those conditions occasion some warranted grievance attitudes even though they don’t presuppose blameworthiness. I’m basically with you on the theory, as I say. But if the attitudes in question are morally objectionable and warrant disapproval by others, then structural conditions they give rise to seem likely to imply a lot of blameworthiness for actions. That’s because (I believe) even people who aren’t to blame for their having morally bad attitudes are often still to blame for wrongful actions stemming from those bad attitudes. But then we can ground grievance attitudes in the blameworthy action. What’s an example where we can’t find blameworthy actions that stem from the attitudes, but can ground warranted grievance attitudes in people’s objectionable attitudes themselves? I mention one special kind of case: people might let their bad but blameless attitudes be known, but not in blameworthy ways. I wonder if there are others.

    In any case, if there are, then this works for me, at least for the sake of argument. By that I mean, I can allow attitudinal culprits in that way and then still argue as I do that there are still target cases that don’t seem to depend on legacy, or superintendence, or even attitudinal “culprits.” (Those hang together, we might say, as being broadly agentive rather than structural sites of grievance.) The example I’ve mostly been using here is some kinds of class hierarchy. That is, I doubt that even adding attitudinal “culprits” helps a great deal, even if some, with the reach problem.

27. Alasia, thanks for following up. This is very helpful in allowing me to engage with you a little more about the framing around our supposedly having “different starting points.” I go on at some length here, since I never see the nest of issues here clearly pulled apart and looked at very carefully (not holding you responsible for that!). I’ll make a try, no doubt not fully successful.

    In part 1 of this post you say,

    “The point [of structural injustice theorists] is that what is unjust about structural injustice cannot (and should not) be reduced to wrongful individual actions.”

    One thing you and Sally seem to think, if I’m understanding, is that I must not realize this. On the contrary, I can assure you, I’m well aware that this is a common view, I understand the view as stated, and that exact view is very much the motivation for this paper. That’s the first thing: I am not somehow missing the fact that this is a view that is widely held by structural injustice theorists. And, I want to raise questions for it.

    Second, you and Sally suggest, I think, that in challenging that very view, I display a difference in basic “starting points.” But that doesn’t seem correct. If that view about justice and wrong is indeed a starting point for some structural injustice theorists—which I take to mean that they don’t propose to go deeper and explain and justify it (or, if I’ve got that wrong, what do you mean by a difference in basic starting points?), it’s nevertheless not the case that I simply start somewhere else. If I did, then what I say about it could be easily dismissed as just rehearsing my own unsupported assumptions. Instead, though, I argue that the view gives rise to some hard questions.

    That said, it’s still possible that you or others don’t recognize them as hard questions at all (this may be what Sally has indicated). Then maybe, after all, they just reflect even deeper unstated basic starting points. But as I read your first post, that doesn’t seem to be what’s going on in your case. For example, you write,

    “Estlund concludes by asking ‘[I}f there is no culprit, as well as no warrant for grievance attitudes, what does it mean to claim that it is wrong at all rather than only bad?’ (p. 342). I believe that many theorists of structural injustice would reply that the difference is that *conceiving of structural injustice as wrong means that we have a responsibility to address it,* rather than simply lamenting its misfortune.” (emphasis added)

    You and I are not coming from different intellectual or normative or metaphysical universes here at all, I think. You suggest, and I agree, that for some conditions that might not in themselves reflect any wrongdoing—and as you say, the case of unavailable women’s health care might sometimes be one—there is nevertheless “wrong,” (as we both call it) insofar as many people have a “responsibility” to address it, but do not do so.

    Absolutely. That is one deontic kind of wrong–wrongdoing by an agent—often committed, as in this case, by many agents. (Young’s ambiguous use of language of “responsibility” is famously unclear on this, but, as you probably know, many have made this point in response to her.) Bringing it in as you do (and as I also do, for this very reason, as part of the duties of “superintendence”: i.e., to help prevent, remedy, etc) does indeed help with the reach problem. So, we both recognize the points here and how yours is a good response to my “reach” challenge (and I treat it as a good response by saying so in the paper).

    But it might yet not settle the matter. I argue that there are further important points as well, ones I expect that you and others to recognize: that is, letting that responsibility point entirely stand, it’s still far from clear, I argue, that even by adding such failures of our responsibilities to remedy such things, this will give the account the reach I and you (I conjecture) would want it to have. Some cases we may want to reach don’t seem to depend on that kind of wrongdoing (that failure of our responsibilities) either. You know the story by now, but quickly: people might not know what would work, and/or they might meet those responsibilities fully and yet not succeed. It’s not as if the problems are easy in either of those ways.

