In the previous post, I noted that there are two important aspects of Scanlonian blame — relationships and meaning — and that extending Scanlon's account to cover criminal blame (punishment) was problematic with respect to the former in virtue of the fact that the necessary sort of relationship (whose impairment prompts blame) was missing in the citizenry/legal case. I want to focus here (in much briefer terms) on the second aspect, meaning.
For Scanlon, what blame responds to is the meaning of what the blamed agent did, which is itself a function of the reasons the agent acted on. These reasons will reveal something important about the agent’s attitude with respect to the blamer as well as the relationship they bear to one another. Meaning, for Scanlon, is crucially distinct from permissibility. Permissibility is simply about the question “What may I do?” and so focuses exclusively on what is actually licensed by the principles we ask one another to use in moral deliberation. What this means is that one may not be blameworthy for performing an impermissible action (as when the reasons one acted on don’t reveal any impairment in the relevant relationship), and one may be blameworthy for performing a permissible action (as when one’s reasons for doing so nevertheless reveal an impairment). Blame and blameworthiness are entirely about meaning: we are tracking the agent’s attitudes as reflected in his conduct.
Criminal punishment, on the other hand, is almost entirely about permissibility: we are simply tracking and responding to the agent’s conduct. First, one could never be punishment-worthy for performing a criminally permissible action: one’s conduct has to be illegal for one to deserve punishment. Second, and perhaps more controversially, one will always be punishment-worthy for performing a criminally impermissible action. Now one might object to this claim by pointing out that justifications or excuses count as ways of legitimately getting one off the criminal hook for impermissible actions, and these are precisely dependent on the agent’s reasons (so they would be about meaning). As Victor Tadros puts it, “[J]ustification depends upon motivation,” so that the “action of a defendant who breaches the criminal law for the wrong reasons is not justified even if he acts knowing all the considerations which make his act justifiable” (Criminal Responsibility, p. 280). And while an excuse may not mitigate all grounds for criticism, it ostensibly serves to remove attributability of the breach to one’s character for purposes of criminal responsibility in the first place, so that one isn’t criticizable for the illegal conduct given that one didn’t intend to do it, say (perhaps one was under duress).
I am suspicious that either justifications or excuses in the criminal law are really about the agent’s own reasons (and so about meaning), but that point actually does not matter for our purposes, and this is because justifications and excuses are both kinds of defences, whereas defendants are ultimately held criminally responsible only for offences, and these are primarily (albeit perhaps not entirely) about impermissible conduct, not meaning. Offences are actual public wrongs, types of conduct against which there are legally sanctioned reasons. They are crimes, in other words, with which defendants are charged. Defences, if successful, offer legitimate grounds for acquittal, where these “are compatible with the defendant’s conceding that the offence charged was indeed committed” (John Gardner, Offences and Defences, p. 141). So in offering a justification, I effectively say that, while I did commit the crime in question, there were good reasons for me to do so, reasons that were in some sense stronger than the legally sanctioned reasons against that conduct. And in offering an excuse, I say that, while there is some sense in which I committed the crime (my finger pulled the trigger, say), nevertheless that criminal action is not properly attributable to me (as perhaps I was provoked into pulling the trigger because I caught him with my wife and was enraged). In the former, I do indeed have good reasons to commit the crime; in the latter, I am justified in believing that I have good reasons, although I don’t.
Given the possibility of successful defences, we cannot say that a defendant warrants conviction only in virtue of having performed a criminal offense; we must say in addition that such warrant also depends on the defendant having no successful defence. But now we have a definition of criminal impermissibility, the violation of which is sufficient for punishment: offence without defence. The defendant’s actual reason(s)—the meaning of his action—is found, if at all, only in his defence(s). But if criminal impermissibility delivers grounds for punishment, and such impermissibility depends on the absence of (successful) defence, then the grounds for punishment may be delivered precisely without the sort of meaning Scanlon takes to be essential to blame. Blaming in the criminal realm is simply a response to people who have violated the law without justification or excuse. If their reasons for acting don’t acquit them, then they are just irrelevant to the sort of response the state appropriately has. In the criminal justice system, punishment is not a response to the defendant’s attitudes at all; it is instead a response solely to what he has done. In Scanlon’s account, however, blame is solely a response to the blamee’s attitudes; what she has done, if anything, merely reveals that real target to the blamer. Scanlonian blame, therefore, cannot be extended to holding responsible in the criminal realm. Punishment does not count as an instance of Scanlonian blame.
