Can one justifiably believe that, on the one hand, agent-regret is a warranted response to certain situations, but, on the other hand, survivor’s guilt isn’t? I do not think so. I think survivor’s guilt and agent-regret stand or fall together. Should they stand, or should they fall? I think they should fall. Below I outline why I think that someone who goes in for agent-regret must also go in for survivor’s guilt. Then I give some reasons to be wary of both. Because space is limited, I cannot fully defend all (or, really, any) of my claims here. But, I hope the picture I paint is at least coherent, and perhaps attractive.
First, I am going to introduce an unfortunately confusing terminological modification. I am going to rename ‘survivor’s guilt’ to ‘survivor’s regret.’ Survivor’s regret (formerly survivor’s guilt) is, as Thaddeus Metz puts it, “feeling bad about oneself for one’s associates having died, for not having died along with them, or for not having saved them, even though one did no culpable wrong in contributing to their deaths.” It is “the feeling of guilt (or something like it) despite not being guilty” (Metz 2018). I use this new label ‘survivor’s regret’ so that we can use the old label ‘survivor’s guilt’ to refer to the similar feeling of guilt when one is in fact guilty. So for instance someone might feel survivor’s regret if they were the only survivor of a shipwreck which was not their fault, and they might feel survivor’s guilt if they were the only survivor of a shipwreck they caused.
With this new terminology, survivor’s regret lines up to some extent with agent-regret, which is a negative feeling towards a situation one was causally involved in but for which one was not morally responsible, and survivor’s guilt lines up with guilt, which is a negative feeling towards a situation for which one was morally responsible.
Survivor’s guilt and guilt quite clearly stand or fall together: survivor’s guilt is merely a subset of guilt. Do survivor’s regret and agent-regret stand or fall together? Perhaps the former is a subset of the latter. Nancy Sherman, although she does not like the term ‘agent-regret,’ more or less thinks that survivor’s regret is a subset of agent-regret, just like survivor’s guilt is a subset of guilt (Sherman 2011). Whether or not survivor’s regret is a subset of agent-regret, they share similarities. Both require us to accept moral luck in order to think that they are appropriate. Both have nearby concepts (survivor’s guilt and guilt) which imply moral responsibility, whereas survivor’s regret and agent-regret rule out moral responsibility. Both are widespread emotional responses to traumatic situations. I think there are no arguments which show that agent-regret is reasonable but that survivor’s regret is not: either these arguments fail to successfully defend agent-regret, or they defend it at the cost of defending survivor’s regret too. To take one example, Jordan MacKenzie argues that we need agent-regret so that we can have a ‘bad guy’ to blame in tragic cases (2017, 109-110). But cases of potential survivor’s regret are just as tragic as any cases calling for agent-regret, and hence just as in need of a ‘bad guy.’
If you object that branding a survivor a ‘bad guy’ is perverse, then you are primed to agree with the second half of my argument, which is that agent-regret and survivor’s regret ought to be discarded. Any issues one might have with survivor’s regret are equally issues one might have with agent-regret. Like survivor’s regret, agent-regret results in someone who did nothing wrong feeling quite bad about themselves. Like survivor’s regret, agent-regret piles these bad feelings on someone who is already in an unfortunate situation. Like survivor’s regret, agent-regret sits ill with the approaches to responsibility taken by classic theories like Kantian ethics (which reserves opprobrium for someone whose will had some defect) or consequentialism (which recommends negative feelings only insofar as these have good consequences). And because they stand or fall together, if we want to get rid of survivor’s regret, we also have to get rid of agent-regret.
