NDPR Forum: Gideon Yaffe’s The Age of Culpability, reviewed by Doug Husak

32 Replies to “NDPR Forum: Gideon Yaffe’s The Age of Culpability, reviewed by Doug Husak”

1. First, thanks to Doug Husak, for his very thoughtful and generous review of my book, and to Dave Shoemaker for organizing this event. I’m thrilled to have this opportunity. I’ve been a PEA Soup voyeur for a long time, so I’m glad to finally participate.

   My book aims to explain and justify the appealing proposition that people under 18 (“kids”) deserve some kind of leniency when they commit crimes (“a break”). As it turns out, to do that adequately requires philosophical analyses of several things of central importance to philosophy of criminal law. It requires accounts, I think, of criminal culpability, desert for wrongdoing, the strength of a legal reason, and having a say over the law. At least, accounts of those things are required to support the following central argument of the book. Here D is a criminal defendant; C is a crime; R(C) is how much reason D is disposed to take the features of C to give him to refrain from C-ing; L(C) is how much legal reason the features of C actually provide D to refrain from C. The argument:

   1. D is criminally culpable for C if: (1) C manifests D’s disposition to recognize a reason of strength R(C) to refrain from C, and (2) R(C) < L(C). The greater the gap between R(C) and L(C), the greater is D’s criminal culpability for C.

   2. L(C) is in part a function of how much say D has over the laws thanks to which there is a legal reason to refrain from C.

   3. Kids have less say over the law than adults. So, ceteris paribus, L(C) for a kid < L(C) for an adult.

   4. So: If a kid and an adult commit the same crime, in the same circumstances, with the same mental states, the gap for the kid between R(C) and L(C) will be less than that for the adult.

   5. So: If a kid and an adult commit the same crime, in the same circumstances, with the same mental states, the kid is less criminally culpable than the adult.

   6. If two people commit the same crime, in the same circumstances, with the same mental states, but one is less criminally culpable than the other, then the one should (morally) be treated more leniently than the other.

   7. So: If a kid and an adult commit the same crime, in the same circumstances, with the same mental states, the kid ought to be treated more leniently than the adult.

   I sometimes summarize this argument by saying that kids are owed a break because of the political meaning of age, rather than its psychological or neural meaning.

   The book offers new accounts of criminal culpability, supporting premise 1; legal reasons and their strength, supporting premise 2; what it is to have a say over the law, supporting premises 2 and 3; desert for wrongdoing, indirectly supporting premise 6; and also some arguments in favor of the disenfranchisement of kids. The book also discusses a variety of connected topics—the import and limits of developmental psychology and neuroscience to the criminal responsibility of kids; the relevance of democratic governance to the justification of criminal punishment; the implications of my argument for the criminal culpability of other disenfranchised groups such as immigrants, felons and the poor; the nature of children’s free speech protections; the justifiability of life-without-parole sentences; and so on. Doug necessarily had to pick what to focus on for the purposes of his review, and so my remarks here are going to center on culpability and legal reasons, since he placed those central topics on the agenda. But I’m anxious to hear anything that anyone has to say about any of the other myriad issues the book touches.

   There is one important inaccuracy in Doug’s description of my theory of culpability. He says that I think culpability involves comparing how much reason the agent takes herself to have to refrain (R(C)) with the sum of all the reasons that he actually has to refrain (moral, legal, prudential, etc.). But that’s not what I think; it also can’t be right. Say that a bank heist D is considering is very aesthetically pleasing; it would be a work of art. So, there are strong aesthetic reasons to perform it. But that does not reduce D’s criminal culpability when he goes ahead with it. The aesthetic reasons are irrelevant to the criminal culpability calculation; they don’t figure in at all. Nor do the prudential reasons; one is not reduced in culpability for more lucrative bank robberies. My view is that the same is true of the moral reasons to refrain. They matter to moral culpability, but not to criminal. (I am unflinchingly positivist: a legal system that predicates differences in treatment on criminal culpability might continue to maintain perfect separation between law and morals.)

   Doug’s example of a spousal rape in 1990 in the UK when there was a marital exemption provides some evidence for my position. The perpetrator has no criminal culpability (although no shortage of moral culpability). How could someone be criminally culpable for doing something that isn’t a crime? The view that Doug attributes to me implies, by contrast, that the spousal rapist in the example is criminally culpable, since he grants far less reason to refrain than he actually has. But any theory of criminal culpability must account for the fact that there is no criminal culpability for perfectly legal conduct, no matter how morally monstrous.

   In my view, there is no such thing as culpability, full stop. There are only types of culpability: criminal, moral, rational, maybe even aesthetic. The type is provided by the relevant body of reasons, often tacitly identified by context: legal, moral, prudential, aesthetic, etc. If you manifest a failure to grant as much weight as you should to the moral reasons, you’re morally culpable; to the legal, criminally culpable. The “should” here is relativized to a particular correlated body of norms: the moral or legal norms concerning what factors one is to weigh in deliberations and how heavily.

   This view gets me in some trouble, however, as Doug notes in his review. In particular, I take there to be moral reason for the government to punish to a degree that is proportional to criminal culpability, even though that might mean issuing a punishment that is entirely out of line with moral culpability. Doug is skeptical of this; he thinks that the theory that supports premise 1 undermines premise 6. I must confess that I do not have as much control over the issues here as I would like. I still persist in thinking I’m right about this, however. One piece of evidence for my position, as Doug is aware, comes from reflection on punishment for mala prohibita offenses. For instance: in many jurisdictions, it’s a crime to sell individual cigarettes, removed from packs. It’s legal for me to buy a pack of 20 and sell it to you for $10; but it’s a crime to sell them to you individually for 50 cents each. Since single cigarette selling is illegal but not morally wrong, offenders cannot be morally deserving of any kind of punishment for such behavior. All punishment in such cases is morally excessive. But now imagine that the state recognizes no excuses for some type of mala prohibita behavior. The person, for instance, who sells individual cigarettes at gunpoint is given the same criminal punishment as someone who sells them of his own free will. In my view, the state is doing something morally wrong in denying the excuse. But that can’t be because it is causing a mismatch between punishment and moral culpability. Neither cigarette seller is morally culpable at all. So the only explanation for the moral pressure the state is under to recognize an excuse in such a case is that there is moral pressure to punish proportional to criminal culpability. I don’t know why the Principle of Legality has moral force; but I’m confident that it does.

   In his review, Doug seems to suggest that the mala prohibita cases should be set aside. They are in need of their own special analysis, he thinks. It seems to me, by contrast, that a theory of criminal culpability has to accommodate the fact that we sometimes find it—not something like it, but it itself—where we find no moral wrongdoing. And, further, that where we find it it places moral constraints on government punishment.

   Doug has another argument against me: he thinks my theory of criminal culpability cannot adequately model the culpability of mala in se offenders, like the murderer Christopher Simmons. Doug’s argument here relies on the idea that when behavior like murder or rape is criminalized the effect is, as he puts it, to “recognize[] and rename[] an old reason that existed all along.” But I don’t see why Doug thinks that’s incompatible with my position. Say he’s right; there are just moral reasons that are institutionally spotlighted and relabeled thanks to the law. Then, on my view, a person’s degree of criminal culpability is a function of the strength of the “recognized and renamed” reasons, and only those. So long as you grant that there are such things as legal reasons, it’s compatible with my position to hold that they come to be thanks to “renaming” of moral reasons, or through some other mechanism entirely.

   But, like I say, there’s lots to talk about here. I’m looking forward to reading what everyone has to say.

2. Gideon, this is really fascinating stuff. I just have what I think is a pretty small question: To what extent does your view of culpability and reasons depend on there being identifiable, hard-and-fast distinctions between the different normative domains (e.g., moral, criminal, prudential, aesthetic, etc.)? You say the type of culpability is a function of the relevant domain of reasons, but it might be that the normative domains aren’t so distinguishable that viable distinctions between types of culpability can be made or, more importantly, even if they can be made it could be that (in)competence in different distinct normative domains is inexorably tied to (in)competence in others due to a common cause. Having a bad sense of humor, for instance (due to one’s inability to see certain kinds of teasing as a reason for amusement) may be a function of a deficit in theory of mind (so one can’t read others’ intentions very well), the very same deficit that may undercut one’s ability to not be rude in social situations (one can’t “read the room”). Is there something to be gained, then, by pointing to multiple types of culpability for multiple domains, as opposed to one type of culpability that may manifest differently in different domains? (This may be an ill-formed question; hope it’s clear enough.)

3. Hey Gideon, really cool stuff. If I remember correctly, you mostly assume that kids should be given a break. One nice thing about your approach, in contrast to those that take underlying capacities to be relevant, is that you can at least aspire to such a categorical break.

   Having not finished the book, I’m not familiar with your arguments against at least some kids getting the vote (or having more of a say over the law). I would think that it’s a place where worries about capacities might again rear up.