Distinctions between various crimes, such as manslaughter, first degree murder and second degree murder are a matter of the agent’s attitudes. This strongly suggests to me that criminal penalties are, among other things, expressions of blame the sense I describe.
Thanks, Tim. I agree there are these instances, and I deliberately made space for them (I said, e.g., that criminal punishment is “almost entirely” about permissibility — I had in mind, in addition to the cases you cite, hate crimes). I believe these to be genuine exceptions, though. Most of the criminal law isn’t articulated in the way that the first-/second-degree distinctions are, and the aim of criminal trials isn’t, most of the time, about tracking the attitudes of the defendants, i.e., the meaning of what they did. I take the cases you cite, then, to be exceptions to the general approach, so even if they are indeed tracking meaning, most criminal punishment doesn’t involve that sort of response, and insofar as this is true, most criminal punishment isn’t an expression of Scanlonian blame.
Hi Doug,
I no little about the law, but wonder about crimes of neglect and negligence? These seems to centrally involve epistemic and moral assessments of the attitudes and psychological capacities of the agents in question.
Brad: Would that I were Doug, so as my faults would be his.
I think that neglect and negligence help support my case. The reason is that, when trying to determine mens rea in such (criminal) cases, the idea is to try to figure out what the reasonable man would do. This strikes me as straightforwardly about permissibility, not meaning. In moral cases of negligence, though, the (in)actions for which we blame are incidents in which we judge that the targeted agent failed to have a sufficient level of regard for us (i.e., it is about meaning). I suppose the distinction may depend on whether or not we take the reasonable man test in criminal law to be an epistemic or metaphysical matter (i.e., whether or not it provides evidence for, or makes it the case that, the defendant violated the law), and I guess my inclination is that it’s the latter, so I’d like to hear more if you think it’s (merely) the former.
I should add to Tim that my point here isn’t that your view of blame is wrong or incomplete at all. Rather, for a variety of reasons, I’ve come to believe that criminal and moral responsibility involve very different elements, and, in particular, that criminal and moral blame are just different conceptions of blame. Indeed, I think your account of the moral conception is, for the most part, right (with some exceptions; see http://www.jstor.org/pss/10.1086/659003 for those). But here’s one straightforward reason to think there’s a difference: the aim of criminal blame is, necessarily, to sanction (e.g., to punish/fine); the aim of moral blame isn’t. It rather involves a response rendered appropriate by the relevant relationship impairment, which varies wildly (depending on the norms and standards constitutive of the relationship), but doesn’t necessarily involve sanctions at all.
David, Apart from strict liability crimes (which are rare) one of the elements of an offense is that one be in a relevant mental state, whether it be a state of knowledge, intention or whatever. And even if the guilt phase of a criminal trial did not turn on such factors, it seems like sentencing authorities often decide on the basis of such factors. So I’m still skeptical of your conclusion. OTOH, I’m not convinced that the criminal law is coherent or admits of overall coherent justification.
Sorry about the identity confusion!
Your response sounds right for the most part, but I was talking with a lawyer friend over dinner and realized that someone would be in legal hot water if he *knowingly* put someone at risk in a case in which the “reasonable man” would not know it was a risk. These seem like cases in which the reasons for which the person acted makes the act criminal while another person who performed an act of the same type, for different reasons, would not have done anything criminal.
Imagine a car manufacturer who knew that some production process would run a risk of harm, but also got info that awareness of this risk was above and beyond the industry norm. If the emails reflecting his “supererogatory”awareness of the risk were part of the evidence, he would be in trouble.
Does that help at all?
Mark: Determining mens rea is a matter of determining attributability, not meaning. In other words, the fact that you intended to do what you did is a necessary component of the action’s being properly attributable to you for purposes of further appraisal. But that you intended to do X doesn’t reveal anything about (Scanlonian) meaning, insofar as it isn’t a fact about your reasons for doing X. So while offences typically have built-in conditions about mental states, those mental states aren’t the ones relevant to the meaning of actions.