Moreover, it looks like we want to get rid of survivor’s regret. One key difference between agent-regret and survivor’s regret is that we tend to have negative feelings towards someone who manages to shrug off agent-regret without much trouble, whereas we tend to encourage someone’s efforts to shrug off survivor’s regret. This demonstrates that far from being deeply embedded in the necessary normative structure of the universe, both agent-regret and survivor’s regret are society-contingent ways of reacting to the world. Whether someone is inclined to feel these ways, and what society does in response to these sorts of feelings, are up to us. We could have a society that mandates survivor’s regret as much as we currently mandate agent-regret, but we could also go in the other direction. We could have a society that tries to help people get over agent-regret, or simply not feel it in the first place, just like we might try to get someone not to feel survivor’s regret. Many sophisticated recent defenses of agent-regret, like Jordan MacKenzie’s or David Sussman’s (2018), already accept that it is constituted by societal norms which in principle could be foregone. I am suggesting we forego these norms. Just like we try to help people get over survivor’s regret as soon as possible, and we would feel happy for someone who managed to avoid it altogether, we should work towards a society which aims to help people get over agent-regret as soon as possible, and which feels happy for someone who manages to avoid it altogether.
*****
[Thanks to Justin D’Arms and Oded Na’aman for discussion of this topic.]****
MacKenzie, Jordan. 2017. “Agent-Regret and the Social Practice of Moral Luck.” Res Philosophica 94 (1): 95-117.
Metz, Thaddeus. 2018. “Survivor’s Guilt.” In Hugh LaFollette (ed.) The International Encyclopedia of Ethics. John Wiley & Sons Ltd.
Sherman, Nancy. 2011. “The Moral Logic of Survivor Guilt.” Psychology Today. <https://www.psychologytoday.com/intl/blog/stoic-warrior/201107/the-moral-logic- survivor-guilt>
Sussman, David. 2018. “Is Agent-Regret Rational?” Ethics 128 (4): 788-808.
Danny, tanks for this post, it’s really interesting! I have lots to say, but I’ll try to keep it brief and focus on one point about agency. I think you’ve made a pretty compelling case that agent-regret and survivor guilt are linked, and that we might want to get rid of them both… but I think a notion of agential responsibility (rather than moral) might cause problems for your position and I’d like to hear what you think!
First, a minor point: I don’t think we need to accept moral luck to accept *either* agent-regret or survivor regret. Take the lorry driver case. Williams says “We feel sorry for the driver, but that sentiment co-exists with, indeed presupposes, that there is something special about his relation to this happening, something which cannot merely be eliminated by the consideration that it was not his fault.” (Moral luck, p 28). We can accept this without thinking that it says anything particularly interesting about the *moral* involved. I like to think of it in terms of the agential, instead.
We can say that the agent is responsible without being morally responsible. Agent-regret might imply that one was *responsible* for something whilst not being morally responsible. The driver *killed* the child–and for agent-regret we always need some sort of action or agency due to which the agent is responsible for something traumatic. Does survivor regret imply that? I don’t think so: I regret that I got lucky, and part of the issue is that I did nothing to deserve it. So maybe they don’t stand or fall together: One might think that responsibility is important and relevant in traumatic situations are important but the desert of surviror regret is misplaced .
And I’m not sure that agent-regret *should* fall. I think agential responsibility is important–Tony Honoré thinks (and I agree) that what we are responsible for is important to our identities. Something like this seems right: the driver, despite not being at fault, is a killer. What we actually do–beyond our intentions–is important to who we are. I’m not sure that MacKenzie or Sussman are right that we could or should forgo agent-regret because I think this sort of responsibiltiy is important no matter what we think about it.
Super interesting topic — thanks, Danny.
I wonder what you have in mind when you say that survivor’s regret and agent-regret ought to be “discarded”. Is that just another way of saying we should believe that they are never fitting responses to any situation?
I wonder, too, about whether they really are parallel. Perhaps there is one variety of survivor’s regret that involves feeling guilt about something one was not guilty of. But isn’t there a variety of survivor’s regret that is something more like a bad feeling about having unfairly received a reward? I’m alive, but it’s not like I’m any more deserving of having lived than my friends who died, so I feel like I’ve been given a gift under false pretenses, and I want to give the gift back (so to speak). I guess that doesn’t strike me as unfitting at all. It seems like the response of someone who is well-attuned to the moral residue of the tragedy.