   No need to recapitulate entire arguments, of course, but I’d be interested in how you defend against those who might accept part of your main argument and conclude not that all kids should get a break, but that some kids should get more of a say over the law?

4. Thanks for jumping in, Dave! First, a confession: I aspire in my work on philosophy of law to metaethical neutrality. My hope is that any metaethical view that saves the appearances (as it were) could be coupled with my views about why kids are owed a break without inconsistency. But I think I’m probably falling short of this aspiration. In particular, I’m probably committed to thinking that there are true and false claims about specific domains of reasons, and you can envision metaethical positions according to which that’s not so. So, to use one example that I discuss in the book, there’s a California law that criminalizes the defrauding of pawnbrokers. You have not committed the crime if you defraud a shopkeeper, or an accountant; it’s got to be a pawnbroker. I think it’s true that the fact that someone is a pawnbroker gives others a legal reason not to defraud them. But this fact does not give anyone a moral reason not to defraud them. The moral reasons not to defraud come from things like the fact that the other person is a human being, pawnbroker or no. If a view of the nature of reasons, or the differences between domains, makes this turn out to be false, then, yes, I’m committed to the rejection of that view of the nature of reasons.

   On your other point: I think that there are lots of cases in which single defects underly different kinds of culpability. That’s consistent with my position, I think (and hope!). Christopher Simmons failed to take another person’s humanity as a reason not to kill her. The manifestation of that failure in his act is a big part of the reason why he’s morally culpable for the murder he committed. But it’s also a big part of the reason why he’s criminally culpable for it. After all, the fact that another is a person gives you not just a moral reason, but also a legal reason to refrain from killing her.

5. Thanks, Matt! A first point, which is easy to overlook, I think, is that who does and does not get the vote has, and should have, almost nothing to do with capacities to vote competently. It also has almost nothing to do with who’s affected by the law. We deny the vote to all the citizens of Mexico, for instance, despite the fact that they are just as competent as Americans and very strongly affected by American law. Capacities to vote are not totally irrelevant; there is good reason to deny the vote to someone, for instance, who is completely out of touch with reality. But the competence bar is very low and even little kids–6 or 7–meet competency standards. Any adult, for instance, who can fill out a voter registration form can vote in California and most kindergartners could get that job done. So, although competencies matter a bit to vote distribution, they can’t do much work in justifying the line at 18. Generally, to explain why it’s a good idea to deny the vote to X, you need to appeal to other values besides the value of competent voting. In the book, I argue that kids should be denied the vote because of considerations of equality–you don’t want parents to have greater say than non-parents–and considerations of the persistence of the state.

   But the other point you raise is this: Why not react to my argument by lowering the voting age (or, at least, allowing in some below 18) rather than being lenient towards the under-18s who commit crimes? My answer in the book is, basically, that considerations of coherence don’t rule this out but there are more coherency considerations to weigh than it might appear. In particular, the vote isn’t the only mechanism through which people have a say over the law. Another important one is free speech protection. Kids enjoy some free speech protection, but it is limited in various ways by parental entitlements. Those too would need to be loosened, if you were going to undo my argument for leniency by granting a kid a say over the law. One of the main ways in which kids’ free speech protection is limited is by parental control over where kids can go and who they can talk to. We give parents a legally-protected entitlement to keep their kids from attending political rallies, if they choose, for instance. To transform a kid into someone who is undeserving of the kind of leniency that my argument supports, then, you’d also have to lift this kind of restriction. Basically, you’d have to make them legal adults in many senses, and not just for the purposes of the vote. The result: my argument does not support leniency for “emancipated” kids–those who achieve the legal status of freedom from any kind of parental, or in-loco-parental constraint–provided they are also given the vote. So to make any large class of kids undeserving of leniency when they commit crimes, you would need to emancipate them in this way in addition to giving them the vote. Any argument for doing that will have to grapple with parental right; it won’t do to note that a lot of kids are capable of voting and voting well.

6. Ok, Gideon, that seems sensible, and I agree that who should and should have a right to vote ought not be settled solely on capacity criteria. I’m not as sold on the political claims though, since they seem most plausible only against a backdrop in which we think kids should be treated differently. For instance, you’re probably right that if even young kids could vote, parents would have more influence than non-parents. But I think the same will be true lots of social group units. Leaders of large organizations (whether churches or social clubs) will have more influence than others, too. Only if we assume that kids are different (say, in capacity) does the difference seem stark to me.

   You raise a good point regarding parental rights, though I should think whatever we’re inclined to think about parental rights will apply to the hardest cases of the categorical “break” that’s at issue here. So, a parent’s right to restrict the movements of their child will seem less justifiable the older the child is, just as leniency for older kids seems less obvious. Or, so it seems to me (of course, I’m not a parent!).

   A related worry might be that the flip side of categorically giving kids a break is that we categorically disempower them as well. Here I’m thinking about states that don’t have mature minor statutes for medical decisions. In Connecticut, for instance, a 17 year old wanted to refuse chemo for a highly treatable form of cancer, and her mom supported her decision. The state intervened and forcibly administered the treatment, keeping her confined to the hospital. Obviously, CT couldn’t have done that if she were 18. There’s more to say about such cases, of course, but it strikes me that giving them a categorical break might also be to not take them seriously enough (to speak very loosely).

7. I would put your main point, Matt, like this: It’s a very bad idea to use the distribution of the vote to equalize natural power imbalances arising from people’s differential capacity to influence each other. And that’s got to be right. Otherwise, as you say, we’d have a reason to deny congregants the vote so as not to give excessive influence to their ministers. The point I was after, however, is not about the power to influence, but about the entitlement to (try to) influence. The difference between the parent and the coach might not be one of power to influence the kid’s opinion; the coach might have more power in this respect than the parent, or less, who knows? The difference is that the parent has a legally-protected entitlement to exert influence over the kid’s opinion. When you have a legally protected entitlement to exert influence over the opinion of a voter, you thereby have greater say over the law than someone who lacks that entitlement. “Having a say”, in this sense, is a matter of entitlement not power. We ought to be distributing the vote in a way that results in equality of entitlement. Equalizing power, which might be a worthy goal, has to be pursued through entirely separate means. Given the parental entitlement to exert influence over the kids opinion, to give the kid the vote is to give the parent greater say over the law than the non-parent. This is true even if the coach continues to have more power over the kids’ opinion than the parent. Since we ought to care about equality of entitlement, we ought to deny the vote to kids even if that step does nothing to equalize imbalances of power. See what I mean?

   Your point about not taking kids seriously enough is a great one. However, I also think it’s important to see that the issues are different in criminal law than in many other domains. As I see it, a criminal punishment, which is a state action, is generally justified only if the person punished is complicit in that state action through having a say over the law. By contrast, justified medical treatment, even if it’s state imposed, need not obey that constraint; think quarantine. So, I don’t think forced treatment has the same normative logic as punishment. And so I don’t think denying kids the vote implies that the state should do the kind of thing that CT did in your example.

8. Gideon Yaffe has written a terrific book. It is full of fascinating and original arguments. I urge everyone to read it who has an interest in the connections between morality and law.

   I continue to believe his argument fails to show why kids morally deserve a break relative to adults who commit the same crime. In my review, I claimed that the most worrisome premise is what Yaffe (now) identifies as (6). Ultimately, we are interested in moral culpability, which is a key component of a moral justification for whether and to what extent wrongdoers morally deserve to be punished. And I do not see why we should think moral culpability relates to or tracks criminal culpability in anything like the way Yaffe believes. A person can have lots of moral culpability with little criminal culpability, and little moral culpability with lots of criminal culpability. Or so it seems to me.

   Clearly, however, the connection between moral and criminal culpability cannot be identified unless we know what each is. Moral culpability is hard, but Yaffe has lots of sensible things to say about it. So let’s return to premise (2), where Yaffe has something to say about criminal culpability. (2). recalls, is “L(C) is in part a function of how much say D has over the laws thanks to which there is a legal reason to refrain from C.” Whether and to what extent D has a “say” over the content of law C seems to me not to be very related to whether D can vote, but is affected more by many empirical variables such as those involved in *Citizens United*.

   But consider the following more direct rejoinder: Do we really believe that in a monarchy or any non-democratic state—in which D has (virtually?) no say over the content of the law—that D has (virtually?) no legal reason to refrain from a given crime? (Whatever “say” D may have over the content of the law certainly does not derive from his having a vote, since no one in this state has a vote). Is premise (2) true even when the crime in question is the horrible malum in se (a gruesome murder for thrill) perpetrated by Simmons? If D has (virtually?) no criminal reason not to murder, and moral culpability tracks criminal culpability in anything like the way Yaffe suggests, can it really be true that D has no moral culpability when he kills in the non-democratic state? I would think D has just as much moral culpability for killing in the non-democratic state as he would have in the good old USA or UK. I am pretty sure Yaffe would think the murderer can be (and is) morally culpable when he kills for a thrill in the case I have described, but it’s not clear to me why he thinks this.