Now I completely agree with your point about sentencing: here is precisely a location in which considerations of meaning come into play. But they are raised typically to determine how long someone is to be punished, not whether or not he is to be punished. The latter — the question of justification for criminal blame — doesn’t involve a response to meaning, at least most of the time (or so I claim).
Regarding your final point, that the criminal law may not admit of coherent justification, fair enough. I myself think there are plural justifications involved which are sometimes in tension. But then so much the better for my general thesis, which is that we should be skeptical of what has become an article of faith amongst many legal theorists and moral philosophers, namely, that the elements of criminal responsibility are either identical to, or analogous with, the elements of moral responsibility.
Brad: I think what you’ve identified would be the mens rea conditions for criminal recklessness, and here they would be knowledge conditions. But as I just noted in response to Mark, those aren’t sufficient for provision of the relevant sort of Scanlonian meaning. Instead, they would simply provide the attributability conditions of the offence to the agent.
An interesting feature about criminal law is that fault elements are contained in offence definitions even when there are independent mens rea conditions sufficient for attributability.
For example, in rape it is standard that the prosecution must demonstrate that D acted either knowing or being reckless as to the lack of consent. It would be sufficient to establish attribution to D that D intentionally had intercourse with V and V did not in fact consent. If D has intercourse with V, V does not consent, but D had no reason to believe that V was not consenting, D has not committed the offence. He does not need a defence. If D has intercourse with V knowing that she doesn’t consent, but D is acting under duress (say a threat to his life), in contrast, D has committed the offence, but has a defence.
This also seems the intuitive view about how rape should be defined (though there are detractors in the literature). One justification for the intuitive view is that the definition of rape provides a public statement of the terms in which rapists are to be condemned when they are not entitled to a defence, which implies that offence definitions are about meaning as well as attribution. But that is not the only justification. Antony Duff, for example, believes that this is to do with what needs to be answered for at trial.
Hi, Victor. Good to hear from you. Thanks for this.
A couple of points. First, I have a hard time understanding how you can get mens rea sufficient for attributability independently of the fault elements you mention. In the example of rape you give, you say that “it would be sufficient to attribute that D intentionally had intercourse with V and V did not in fact consent.” But what precisely is being attributed to D? If there are additional elements that must be proven — e.g., whether or not D had a reason to believe that V wasn’t consenting — then the offence isn’t yet attributable to D. So what is? Just the act of having intercourse with V? Having intercourse with V without V’s consent? If these aren’t what the offence is, then neither could be enough for the criminal justice system to bring down the hammer on D. It seems to me that, for attributability to be relevant to the criminal justice system, what’s attributable to D must be precisely what D is punished for. To the extent that there’s an action attributable to D that isn’t an offence, it’s not something for which he can be blamed.
Now perhaps you mean to say that, once we add some “fault” considerations (e.g., he had a reason to believe V wasn’t consenting) then (and only then) is the offence attributable to D. And perhaps these “fault” considerations are about meaning. But I’m not yet seeing this from the examples you’ve given. With respect to the excuse example (acting under duress), the relevant reasons the agent has are about his defence, not the offence, and so if successful, he’s off the hook, and if unsuccessful, he’ll be punished for the offence. But then his reasons aren’t relevant to the nature of the offence (which is what he’d be punished for), as I noted in my original post.
Regarding the justification example (where D had no reason to believe that V was not consenting), what gets D off the hook is a theoretical belief (he believes that V was consenting and has no reason to believe otherwise), whereas I don’t see Scanlon’s account of meaning being about such beliefs (so again, I see them as relevant only to mens rea). Instead, the reasons relevant to meaning are practical, reasons for action (see, e.g., p. 52 of Moral Dimensions).
Finally, that there are grounds for public condemnation (when not entitled to a defence) doesn’t yet imply to me that offence definitions are about meaning as well as attribution. Indeed, this seems to be precisely the issue at stake here. One could deserve condemnation for some offence for a variety of reasons (including answerability reasons) in the public realm without that condemnation being a response to one’s reasons for action.
