Blog Coordinator

« Optimistic Skepticism and Forward Looking Blame | Main | Billboards Now Promoting Moral Responsibility Skepticism »



Feed You can follow this conversation by subscribing to the comment feed for this post.

This is a very intriguing and thoughtful post, thanks Justin. I think I agree with much of what you say here. I find it plausible that forgiveness is central to our close personal relationships, and that apology is a central component of (at least typical cases of) forgiveness. And I think I agree that it is difficult for a moral responsibility skeptic to provide an adequate account of apology. It seems to me that this is because a central component of apology is accepting responsibility for what you have done - accepting not only that you have done something wrong (that you have violated a moral obligation), but also that you deserve blame for it, and are asking to be 'released' from that blame (to put it in Nelkin's terms).

Where I think I disagree is about the suggestion that genuine apology is not tenable if we lack freedom in the sense of being able to do otherwise than we actually do. Now, if you think that freedom in the sense of the ability to do otherwise is necessary for moral responsibility or even for moral obligation, then I can see why you would arrive at that conclusion, since a lack of such freedom would preclude the acceptance that you had done wrong and deserved blame for it. But as I argued on the other thread, I think we can make sense of moral obligation and moral responsibility even if we lack that kind of freedom.

I don't want to wade too much into that debate again here. I just want to suggest that someone who holds the kind of view that I hold is not in any worse position to account for apology than someone who holds your kind of view is. We disagree about what the necessary conditions for genuine moral responsibility and moral obligation are, but unlike the moral responsibility skeptic, I think we agree that typical agents can meet those conditions - and I think that's enough to give a satisfactory account of apology (and of forgiveness as well). Would you agree with that, or do you think that there is something essential that someone who holds my sort of view is missing that is needed to account for apology?

Thanks, Ryan! And nice summary in your first para! Now a brief remark about the source of our disagreement.

You say: "Where I think I disagree is about the suggestion that genuine apology is not tenable if we lack freedom in the sense of being able to do otherwise than we actually do."

I think an apology will remain but not a very meaningful one. How can I, as a FW skeptic (or anyone who thinks determinism is true) accept responsibility for something I know I could not have prevented? So, sure, I can apologize to minorities in the U.S. because old rich white men held down their ancestors and stacked the deck against them succeeding in life. It *could be* and often is meaningful. I could apologize to the mother of Micheal Brown as well. But that apology is not what I mean by genuine apology. If you smacked me and your mom apologized for it it would seem, well, sort of generic. And in what way could one truly take accountability if they also believed it "wasn't their fault" so to speak. I couldn't have not done what I did. So, ya, I'm sorry I did it and that things turned out bad for you but....

To you last para: I think a libertarian view is necessary to make sense of a genuine apology. The agent must be the cause of their wrongdoing in a way that allows them to reflect on their action and feel as though they could have done something different in a real sort of way.

So, I'm not saying that a compatibilist or a FW skeptic suddenly loses all ability and desire to apologize given determinism. Only that the emotion that would drive such an apology would be different than what I think is needed for the apology to have deep meaning. So no deep remorse or guilt. One could acknowledge what they've done and hate that they did what they did but they couldn't regret doing what they couldn't help but do. Not after deep reflection and the belief that determinism is true...

Anyway, I think you see the worry.

If motivated by a sense of justice, such a free will skeptic may have sincere ongoing forward-looking regret that they cannot repair the harm that they have done. This seems to me to be a significant component to a sincere apology.

Thanks for the comment, David. But what sense of justice could the free will skeptic have other than purely consequentialist?

When I hear you say "motivated by justice" I can't help but think that basic-desert is creeping in.

Consider the situation in France. I cannot repair the harms that these people have felt and I feel bad for them. I don't regret it though because I could not have prevented the attacks (the 'can' here seems to be the strong 'can' as well) and if I were to regret it and apologize for it in some deep way it would seem out of place, and for good reason, I could not have prevented it. Much like the wrongdoer could not have prevent the attacks themselves if determinism, then so regret seems misplaced in there case as well.