9. As I said to Doug in conversation recently (and thanks for jumping into this, Doug!), sometimes when I stand back from my argument I want to give myself the incredulous stare. How can this possibly be the ground for leniency towards kids? But then I’m able to talk myself off the ledge by reminding myself of the arguments, which, for better or worse, I find convincing (at least when alone in my study). So, with that in mind, let me just charge right at Doug’s point about monarchies.

   First: my view is not that moral culpability varies with political arrangements. It doesn’t because the strength of moral reasons doesn’t vary with political arrangements. The strength of legal reasons does, which is why criminal culpability does vary with political arrangements. So I balk at the claim that for me “moral culpability tracks criminal culpability”. That’s true whenever, and only whenever, the strength of an agent’s moral reasons tracks the strength of her legal reasons. That’s the case when the law picks out moral reason-giving features of acts in such a way as to make it the case that those same facts are legal-reason giving. And that happens frequently. But it’s an empirical matter how frequently it happens. So: Christopher Simmons’ moral culpability would be no different were he living under a regime in which he had as much, or as little say over the law as adults have.

   To see what I’m up to, maybe it helps to remind ourselves of Hart’s discussion of the gunman. The gunman demands that you refrain from C. You defy the gunman and C. Now, as I see it there are two things that the gunman cannot say to you in justification for doing something bad to you for C-ing and both are things that a well-functioning state can say, by contrast. First, the gunman can’t justify the punishment by noting that C-ing was illegal; he lacks legal authority and so his demands aren’t law. Second, the gunman cannot say “You brought this on yourself.” He cannot cite the punished person’s complicity in the punishment. Since you had no say over what the gunman would demand, or over the fact that he had power over you, or anything, you were in no way complicit in the demand and so in no way complicit in what the gunman is doing in response to the failure to comply with the demand. And notice that the gunman has no reason to vary the severity of his response to defiance with the amount of say over his demands his victims have. None of them have any say, and he doesn’t care. And further this gunman has no reason to be more merciful to children than adults, when they really are the same in their behavior and their psychology.

   If a monarch is just a gunman wearing a crown, then she has no special reason to treat kids any differently from otherwise identical adults who defy her. Of course, kids will often have excuses that adults don’t have; a lot of them don’t know things that a lot of adults know, or can’t control themselves in ways that a lot of adults can. These things are morally relevant. But when kids don’t have any kind of excuse of this kind, they aren’t owed any kind of leniency by the monarch-qua-gunman. If the state is to Christopher Simmons just a benevolent gunman demanding that Simmons not bind people up and throw them off bridges then the state owes him no leniency in comparison to what it would give to an otherwise identical person who happened to be a bit older.

   But the state, of course, can aspire to be, and often is, something a lot better than a gunman writ large. When it can justify its punishments by appeal to the fact that the demand the offender defied had legal authority, and when it can point to the offender’s complicity in the punishment, it is better than the gunman writ large. But what comes along with that form of justification of punishment is attentiveness to the amount of say over the law that the punished have. The more say they have, the more complicit they are, and so the firmer ground the state is on in punishing.

   If the state is to justify its punishments by appeal to its legal authority and the offender’s complicity in the punishment, then it must grant offender’s a say over the law. But what comes along with that grant is exactly as much variation in the strength of a citizen’s legal reasons to comply with the state’s demands as there is variation in the entitlement to exert influence over the law. And given what criminal culpability is–manifestation in conduct of a disposition to improperly recognize, weigh or responds to legal reasons–it follows that there will be reduced criminal culpability among those who have reduced say over the law.

   I fear I’m repeating myself (and at excessive length!)…but sometimes a mantra makes the demons, with their incredulous stares, go away.

10. Congratulations on this very fine book, Gideon! I admire the coherence of the vision and, as always, the rigor of the argumentation. I have lots of questions, but let’s stick to criminal culpability.

    As you recognize, your view that culpability is always relative to a particular normative domain is strikingly heterodox. I’m a big believer in multiple normative domains, so I think that I’m more open to the account than are many. But even I am skeptical of the particular way you carve things here, and the use to which you’d put criminal culpability. So, with that background, would you say more about your arguments for criminal culpability?

    I can envision two ways to argue for criminal culpability, what I’ll call metanormative and bottom up. The metanormative route would try to show that domain culpability (as you conceive of it) forms part of the architecture of any normative domain. The idea, more or less, would be that it’s part of the nature of any normative domain D, that D-culpability plays a certain role in its operation. The bottom up route posits criminal culpability as necessary to justify particular doctrines or casuistic outcomes in the criminal law that strike us as intuitively correct but are otherwise hard to justify. Do these routes make sense to you? Do you adopt one or the other or both, or something else?

    In your original post, you make at least one bottom-up move, when you invoke mala prohibita offenses as one piece of evidence for your position. But I think that I do not understood that argument correctly. Adam sells individual cigarettes in knowing (?) violation of a criminal ban, and without justification or excuse. Eve sells individual cigarettes in knowing violation of a criminal ban under duress (and as a lesser evil, it seems). I’d think that the judgments we want to explain are: (1) it’s morally wrongful to punish Adam; (2) it’s morally wrongful to punish Eve; and (3) it’s morally more wrongful to punish Eve than to punish Adam. Are you contending that the only explanation for (3) is that there is moral reason to punish proportional to criminal culpability? I’d reject that contention if it’s yours, but I suspect that I’m not understanding your argument correctly.

    Insofar as you go the metanormative route, I’d want to hear more about how culpability works in some additional domains. What does this “aesthetic culpability” that you mention look like? What function does it serve in aesthetic judgments and practices? Is there baseball culpability or fashion culpability? Moreover, is “the law” a single domain or many domains for culpability purposes? That is, if there is criminal culpability, then shouldn’t there also be tort culpability and property-law culpability? Is this your view? Can there be distinct normative domains without it being the case that each has a place for culpability notions at all?

    Too much for an initial post, but those are some questions.

11. I think that Gideon’s positive argument for why kids deserve a break depends on his reliance on there being legal reasons (of kids not to do wrongful actions) and on those legal reasons having sway over whether legal institutions act (as by punishing). Looks like it anyway, given his responses in this exchange. What does Gideon mean by “legal reasons?” What might he mean? Consider three possibilities.
    1. There are all kinds of things that have “have-to-be-doneness” as their nature (Mackie’s name for reasons for action): moral reasons, prudential reasons, legal reasons, etc. Yet each gives its own distinctive kind of normative oomph; each is not the species of some one genus but is a motivator of rational actors in its own unique way.
    This may tempt Gideon. It tempted Hart about legal obligation. But I think that it is a non-starter. We can see this better by moving to the kinds of reasons the flouting of which gives rise to wrongdoing, viz, obligations. To make legal obligations distinct as a kind of normative force leads to Bentham and Austin, where legal obligation just is liability to sanctions (i.e., no obligation at all but only a reason of prudence). Yet to avoid that requires that legal obligation be a kind of moral obligation, one created by law. This is to deny that in this sense there are distinctly legal obligations or reasons in this sense.
    2. So make “legal” identify the source of the obligation, not some supposedly distinctive kind of normative oomph. Then it becomes like “promissory obligation” — a moral obligation just like any other moral obligation in its normative pull, but a moral obligation that arises because of the historical fact that a promise was made.
    This makes much more sense. Yet Doug’s objection (that I share) is that “law adds nothing” in the sense that law (“qua law”) does not create new moral obligations nor even new reasons for action of a non-prudential kind (neither do promises but that is a different story). If one is a so-called “new anarchist” of the last 60 years or so (Smith, Raz, Hurd, et al) then there are no legal reasons of this kind either.
    3. Of course, laws can solve co-ordination problems, be relied on by others, further equality, predictability, and all the other Fullerian virtues of the rule of law; and in that sense create a kind of moral obligation. A legal reason is then a moral reason not to frustrate expectations, disappoint reliance, treat people unequally, etc., by diverging from what the law requires, even when what it requires goes beyond or even against what morality without the less-than-ideal institutional history would require. That is the sense in which I think Gideon must be speaking of “legal reasons.”

    But in this sense of “legal reasons” it is hard to make Gideon’s argument go through. Such rule of law based moral reason create a little space between the content of moral norms simplicatur, and the rule of law adjusted moral norms formed out of legal rules. But not much. Most (damn near all?) of the moral reasons that bound Simmons was the non-law-based moral norm that one shouldn’t kill innocent young mothers just for the thrill of it. The law’s adition was small potatoes, morally speaking.

    Gideon urges that it matters that the law mark out the moral norm against what Simmons did, as a legal norm as well as a moral norm. But does such marking out by the law increase in any way the stringency of the obligation that Simmons breached? By Gideon’s hypothesis it doesn’t but it still affects Simmons desert. I don’t see how.

