Dear Fellows,

What follows are some notes for a paper I mean to write. Since this comes in rather late, I wanted to keep it short. You'll see that it fizzles out towards the end. I'll say more about how the argument concludes when we meet. Until then …

– GR

Responsibility and Moral Ignorance

For the ODFF, September 1999

Gideon Rosen

I. Skepticism about Moral Responsibility

The aim of this note is to explore a neglected argument for skepticism about moral responsibility. The skeptic I have in mind is not a skeptic about morality as such. So she need not deny that much of what we do is morally wrong. But there is a distinction – the crucial distinction in this area – between the claim that an act was wrong and the claim that the agent is morally responsible (i.e., blameworthy or culpable) for having done it. When the kleptomaniac swipes a candy bar she clearly acts wrongly, and our first response may well be the sort of hot, agent-directed indignation or resentment that constitutes blame in a case of this sort. But when we find out that she suffers from a psychological disorder that interferes with her capacity to control her conduct in certain circumstances, then it’s natural to suppose that our indignation should abate. Perhaps this conclusion is not inevitable in this case. But the fact that the question arises at all is enough to show that it is one thing for an agent to do something wrong, and another thing – a further thing – for her to be responsible for the wrong she does.

There is nonetheless a standing presumption that we are responsible for what we do. If we call any fact that defeats this presumption in any particular case an excuse, then skepticism about responsibility amounts to the view that everyone who acts badly has an excuse of some sort. There need not be a single master excuse. It would be enough if there were a range of excuses that together cover the field. The challenge for the skeptic about responsibility is then two-fold: to say which factors should count as excuses and why, and then to argue that common sense morality and legal practice notwithstanding, the fact is that everyone invariably has one.

II. The Usual Strategy

One venerable tactic for supporting skepticism about responsibility is to argue that responsibility for present actions always presupposes responsibility for some past act or state of affairs. Since it is incontrovertible that no one can be responsible for what happens before his birth, this principle virtually entails that no one is ever culpable for what he does.

Say that X’s responsibility for A is derivative when it is contingent on X’s having been responsible for some prior act or state of affairs. The crucial premise in the standard argument is then that all responsibility is derivative responsibility. Why should any one believe this? The traditional case for this view is the case for hard determinism. Very roughly: (a) You are morally responsible for A only if you are responsible for every causally sufficient condition for A – every condition given which A had to happen. But (b) under determinism, every action is determined by sufficient conditions extending backwards indefinitely in time. So if determinism is true, you are responsible for doing A only if you are responsible for a series of arbitrarily remote prior happenings. But that’s impossible. So you’re not responsible for anything.

But of course this will not do. It’s not just that determinism is a disputable bit of physics. Even if it were true, the argument would still be unconvincing. The "transfer principle" (a) derives its superficial plausibility from a mistake – or better, two mistakes.

First, as compatibilists have been saying for centuries, the intuitive case for the transfer principle trades on an illegitimate assimilation of causation to coercion or constraint. The standard argument takes it for granted that if C is causally sufficient for A, then given that C obtains, the agent had to do A; had no choice about A; was forced or compelled to do A; could not have done otherwise. This assumption is crucial, since it alone forges the link between the metaphysical thesis of determinism and the moral language of excuse. Once this link is granted, the argument proceeds in what appear to be moral terms: If C makes it the case that X had no choice about doing A, then X is responsible for A only if he was responsible for C. Otherwise he has a legitimate excuse: the excuse normally expressed by phrases like "I couldn’t help it (and moreover, it’s not my fault hat I couldn’t help it)".

The standard argument fails to convince mainly because the modal idioms whose application would be licensed by the fact of determinism are distinct from the morally relevant modal idioms from the theory of excuses. From the fact that an action was caused, and in that sense had to happen (given its causes), it simply does not follow that the agent had no choice, or that he lacked the capacity to do otherwise. A sugar cube in an airtight jar may be soluble even though in the circumstances it is physically impossible that it dissolve. Similarly X may possess the capacity to refrain from (say) stealing, even though in the circumstances it is physically necessary that she steal. It may be somewhat unclear what it means to attribute a capacity to a thing in circumstances in which the exercise of that capacity is ruled out by the laws of nature. But it seems clear that the "coulds" and "cans" that figure in the theory of excuses are modal idioms of precisely this sort. And if this is right the standard argument is based on a mistake.

Beyond this we may add – and this is the second defect in the standard argument – that even if determinism did imply that we invariably lack the capacity to do otherwise, it still would not follow that we are never responsible. For as Harry Frankfurt has shown, it is a straightforward mistake about the content of ordinary morality to suppose that "I could not have done otherwise" is always an excuse.