Thanks David. Here was something that confused me I think.
You attributed to Scanlon these two views
1) D may act permissibly and yet be blameworthy. This is so where D does the right thing for the wrong reason.
2) D may act impermissibly and yet not be blameworthy. This is so where D does the wrong thing, but ‘when the reason for which D acted did not impair his relationships’.
Suppose that impermissible is to be understood relative to the evidence available to D (which Scanlon thinks is the right understanding of impermissible). I then don’t see how 2) can arise without adopting an understanding of blameworthiness that is clearly false.
For example, suppose that relative to the evidence there is a decisive reason against Ving in virtue of the fact that X will have his arm broken as a side-effect. D Vs, breaking X’s arm. He knows that this will happen, but regrets it happening. The breaking of X’s arm is not an explanatory reason for V’s action. But clearly D is blameworthy, and D’s relationship with X is impaired. Failing to respond to a feature of a situation that provides a decisive moral reason against acting is clearly blameworthy. One need not act in order to bring that feature about to be blameworthy. I take it that you don’t wish to saddle Scanlon with the view that these judgements are false (and that’s not how I read him).
Because of that, I took you to mean that attribution could come about when a person does something that is wrong relative to the facts, but not relative to the evidence, as is implied by strict liability offences.
If this is right, the better reading of Scanlon’s view holds that blameworthiness and permissibility come apart only in 1). On all cases, where there is a decisive reason, relative to the evidence, against D Ving in virtue of some feature, f, and D Vs, f is attributable to D and D is blameworthy for Ving in virtue of f.
I believe that Scanlon’s views about 1) are defective, but that is a much longer story.
Victor: Here’s the relevant quote, from pp. 124-125: “The blameworthiness of an action depends, in ways that wrongness generally does not, on the reasons for which a person acted and the conditions under which he or she did so. So it can be appropriate to say such things as, ‘Yes, what she did was certainly wrong, but you shouldn’t blame her. She was under great stress,’ or ‘You can’t blame him. He thought he was acting for the best.’ Good intentions, and conditions such as stress, can be relevant to blame even when they are not relevant to the rightness or wrongness of what the person did.”
Now here’s another crucial quote (pp. 152-153, emphasis mine): “The agent’s reasons for acting (and the fact that other considerations did not count for him as reasons against so acting) are what constitute his attitude toward others, and what have the implications that blame involves, in the account I am offering.”
OK, so in the example you’ve just given, if D knows that his action will cause V’s arm to break, then even if he didn’t perform the action in order to break V’s arm, the fact that he knew it would do so and this didn’t count for him as a reason against doing so are among the reasons constituting the attitude that impairs his relationship with V. So you’re right: I don’t want to saddle Scanlon with the view in question.
That said, I don’t also have to take on the view of attribution you took me to mean. I take impermissibility to be relevant to evidence. (Indeed, I have elsewhere tried to argue that there’s not a plausible story one can tell about strict liability that allows for genuine attributability). Attributability has to occur relative to the evidence, as you say.
Do you still think, then, that Scanlon can’t defend (2), the view that blameworthiness and permissibility can come apart, such that one could act in an impermissible fashion yet not be blameworthy? When under great stress, one’s doing something impermissible may not have any significance for any relationships. And even when you should’ve known better, your intentions may be pure enough in doing the wrong thing that we don’t think it appropriate to hold it against you. (I should stress: this is all, on my view, true only of the moral/interpersonal realm.)
I hope I’ve understood you here, Victor.
Thanks David.
As I see it, the possibility that a person might act impermissibly but not be blameworthy in the criminal law is to be found in the law of defences, which are not your concern in your original post. So it is true that a person can act wrong but not be blameworthy (or ‘as’ blameworthy) in what are sometimes called ’emotional excuses’ (such as duress or provocation). I was only addressing circumstances where these defences are not in issue.