    Gideon urges that a state does more wrong when it punishes an individual who is not culpable with respect to the crime charged than one who is so culpable. This is where Gideon speaks of “legal culpability.” I would understand that phrase this way: this is the moral culpability one would have if the act one did was actually not just legally prohibited but morally wrong to do. And Gideon is right: the state punishing under such an unjust law does two wrongs rather than one: it was wrong to punish for the morally innocent act, and it was even more wrong to punish someone who didn’t even meet the requirements of culpability that would exist if th3e act punished were actually wrong. But the wrongs of the state in punishing do not directly correlate with the desert of the offended; it may well be that the state has less right to give the offender what he deserves, although what he deserves does not vary.

    Hasty thoughts on my way to dinner. Hopefully of some help. Michael

12. Hi Gideon – Thanks for the book – great to see this material in print! (And thanks to you, Dave, for coordinating the symposium.)

    Like others, I must confess to being mystified by the idea that a legal reason can have bona fide normative force – i.e., can carry genuine pro tanto weight in our practical reasoning about what we all-things-considered ought to do – unless it is also a moral reason (or maybe a prudential one, but set that aside). Sure, we can think that the fact that x is illegal gives me a legal “reason” not to do it, but for this legal “reason” to be a *real* reason – for it genuinely to count in favor of a certain course of action (not from the internal perspective of the law, but from *our* perspective, as agents), it must also qualify as a moral reason. I’m still not sure why you deny this, Gideon. (Note that this is consistent with the claim that legitimate legal authorities can create new moral reasons, not simply “recognize and rename” old ones, and it’s also consistent with the claim that the state is morally disallowed from punishing wrongful conduct that has not been criminalized, thereby making the category of “criminal culpability” dependent on positive law as you rightly wish it to be.)

    In any case, I want to suggest that your distinctive normative contribution in the book, Gideon, simply does not depend for its justification on any of your controversial positions on the nature of legal reasons. I find myself wholly in agreement that those who participate in the co-authorship of legal rules thereby have an additional *moral* reason to comply with those rules. The moral phenomenology of democratic citizenship seems to corroborate this; we experience pride when the institutions that act in our name accomplish morally laudable ends, and we feel not merely indignation but guilt or even shame when our institutions commit heinous injustice. These are not feelings we experience when other nations’ institutions morally misfire, and they testify to a special moral connection between democratic citizens and the positive law they cooperate to create. (I think Eric Beerbohm’s work in democratic theory is a compelling elaboration of a similar idea.) Your contention that citizens who author their own criminal law, and then break it, are in an important sense bringing their punishment upon themselves is highly attractive, and seems to have exactly the implications for children and other groups that your novel book identifies.

    But I have two worries about this idea in particular. The first, as I’ve already intimated, is that your distinctive claim (about citizen complicity with punishment) seems not to depend on the kind of controversial view of legal reasons you defend in the book (and on which people are pressing you in this blog). Your claim, to my mind, is most naturally expressed as a moral argument about the weightier moral reasons that democratic citizens have to comply with a legal order that they have co-authored. Such an argument need not get entangled in messy disputes about the nature of legal reasons. Do you agree? If so, why do you encumber your own contribution in this way?

    Second, consider a citizen who protests mightily against the enactment of a proposed criminal statute. She does everything she can to explain to her fellow citizens why this statute is a moral abomination. Alas, the statute is enacted. The citizen then commits the crime in question. It seems farcical, or at least counter-intuitive, to suggest that she has brought the punishment upon herself, when she fought so hard to prevent the relevant law’s enactment. The mere fact that she participated in a democratic process is not sufficient to render her complicit in the outcomes of that process. At the very least, she is much less complicit than someone who enthusiastically supported the criminal statute and then broke it. (Likewise, suppose we have a referendum on whether to restrict the religious freedoms of a certain disfavored minority. Members of the minority show up to the polls en masse to vote against the heinous proposal. They lose. Surely they are not complicit in their own subsequent subordination, simply because they participated in the election that led to it.) So the question is: if complicity can vary from citizen to citizen, as it clearly can, shouldn’t criminal culpability, on your account, vary with it?

    I have some further questions, but I’ll leave it there for now. Thanks again to all.

13. Thanks for joining the fray, Mitch! Your distinction between metanormative and bottom-up arguments for my domain-relative view of culpability is really helpful, so thanks for that. I employ both kinds of arguments–or, rather, I have one metanormative argument and several bottom-up arguments.

    The metanormative argument runs like this: To be responsible for an act is for the act to manifest the agent’s dispositions to recognize, weigh and respond to reasons. (I have views about what manifestation is and I eschew the term “disposition” in the book for various reasons, opting for “modes” instead; but set these finer points aside.) To be culpable is to be responsible and for the manifested dispositions to be faults–dispositions to recognize, weigh or respond to reasons improperly. The standard for propriety is accuracy: if the agent’s manifested disposition is to give greater reason-giving weight to a particular fact than it actually provides, then it is a fault, for instance. (Improper weighing is, perhaps, the simplest kind of fault, but there are others.) But accuracy is domain-relative. X really enjoys seeing Y in pain. And so the fact that action A, under consideration by X, would cause Y pain provides X with prudential reason in favor of performance; but, of course, that same fact provides X with moral reason against performance. When X manifests his disposition to give positive reason-giving weight to that fact (not too much, but just as much as it provides prudential reason), is he culpable? It depends whether we are assessing his disposition by appeal to prudential or moral reasons–if prudential then no, if moral then yes. To turn these reflections into a metanormative argument, we need an additional claim: normative assessments, entered into with a potential punitive response in mind, always include responsibility assessment. Part of what it is to normatively assess for a punitive purpose is to assess wrongdoing and assess responsibility for wrongdoing. But the only reason we care about responsibility is because we care about culpability. And since culpability is necessarily domain-relative, we should expect to find appeals to it wherever we are engaged in normative assessment of this kind.

    I’ve mostly described the bottom-up arguments, such as the one about the mala prohibita that you mention, in previous posts. Basically, we have a 2×2: non-criminal/criminal x morally wrong/morally permissible. It is my contention that if we reflect on what the right thing to say is about criminal culpability in each of the four boxes we will reach the conclusion that criminal culpability is relativized to legal reasons and not to moral. The point about the mala prohibita (which is in the criminal-permissible box) was not well-made with the duress example since duress is controversial in its status as excuse or justification or some kind of hybrid. Better: Adam sells individual cigarettes knowing that what he is selling are cigarettes; Eve sells individual cigarettes in the reasonable but mistaken belief they are cigars (which are legally sold individually). The intuition: Eve is not criminally culpable, Adam is. There’s no way to explain this, I don’t think, without asserting that criminal culpability is domain-relative in the way I have described. (Or, anyway, it is best explained that way.) My further contention is that the only way to explain why it is morally more wrongful for the state to punish Eve than Adam is that the state has moral reasons to proportion punishment to criminal culpability, even though criminal culpability involves manifestation of faulty dispositions in one’s dealings with legal, rather than moral reasons. Are you still skeptical, or did your skepticism arise from noise introduced by duress?

    You point out that metanormative arguments, like mine, bring with them a burden to show how the concept of culpability functions in various domains. In most domains, we don’t ordinarily engage in normative assessment with the aim of determining whether a particular punitive response is appropriate, but with other aims. But there’s nothing necessary about this. There are philosophy professors floating around who view question-and-answer sessions after talks to be largely about issuing the punishment predicated on the assessment of the talk. (We all know people like that, don’t we?) We can imagine two such professors arguing over whether the speaker’s egregious performance was excused. What they are arguing about is whether the talk manifested a disposition to improperly recognize, weigh or respond to philosophy-reasons, as it were. Not moral reasons, not legal reasons,. Of course, most of us aren’t like this. We asses the talk without having any potential punitive response in mind and for our purposes in engaging in the exercise of assessment, culpability might not matter. So, yes, there are such things as tort culpability and even fashion culpability. But they matter less frequently than criminal culpability matters because we aren’t usually assessing harm-causing acts and fashion faux pas with an eye towards the possibility of punitive response.

14. And to return again to legal reasons and their relation to moral…Thanks, Michael, for your remarks about this, and yours, Jeff, too. (In my next post, I’m going to pick up the other thread in Jeff’s post about complicity. Spoiler alert: I think what matters is complicity-by-entitlement not complicity-by-causal-influence.)

    The position that Michael says “may tempt” me does tempt me. Actually, it more than tempts me; I’m going to devour the cake whole. (Sadly, I think it likely that no one will share it with me!) But I should issue a disclaimer: the topic of reasons remains, it seems to me, one of the most slippery in all of philosophy, despite its centrality to so many issues in ethics, in philosophy of law, in political philosophy, in epistemology, in the philosophy of science, etc. The topic has become its own sub-discipline and I’m ashamed to say that I am far from being able to claim mastery over it. So, there’s a decent chance that most of what I think about this is false. Doubtless it is too simple. But, nothing ventured nothing gained, so here goes…

    I would describe the first view on Michael’s list slightly differently than he does. A first step, which I find helpful, is to stop talking of “reasons” and start talking of “facts providing reasons”. These two things mean the same thing, but the second locution avoids encouraging a certain kind of reification of reasons that the former term does encourage. To label a particular fact a reason is to assert that there is a certain relation between it and something else; the fact stands in the “providing reason”-relation to the something else. Here’s a fact: an act that X is considering would be painful to Y. Now there’s a question about whether that fact provides a reason for or against the performance of the act by X. When the fact does provide a reason for or against the act, it (at least ordinarily) does so thanks to the existence of some valid norm. If X gets a kick out of Y’s pain, then the fact provides a reason for the act thanks to the prudential norm that says, roughly, “Do what provides desire-satisfaction.” The fact also provides a reason against the act thanks to the moral norm that says, roughly, “Don’t cause other people’s pain.” Valid norms come to exist in a variety of different ways, and there are well-known debates in metaethics and in analytic jurisprudence about how this happens. But it’s clear, I think, that there are different kinds of norms; there are legal- and moral- and aesthetic- and so on. To say that a norm falls into one category or another is to say something about the grounds of its existence and validity. As I sort of indicated in response to Dave’s initial post, I try to remain neutral about what exactly is involved in such grounds. Maybe a particular moral norm like “Don’t cause other people’s pain” exists and is valid thanks to the fact that it has the form of the categorical imperative; I don’t know. I believe that there are legal norms. And some kind of positivist theory about their grounds is probably right. Broadly speaking, legal norms exist and are valid thanks to certain social facts; which ones, I’m not sure, or, anyway, I hope not to be committed to any particular account. What I mean, then, by saying that a particular fact provides a legal reason is this: it provides a reason thanks to the existence and validity of some legal norm. The fact that the person X is considering defrauding is a pawnbroker provides a legal reason against defrauding him. Why? Because there’s valid norm of California law that says, roughly, “Don’t defraud pawnbrokers.” That norm isn’t existent and valid because it has the form of the categorical imperative; it’s existent and valid because of various social facts concerning joint acts performed in accordance with the procedures of the California legislature, and the attitudes of various people towards those procedures, and probably other social facts too. (One of the things I try to do in the book is to describe the social facts that matter to the strength of a legal reason, and not just to the validity of the relevant norm.)

    Michael charges that anyone who adopts this position will ultimately have to say that legal reasons are really just prudential reasons. But they only have to say that if they think that legal norms exist and are valid thanks to the fact that it is prudentially rationally to obey them. But the lesson of the demise of Austin’s command theory of law is, exactly, that this is not why legal norms exist and are valid. In assuming, as I am, that some kind of positivist theory can be made to work I am thereby assuming away the kind of theory that Michael claims to be the only defensible sort. He might turn out to be right; the positivist research program might fail. But it’s hardly a “non-starter”. It’s more than started. In fact, it’s looking pretty good for the good guys.

    Here’s where Jeff’s intervention about legal reasons comes in. He suggests that I don’t need this (crazy, everyone seems to think) view about legal reasons in order to make my argument work. I think he’s right, but with an enormous caveat. As Jeff fully recognizes, my argument does require that the strength of a legal reason for a particular person is in part a function of how much say that person has over the law in virtue of which there is a legal reason in the first place. Jeff’s opinion, I take it, is that that can be true even if the legal reasons are a species of moral. And he supports that idea with the claim that governments can create new moral reasons. To me that seems like alchemy. But whether it is alchemy will depend entirely on what the right account is of the grounds for the existence and validity of moral norms. For instance: if a divine command theory of morality were correct, then it would be possible for government to make moral norms only if the government issues divine commands. Not likely. But if the correct account, a different account from the divine command theory, of the existence and validity of moral norms is such as to support the idea that the strength of a legal reason can vary with an agent’s say over the law, then, absolutely, I do not need my view about legal reasons. When I step back from it, it’s hard for me to say for certain which thing is harder to accept: my view of legal reasons, or Jeff’s view of moral. But Jeff’s remarks have helped me see that either could serve to support my main line of argument about why kids deserve leniency.

15. Thanks, Jeff, for bringing up the issue about complicity in one’s punishment. Mitch described my view of culpability as heterodox; so is my view about complicity. The book sketches the following idea, but I may write more about it soon. The idea is that there is a kind of complicity that you might call “complicity-by-entitlement”. Say that X does something bad. Y is a complete stranger with no right to try to exert influence over X’s behavior. Z, by contrast, is entitled to do that. I believe that the following comparative claim is implied by only what has so far been described: Z is more implicated in X’s wrongdoing than is Y. Put another way: Say someone looks at Y and Z accusingly and says, “Look what X just did!” Y can defend himself fully by saying, “It had nothing to do with me!” Z, by contrast, cannot defend himself that way. X’s wrongdoing did have something do with Z; Z was entitled to exert influence over it. Parents should be chagrined when their children do wrong, and they should be more chagrined than strangers who have no entitlement to exert influence over the children. Similarly, US citizens should be more chagrined about the atrocious effects of the recent, brief but terrible, “zero-tolerance” border policy than should citizens of Mexico or Canada. US citizens are implicated in the US government’s behavior thanks to their entitlement to exert influence over it.

    Notice, and this is the important point, nothing whatsoever has been said so far about causal influence. Causal influence is, of course, a central type of complicity. It’s a central source of on-the-hookedness, as it were. But entitlement is another source. My claim is that the strength of a person’s legal reasons varies with her degree of complicity-by-entitlement in that in virtue of which there is a legal reason for her. It does not vary with her degree of complicity-by-causal influence. This is why the person who votes against the policy and the person who votes for it have a legal reason of exactly the same strength (ceteris paribus) for following the policy; they have equal entitlement to exert influence over it, despite the fact that they differ in the degree of causal influence they had over it.

    It could be that “complicity” is the wrong word for what I am after. I’m not sure. But what matters is the phenomenon, rather than the word, and I think the phenomenon that I’m tagging with the phrase “complicity-by-entitlement” is a real one and common.

16. Let me push back against the “complicity-by-entitlement” claim, Gideon. I’m not sure I buy it. Or, more weakly, it seems a lot is going to hang on the nature of the entitlement and pretty deep questions about political/legal authority.

    Let me put the challenge this way: to my mind, entitlement to influence without any real causal influence looks like sham influence to me. Especially if that entitlement is a bit of a sham itself. This is how I’m starting to think of political influence in US. Suppose a minority group was given the vote but the voting structure was setup to ensure that they would have almost no causal influence over the outcome. A member of that group would justifiably chafe at the idea that she still has an “entitlement” to influence the law. And, similarly, such a person would arguably have grounds for not being anymore chagrined than a Canadian at the result.

    Now, one way of interpreting my challenge is as asking you to justify a whole theory of democratic state authority. But I guess I am asking for a glimmer of that, since I, at least, am rather skeptical about the actual political power of lots of groups, certainly individuals, which is the level at which the claims seem to matter, since it will, after all, be individual kids getting the break (even if that break is categorical in nature).

17. Thanks for your reply, Gideon. It’s very helpful. I have questions about the metanormative argument that I’ll hope to get to in a later post. Here I’ll react to your bottom-up argument.

    To answer your question in the paragraph that starts “I’ve mostly described . . . ,” yes, I’m still skeptical. I’ll try to explain. To start, I’m a little worried about the structure of much of what you say in this paragraph before your “further contention.” I’m imagining a critic who doesn’t own the concept of criminal culpability and who doubts its value. I’m not sure how that critic can accept your invitation to reflect on what is “the right thing to say about criminal culpability” in different contexts. For the same reason, I don’t think you can appeal to any shared intuitions about whether Adam or Eve is “criminally culpable.” I think that criminal culpability has to be part of your explanans, but not part of the explananda. (Of course, it is true that, in most of the jurisdictions we’re familiar with, Eve is not “criminally responsible,” whereas Adam is. But this is just the upshot of well-understood doctrine; this doctrinal fact isn’t a puzzle in need of explanation.)

    If what I’ve just said is right, then it follows, I think, that your “further contention” is the whole ball of wax for this particular bottom-up argument. I agree that it’s morally more wrongful for the state to punish Eve than Adam. But I don’t think that the only way to explain that fact is yours. One explanation is that Adam is more blameworthy than Eve. Eve isn’t blameworthy at all. But some people will believe (given certain background assumptions) that Adam has some moral reason to obey the law on the sale of cigarettes notwithstanding that the law forbids conduct that morality permits. A second explanation is that, even if Adam, like Eve, isn’t blameworthy, he differs from Eve in that he voluntarily assumed the risk of being charged and convicted of a crime, whereas she didn’t. It seems very plausible to me that the state acts morally wrongfully (a) in punishing people who aren’t blameworthy, and (b) in punishing people who lacked a reasonable chance to avoid running afoul of the criminal law (see Hart). The state commits both wrongs when punishing Eve, but only (a) when punishing Adam. I’m not saying anything original in suggesting these alternative explanations, so I wonder what you think I’m missing.

18. On a different note, Gideon, I am in very substantial sympathy with almost all that you say in your reply to Michael. Indeed, I’m working just now on what reasons are, and on how the norms delivered and maintained by artificial normative systems (law, fashion, baseball, etc.) are metaphysically determined. (Incidentally, you should know that it’s not clear–at least it’s not uncontroversial–“that there are different kinds of norms; there are legal- and moral- and aesthetic- and so on.” “One-system theorists” such as Dworkin and Scott Hershovitz are challenging precisely that. So is Mark Greenberg, though I think his view is somewhat less radical.)

    Anyway, as I say, I’m largely with you, even up to your counsel that we should “stop talking about reasons.” But I was surprised to see you say it, because I’ve been struck by the extent to which your analysis depends precisely upon reasons-talk. More significantly, I worry that you don’t take your (warranted!) wariness about “reasons” far enough. You propose that, instead of talking about “reasons,” we should talk about “facts providing reasons.” But because many people believe that reasons are facts, and that only facts can be reasons, it’s not clear to me that your preference is substantive, not just notational. So maybe you could elaborate?

    Much more importantly, I’d urge you to consider jettisoning or sidelining reasons-talk more thoroughly. When I think about multiple normative systems or domains, I’m often put in mind of Philippa Foot’s “Morality as a System of Hypothetical Imperatives.” Putting aside her view of morality (which I happen to think was basically right, although she abandoned it), I think that she presents in clear fashion a widely held picture of multiple normative systems. Perhaps you think so too? In any event, recall that she always speaks in terms of domain-specific oughts and shoulds. She doesn’t speak of legal reasons or etiquette reasons. For Foot, reason is a domain-independent normative notion. I think that’s the more natural vocabulary. We naturally speak of what you are “legally required” to do, or what you are “legally obligated” to do, or what you “legally ought” to do, and what our “legal rights,” “legal duties,” and “legal powers” are. But only (some) legal philosophers speak of “legal reasons.” I wonder whether you might at some point try translating your substantive view into terms that don’t depend upon the notion of “legal reasons.”

    To return to the start, though, you’re right to resist Michael!

19. To disclose a bit more in response to your query about complicity, Matt, in the book I equate having a say over the law with a trio of conditions. The first two are entitlements: the entitlement to exert influence over the law, and the entitlement to be free of a set of obstacles to the exertion of that influence (although the question of where the boundary of that set lies is non-trivial). The third condition is an empirical one: actual absence of the obstacles that one is entitled to be free from. I take your (pessimistic, but maybe not unduly so) view to be that a lot of people today, in the US and elsewhere, are not free of the obstacles to exertion of influence that they are entitled to be free of. I don’t disagree, although it’s an empirical question how wide-spread and systematic the problem is. And there’s no way to answer the empirical question without an account of what obstacles people are and ought to be entitled to be free of. (In an earlier post Doug mentioned Citizens United in passing. This is the issue that case touches.)

    The observation that a lot of adult citizens lack a say over the law, because they aren’t free from obstacles to the exertion of influence over the law that they are entitled to be free of, has the implication, when conjoined with my argument for leniency towards children, that there are a lot of people who, for principled reasons, are owed leniency when they commit crimes. (I discuss this at some length in a late chapter of the book.) This is an implication that I’m willing, even eager to accept, with the qualification that those who fall into this category need to show that they do and shouldn’t, like children, benefit from a blanket policy of leniency.

    But here’s the point I really want to make in response to your last post: we can’t extend my argument to other groups who are diminished in their say over the law unless we accept that their diminished say over the law results in diminished complicity with legal demands. That is, you have to accept my claim that there’s such a thing as complicity-by-entitlement. Saying that there is such a thing is independent of the question of who has the kinds of entitlements involved in having a say over the law, and who doesn’t.

20. Thanks for following up, Mitch. Let me post first about your remarks about the bottom-up argument we’ve been discussing. First, I think you’re right that intuitions about criminal culpability–which hypothetical people have it and which don’t–are flimsy things on which to rest much. And if that was all I had in support of my interpretation of the examples, like the mala prohibita one we are discussing, that wouldn’t be much to go on. But they aren’t nothing, and they corroborate what other arguments support. Second, and more importantly, let me explain why I think the alternative explanation for why the government is doing something morally worse to Eve than to Adam in the example (Adam: knows they are cigarettes, sells them individually in violation of law, Eve: reasonably thinks they are cigars, which are legal to individually sell, sells them individually) is inadequate. Your first suggestion: Adam is blameworthy, Eve is not. But that’s not true. Nobody is worthy of blame for doing something permissible, whether they do it knowingly or not. Blame is appropriate only where there’s moral wrongdoing and there isn’t any here. Your second suggestion: Adam is forewarned of the punishment, Eve is not. First, that’s not necessarily true. We can’t conclude from the fact that Adam knows they are cigarettes that he knows that it’s illegal to individually sell them. Even if he doesn’t know that, the government’s punishment of him is still more morally wrongful than that of Eve. Second, there’s no reason to think that the government is on firmer moral ground in doing morally objectionable things to people after warnings than without warnings. If they are, then the government enjoys a privilege here that individuals don’t. “I warned him!” is no defense to a charge of assault, nor even a ground for mitigation.

    Since we’re talking about this, let me pick up a left-over thread for Michael’s post. Michael suggests that the reason the government’s punishment of Eve is worse than that of Adam is because of the truth of the following counterfactual: If it were morally wrongful to sell cigarettes individually, then Eve would be reduced in blameworthiness and Adam would not. There are well-known problems of imaginative resistance with counterfactuals of this kind. It’s not easy to imagine that permissible behaviors are wrongful, or vice versa. But setting that aside, I think the only reason to think that this counterfactual is morally significant is that it’s true only if the agent is criminally culpable in my sense. What the truth of the counterfactual shows is that if the reasons in question were moral and not legal, she would be morally culpable. That’s true when she’s criminally culpable, and that’s what’s morally significant. So, as I see it, Michael is tacitly accepting, rather than rejecting, the claim at issue.

21. On your other most recent post, Mitch, concerning Foote and the diversity of normative domains: I don’t mean there to be more than a notational variant between “F is a reason not to C” and “F provides a reason not to C”. I think they mean the same thing. The former way of speaking, however, makes it appear that you are asserting contradiction when you say that, for instance, “The fact that C-ing will cause Y pain is a reason for X to C (because he likes watching Y writhe)” and “The fact that C-ing will cause Y pain is a reason for X not to C (because nobody should cause others pain)”. But this isn’t a particularly important point, I don’t think. Your more substantive point–namely that it should be possible to refigure my view without mention of reasons–is surely correct; thanks for that. I’ve inherited that way of thinking about responsibility and culpability from reasons-responsiveness theories, such as John Fischer’s, which I think get things mostly right about moral responsibility (although they are in need of extension to the legal case, which is part of what I think I’m doing). But I bet you’re right that the ideas here could be expressed in a Foote-ian way without loss.

22. Thank you, Gideon, for this wide-ranging and thoughtful work. I’ll offer a couple thoughts here that pick up on your initial comments above – and then I’ll end with a question.

    I’ll focus on a version of the selling single cigarettes example. To avoid controversies re: selling carcinogens, undermining regulatory systems designed to protect health, fair markets, etc., I’m going to switch your example to something more obviously innocuous: whistling a happy tune whilst walking down the street, with no one within earshot (hereinafter, simply, whistling).

    *Stipulate whistling is not morally culpable.
    *Stipulate whistling should not be a crime anywhere.
    *Ex hypothesi, whistling is a strict liability crime in state X and no excuses are recognized in the law of state X for the crime of whistling.

    Compare two cases:

    Whistler_1: W1 wants to whistle, and does so, of his/her/their own free choice.
    Whistler_2: W2 does not want to whistle, but does so because he/she/they are threatened with death if he/she/they do not whistle.

    Aside from being wrong about making whistling a crime in the first place, is state X even more wrong because it refuses to recognize an excuse in defense of Whistler_2?

    Yes, because condemnatory responses are only properly targeted at culpable wrongdoing (that is, wrongdoing which is both unjustified and unexcused).

    Neither whistler is morally culpable – and thus neither should be treated as culpable.

    Both whistlers are wronged by the state by being told (by the state, with threat of sanction) not to whistle. (Let’s call this a prohibitory wronging.)

    W2 is wronged by the state in another way – because, even if W2 had committed a moral wrong, W2 would have had an excuse, and thus should not be condemned/punished. (Let us call this a condemnatory wronging.)

    So, both W1 and W2 suffer prohibitory wrongs in virtue of the general prohibition against whistling, but W2 suffers the additional wrong of being condemned under circumstances where W2 had an excuse for doing what he/she/they were told not to do.

    We agree that neither W1 nor W2 is morally culpable at all. Your view is that “the only explanation for the moral pressure the state is under to recognize an excuse in [W2’s case] is that there is moral pressure to punish proportional to criminal culpability.”

    Is that the same as saying that the state should not arrest, prosecute, condemn, and/or punish W2 because doing so would inflict an additional wrong against W2? That is, W2 would suffer a condemnatory wrong at the hands of the state, above and beyond the prohibitory wrong that both W1 and W2 (and everyone else in that jurisdiction) suffered in virtue of the original anti-whistling law?

    I gather that you think the answer to my question is “no” – you think legal reasons (grounded in the mere fact that State X prohibited whistling) add weight to the reasons why any W (or any W subject to State X’s law) has reason not to whistle.

23. This is all extremely helpful, Gideon. I’d like to make just one further comment about complicity (though feel free to refer me to the book, which I admittedly haven’t completed, if this is sorted out there). One worry about your entitlement account of complicity is that it cannot play one of the crucial roles that complicity tends to play in our moral thinking. To be complicit with an injustice is to be involved, somehow, in the wrongdoing of others. To claim that an agent is complicit is to make a claim about that agent’s objectionable relationship to the principal wrong in question — paradigmatically, by causally contributing to its occurrence, either by commission or omission, but I grant that there can be other ways to relate objectionably to the wrongdoing of others (e.g., sharing the intention to bring it about, even if one is causally inefficacious).

    How does the entitlement account of complicity fit in the context of these kinds of complicity judgments, which are by far the most common? Based on what you say above, here is one possibility: One is complicit in the wrongdoing of another just in case one is entitled to influence the conduct of another. But this is very counter-intuitive. A more intuitive interpretation of the entitlement account is this: One is complicit in the wrongdoing of another just in case one is entitled to influence the conduct of another, *and yet refrains from doing so* (or, to put the point another way, exercises the entitlement wrongly). Suppose Alfred has the entitlement to try to persuade Jim not to commit some crime, but Alfred fails to exercise that entitlement (or, alternatively, exercises it by encouraging Jim to commit the crime). Either way, Alfred is clearly complicit in Jim’s subsequent wrongdoing. But now consider Beatrice, who also has the entitlement to try to persuade Jim. She rightly exercises it for good, trying her best to convince him not to commit the crime. Is Beatrice still complicit when Jim nevertheless commits the crime? Your account seems to say yes. That is what I find counter-intuitive. It is very odd to say that we should view Alfred and Beatrice as equally complicit, and therefore equally liable to blame.

    Note that I am not slipping into a crude causal view here; we can stipulate that Jim is so stubborn he couldn’t be talked out of the crime, rendering others’ persuasive efforts causally futile. My case still uses the entitlement account, since only those with the entitlement to (try to) influence others are even eligible candidates for complicity. (Note that your notion of influence still has a causal dimension, raising a question about just how much it actually eschews causality, but set that aside.) My suggestion to you is that, on the most plausible version of your view, at least as applied in this context, entitlement is necessary but not sufficient for complicity. If one exercises one’s entitlement in the right way, disowning and/or working to prevent the relevant wrongdoing, one can free oneself from any concomitant blame and guilt, since one has done everything one could reasonably be expected to do. Hence the salience of “Not in Our Name” posters at political protests.

    Perhaps you think different accounts of complicity are appropriate for different contexts, which is, of course, possible. Perhaps my variation of the entitlement account, which focuses not merely on one’s entitlements but on one’s exercise of them, rightly figures in our theories of accomplice liability, whereas your variation, which focuses on entitlement simpliciter, is the right theory when reflecting on the strength of our legal reasons to obey laws generated through a process in which we participated. Given your interest in developing your thinking on this topic, it would be welcome to hear more about what determines which sort of complicity standard is the one applicable for a particular domain, and why.

24. Thanks so much for jumping into this, Michelle! My answer to your question is “yes” (not “no”, as you suggest). Here’s why. A first tiny point is that it doesn’t follow from the fact that W2 is excused that the state commits a greater wrong in arresting and prosecuting her than it commits in arresting and prosecuting W1. It can be permissible to arrest and prosecute people who are excused (or justified, or have some other full defense). I also think we’re on the same page in recognizing that what you call the prohibitory wrong (a useful term, thanks for that) is the same with respect to both W1 and W2. So let’s focus on the condemning and punishing of W2 in contrast to W1. I think it’s morally worse for the state to do the former than the latter (although both are bad). W2’s act does not manifest a disposition to improperly recognize, weigh or respond to the legal reason not to whistle, while W1’s act does manifest such a disposition. Equivalently: W2 isn’t criminally culpable while W1 is. But my primary point here is comparative: there is a moral difference between what the state does to W1 and what it does to W2 and the best way to explain that difference is by appeal to the idea that the state has moral reasons to proportion punishment to criminal culpability. This way of putting it (“there is a moral difference”) does not appeal to the idea that the punishment and condemnation of either W1 or W2 is a wrong. I think it is, but my argument here does not depend on that. The idea that criminal culpability is relativized to legal reasons would follow just as well if what the state does to W2 is less morally laudatory than what it does to W1.

    This brings us, then, to the bite of your question, which I would put like this: If there’s a reason (I claim a legal one) for any W to refrain from whistling, then doesn’t it follow that the state has done no wrong in condemning and punishing those who whistle? I don’t think that does follow quite. Consider the following reasoning:

    a. If the state punishes whistling it ought to give greater punishments to unexcused whistlers than excused.
    b. The state does punish whistling.
    c. So: The state ought to give greater punishment to unexcused whistlers than excused.

    I think this is a good line of argument. And it’s conclusion is true only if “excused” is interpreted in such a way as to involve the negation of criminal and not moral culpability, which in turn involves granting that there’s a legal reason not to whistle (when the state prohibits whistling). But it doesn’t follow from the argument’s conclusion that the state ought to give any punishment at all to whistlers, or that it is not wronging whistlers in punishing them. Whether that’s so will turn on the outcome of the general debate over the justifiability of punishing mala prohibita offenders. My main point–the separation of criminal and moral culpability–should survive any outcome of that debate.

25. A very quick addendum to my post in response to Michelle: One might think that it follows logically from “The state ought to give greater punishment to unexcused whistlers than excused” that “The state ought to give some punishment to unexcused whistlers.” But we know from reflection on the paradox of gentle murder (Forrester’s paradox) that that isn’t so. That is, such an inference is barred in a defensible deontic logic. Something about deontic logic blocks the conclusion that “You ought to murder someone” from the premises “If you murder someone, you ought to murder him gently” & “You murder someone”. Whatever it is about deontic logic that blocks this inference is what matters to the case at issue also.

26. On a new thread, there was an exchange between Doug and Gideon that interested me. This was Doug’s view that Gideon’s general account of culpability was incomplete until Gideon showed why the factor that was culpability constituting for Gideon — non-responsiveness of the actor’s will to the facts that give moral reasons for action — was itself culpably caused by the actor.

    I see this as part and parcel of a view Doug generally has about desert-bases, what might be called the “turtles all the way down” view. It is a reverse tracing strategy: whereas the tracing strategy for responsibility assessments regards it as sufficient for responsibility if one can trace one’s otherwise excusing condition to some culpable act, Doug regards such traceability as necessary if one is to be justly held responsible for an otherwise culpable act.

    The turtle syndrome shows up in Doug’s Ignorance of the Law book too. There it takes the form of requiring that an actor know not just what he is doing but that the constitution of culpability by knowledge be extended to include knowledge also that what he is doing is morally wrong. (It also shows up in Phil Pettit’s demand that we not only have the capacities to act in light of our choices and to choose in light of what we most desire, but that we also have the capacity to choose what we desire.)

    Gideon responds, rightly it seems to me, that the turtles all the way down approach leads to an infinite regress. Doug denies this but I don’t see how. At some point the compatibilist about responsibility has to believe that some natural property possessed by the actor gives rise to blameworthiness even though that factor was itself not culpably brought into existence by the actor– where “culpably brought into existence” here means that the actor did some earlier act itself having that culpability constituting property. Maybe’s Doug’s denial only comes to this: although one has to stop somewhere (on pain of infinite regress), that is no argument for stopping where Gideon stops.

    Fair enough. But isn’t Gideon’s the intuitive stopping point? That Simmons’ will was not spoken to by the reasons not to do what he did — reasons based on the horrorible harms he was knowingly and intentionally causing — is all I would need to find him extremely blameworthy. It seems to me that to ask that Simmons himself be culpable in producing such a defective will as that, is to start down the road to a skepticism about responsibility that is as hard as hard determinism.

    I’d love to hear from the patient here, or from others who are also overly fond of turtles.

27. Thanks, Jeff, for your very perceptive follow-up about complicity. I’m afraid that the issues are not sorted out in the book. I say some things, but I mostly rest on the undefended claim that there’s such a thing as complicity-by-entitlement. More, definitely, needs to be said.

    I want to push back a bit against your suggestion that you don’t find the kind of complicity I’m talking about without a failure to exercise an entitlement to exert influence. To be clear, I want to grant that there may be some forms of wrongdoing by Y that X is complicit in by entitlement only if X fails to exercise her entitlement to exert influence over Y. I don’t want to exclude this possibility. But what I care about is that such a failure is not necessary for complicity in the authoritativeness of a legally authoritative act. This is dark, I’m sure, so let me try to explain: Say that a legal official performs act A and that act is an exercise of the official’s authority. The act has many properties, but the one that I care about is that it is an exercise of authority. In general for a person to be complicit in another’s act he must bear some kind of relation, of which causal contribution is just one example, to one of the features of the act. What kind of relation need he bear to the legal authoritativeness of A to be complicit in A? Well, my claim is that it’s enough that he has entitlements thanks to which A is an exercise of legal authority, whether or not he exercises those entitlements. Let me say that again: Question: why is A-ing an exercise of legal authority? Answer: in part because citizens like X have a say over the law (where that’s understood as consisting most centrally in entitlements of certain kinds). That, I think, is enough of a relation between X and the legal authoritativeness of the legal official’s A-ing to make X complicit in A-ing. So entitlement all by itself, even when unexercised, suffices for complicity whererever we have governments that have authority over citizens thanks to the fact that those citizens have a say over the law. In short, what I need is that entitlement can suffice for complicity in this one way, even if it needs to be coupled with more–such as a failure to make an effort to influence–in order to constitute complicity in other kinds of conduct.

    (By the way, I think that the role of entitlement in constituting complicity is important in many contexts in criminal law. I have a paper about criminal liability for possession where I claim that part of the reason why there is no violation of the act requirement in such prohibitions is because of the entitlements of possessors, even when they are not owners, such as protection against theft. Thanks in part to those entitlements, I claim, not dispossessing yourself is an exercise of agency on a par, for responsibility purposes, with voluntary action or intentional omission. That’s just one other example.)

28. On Michael’s most recent post about turtles: Hear hear! I’m also anxious to hear what Doug, or others, have to say about this, if anyone’s inclined to weigh in. I move past this issue fast in the book, but I do require rejection of the kind of view that Doug, among others, espouses.

29. Thanks, Gideon – this is very helpful. I certainly agree that it doesn’t follow from a legal reason not to whistle that the state has done no wrong in condemning and punishing those who whistle.

    I think we’re in agreement re: Principle of Legality issue. Just to confirm, consider this rephrasing of your original post. You wrote:

    (1) “I take there to be moral reason for the government to punish to a degree that is proportional to criminal culpability, even though that might mean issuing a punishment that is entirely out of line with moral culpability.”

    Am I right to think you accept the following rephrasing?:

    (1.1) “[If the government is going to issues punishments that are entirely out of line with moral culpability], [then there is] moral reason for the government to punish to a degree that is proportional to criminal culpability, even though that might mean issuing a punishment that is entirely out of line with moral culpability.”

    If so, then we’re in agreement – and I find your account a helpful way of thinking about the moral force of the Principle of Legality.

    My rephrasing simply makes explicit the stipulation that certain action by the government (issuing punishment out of line with moral culpability) is inevitable, ex hypothesi. That stipulation is important to the formulation, because the inevitability of that action by the government is an essential part of the reason for the original prescription (the prescription that the government should punish to a degree that is proportional to criminal culpability).

    Just as in Forrester’s paradox, the fact that there actually will be a killing is an essential back-ground feature of the prescription that there should be a gentle killing. If we leave out these kinds of stipulations, it starts to sound as if one is suggesting that the mere fact that ϕ-ing is made a crime gives the government moral reason to arrest, prosecute, punish folks for ϕ-ing (the Sessions-Doctrine?) – and that would be a very odd suggestion indeed, especially coming from an unflinching positivist.

    Thanks for the book and discussion. I’ve learned a lot. (Or I’m still confused. …awaiting your response to see which you think it is.)

30. I indeed hold that the factor that constitutes culpability for Gideon — non-responsiveness of the actor’s will to the facts that give moral reasons for action — must itself involve culpability in the actor. If he lacks culpability in failing to respond to reasons, he is not culpable—or at least not fully culpable—when he acts. I certainly don’t want my position to give rise to a regress, much less to an infinite regress. But one extra turtle is a far cry from “turtles all the way down.”
    I am not entirely sure how to characterize the extra turtle that is needed before the actor is fully culpable when his will fails to respond correctly to the facts that give him moral reasons for action. So let me begin by describing my motivation for believing that something more is needed. As Michael notes, the kinds of case that convince me indeed came to my attention when thinking about the difference in culpability between two wrongdoers who commit the same wrong, when the first but not the second is unaware of the norm he violates. When the ancient Hittite enslaves Joe who has been conquered in battle and believes persons so conquered forfeit their rights not to be enslaved, the Hittite fails to respond correctly to the facts that provide reasons not to enslave. When I enslave Joe and believe no such thing about Joe’s rights but can’t be bothered to do the hard work of picking cotton on my plantation, I am convinced my culpability is greater than that of the Hittite. Since both of us fail to respond to the same facts that give us the same moral reasons not to enslave, there must be more to culpability than a mere failure to respond correctly to the facts that provide moral reasons for action.
    What is the added factor that makes me more culpable for enslavement than the Hittite? Crudely, I know better than to do as I did. When I tell my grandson that failing to share his toys with his sister is wrong, and he realizes I am right but proceeds to do the same thing again and I catch him, we all think he is more culpable when he commits the same wrong and knows better than when he did not. Don’t we all think this? I am aware that some philosophers deny it, but that denial seems radically unintuitive to me. I sometimes suspect that these philosophers deny it in order to cling to a theory that the example is designed to undermine. When these same philosophers catch their grandsons the second time, don’t they blame them more than they did the first time? (In case this case comes to close to Gideon’s turf in using kids, I can cite similar examples that involve adults).
    Like Michael and Gideon, I locate culpability in choice. What is the content of the choice that makes one fully culpable? Full culpability requires that I choose to commit a wrong (both de re and de dicto). I am aware that few philosophers seem to agree with me about this. Admittedly, he knows better is not a very sophisticated label for the added culpability of the knowing as opposed to the unknowing wrongdoer. But even if the view needs to be refined, it certainly does not lead to a regress. It simply adds a further dimension to culpability that quality of will theories seem unable to capture. It’s the bottom turtle on the pile.
    I am aware of some weaknesses of my position and some moves that can be made to resist it. But I remain convinced that quality of will theories leave something out that is crucial to culpability.

31. What a pleasure it is to be understood, Michelle, not to mention to have a comrade in arms! That’s a very helpful rephrasing and accurately expresses my position.

32. I fear, Doug, that if we aren’t careful we’ll spend several posts repeating large sections of the free will debate. There is really no contributor to that debate–contemporary, modern or ancient–who hasn’t thought about the basic issue that Michael invited us to pick up. Practically no one who has thought it through at all has failed to offer some strong opinion about which views are “radically unintuitive” (to use your phrase) and which are conceptually incoherent, and so on. But let me circle the issue back to the one that I use my theory of culpability to address in the book–namely the diminished responsibility of children.

    It’s pretty common for me to hear the following from non-philosophers when I pose the question of why kids are owed a break when they commit crimes: “Because it’s not their fault that they are like that!” This is to assume that kids manifest in their conduct the same problematic dispositions to recognize, weigh and respond to reasons as adults, but to further claim that while adults are at fault for having and manifesting such dispositions, kids are not. The claim seems to be that your last turtle is somehow missing in kids, but not in adults. I don’t think there’s any need for this extra turtle, or an extra condition of fault for one’s problematic manifested dispositions. I don’t think it’s necessary even in the modest form in which John Fischer and Mark Ravizza proposed some time ago (their condition of “taking responsibility for” one’s problematic reasons-responsive mechanism). But I also don’t think this needs to be insisted upon in order to see what’s wrong with this particular way of explaining why kids are owed a break. What’s wrong with that is that even if there is an extra condition, an extra turtle in the stack, there is no reason whatsoever to think that adults meet it and children don’t. Or, put another way, children would be owed leniency even if they did meet that condition on top of the other conditions for culpability that I’ve outlined. In your NDPR review, I took you to agree with me about this when you said that my negative arguments in the book–my attacks on alternative ways of explaining why kids are owed a break–were successful, in your mind. But I’m sure you’ll correct me if I misunderstood you.