III. Some Cases of Derivative Responsibility

So much for the standard case for skepticism about responsibility. Is there a better way to make the same point? I think there may be, although the case is hardly straightforward. The argument I want to explore mimics the standard case in this resect: First a genuine excuse is identified; then it is argued that because responsibility for the present depends on responsibility for the past, this excuse applies generally, or at least in a vast range of normal cases. In the standard argument the excuse was "I just couldn’t help it (and moreover, it’s not my fault that I couldn’t help it) and the generalization depended on the assumption of determinism. In the version I wish to discuss the excuse is rather different: very roughly, "I did not know that what I was doing was wrong". Moreover the case for its generalization does not depend on determinism or any other metaphysical hypothesis, but rather on certain normative views about our intellectual obligations.

Before I present the argument in earnest, it will help to consider a range of cases in which it should be uncontroversial that responsibility for the present depends on responsibility for the past.

(1) Gus is presently depraved. Not only is he prone to unprovoked malicious violence; he lacks the capacity to control himself. When he slugs an old lady for jostling him inadvertently on the subway, there is no doubt that he acts wrongly. But whether he is blameworthy depends (simplifying slightly) on whether he is responsible for his depravity. Suppose that it is the result of his having ingested a drug that temporarily destroys the capacity for self-control while amplifying one’s normal aggressive impulses. In that case, whether he is responsible for the assault depends on whether he is responsible for ingesting the drug. If he did it deliberately because he’s a blocked novelist who wants to experience the Dark Side for himself, then of course he’s responsible despite his incapacity in the moment of action. On the other hand, if he has been drugged against his will then he may well be off the hook. He may legitimately say in his defense, not just "I couldn’t help it" but "it’s not my fault that I couldn’t help it". Jones’s past is relevant because this last remarks is an essential part of his excuse.

(2) Ed is walking through the woods and inadvertently trespasses on your land. He has no right to be there, so what he is doing is wrong. Of course, he doesn’t know that he’s trespassing, so he can’t be blamed for knowingly violating your property rights. But there is still a question about whether he’s culpable simply for trespass, which is wrong (I suppose) whether done knowingly or not. And that question turns on whether Ed is culpable for his failure to know that the land was privately owned. If the property line was well marked, and Ed was reckless or negligent in failing to notice the signs, then he is culpable for his ignorance and hence for his trespass. On the other hand, if Ed has been careful to watch for signs but has somehow missed them nonetheless, then his ignorance is non-culpable and so is his act.

Note that the question is not whether Ed could have known that he was trespassing. We may suppose that the property lines were duly recorded in the municipal office, and that nothing prevented Ed from clearing his route in advance. Still, in failing to bother with this sort of research Ed is hardly reckless or negligent. There is a distinction between the man of ordinary prudence who does what is ‘normally’ required to minimize the risks to others from his actions and the man of maximal prudence who does everything in his power to minimize risk. The standard for negligence in tort law is the former and not the latter, and this reflects a moral distinction. You are morally responsible for your ignorance only if it derives from a failure to do what is morally required of people like you in your circumstances. Ed has not done everything he could possibly have done. But for all we have said, it may be that he has done everything that he was required to do, in which case his ignorance is non-culpable and he is off the hook.

The two cases are very different, needless to say. Gus is incompetent, at least temporarily. In one sense of the phrase, he is not a free agent. Ed, by contrast, is a normal adult with the abilities and capacities that normally suffice for free agency. Still they have something important in common. In the paradigm cases of morally culpable bad action, the agent satisfies certain "synchronic" conditions: conditions that characterize the agent around the time of the act in question and which are normally taken to suffice all by themselves for culpability. The paradigmatic bad agent not only does something wrong: he does it deliberately and he does it freely in the knowledge that it is wrong; he is neither forced nor coerced; he possesses the capacity to deliberate and to act in light of his deliberations; etc. Ordinary morality seems to hold that when these synchronic conditions are satisfied, the agent is automatically responsible, in which case there is no need to look backwards in order to determine whether he was responsible for some past act or state. In the exceptional cases illustrated above, by contrast, the agent fails to satisfy the normal synchronic conditions for responsibility. Ed is unwitting; Gus is unfree. And what the cases suggest is that a failure to satisfy the synchronic condition constitutes an excuse if – but only if – this failure was not the foreseeable upshot of some past wrongdoing on the agent’s part for which he is independently culpable.