When it comes to the definition of offences (assuming that they are legitimately formulated), on the evidence-relative reading of permissibility the conditions of attribution of a wrongful act and the conditions of blameworthiness seem identical. The definition of any offence must include a feature that, in the absence of a defence, renders perpetrating it impermissible all things considered. And in fulfilling the conditions of the offence, in the absence of a defence, the defendant must at least have failed to give due weight to this feature. On the quotes above, that is sufficient to render the action blameworthy on Scanlon’s account. So, in the absence of a defence, if D has acted impermissibly in offending she is also blameworthy.
It follows that when we punish a person for a legitimately defined offence, where he legitimately lacks a defence, it will necessarily be the case that the defendant has acted impermissibly and that he is blameworthy. Hence it seems wrong for you to say in your post that punishment could be meted out without blameworthiness in Scanlon’s sense.
The trickier issue involves cases where D does ‘the right thing for the wrong reason’. Such cases, for Scanlon, are blameworthy but not impermissible. Translated into the criminal realm, we would then face unpalatable options. Scanlon might believe that it is permissible to convict a person for acting in a way that permissible but blameworthy. But that seems highly revisionist – surely it is a necessary condition of a criminal conviction that D has acted wrongly. The alternative is that D can be convicted only if he has acted in a way that is wrong all things considered. But on Scanlon’s view about wrongness, that has problematic implications.
For example, suppose that a criminal gang robs the local post office. It is the case, and each person knows it, that failure to participate will certainly result in severe reprisals – the killing of a family member of the person who fails. All members of the gang except one, Beth, act only for the money. Beth acts only in order to protect her family member.
I take it that it is permissible for Beth to rob the post office to protect the life of her family member (we can stipulate that there was no prior fault in joining the gang). Scanlon’s view seems to imply that all members of the gang act permissibly, which seems wrong. His view also implies that members of the gang could be convicted of the robbery only if it is permissible to convict a blameworthy person who has done nothing wrong. That seems wrong too. Surely the right view is that those who did it for the money wrongly robbed the post office and ought to be blamed and punished for doing so. Beth, in contrast, acted permissibly and ought to be acquitted.
I’ve written about this at greater length in a forthcoming book and would happily send you the chapter.
Excellent, Victor. This is all very helpful. I have a few things to say about this.
First, yes, the examples Scanlon himself gives for impermissibility without blameworthiness (e.g., stress) have their most comfortable fit within the law of defences. So I’m trying to argue that you get impermissibility without (Scanlonian) blameworthiness in the law of offences alone. Here you suggest that the quotes I pointed to yesterday render punishment for offences identical to (Scanlonian) blaming. I agree that it might look like that (I thought about immediately noting that point with the following explanation, but left it for today). That something doesn’t count for me as a reason is ambiguous: it could mean that it didn’t occur to me, and it could mean that I judged it not to count. It would be extremely difficult, in the former case, to distinguish between insensitivity and mere forgetfulness or absent-mindedness, and forgetfulness/absent-mindedness surely wouldn’t count as having any significance for our relationship. So I’d want to push for the latter as the interpretation relevant to meaning. That said, “and the defendant judged that the relevant feature of the situation didn’t count as a reason for him” isn’t part of the law of offences, as I understand it. Perhaps it’s implicit? In other words, perhaps it’s just understood to be true of the defendant if he lacks a successful defence? I don’t see this as a wise assumption, given the other interpretation of “didn’t count as a reason for him.”
As for those acts that are blameworthy but not impermissible, why would Scanlon have to maintain that blameworthiness is sufficient for criminal blame? What counts as the appropriate response to blameworthy actions may depend on pragmatic features of state machinery and the like.
Yes, please send me that chapter.
that’s interesting david. So you think that in the case of the forgetful, the wrongful action is attributable to the defendant, but that he is not blameworthy, and yet he is liable to conviction and punishment in criminal law.
Many people (myself included) criticize the reasonable man standard on the grounds that it does not adequately pick out fault. It criminalizes the forgetful/absent minded as well as those insufficiently motivated by the interests of others (of course, some forgetfulness might issue from inadequate motivations, but let’s leave that aside). Do you think that these criticisms are wrong, and that a person can be liable to conviction and punishment by being forgetful/absent minded?