    Now I don’t think that’s plausible about structural racism in the U.S., since I don’t see how our racialized structural pathologies could have occurred without brutal individual racism at least in the past. But are we so sure it’s also not plausible about any kind of class hierarchy that strike us (you choose what strikes you…) as structural injustice? Are we sure that we can find the hierarchy to be wrong, and explain that wrong in terms of people not living up to their “responsibility to address it?” (You indicate otherwise, as I’ll come back to in a moment.) Or doesn’t it seem perfectly possible that some class hierarchy might be more intractable than that, impervious even to full satisfaction of our responsibilities, past and present, were we somehow to manage that? If that’s possible, then there’s a clear question: is that still any kind of wrong? We might disagree about the answers, but it seems that we both know what we’re discussing there, and are not somehow just rehearsing our incompatible basic starting points.

    Now, you also write,

    “I don’t think that the nature of the problem [in Italy] can be fully illuminated through the lenses of citizens’ failure to fulfil duties of superintendence. In a sense, I think that approaching the problem in this way would miss the workings of a complex net of structures (e.g., patriarchy and the relationship between respectability in the medical profession and acting as a ‘good’ Catholic) that must be changed to improve women’s condition.”

    That’s fair as long as you understand that I didn’t claim that “the nature of the problem can be fully illuminated through,” my point, our point, about what I call superintendence. Of course not. I was just discussing the point you did make: that there seems to be a wrong in people’s failure in their responsibilities to address the problem. No doubt there is also much more to it.

    Finally, you write that,

    “this comes down again to your different starting point from structural injustice theorists. I read structural injustice theorists as being deeply political in their goals – the aim is to change unjust structures, not to find culprits.”

    This point about the work being “deeply political,” too, is often said, so I want to report that it is usually very unclear to me how it is meant to be relevant in the exchanges it appears in. It doesn’t seem to add anything to the arguments and response the theorist offers about injustice and so on.

    Now, here’s one way that response could be relevant. Some structural injustice theorist might say this:

    “I’m not interested in your questions about whether it is “wrong,” or “unjust,” or whether there are culprits (not interested in “finding culprits”), or whether grievance attitudes are warranted. I’m just not interested replying, or in addressing any of those. All I’m interested in addressing is how to change the structures that seem to me to need changing.”

    Fair enough. Who can object to someone simply being interested in a different thing? (That point, if it’s right, should presumably function reciprocally, though.)

    But, I don’t see anyone saying this. I’m generalizing but see if you agree: Even if it sometime sounds like that’s what they say, that is typically belied by their explicitly offering theories, arguments, and objections to same, about what is unjust, wrong, whether it depends on culprits, and so on (as do Sally, Young, you and most others we’re calling structural injustice theorists). It’s perfectly possible to have strong political motives for one’s work and also be led to offer and critique views about all those things—to have an interest in them after all, perhaps because by helping to get clear about them one might think philosophy can be in service to political progress, which may one’s overarching goal. But for whatever reason, structural injustice theorists do show an interest in all those things mentioned in my imaginary disavowal, so few if any really seem to take that view.

28. Sorry, Alasia, I might have misinterpreted you in that reply. Thinking back over it, I see that it’s not clear whether you meant, as I thought, the wrong is in the failure of individual responsibility to address the problem, or whether it simply lies in the situation’s needing addressing by us. I spoke to the first way of reading you. As for the second, something’s needing addressing seems often to be a way of showing that it’s bad, but why wrong? If a hurricane devastates a city, we have a responsibility to address that damage. That wouldn’t be much of an argument that the devastation is wrong. It’s not clear to me how it would be a better argument that some problem which we have a responsibility to address is wrong because, instead of a hurricane causing it a combination of social factors caused it but without any wrongful or blameworthy acts, omissions, or attitudes. Grant that it’s bad. What would be meant by adding that it’s wrong in that case? (On the other reading of your remarks, I granted that there was a genuine wrong in the failure of a responsibility to address it.)

29. Before the comments shut down shortly, let me just thank you all for such a frank and constructive discussion about structural injustice. We produced nearly 20k words (probably too many of them mine) that I’m glad I’ll have the chance to look back at and think more about, especially the honest hard-working posts from those of you who see these things very differently from me.

Comments are closed.