So we have identified an excuse – perhaps the master excuse: non-culpable failure to satisfy the normal synchronic conditions for responsibility. The availability of this excuse yields a class of cases in which responsibility for the present turns on responsibility for the past. Now the skeptical strategy was to argue that every case falls into this category. But the considerations we have so far adduced seem quite useless for this purpose. After all, the excuse as we have characterized it only kicks in when the agent fails to satisfy the normal synchronic conditions for responsibility. And the trouble (for the skeptic) is that on the face of it a great many agents clearly do satisfy those conditions: they knowingly and wholeheartedly do what is wrong, neither forced nor coerced, in full possession of their faculties, etc. Nothing we have said so far points to an excuse that might apply to ‘normal’ agents of this sort. Cases of derivative responsibility therefore give us – for all we have said– no reason to suspect that the ordinary bad actor must inevitably have an excuse that neutralizes the normal presumption of blameworthiness. But that’s what the skeptic needs, if her position is to be at all surprising.

IV. Action from Moral Ignorance.

Against this, it seems to me that the phenomenon of derivative responsibility is much more widespread than is normally supposed. Consider another case.

(3) In the ancient Near East in the biblical period it was universally acknowledged that there was nothing wrong with owning or trading in human slaves. No one denied that it was bad to be slave, just as it was bad to be sick or deformed. But it never occurred to anyone to object to the institution of chattel slavery on principle. So consider an ordinary Sumerian lord. He buys and sells human beings who have defaulted on their debts or been captured in battle; he confines them against their will, forces labor without compensation and separates families to suit his purposes. There is no doubt that all of this is morally wrong. But it is not at all clear that he is morally responsible for what he does.

The Sumerian lord acts from ignorance, not of fact, but of principle. And it seems to me that whether he is blameworthy ought to depend, as with the unwitting trespasser, on whether or not he bears responsibility for his ignorance. But when we put the point in this way, it seems reasonably clear that he should not be held responsible. There may be principled arguments against slavery. But if there are they are not easy or obvious. The Sumerian lord need not have been negligent or reckless for failing to hit on them. Indeed, he may have been every bit as careful and reflective and concerned to do what is right or noble as anyone can reasonably be expected to be. It’s just that given the cultural and intellectual resources available to him, it would have taken a moral genius to see through to the wrongness of a universally accepted social practice. It would be different if the case against slavery had been staring him in the face all along, and if his ignorance reflected a negligent or perhaps even a willful refusal to face the facts. But the average Sumerian lord is not Thomas Jefferson. The case against slavery was simply unknown in the ancient Near East, and he can hardly be faulted for failing to construct it from whole cloth.

This observation seems to me to temper if not to neutralize entirely our capacity to hold ancient slaveholders fully responsible for their actions: to feel anger or resentment towards them for what they did. To the contrary, it seems to me that in view of the historical facts the proper moral response to ancient Near Eastern slavery is not the engaged attitude of moral blame, but rather something more like the unfocused ruefulness characteristic of what P. F. Strawson calls the "objective stance". You may condemn the act. You may resent the fact that the universe so reliably serves up injustice on this scale. But what you cannot do – or better, what you should not do – is to hold their injustice against them in a case in which it would have taken a miracle of moral vision for them to have seen a reason for doing things differently.

It may seem that our intuitions about this case reflect what Bernard Williams has called the relativism of distance. Biblical Sumer is so remote from us both in time and culture that we are not really called upon to feel one way or another about it; and this may be thought to explain our unwillingness to engage the Sumerians across time with moralized reactive attitudes like anger and resentment. We can correct for this to some extent by imagining the discovery of a traditional slave-holding society in the modern world. Such a discovery would raise genuine questions about what do and what to feel, and the reflections in the previous paragraph would still apply. The right moral/emotional response would not be the sort of engaged contempt directed at the wrongdoer himself that constitutes moral blame in a case of this sort. The fact that their ignorance of their own wrongdoing is (as it were) innocent deflects anger of this sort onto a target to which it cannot properly adhere; history, human nature; the social conditions that made such monumental moral ignorance possible, and so on.

So our disinclination to blame the ancient slaveholder after we have taken his non-culpable moral ignorance into account is not just a reflection of the relativism of distance. And to reinforce the point, consider a case closer to home.

(4) Smith is a run-of-the-mill American sexist ca. 1952. Like any conscientious father he encouraged his sons to go to college and set aside money for the purpose. But like any run-of-the-mill sexist he did nothing comparable for his daughters. He is clearly in the wrong. He has treated his daughters unfairly (though not maliciously) and this violates a duty. But again, it is unclear whether he is responsible for his violation.