Thanks for your comment, Jake! There’s a lot there which I can’t cover without writing too much, so here is just one sort of reply (I have many more if you’re interested!):
You suggest that agent-regret is important for acknowledging our identities. I haven’t read Honoré on this, so I’ll have to look into it. But my initial response is that my identity as a “killer” might not be very interesting. Certainly it might not be important enough for me to feel bad about it. Maybe I’m a killer – so what? I think I need a further argument to think that my acknowledgement ought to take the form of feeling quite bad about this.
Now, “killer” is a loaded term. We usually use it for people who maliciously or perhaps negligently kill. And of course in those circumstances the identity is very interesting and I ought to feel bad about it. But agent-regret cases are special. After all, I haven’t done anything wrong. I’m not even sure I’m comfortable with this sort of thick language to describe who I am! But even granting the language, I don’t see how this automatically means I ought to feel a certain way about it (namely, agent-regretful).
(The same, I might add, applies to survivors. I may be a sole survivor – so what?)
That’s not to say that no further argument is waiting in the wings. I think (for instance) MacKenzie and Sussman both have a good go at trying to explain why we ought to feel agent-regret. And there are other arguments too. But I have lots of doubts, and most relevantly to this blog post, I’m not sure there’s a great argument out there which wouldn’t also try to get us to feel survivor’s regret.
Thanks Neal! Yes, my talk of “discarding” agent-regret and survivor’s regret is another way of saying these are never fitting. Or, alternatively, it’s another way of saying that, insofar as they are fitting, it’s because we’ve made them fitting, and it’s in our power to change this, and we ought to change things so that they are no longer fitting. (For some this will be a distinction without a difference. For others it won’t. I want to be maximally inclusive.)
You’re right that the kind of survivor’s regret I focused on is not the only thing involved in survivor’s regret broadly speaking. So one route I could go is just to narrow my focus to the kind of survivor’s regret that is most relevant – feeling guilty, rather than feeling like an unfair recipient of a boon. The former’s bad, the latter’s maybe fine.
Another route (which is the one I prefer, actually) is to say that of course from a disinterested moral perspective, one ought to be unhappy about the unfair situation. But, third parties can make that exact same evaluation. The suspect thing is the attached first-personal baggage which the survivor feels, over and above their judgment that this was a crummy, unfair outcome. I think all of that extra stuff ought to be thrown overboard (and I think a therapist would rightly try to help someone accomplish this). The survivor should feel as bad as third parties feel. How bad should that be? Well, I think probably not incredibly bad, otherwise we’re all going to feel very bad all the time. Probably a little bad, otherwise one risks not feeling bad about anything. But, certainly not survivor’s regret bad.
I think both routes are convincing, but obviously the second is more controversial!
Hi Danny,
Interesting post!
I question this: “Like survivor’s regret, agent-regret results in someone who did nothing wrong feeling quite bad about themselves. ”
You might think that the wrongness of killing an innocent person is independent of one’s motives (/quality of will), such that one can blamelessly do something wrong (children seem to do this all the time!). For, one way to think about blameworthiness is as wrongdoing performed without justification or excuse (perhaps plus some other conditions). On this way of thinking about the relation between wrongdoing and blameworthiness, the lorry driver does do something wrong (killing an innocent person), while the survivor does not, since *not being killed* (or surviving one’s friends) is neither an action nor wrong.
Here’s one normative consideration we can point to to substantiate the idea that the lorry driver does something wrong: he *harms* the innocent person. That the agent of agent-regret causes harm is intuitively an essential component of the thought constitutive of agent-regret (and grounds for motivation to personally make amends). By contrast, the person who feels survivor guilt/regret need not, and is not ordinarily understood to, have caused anyone harm.
[This point about harm is separable from the above point about wrongdoing; even if we deny that one can blamelessly wrong another, one can surely blamelessly cause harm, and the survivor doesn’t do this.]
Hi Dan. Thanks for the reply!