On blameworthy but not impermissible actions. It seems unpalatable, in the case above, either to convict Beth, or to acquit the other robbers. But they differ only in their motivations. Given that Beth’s action is permissible, if Scanlonian’s wish to see the robbers convicted they must warrant convictions for those who act permissibly but in a blameworthy manner. But it seems an attractive principle of criminal responsibility that a person can be liable to be convicted and punished only if she has done something wrong.
Interesting because probably false, I’m sure.
First, let me say that I now see the force of your Beth example. I think there are various recourses the Scanlonian has (based on the first three chapters of the book), but I’m not as fluent on that material as I’d like to be and so may not be able to respond as effectively as a true Scanlonian (or Scanlon himself!) might.
Regarding the criminal blame issues, I also want to make sure I’m understanding the dialectic, so I’m going to think a bit on this and then reply shortly.
Victor, here’s what you said that I was focusing on: “The definition of any offence must include a feature that, in the absence of a defence, renders perpetrating it impermissible all things considered. And in fulfilling the conditions of the offence, in the absence of a defence, the defendant must at least have failed to give due weight to this feature.” In my previous comment, I was trying out the idea that there are two ways in which such a comment could be read (although I put it in Scanlonian terms, the same could apply to how you put it). On the one hand, I might have failed to give due weight to this feature by its not occurring to me. And its not occurring to me could itself be the product of two very different sorts of conditions: I might be insensitive to such reasons, or I might merely be absent-minded/forgetful. On the other hand, I might have failed to give due weight to this feature by my (actively) judging it not to have due weight. Of these three ways (two in the first hand, one in the second) of failing to give the feature due weight, only the first (insensitivity) and the last (judgment) strike me as having anything to do with Scanlonian meaning.
My second point was perhaps more dubious. The basic thought was that, as a practical matter, it would be a nightmare to try to distinguish between the two conditions of the first handed reading, and so to the extent the criminal justice system was really tracking meaning (and to the extent that offences, as written, must meet basic practical demands), its safest strategy would be simply to include the active judging component in its articulations. But as far as I know, this ‘hand’ isn’t actually included in offences and isn’t the object of pursuit by prosecutors. So this sort of “meaning” can’t be what defendants are being punished for.
So what I’m saying is that I agree with your criticisms of the reasonable man standard for just the reasons you give: it couldn’t possibly distinguish between cases of insensitivity and absent-mindedness. Now are there replacements for this standard that do? If so, and if these are incorporated into offences and/or prosecutorial strategies, I’d love to hear about them, for they may establish Scanlonian meaning after all.
I hope this makes sense.
Thanks for the clarification David.
It’s important to distinguish between the question of what a person is punished for and what has to be proved in order to convict them of that thing. There are some offences which are defined more broadly than the class of actions that they are intended to target. These offences may violate the best understanding of the presumption of innocence. In your example, the reasonable man standard would be a crude way of targeting the insensitive/indifferent. We aim to punish only the insensitive/indifferent by permitting conviction of a broader class of agents for the reason that it is too difficult to distinguish the insensitive/indifferent from the merely absent minded. But in this case, we may still believe that the moral liability to be punished depends on fault, and the hope of legislators is that only the insensitive/indifferent will be punished.
The normal way to go when it is too difficult to discriminate between wrongdoers and non-wrongdoers is to define offences in a way that is underinclusive – let some wrongdoers off the hook in order to avoid convicting the innocent (for example, the House of Lords, in R v G, a leading case on recklessness explicitly took this route in defining recklessness subjectively – they recognised that there may be some people who are unaware of the risk that they imposed that may be sufficiently at fault to be punished, but in trying to punish them too many innocent people would be convicted). Sometimes, though, indifference is explicitly included in the mens rea definition. For example, in the Scottish law of murder the defendant must have acted in a way that demonstrates at least indifference as to whether the victim lived or died (this definition may be underinclusive as well, but that’s another story). In other words, there are many different ways of formulating the mens rea of a criminal offence. Often effort is made to distinguish those who are at fault from those who are not. When this is not done, the law is typically subject to a great deal of academic criticism. That seems to vindicate the idea that, in the minds of many people at least, the law ought to be scrupulous in ensuring that only those who are inadequately motivated by the interests of others are convicted of an offence.
Victor: I can’t tell you how helpful all this has been. At the very least, you’ve forced me to rethink my position on these issues, especially with an eye more toward actual case law, and in particular w/r/t offences as written. But I’m not yet ready to concede the main point, so here’s another question for you. I had originally been thinking of mens rea as establishing attributability, not meaning. In other words establishing that the defendant had the “guilty mind” just establishes that the action is properly attributable to him qua practical agent (to use your favored language). I still believe this. But now it looks like you think that the reasons the defendant may have had (those that amount to his being insufficiently sensitive to the interests of others), when they are part of offences, are packed into the mens rea aspect of the offences. If this merely establishes attributability, though, then it may not yet establish Scanlonian meaning as the target of punishment. That is to say, if investigation into the agent’s reasons serves merely to establish that the criminal conduct belongs to him for the purposes of criminal sanctioning, then it’s still just the conduct alone that is the target of punishment, not the meaning of the agent’s conduct.
Excellent.
I take it that your question is not with criminal justice as it is (who knows what officials take to motivate the actual system) but on criminal justice as it ought to be.
What is needed here, I think, is a theory of punishment as part of a more general theory of the functions of criminal justice. My own view of punishment would treat your approach quite sympathetically – I think that liability of a person to be punished depends on the duties that a person incurs as a result of their wrongdoing (I defend this in a forthcoming book, The Ends of Harm). These duties, it is plausible to say, depend primarily on attribution of a wrongful action rather than meaning. The most important duty, in my view, is the duty to protect the victim and others. This, I think, grounds the permissibility of general deterrence. However, I think that there are some duties that offenders incur that may be primarily meaning based – namely the duty to recognise the character of one’s wrongdoing.
Others, with a different view of punishment, may see things differently. Standard retributivists, who think that punishment is justified in order to give wrongdoers what they deserve, might argue that desert depends on motivations, or even on the vices underlying those motivations. Their view, which I think mistaken, has been dominant amongst criminal justice scholars over the last two or three decades.
Finally, it might be argued that offences play two distinct roles – they provide the terms in which offenders can be publicly condemned, and they determine the conditions under which a person is liable to punishment. It might be argued that meaning has central importance in the former role and attributability in the latter.
At any rate, I think that your account will ultimately rely on an account of the functions of the criminal justice system.
Again, very interesting and helpful, Victor. Yes, I have in mind a normative view, and yours strikes me as initially very plausible. I can’t wait to see (and likely teach) the book. I doubt I could come up with a theory of the functions of the criminal justice system on my own, so I will be happy to glom onto yours!
For now, though, I’m interested in the distinction you run between public condemnation and liability to punishment. I take it that you think condemnation is a function of the labeling that occurs with a guilty verdict alone, or do you also think that condemnation attaches as well to any punishment that occurs as a result? I have a hard time thinking that all guilty verdicts alone constitute labeling sufficient for condemnation. Isn’t that a little strong to describe the guilty verdict assigned in a lonely courtroom to a petty thief, say?
Did you think that the condemnation of the thief in the courtroom is too weak given the seriousness of his crime? I’m not sure – don’t forget that the person has a criminal record, which is a public document published to, and accessible by, all the world that he has offended.
You might think that punishment has an expressive function. I don’t think that the expressive function can justify punishment (as Scanlon excellently puts it, why not say it with flowers, or rather with weeds). But if punishment is justified on another basis, say the basis that the offender incurs duties as a result of his wrongdoing, in punishing the person we can also recognise the seriousness of the wrongdoing – your wrong was so bad, we say, that it is permissible to treat you in a way that it would be seriously wrong to treat innocent people.
Now the question about the relationship between meaning and attribution returns. Should we think that the expressive dimension of punishment just outlined relates attribution for a wrongful act, or is it also related to the offender’s reasons for acting, assuming that Scanlon’s distinction between attribution and meaning can be rescued in some form. I’m not sure about the answer to that. I’d need to work out a bit more clearly than I have the best way of understanding Scanlon’s distinction.
No, I had in mind that “condemnation” seemed rather excessive (too strong) as a description of what applies in the case of the petty thief in the lonely courtroom.