If you had asked Smith at the time why he had not encouraged his daughters to go to college, we would have said, "because they’re girls", as if the sufficiency of the remark were obvious. The casual sexist takes it for granted that sons qua male have certain legitimate expectations to which daughters qua female are not entitled (and vice versa). Let’s suppose that this assumption is not based on reasoning, and in particular that it does not depend on bad science or bad religion. Let’s assume – and this is not unrealistic – that our sexist believes what he believes because he was raised to believe it and because everyone he takes seriously also believes it. The principle that gender matters in this way thus functions for him as an undefended axiom of moral common sense. There may be arguments against it – a priori from first principles, or a posteriori from the consequences of sexism for women’s welfare. But if there is such a case, it is not easy or obvious. To the contrary, the emerging consensus on the moral irrelevance of gender is hard won – the work of geniuses and activists over decades. That Smith has failed to see through a pervasive and well-protected ideology need not be a sign or recklessness or negligence or stupidity on his part. And if that’s right, then it strikes me as simply unreasonable to hold him responsible for the wrong that he has done.

To be sure, when the patent wrong of his sexism comes fully to consciousness later on, it will be natural for Smith’s daughter’s to resent him for treating them so unfairly. It will also be natural for Smith to feel guilty, and in this sense to hold himself responsible. Certainly, when his daughters reproach him it would be unseemly for him to do anything but apologize, beg forgiveness, and make amends if he can. But however natural and even laudable these responses may be, the fact is that Smith has a cogent defense against the charge of unfairness, whether brought by his daughters or by his own conscience, even if it would be unseemly for him to mount it. He (or his advocate) can say quite correctly: "Consider what it would have taken for me to see a reason for acting differently. I would have had to invent form myself a new sensibility. I wish I had. I regret that I didn’t. But it is frankly unreasonable for you to blame me or for me to blame myself for doing what seemed perfectly reasonable at the time given everything that I can plausibly be expected to have known."

This is not to say that Smith should not feel badly about what he’s done when he realizes just how wrong it was. But the proper response is not guilt or self-reproach, but rather a sentiment more akin to what Williams calls "agent regret". The scrupulous driver who injures a pedestrian through no fault of his own may feel miserable and perhaps he should, not just because something bad has happened, but because it has happened "through his own agency". But if he is clear that he has been neither reckless nor negligent, then no one should hold the bad consequence against him. This means, in particular, that that his victim is morally obliged to forgive him. He is not morally responsible for the accident, so his victim has no choice but to foreswear resentment.

Smith’s treatment of his daughter’s is in some ways analogous. There is a difference, of course. There is a sense in which the driver has done nothing wrong. Insofar as moral constraints are constraints on the quality of will with which we act, the driver has not violated any moral constraint, whereas Smith in his invidious treatment of his daughters clearly has. So a better analogy might be with Gus’s depraved violence as a result of involuntary intoxication. When he comes to his senses Gus should feel badly about slugging the old lady. But he is not responsible for his malicious act; he should not feel guilty for it, and his victim is morally obliged to forgive. My claim is that the same goes for the actions of the casual sexist if it really is true that in his circumstances it would have taken supererogatory moral insight to see a reason for acting differently.

Examples of this sort could be multiplied indefinitely. Thus we might consider the scrupulously reflective agent with sound fundamental principles who reaches a mistaken verdict about a difficult case. It may be, for instance, that Truman was wrong to order the bombing of Hiroshima. Suppose he was, but suppose further that he not only believed that the act was right, he believed this after a searching review of the moral considerations on both sides – a review every bit as scrupulous as the gravity of the decision would require. Under these assumptions, it seems to me that Truman is not morally responsible for his action. It is morally wrong, perhaps monumentally so. But he has done everything that a person in his position can reasonably be expected to have done and yet failed to see this. His proper response in retrospect is the ‘agent regret’ mentioned above: the otherwise nameless emotion one feels when one realizes that one has blamelessly committed an egregious moral wrong. And the proper response on the part of others – on the part of his victims and on the part of history – is not the sort of agent-directed antipathy that makes for blame, but something else for which we also have no standard name.