First, you suggest that the lorry driver does something wrong, whereas the survivor does not, both because surviving is not an action, and because it’s not wrong. I think surviving often is an action, like when the survivor clings to a piece of driftwood to keep from drowning, and I think I could feel agent-regret for an omission, depending on the circumstances. So I don’t think the action/non-action distinction will help.
As for doing something wrong, I’m happy to say the driver does something bad (or, at least, they do something with bad results), but not that they do something wrong. It’s bad to cause someone harm, but is it wrong? It’s certainly wrong to intentionally harm (at least in most cases), but is it wrong to harm tout court? If we have a moralized account of harm, then maybe, but that begs the question. If we have a non-moralized account of harm, then there are certainly harms that aren’t wrongful (like if I open a better restaurant than yours, which harms your business). So now we just have to have an account that separates the wrongful harms from the non-wrongful harms. Enter, say, Kantian ethics. But something like that is famously incompatible with agent-regret. What we need is an account of wrongful harming that doesn’t just beg the question by assuming the driver wrongfully harms.
Hi Danny,
Thanks for your response.
I agree that agent-regret can be felt for omissions, e.g. Tom’s not retrieving his child, a fledgling swimmer, from the backyard swimming pool, leading to the child’s death. So, I shouldn’t have suggested that agent-regret’s fittingness is tied to a contrast between action and omission. Suppose the above omission is not traceable to some culpable act/omission of Tim’s. Nonetheless, there is an explanatory connection between Tim’s omission and the harm that, to my mind, renders fitting his agent-regret, but not the survivor’s. (His omission explains, or is an integral part of the explanation of, the death.) Take, for contrast, the survivor who survives simply because he *is not killed* by his captors, while his friends are. His ‘not being killed (along with his friends)’ does not explain why they are no longer alive. We can imagine a different case in which his ‘not being killed’ *does* explain why they are no longer alive, e.g. the captors decided to arbitrarily spare one and only one of the friends. In this case, he *does* take their place, as it were, and so, his surviving is an integral part of the explanation of their deaths. Perhaps this kind of thought is assumed by survivors in feeling survivor’s guilt (Primo Levi’s description of survivor’s guilt suggest this), but in that case, we can identify why some such cases of survivor guilt/regret are unfitting, namely because they involve the incorrect thought that one’s survival explains others’/another’s death.
Thanks Dan! Starting with the last point: it sounds like you think survivor’s regret would be appropriate for the person who gets arbitrarily spared, but not for the person who merely isn’t killed? I think our intuitions come apart here: if I were a therapist and both of these people felt an equal amount of survivor’s regret, I think I’d rightfully try to get each of them to feel less or no survivor’s regret, for exactly the same sorts of reasons. I’d be interested in how many people feel the way you do about this case, vs. how many feel the way I do. It would be a problem for me, I think, if people tend to lean in your direction.
As for the point about the explanatory connection between the act or omission and the harm: a lot of work gets done by saying there’s an “explanatory connection” that “renders fitting” the agent-regret. I want to hear more about what makes the explanatory connection normatively interesting. There’s also an explanatory connection between someone founding a city and the lorry driver killing the kid in that city centuries later, but I don’t think that’s supposed to occasion agent-regret. But in order to limit things so that only some explanatory connections are relevant, we have to wheel in considerations which will work just as well for survivor’s regret (or so I’ve claimed). Maybe there are accounts of the “explanatory connection” which don’t entail this, but that’s what I’m denying. Of course to substantiate this fully we’d have to go through each potential account, etc.
Hi Danny,
Interesting Post! I wonder what you think about this attempt to justify agent regret but not survivor regret:
(1) Agent regret is warranted in the wake of a moral dilemma because one has performed a pro tanto wrong (or wronging) action that was permissible because there was no available action that was not pro tanto wrong(ing).
(2) The ground for agent regret in the wake of moral dilemmas is not present in standard survivor regret cases.
(3) More strongly – in standard survivor regret cases, agents do not perform pro tanto wrong(ing) actions, so there is no warrant for survivor regret.