Many skeptics admit that guilt would also be ill-founded if determinism but they try to hold onto the sincere apology and this seems problematic to me.

A consequentalist sense of fairness and justice is surely the main kind consequentalists have! And we can still have a role responsibility (qua Waller) and a desire or a duty to reverse the effects of a previous deleterious interaction with a specific individual. Even crude determinists believe that humans can see a past action as less than optimal, actually learn from the earlier failure, and alter behaviour - the associated emotions are part of this process. For the apology to be sincere, we expect that the probability of a recurrence of the behaviour to have changed - why we prize evidence of an emotional commitment.


You seem to assume throughout that regret, forgiveness, and feelings of injustice cannot be given a purely consequentialist justification or grounding--that is, on your view, when skeptics appeal to these sorts of attitudes and behaviors, "basic-desert is creeping in." I think this assumption is problematized by the literature from comparative psychology on the so-called "proto-moral" attitudes and behaviors that are often displayed by other highly complex social creatures (e.g., dogs, wolves, elephants, monkeys, and other primates). Look at de Waal's work on monkeys and responses to injustice and unfair distribution of resources. When I read his accounts and see his videos, I get the strong sense that these monkeys are experiencing injustice, that they display "regret" when they misbehave, and that post-supplication, violators are often "forgiven" without needing to be punished. I take these findings to show that all of these attitudes and behaviors are grounded in forward-looking considerations that don't depend on free will or basic-desert. These considerations are driven by game-theoretic mechanisms having to do with the balance between individual interests and group interests. Because free-riding is a problem for any complex social group, it is something that nature must solve. And I believe the kinds of attitudes and behaviors we're discussing are part of the solution. Without them, the stability of the group would be undermined. The same goes for punishment. But given how far back these attitudes and behaviors can be traced--to our pre-verbal days--it's hard to see how they could require something as complex and sophisticated as notions of free will and basic-desert. Or, so it seems to me...

p.s. I should also have said that I believe our theories of free will and basic-desert are simply post-hoc rationalizations for attitudes and behaviors we engage in (and care deeply about) for purely consequentialist reasons. It might even be said that these beliefs and theories--however post hoc--may themselves confer benefits. That is, backward looking theories like the ones we have may themselves be grounded in forward looking considerations! :)

Thanks for the reply, Justin. The point I wanted to make is that someone who is inclined towards semi-compatibilism - who is skeptical about free will in the sense of the ability to do otherwise than we actually do, but believes that we can nonetheless be morally responsible for our actions, fully deserve praise and blame, etc. - can give a deeper account of apology than you suggest. I wouldn't say some of the things you say. I wouldn't say an immoral action that I committed "wasn't my fault", for example - the point is I think that it *is* my fault, that I failed to live up to a moral obligation that I had, that I fully deserve blame for that, and that is what I have in mind when I apologize and ask forgiveness. That strikes me as a deep sense of regret and remorse.

I understand of course that you are skeptical that we can account for moral obligation and moral blame if determinism is true, and maybe ultimately you're right. All I wanted to suggest is just that the position a semi-compatibilist is in is different from the position of the moral responsibility skeptic; given what they reject and what I still accept, I think I have more resources to provide an account of apology.

In reply to Thomas - that's very interesting. I like those examples, but I would be inclined to draw exactly the opposite lesson from them. I would say something like 'given how far back these attitudes and behaviors can be traced--to our pre-verbal days--it's hard to see how they could require something as complex and sophisticated as a calculation of the long-term consequences of our tendencies and actions'. That is, if anything, I would think that the evidence that our inclinations towards 'regret' and 'forgiveness' (and praise and blame and love, etc.) are innate seems to fit more neatly within a Strawsonian framework than a consequentialist one (though really I would be wary of drawing any very substantial moral conclusions at all from the tendencies of animals, or naturalistic explanations of our behavior, etc.)