I will not pursue the matter. If what I have said thus far is sound, we have a case for regarding action done from moral ignorance as on a par with action done from factual ignorance, and more generally, with prima facie non-culpable action. When an agent fails to satisfy the normal synchronic condition for responsibility, he is culpable only if his failure is the foreseeable upshot of some prior culpable wrongdoing. It has emerged (perhaps unsurprisingly) that one of these synchronic conditions is a condition of moral knowledge. So if you don’t know that what you’re doing is wrong, then you’re culpable only if you should have known – or more precisely, only if you are culpable for not knowing. The traditional tests for insanity incorporate a restricted version of this principle into the criminal law. If the defendant’s ignorance is the result of mental disease or defect, then it is plausibly non-culpable, and hence exculpatory. But note that as with factual ignorance, what matters is not whether the agent could have known, but rather whether he at fault for not knowing. In insisting that non-culpable moral ignorance excuses, I am not rejecting the rationalist view that moral principles are knowable a priori by any rational being. After all, there is a sense in which Fermat’s Last Theorem is knowable a priori by any rational being. But that doesn’t mean that you and I – nor even Gauss and Euler – were in any sense at fault for not knowing it until recently. It is knowable, but it is also very hard to know. For the purposes of the present excuse, what matters is what the agent has a duty to know, and this comprises much less than what the agent could "in principle" have known.

It is also worth noting that what the agent has a duty to know cannot be identified with what it would have been relatively easy for him to know in the circumstances. When an expert has been entrusted with a high-stakes decision based on his special competence, he may be faulted for not knowing what it would have been extremely difficult to know. And there are also cases on the other side. Suppose that unbeknownst to Ed (the unwitting trespasser), a detailed map of the local property lines has been placed in his backpack. He trespasses on your unposted land, not knowing that it’s yours, and presumably he’s off the hook. And yet it would have been easy for him to know that he should not have been where he was – as easy as opening his pack and consulting the map. In this case, his failure to know the pertinent fact was directly due to an omission – failing to look for the map. But since he has no reason to believe that the map was there, the omission is non-culpable, and neither, therefore, is the ignorance to which it leads.

V. The Extent of Moral Ignorance

We are still some distance from the skeptical conclusion that no one is ever responsible for what he does. I've argued that action from moral ignorance is often non-culpable. But I haven't argued that it is always non-culpable, and (perhaps more importantly), I haven't said anything about bad action done in knowledge of the moral facts. Let me say –very briefly now – how I think the argument might continue.

Consider the following sort of case. It's 4:30 on a rainy Sunday afternoon in NYC. You've spent the past twenty minutes waiting in the rain for a cab, loaded down with packages and two whiny children. When a cab finally pulls over to pick you up, a stranger races across the street, elbows you into the gutter, hops in and takes off. You are (presumably) speechless with indignation.

A week later you run into the stranger at a cocktail party. You still resent her for what she did, but you're no longer speechless. So in the interest of moral science you decide to ask her how she could possibly have done such a thing.

"Remember me?"

"Yes, of course. I took your cab last weekend. Tough luck"

"But why did you do it? Didn't you see that I was standing there in the rain with my children?"

"Of course I did. I knew that cab had pulled over for you. I knew I had no right to take it. I knew you'd be angry and miserable if I did. But the fact is that I didn't really care much about you or your children or about doing what's right. I was in a rush and just didn't feel like waiting for another cab. There very hard to come by at that time of day, as you know."

"Are you insane? Do you have trouble controlling yourself"

"No, none of that. I did it deliberately and I knew exactly what I was doing. I was wholehearted, reflective, in full possession of my faculties. I gave the matter some thought, taking your interests into account, and I decided that they just didn't count for much in the circumstances. I'm bored now. Go away."

You're probably even angrier now than you were before. But on reflection it's not clear to me that you should be. The stranger – call her Bonnie – does not have the excuse we identified in the previous section. She was not ignorant of moral principle, at least not in the following sense: She knew that you had a right to the cab, and that it was wrong for her to steal it. As described, her problem is that she just doesn't care about this sort of thing. But now I think we can redescribe this misguided lack of concern as a species of moral ignorance. She may know perfectly well what is morally required of her in the circumstances. But to say that she doesn't care very much about doing what's required is to say that she doesn't think that your rights and your welfare matter very much. But they do matter. And this is just to say that Bonnie is ignorant of the relative weight of these moral reasons. This is a kind of moral ignorance - or perhaps better, a more general normative ignorance. You need not think that moral considerations are always inevitably over-riding to think that they very often bear considerable normative weight, and that anyone who fails to assign them something like their appropriate weight is making a mistake – a mistake about what really matters. But to say that this is a mistake is to say that Bonnie is ignorant.