(4) There is a defense of agent regret that does not commit one to defending survivor regret.
Danny,
Thanks for your reply. I see what you’re saying about the identity stuff… but I’m also doing a blog for the agency and responsibility month, and I think I’ll just end up repeating that here if I try to say more.
Here, I just want to weigh in on your response to Dan and see if this helps Dan’s point. You (Danny) suggest that any explanatory connection between a traumatic event and a person will be similar in cases of agent-regret and survivor regret (unless I’ve read you wrongly?). But I don’t think that’s true: the person who feels agent-regret is responsible, whether through action or omission, for the traumatic event. There is an explanatory connection between, say, the lorry driver’s mother giving birth to the driver and the child being ran over by the child three decades after. But that’s not an interesting explanatory connection insofar as we don’t want to say she *caused* the child’s death or was *responsible* for it. The driver, however, did cause it, was responsible for it. I think this can be absent, too, in cases of survivor regret. Perhaps there were other planks of wood and I survived because I grabbed a plank but you couldn’t reach any (we weren’t competing). I might feel survivor regret for *surviving*, and maybe I survived because of my actions, but my actions need not explain why you died.
But even if there is an explanatory connection between my survival and your death, there are further grounds for a difference between a case of survivor regret and agent-regret: I won’t be able to feel agent-regret if your death is not down to my *agency*. The driver is responsible for the child’s death because of what he has done. Perhaps one might feel agent-regret in an arbitrary killing case if one escapes because one, say, draws the long straw. But if one is spared becuase of, say, a feature one has (light coloured eyes, maybe), one’s survival isn’t predicated upon any of one’s actions. Thus no room for *agent* regret. But perhaps there is room for survivor regret…
Danny,
Thanks for this fantastic piece. It’s had me thinking about survivor’s guilt (or rather, survivor’s regret) all day. I agree that survivor’s regret shares a few deep commonalities with agent-regret—they both are highly moralized responses to bad outcomes towards which we are not morally responsible. I’m not sure that it follows from these commonalities that our justifications for the former emotion will have to be the same as our justifications for the latter. One salient difference that has already been pointed out in the comments is that only agent-regret involves agency—I still did something that lead to a morally bad outcome in cases of agent-regret, even if I did so in a way that doesn’t involve standard moral responsibility. This might make a difference to the sorts of roles that I can step into in the aftermath of morally unlucky tragedies. But I think this sort of worry has been touched on by other commentators.
I actually just wanted to chime in with a quick defense of survivor’s guilt—it’s not at all clear that this is an emotion that we should want to throw away. I agree that it can go out of bounds (agent-regret can also become overwrought), but there seems to be something admirable about a person who responds to their comparative good fortune with some recognition that it’s simply a matter of moral luck that they didn’t end up in the same position as the unlucky ones. The third-party reactions in the two cases are actually also not as dissimilar as you make them out to be. You note that we’ll try to talk people out of feeling survivor’s guilt—but we actually do that for agent-regret as well (imagine the lorry driver’s friends saying ‘You didn’t do anything wrong—you couldn’t have avoided the child. You have nothing to feel guilty about.’). Even though we try to talk people out of feeling agent-regret, we’d typically be unnerved if they weren’t feeling it. There would be something off about the lorry driver if he didn’t feel agent-regret–but that won’t stop us from trying to talk him out of it once he’s feeling it. And I think the same story is true in the survivor’s regret case—I’ll try to talk my friend out of feeling survival regret after she finds herself the sole survivor of a car crash. But I’d also find it unnerving if she did not feel some degree of anguish about her good fortune.
Anyways–thanks again. This has given me a lot to think about.
Thanks for the comment Brad! My preferred analysis of moral dilemmas is either that guilt (rather than agent-regret) is called for; or, one ought not to feel bad about having chosen the correct horn of the dilemma, or at least one ought not to feel worse than a third party would feel about the situation. So, if either of those responses works, then I’m good to go.