I am not persuaded by your appeal to the complexity and sophistication of the judgments, attitudes, and behaviors in an effort to undermine my suggestion. There are tons of very complicated cost-benefit analyses being undertaken throughout the animal kingdom. Sometimes these analyses have even been built into us via instincts, etc. So, the fact that we find complex and sophisticated attitudes and behaviors in no way undermines the claim that these attitudes and behaviors are ultimately grounded in forward looking considerations--e.g., mate selection, protection of offspring, methods of hunting, methods of evading predators, etc. What I find implausible is that these forward looking attitudes and behaviors depend upon complex metaphysical entities such as free will, the unconditional ability to do otherwise, etc. Here's another way of putting it: The innate attitudes and behaviors you mention were selected for based on the fact that they conferred a benefit to those who had them (or, they could be spandrels). But this means they are grounded in forward looking considerations rather than being grounded in some deeper, backward looking metaphysical postulates or entities.


Thanks for your reply. I don't think I was very clear, let me try to explain again what I was trying to get at and see if it makes any sense.

You say: "The innate attitudes and behaviors you mention were selected for based on the fact that they conferred a benefit to those who had them". I agree something like this might be right. That gives us a possible kind of causal explanation for where those attitudes and behaviors came from, but I don't think anything follows about the nature of those attitudes, or what grounds or justifies those attitudes or behaviors. In other words, animals may love, praise, forgive (or have proto-versions of those attitudes and emotions) because those of their ancestors who did so were more successful, but it doesn't follow that any consequentialist considerations whatsoever are built into those attitudes, or that such considerations would be relevant to justifying them for us. That's the part I find implausible.

Let me suggest an analogy (maybe this isn't very apt, but it's just a thought that occurred to me). Suppose it's true that a capacity for arithmetic is innate, and exists in the animal kingdom (from what I understand, there is some evidence that suggests this). And suppose that capacity exists because those of our animal ancestors who could add and subtract were more successful in various ways than those that couldn't. Does that mean that our basic arithmetical intuitions are 'grounded' in consequentialist considerations, or that they have anything to do with such considerations? Or that they are justified by them in any way? That doesn't seem right to me - even though some sort of consequentialist story may be part of the explanation of where they came from.

I do think I agree with you though that it is also implausible that complex metaphysical views (about free will, or the unconditional ability to do otherwise, etc.) are built into attitudes like blame or forgiveness or love, or that such attitudes depend on such views. That's why I suggested that maybe all this fits more neatly within a Strawsonian framework - the idea that what justifies moral reactive attitudes is not complex calculations about future benefits, and not any complex metaphysics, but rather just the nature of the attitudes themselves and the various essential roles that they play in our (emotional, moral, interpersonal, etc.) lives.

p.s. I should add that I am still wary of drawing any strong conclusions from any details of the attitudes that exist in animals, or even that exist in us innately. We still might, at the end of the day, find good reason to conclude that maintaining the moral reactive attitudes can only be justified by consequentialist considerations, or given certain metaphysical truths, or whatever

Not surprisingly, I agree with Ryan that someone inclined toward semicompatibilism can give a more sophisticated, and, I would venture, plausible account of the conditions for apology.

I would say that when someone acts freely, i.e., exercises guidance control, she may well need to apologize for what she has done (given certain additional conditions, but not a freedom to do otherwise condition). Again, one's general views about the requirement of alternative possibilities are relevant here; I don't see anything special to apology that would make it more plausible that we need alternative possibilities.

Thanks for the response, David. but I don't know how to make sense of strictly consequentialist anything, sorry. Maybe you can help.

We don't normally think in such ways, at least I don't. I don't ask whether feeling a certain way produces the best consequences and if it does, then it is justified. I ask whether or not it might be warranted to feel X now, and even if some of those considerations are consequentialist it wouldn't follow that *only those* considerations make up the overall judgment of appropriateness or not. Given that, I think it's worthwhile to ask what is lost given that our judgments frequently include considerations beyond *just the consequences*.