Once we have identified this sort of moral ignorance, it should be clear that action from moral ignorance is really quite widespread. Setting aside cases of akratic bad action – an important class of cases, about which I have nothing new to say – the rational bad agent who acts in accordance with his deliberation and his judgment about what he has most reason to do all things considered must be acting from moral ignorance of this sort. After all, if he knew how much weight the relevant moral considerations really deserved, he would not have acted badly (other things being equal).

This is all highly tendentious, but if it can be defended it will entail that almost all bad action is action from moral ignorance. But of course that's not enough. What the skeptic needs to show is that the clearheaded bad agent is not only ignorant of moral principle, but also that he is non-culpable for that ignorance. And you might legitimately wonder how that could possibly be.

Consider Bonnie again. As you're fuming and wondering after your encounter at the party, you're approached by another stranger. "I couldn't help over-hearing," he says. "I know you're angry, but you really shouldn't be. I'm Bonnie's (appallingly indiscreet) neurologist and her case is really very interesting. Until recently she was an ordinary person with an ordinary concern for other people and for morality. A month ago she contracted a rare virus that acts directly on the brain, rewiring the neural circuits that underlie the concern for morality, and more generally, one's sense of what counts as a reason for action. In this case, the rewiring had the effect of turning a good person into an unrepentant selfish creep. While the virus is working, its effects are ineluctable. The virus works over the course of a week and then disappears. When she stole your cab, Bonnie was an otherwise healthy human subject, and she is still perfectly healthy. She has the same capacities for moral thought and reflection – the same rational capacity to be moved by cogent argument – as you or I. But I've tried to argue with her, and the fact is that her dismal new view of things is surprisingly well defended. If you ask her how she feels after her moral transfiguration she will say that it is as if the scales have fallen from her eyes. She used to think that other people had a claim on her and that there were all sorts of impersonal constraints on the pursuit of her self-interest. But now she sees – that's how she puts it – that all of that was a crock, and that in fact we're all in it for ourselves in this crummy world. We've been over the arguments from Plato to Kant, and she's a very good philosopher. She finds them all unconvincing."

And now you think: Bonnie was ignorant of the relevant moral facts, in this case the fact that the rights and welfare of others make a difference. But her ignorance is not due to any prior culpable wrongdoing on her part. It was not her fault that she contracted the virus. And she was anything but negligent in policing her moral and practical opinions after the fact. Through no fault of her own she found herself in a position from which there was no easy dialectical route back to orthodox moral concern. This is not to say that there was no route whatsoever. Maybe there is a cogent philosophical argument for the view that moral considerations matter immensely. But if there is, Bonnie can hardly be faulted for failing to find it. We are obliged to reflect to some degree on our moral view from time to time. But Bonnie has met that obligation and she has met it brilliantly. She is not culpable for failing to get things right after such a scrupulous attempt to do so, any more than mathematicians before Wiles are to be faulted for failing to hit on a proof of FLT. So Bonnie satisfies the condition for the excuse we have identified: she acts from ignorance of moral principle, and because her ignorance is non-culpable, she is off the hook.

Now this is still an odd case. What needs to be shown to cinch the largest part of the skeptical argument is that everyone who acts from moral ignorance is in effect in Bonnie's shoes. You can see how the case might run. The normal causes of bad character, whatever they may be, are like Bonnie's virus in every relevant respect. They operate independently of the agent's will, and their operations are largely unnoticed. So an ordinary selfish creep who finds himself with Bonnie's well-defended moral view as an adult is no more culpable for ignorance than Bonnie is. Like Bonnie, he finds himself through no fault of his own in a state of ignorance from which there is no accessible route back to orthodox moral concern. If this is right – and of course I've just asserted it without argument – then we finally have the beginnings of a surprising skeptical thought. Take the bad actor, and ask for the reasons that underlie his behavior. They will typically involve a mistake of some sort: a mistake about right and wrong, or about the relative practical weight of moral considerations. Ask what he could have done at some point in the past to remedy this mistake. What sort of reflection or moral inquiry would it have taken to correct his moral view? Very often it will be anything but obvious what he could have done. But even when it is obvious it will very often be the case that the agent is not culpable for failing to take the necessary measures. We are not obliged to rethink the foundations of morality whenever we act. If that's what it would have taken, then the agent's ignorance, though remediable, may still be non-culpable. Indeed it is hard to see how ignorance of moral principle could ever be culpable. There is a general argument for this surprising view from materials we have already considered. But I'll leave it for discussion. You have enough to see the rough outlines of what I have in mind.