Thanks Jake and Jordan! Both of you are pointing towards agency as one key difference between agent-regret and survivor’s regret. In fact before I wrote this blog post I almost wrote a different one, tentatively titled “where is the agent in agent-regret?” The short version is that I think it’s very hard to actually get any sort of normatively interesting agency into the agent-regret story. Either you go for a full-blooded Kantian notion of agency at the cost of ruling out agent-regret, or you go for a thinner notion of agency at the cost of losing any reason for us to care about what our agency is involved in. This is a large topic all on its own which needs its own defense, but you can hopefully see how, if it works, it would answer these questions.
Jordan, about the appropriateness of survivor’s regret: that’s interesting that you think we’d (rightfully?) look askance on someone who didn’t feel any survivor’s regret. I do think we’d be unnerved if someone shows no sadness or other negative feelings at all with respect to a disaster they are intimately acquainted with. I also think this often manifests in survivor’s regret. But as long as I’m sure that someone is feeling appropriately bad about the disaster, I don’t think I’d be particularly unnerved to learn that they’d dodged survivor’s regret. (I might be misremembering, so I don’t want to slander him, but I think Justin D’Arms suggested that lots of people don’t feel survivor’s regret, and that it’s a good thing that they don’t.) So (similar to my response to Dan above) my intuitions here might just be weird. I do think we need to be very careful thinking about this stuff though, because it’s easy to mistake the “people should feel bad” intuition for “people should feel survivor’s regret in particular” intuition.
Hi Danny,
I can see that if one grants you those normative assumptions then you are good to go; but those are contentious assumptions that conflict with the ideas in the argument i gave so I am curious how you would defend them.
Maybe an example will help me understand how you justify your approach. Take a Sophie’s choice case. If I had to leave one of my kids to die in the desert and could only carry one out I would certainly save one rather than sit down and let us all die. In leaving one to die I don’t think I’m culpable for wrongdoing but I do something horrible (vicious some virtue ethicists would say), which would in other circumstances be gravely wrong. This in not a case of culpable wrongdoing because there was no available option that was less morally bad (or vicious). Your approach would of course then say I shouldn’t feel guilty. I accept that because I’m not culpable.
So now to avoid the general argument you would say I should not feel worse about leaving my kid to die in the desert “than a third party would about the situation”. Can you explain why? Intuitively I think I should and would feel worse. Moreover I have explained how one might justify this intuition, namely that I have done something pro tanto wrong (or vicious) but the third party has not. Why then should this third party feel just as bad as me?
Thanks again Brad! Yes, those are both very contentious assumptions which would each need a separate defense. That project (or rather those projects, since they’re two separate projects) are much bigger than anything I can say here. (For instance, I would defend the “guilt is called for” response by ditching ought implies can.) So, just two thoughts:
First, I think we have to be very careful about examples and the intuitions they call for. Of course you’d feel worse than a third party for leaving your kid to die in the desert, but there are lots of confounding variables here. It’s your kid, not the third party’s kid. In normal cases, you’re culpable for the kid dying, because you didn’t plan your desert trip well enough. (Of course, this is false by stipulation in your example, but our intuitions are trained on cases where it’s not false, and anyways we can never know for sure in real life whether it is in fact false.) Many of us (all of us?) have grown up thinking that agent-regret is what we ought to feel, and so it’s natural for us to feel agent-regret. And so on.
Now, it’s one thing to say that we’re ignoring these variables, but it’s another thing to in fact ignore them. Weird hypothetical cases like this are tougher to think about than we give them credit for, I think. If we look at actual cases of people who feel bad for things we think they ought not to feel bad about (like people who feel bad about being queer, or for menstruating) then I think we can start to get into the mindset I’m suggesting we adopt, which is one according to which we take feelings which seem natural to us and then realize that maybe those natural-seeming feelings ought to be done away with. It seems natural to feel bad about leaving one’s child in the desert (above and beyond feeling bad about losing a child!) but what seems natural isn’t always what’s called for (or what ought to be called for).
Second, that something is pro tanto wrong is not to say it’s all things considered wrong. After all, that’s what the pro tanto label is for. Bad feelings for doing something all things considered wrong are an easy sell. That’s guilt, after all. Bad feelings for pro tanto wrongs are not. Often those would be entirely inappropriate. It’s a pro tanto wrong to slice someone open, but surgeons shouldn’t feel bad at the end of the day. So we need to be more sophisticated about what calls for bad feelings beyond just pointing out that a pro tanto wrong has been committed. The challenge, I think, is coming up with something more sophisticated to say which won’t also rope in survivor’s regret somehow.
Hi Danny,
In reply to your claim, “if I were a therapist and both of these people felt an equal amount of survivor’s regret, I think I’d rightfully try to get each of them to feel less or no survivor’s regret, for exactly the same sorts of reasons”, I don’t see how this could be the case, at least if you were cognitive behavioral therapist, for a key difference between the two survivors is that only one’s pained attitude involves a thought that is true, (the thought being ‘I took his/her place in living’).
More importantly, what you’d do as a therapist– or from the perspective of well-being, generally– isn’t obviously relevant to the fittingness of the patient’s attitude, where fittingness is understood roughly as ‘correctness’. My sadness might impede my well-being or get in the way of my productivity, and while there may be prudential reasons for me to overcome the emotion, it remains fitting given that it correctly represents some event as a serious loss to me. My sadness feels bad and it is a response to nothing I did wrong (suppose it’s about the death of my elderly cat)– so I don’t ‘deserve it’–but it’s fitting nonetheless.
Thanks Dan! The “therapist” shorthand was a crummy way of putting my point – I should have talked about what I’d do with a magic wand or something, for the reasons you note and for other reasons. For instance, you suggest the “perspective of well-being, generally” and that is the sort of thing I ought to have said. But, well-being is not quite the right perspective, I think. Some good replacement perspectives are the perspective of what would be good, simpliciter; or, the perspective of what a disinterested agent has most reason to promote.
However, the perspective I’m most interested is different. I’m interested in what we, collectively, as a society (say, of human beings, although we can go smaller if need be) should aim to achieve. Should we set up our society such that, typically, arbitrarily spared people feel better or worse than merely surviving people? I’m urging “no.” Insofar as our society is not presently like this, we should do our best to change it.
If change isn’t possible, my proposal is misguided, but I think it is, so let’s set that aside. If instead one suggests that change is ill-advised because a difference in feeling correctly represents the fact that the two people are differently situated with respect to what matters, my claim is that actually they are not: whatever differences exist between these two are not differences that matter to how we ought to feel. It might be true that one’s survival explains the deaths of others in one case, but not in another. But is that difference relevant to how we ought to feel? Not all differences are relevant like this. If one survives because one was seated at the back of the plane rather than the front, this shouldn’t influence how one feels compared to someone who survives because they were seated at the front rather than the back. That’s a normatively inert difference.
The question, then, is whether the difference in the two survivor cases that you highlight is a normatively inert difference, or whether it’s normatively ert. And ditto for agent-regret cases. Is it normatively interesting that I happened to be the unlucky driver that killed the child, as opposed to the drivers of the truck just ahead of or just behind me? I think people are inclined to say “yes,” and so I have to do a lot of work defusing that sort of intuition, but I think (with zero evidence!!!) that I’m on firmer ground with respect to the two survivor cases.
Ultimately I don’t want to rest my case just on intuitions. I think that the intuitions themselves are a result of enculturation and thus we can’t treat them as fixed points to reason with. I also think that whatever intuitions we start with, we also have to judge how coherent an overall picture we can tell, and in order to build a coherent overall picture we’ll have to dump some intuitions or another. But, at the very least, when it comes to starting points, I hope I’m not entirely alone when I say that the two survivor cases you mention strike me as ones about which those people should feel equally bad. So, if we accept that some survivors shouldn’t feel bad, then neither should feel bad.
This comment got pretty long – sorry!