To try and make sense of what I'm gesturing at, let me try and give an example. Let's say that I'm with 2 friends, Al and Lee, and Al steps on my foot with the intent to injury it because I beat him in a game of squash, it seems warranted to blame him. Though consequentialist considerations may help me to make sense of why it might be okay to do so, to make it less likely that Al does so in the future". But, as I mentioned in my example in the last thread to Tamler and Thomas, there could be ways to justify blaming Lee (instead of blaming Al who seems to be the fitting target of the blame) if in doing so it produces the effect that is best. And this seems anything like "justice", blaming someone else for doing an act that they clearly could not have prevented. So, imagine blaming Lee for what Al did *because* Al's future behavior is better trained when others, innocent friends in particular, are at fault. It seems weird. Likewise, on this forward looking account, to say that Lee OUGHT to feel regret for what Al does because doing so will produce better consequences (overall for everyone involved) seems bizarre. It seems that any account that fetishizes consequences will have super counterintuitive results, at least in practice.

Is the view (forward-looking regret or guilt) then that one ought to feel guilt when it produces best consequences? Is the view that a sincere apology exists when we feel bad about x occurring and feeling bad produces the best consequences? This seems odd to me but I'd be curious to hear more about how to make sense of it.

The point in discussing forgiveness is to try and unravel how this new world would be if one held the view that basic-desert was imperiled, as many skeptics believe. I'm confident that we would have the ability to apologize and accept apologies even if determinism were true. I'm less confident that the deep understanding of apology would remain though given that to genuinely apologize it seems to require some deep admission of guilt beyond recognize that I am causally responsible for the act I'm being asked to apologize for.

John (and Ryan), thanks for the comments.

I'm curious. Why does one need to exercise guidance control in order for an apology to be appropriate? On a strictly forward-looking view it seems that if what we care about is producing the best consequences, then there are going to be cases where no action (or control of any sort) is needed in order to justify the practice of blaming, forgiving, or apologizing.

Many philosophers (including John) have a historical condition to their view on moral responsibility but it seems that such a condition would seem out of place on a forward-looking model. If what warrants blame, apology, or forgiveness is strictly forward-looking then why care about a historical requirement, or an actual-sequence view to make sense of how I should handle you at the current moment? The historical requirement resonates with me but then I can't seem to make sense of it when we suggest what is appropriate to do in any given scenario (whether to blame, apologize, feel guilt, forgive, etc.) whatever it is that produces the best consequences. It seems that when we make such a leap we give up caring about the agent's specific role in past acts/trangressions unless such considerations themselves help with producing such consequences.

Thomas, thanks for your very challenging comment!

I'm inclined to respond in a fashion similar to Ryan.

On a side note though, you say some things that strike me as at least contentious. The appeals to the evolutionary process to ground "forward looking" attitudes seems fishy.

For instance consider two traits that are competing to be passed on. One trait (trait A) is better fit for the job as it causes the animal in question less harm, or plug in your best reason. (trait B) seems to be all but vanished as of time T. Then, all of a sudden, a meteor hits and wipes out all of those who had trait A because those who have trait A were closest to the meteor. The way you seem to be spelling out your story is that trait B then won out because of forward-looking "reasons". But, due to drift and other related issues that arise in evolutionary theory, I don't know if that claim is warranted. The reason that B won out is because of luck, not because it was better than A or because our species had better reason to adopt A...

Anyway, I'm no philosopher of biology, or science for that matter, but I have enough background (I think) to claim that how the evolutionary process unfolds is very much an open question. And, until we can figure out where the level of selection is taking place for the traits we're talking about here, I'm skeptical that an appeal to the evolutionary process itself is going to be the right approach to ground these forward-looking practices as you seem to be suggesting, but it's possible I have missed the point if not.


I do not defend, nor am I inclined toward, a purely forward-looking view of the relevant range of phenomena (including blame, responsibility, and apology). On my view, there is a freedom condition--just not a freedom-to-do-otherwise condition. Despite my considerable respect for optimistic responsibility skepticism, I'm not a responsibility-skeptic.

Hi Justin, glad you've written about forgiveness. In his book on apology, Nick Smith identifies what he calls the “categorical apology” as the “benchmark of apologetic meaning” (2008, p. 142). Categorical apologies possess three key features: (1) they identify the harm that was done to the victim with some precision and detail; (2) they express the fact that the wrongdoer accepts blame and explicitly acknowledges the moral principles that were violated; and (3) they issue from sincere regret and affirm the fact that the victim did not deserve to be so treated.

Do you think the semicompatibilist could consistently endorse a view like this?

Hi Brandon, I was hoping you would chime in on the discussion.

I like that view, and thanks for the reference, I haven't read the piece nor have I come across Smith's view.

I think (1) could easily be adopted by semicompatibilists and skeptics alike. I think (2) and (3) would be harder to accept though (especially for the skeptic who would seem to be forced to reject 3 (and 2 if they endorse OIC)).

Re: (2) - I'd have a hard time accepting blame for something you did, because I didn't settle on the decision to do it in the right sort of way. Likewise, if it turned out that I couldn't prevent what I did from occurring (and I was the cause of it) because determinism, then I'd have a hard time accepting the blame for much the same reason. If it was already determined to be done because of the past and the laws, then I have a hard time understanding why I should accept that I ought not have done something that I couldn't but do. So, a semi-compatibilist who also endorses OIC would have a tough time to accept (2).

Re (3): - I don't see how one could regret something, in a deep way, if one recoginzed it wasn't *up to them* to prevent. It's like having regret that you were born, it wasn't up to you therefore it seems weird to regret it, at least in a deep sort of way. You could wish it never happened but that's different than regret, or so it seems.

Also, it would seem that (3) would have to be rejected by MR skeptics, who have been the source of much of the project I'm working on. So admittedly, I haven't given as much thought as to whether or not the semi-compatibilist could endorse certain concepts, though, many have been pressing on just that connection, which I am very thankful for.

So, I guess the semi-compatibilist could endorse the view if they rejected OIC (for considerations of 2) and revised traditional concepts like "sincere regret" and desert to accept (3).

Hi Justin. Thanks for your thoughts! This could spin off indefinitely given that you are now bringing up various traditional objections to consequentialism, not to mention the genetic bases of reactive attitudes (seen in other social animals as per Thomas) that may give rise to post-hoc desert based justification. Briefly, things in a society may go better if members have internalised a rule to clean up their own messes first, and that others around them are aware of this fact. Your example of Al and Lee doesn't work too well if there is mutual transparency (Al is going to pick on Lee as well to maximize total utility?). Anyway, I will just reiterate the perception I have that contrition must have some forward-looking component of "will do better" to be taken seriously - surely the idea of forgiveness is precisely the realization that the past act is immutable, and we have better things to concentrate on.

I certainly see no reason why a semicompatibilist could not accept all three components of Smith's view.

I think an agent could deeply regret something, at least in part in virtue of believing that she did the thing in question knowingly and freely. She could, on my view, fully recognize that she was not coerced, compelled, forced, or tricked into doing the thing in question--she did it freely. Under such circumstances, why would she not feed deep regret?

Hi John, and thanks for your continued engagement. I know you're not a skeptic but I thought you were sympathetic to the forward-looking accounts that they have advanced given your rejection of OIC and some of the ways you've pressed me in previous threads. But, if my memory serves me right you did appeal to desert in your piece for the Nadelhoffer collection on Punishment, a piece I remember liking a lot. Anyhow, I share with you the belief that we need more than strictly consequentialist understanding to make sense of our punishment practices (if that's what you argued). I'm not convinced that guidance control is robust enough to ground it though and that's where we part ways.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan