The Secret Politics of the Compatibilist Criminal Law

I am curious to hear what Gardeners think about a recent paper by Anders Kaye entitled, "The Secret Politics of the Compatibilist Criminal Law."  Here is the abstract:

Many criminal theorists say that we have a 'compatibilist' criminal law, by which they mean that in our criminal law a person can deserve punishment for her acts even if she does not have 'genuinely' free will. This conception of the criminal law harbors and is driven by a secret politics, one that resists social change and idealizes the existing social order. In this Article, I map this secret politics. In so doing, I call into question the descriptive accuracy of the compatibilist account of the criminal law, and set the stage for a franker discussion of criminal punishment - one that recognizes that the perpetual struggle to say just who 'deserves' punishment is driven as much by brute politics and the competition to allocate power and resources in society as by any independent moral logic.

I am particularly keen to hear what the compatibilists who frequent the Garden have to say about Kaye's provocative line of argumentation.

Comments

[Kip]

What a brilliant article.

This fits with the other two law review articles that I've mentioned here before.

All three defend a revolutionary form of free will anti-realism. With the emphasis on "revolutionary". That revolutionary spirit is missing in many anti-realist philosophers (namely, Smilanksy, Nichols and Double).

Some points:

1. the article has a great summary of the "origination" idea of free will
2. the article also has a great summary of how compatibilists seem to be motivated to make the world safe for blame
3. don't let the pomo-ish talk of "secret politics" throw you. The article has some great ideas in it.
4. If the article's thesis is right, we should expect compatibilist to be more politically conservative than anti-realists about free will (or at least revolutionary anti-realists). I'm not sure this would pan out, because most academics (including compatibilists) are pretty liberal. It would also be hard to find a big enough sample of anti-realists.

[Eddy Nahmias]

I haven't read the whole article yet, but could someone who has (e.g., Thomas or Kip) explain two things:

1. Why is the target compatibilism rather than libertarianism? I would have thought that libertarians are more likely to be perceived as neglecting mitigating circumstances than compatibilists (though there is absolutely nothing about either position--compatibilism or libertarianism--*in and of itself* that suggests you should neglect mitigating circumstances).

2. On the contrary, why isn't it that compatibilists are best situated to attend to mitigating circumstances, since they think the origination issue is a red herring so they can focus on whether defendants possess the proper capacities to control their behavior (the capacities required by the law) and, if so, whether there were circumstances that made it difficult to exercise them?

Compatibilists can emphasize that free will (the control required for MR) comes in degrees and the law should pay very close attention to the degree to which a defendant possesses the relevant capacities and had the opportunity to exercise them. Seems to me like those who think origination is essential to MR (including skeptics about origination--which I assume the author is) are more likely to throw us all in the same bathwater--we either have these magic powers or we don't--and then to treat us all the same: we've got it so we all deserve retribution or we don't have have it so throw out desert entirely (remember that a forward-looking punishment system can be quite harsh if you think harsh punishment is required for deterrence).

[Kip]

Eddy,

I'll take those up. I had similar questions while reading the article.

1. The article targets compatibilism, instead of libertarianism, because the author is not worried about science falsifying libertarian free will. He trusts that, as science progresses, we'll see more and more that libertarian free will doesn't exist. (This, at least, is what I read between the lines of the article.)

In contrast, the author sees compatibilism as being an evolved version of the free will doctrine that is *resistant* to the advances of science. Like a doctor who is worried about superbacteria that have evolved to resist antibiotics, the author is worried about a compatibilism that has evolved to resist the discoveries of science.

2. Compatibilism per se could attend to the proper antecedent excusing conditions. However, the author argues that compatibilism in practice does not. Rather, compatibilism in practice focuses on mesh or time-slice views of the agent, rather than holistic, gradual changes in the agent. Accordingly, compatibilism in practice turns a blind eye towards the holistic, gradual excusing conditions that, in the author's view, lead to most crime: "hard" social conditions (poverty, inequality, discrimination). The author makes this point in section III starting on p. 380.

You might respond, "well, compatibilism can respond to that challenge by starting to address hard social conditions, etc." But, the author goes on to argue, compatibilists are unlikely to do that. They are unlikely to do that because, in the author's view, compatibilism is motivated by a desire to make the world safe for blame---but if you start considering hard social conditions to be excusing, you'll get rid of much or most of the blame in the world.

Here's one quote that helps explain the author's view:

If this is the way hard social conditions most commonly influence
conduct, it is no wonder that contemporary compatibilism cannot “see”
it. Contemporary compatibilism has a very particular taste when it
comes to inquiry into desire and disposition formation. It is interested in
circumstances and events that influence desires and dispositions in a
distinctive and bizarre way—by sparking to life anomalous and
transitory desires and dispositions, as guns-to-the-head and supernatural
neurosurgeons do. But it has less interest in phenomena that shape
desires and dispositions “normally,” in a gradual, holistic way.
Momentary invasions of the desire/disposition formation process matter,
but otherwise, “the source . . . of the actor’s goals, desires, and values
that motivate his choice are [not] relevant . . . .”135 “The sort of character
a person has is relevant to assessing his moral responsibility for an
action, but not how he came to have that character.”136 In short,
contemporary compatibilism’s interest in desire/disposition formation
extends only to a certain sort of abnormal and transitory
desire/disposition acquisition. Because there is nothing abnormal or
transitory about the way the sorts of hard social conditions found in
modern Western society today shape desires and dispositions,
contemporary compatibilism does not include most real-world hard
social conditions in its responsibility determinations. It cannot see them."

All of this is very armchair, and not very charitable to compatibilists. But it resonates with me (I've never been very charitable to compatibilists).

[tnadelhoffer]

Eddy,

When I said, "I am particularly keen to hear what the compatibilists who frequent the Garden have to say about Kaye's provocative line of argumentation," I should have added "after they have read the article." :)

On a more serious note, I just wanted to disagree with Kip in that I don't think what Kaye says is either uncharitable or done from the armchair. Instead, he provides what I believe is a fair assessment of the uses that legal theorists have historically made of compatibilist theories of responsibility to shield retributivistic desert from developments in science and he tries to show why compatibilism is particularly well-suited for these purposes.

For instance, nearly every one of the contemporary neo-retributivists in legal theory (e.g., Moore and Morse) adopt some (if not all) of the following: (a) a compatibilist account of free will, (b) a deflationary and purely cognitive conception of the insanity defense, (c) a denial that volitional factors are in themselves mitigating/exculpating, and (d) a definition of legal liability and desert that is cast in such minimal compatibilistic terms that most sociological factors could not, even in principle, be mitigating/exculpating.

All that is required for blame and punishment on this view is that you have minimal powers of rational deliberation such that you know how to navigate the legal norms--i.e., all that is required is something like a legal-reasons-responsive mechanism. That it is perhaps understandably hard for you to navigate these norms based on poverty, childhood abuse, drug addiction, etc. is irrelevant as far as responsibility is concerned. All that matters, is that you have the threshold rational capacity to appreciate that your actions are illegal. Of course, in the event that you fall below this threshold, you will cease to be responsible--but only because your reason has been compromised. In short, all threats to free will operate by either by-passing or undermining an agent's powers of rational deliberation.

That being said, I will happily write up an overview of Kaye's position later on today. For now, I just wanted to briefly respond to your query about compatibilism vs. source incompatibilism. The first thing worth pointing out is that the source-hood skeptic can bring all of the compatibilistically grounded cognitive and conative capacities on board in an effort to determine where people fall on what might be called "the continuum of control"--which everyone agrees we have to be able to do to have a system of criminal laws at all. Sourcehood skeptics don't deny that some people are more rational than others nor do they deny that some people have more volitional control than others. What they deny is that we are the types of creatures who could be full-blown desert-based responsible for our actions--i.e., that we could deserve to suffer for sufferings sake. As such, both compatibilism and source skepticism have many of the very same tools at their disposal for separating people who are apt targets for criminal sanctions from those who are not. It's just that the sourechood skeptic has the theoretical machinery in place for additional mitigating and exculpating conditions.

At the end of the day, I believe compatibilists and originationists often part ways when it comes to both how and who we punish. It is this difference that Kaye is interested in--but more on that later...

[Sofia Jeppsson]

I don't seem to be able to open the document, unfortunately.

But shouldn't someone actually do some experimental research on this? I e see how philosophers engaged in the free will debate view crime and punishment, and if there are any interesting correlations between crime and punishment views and free will views.

[Eddy Nahmias]

Kip says Kaye suggests: "Compatibilism per se could attend to the proper antecedent excusing conditions. However, the author argues that compatibilism in practice does not." If so, that's the part I'd like to see explained. I read enough of the paper to see that he cites a lot of philosophers. But does he say that those philosophers are the ones who don't attend to excusing conditions or does he just mean the legal scholars who call themselves compatibilists (Morse, Moore, others?). If it's the latter, then shame on him, since he's read the philosophers and should know that it's the compatibilists who tend to analyze the conditions for control--and hence the "continuum of control--more than the hard incompatibilists or libertarians since they are spending more of their efforts explaining why we have or do not have "genuine free will." Indeed, part of the reason I think we should get away from the issue of determinism (or whether FW and MR require the ability to do otherwise holding fixed past and laws or the ability to do the impossible and be the ultimate source of one's self) is that if we refocus away from that debate, we can focus on the more important conditions for control, how to inculcate them, how to determine whether they are met (e.g., for legal reasons), and how much scientific discoveries might threaten them.

[Kip]

Eddy,

He discusses the theories of Arenella, Wallace, Wolf and Fischer and Ravizza at least briefly, at pages 382 and 383. Various aspects of these theories are discussed throughout the paper. I don't think he provides (or intends to provide) an exhaustive overview of any given theory of compatibilism, though.

[Kip]

Thomas:

When I say he's being uncharitable to compatibilists, I don't mean he's wrong. I tend to think he's right. All I mean is that: in the absence of a proof of compatibilists' motives, he tends to assume that they are improper. There are lots of possible good motivations for compatibilism, other than "a secret politics to maintain the status quo of inflicting tremendous punishment and suffering on the state's citizens." But that's the motivation he goes after, without much evidence to support it, which leads me to the second point:

How are the ideas in the article anything *other* than armchair philosophy/psychology? I don't recall seeing a single point of data that he collected in his research. Did I miss something?

[tnadelhoffer]

Eddy,

Since you are apparently unwilling to read the article yourself :), perhaps it would be helpful if you took a few minutes to say how you think your own preferred version of compatibilism could be used to argue for a "rotten social background defense"--i.e., how it might be used to mitigate/exculpate desert in cases normally treated by the law as non-mitigating/exculpating. If you don't think that this kind of factor ought to be mitigating/exculpating, then your views about free will have precisely the kind of effect Kaye suggests. Of course, it doesn't follow that Kaye is correct in thinking that this outcome is morally sub-optimal, regressive, etc., but it would nevertheless show that his claim is descriptively correct (even if his subsequent normative judgment would still be open for debate).

If, on the other hand, you do think that having a "rotten social background"(RSB) could be mitigating/exculpating, then I, for one, would like to hear more concerning why. I suspect you will just say that people with RSB may have cognitive and volitional impairments such that they are less free and hence less responsible. But then how do you think the law should handle this difference--especially given that most people who commit crimes had a RSB? An originationist can simply evade this issue altogether since the point of the criminal law will be to prevent crime and rehabilitate offenders not to mete out suffering to the deserving. So, fixing blame need not be part of the process. For a compatibilistic desert-based theorist such as yourself--where ever you happen to fall on the retribution-continuum--you owe an account of how your views about free will and responsibility will be applicable in the real world. Do you think that a violent offender deserves less blame and punishment just because he was severely abused as a child? If so, why? If not, why not?

While you work on answering those questions, I will try to piece together the summary of Kaye's view you've requested!

Kip,

I suppose it depends on what you mean by armchair theorizing. I usually take it to mean appealing to one's own intuitions about particular cases. But I don't see that as an apt description of what Kaye is doing. Instead, he takes a historical look at (a) the sorts of reasons compatibilists have traditionally given for going the compatibilistic route, (b) the historical use that has been made of compatibilist theories of FW and MR by legal theorists (especially retributivists), and (c) the real world effect this has on enshrining the retributivistic status quo in the legal system. Now, he admittedly approaches the entire issue from a critical-theory perspective, but I wouldn't say that based on that alone, he is engaging in pure armchair speculation. At the end of the day, I take his target not to be compatibilism per se, but rather the way it has been used to defend the citadel of retribution in the criminal law against developments in the behavioral sciences.

If compatibilists read his article and respond by (a) showing that Kaye has merely identified ways that compatibilism has been *misused*, and (b) showing how and why the uses Kaye identifies do not follow from compatibilism per se, I think he will have advanced the debate.

[tnadelhoffer]

Sofia,

I couldn't agree more with your call for further empirical investigation! For instance, a recent study by Shareef, Schooler, and Greene suggests that people who are primed with deterministic cues are less retribuvistic. Hopefully, their paper will be available soon so we can examine it here at the Garden!

[Tom Clark]

I sympathize with the general thrust of the paper but don’t think Kaye is correct in saying that hard social conditions should count as excuses in a reformed criminal justice system. Some consequentialist, non desert-based grounds for involuntary incarceration (e.g., deterrence, reformation) hinge on the agent being minimally rational, in which case compatibilist criteria for responsibility are still centrally relevant for non-retributivists. But even if hard social conditions and other criminogenic conditions outside the offender aren't excuses, they are still very much *causes* of crime. Inquiry into the role of such conditions helps distribute causal responsibility for crime outside the agent (but leaves the agent still morally responsible), which has two effects: it undercuts the retributivist intuition (which thrives when situational causes are ignored) and widens the response to crime (or should anyway) to include not just agent-focused interventions such as incarceration but amelioration of the conditions themselves, reducing the need for criminal justice.

The defect of compatibilism that the article highlights, imo, is that by narrowing our focus to individual rationality and control (if compatibilism indeed does that), it wrongly assigns causal responsibility and draws attention only to individuals, not situations. Kaye says compatibilists are “corralling excuse” in service to “salvaging [retributive] blame” in an era of causal explanation (p. 399). More accurately, I’d say they are (if he’s right in his accusation) corralling *causation* in service to restricting *causal responsibility* to individuals, after which most compatibilists go on to say that retribution is warranted, why I’m not sure. But not all compatibilists about moral responsibility (such as myself) go on to say this, and widening the scope of causal considerations, as the author advocates, can help to motivate consequentialism without posing any danger to MR compatibilism itself. You don’t have to be an originationist (incompatibilist) about MR, or suppose that hard social conditions count as excuses from being held (consequentially) responsible, to be interested in situational causation and be against retributive punishment.

[Mark Smeltzer]

"But not all compatibilists about moral responsibility (such as myself) go on to say this..."

When did this happen? I imagine that there should have been a party!

[Tamler Sommers]

Following up on Eddy's apt (if scandalously uninformed) comments, here's what seems to be a tension within the article. Kaye claims that compatibilist interpretations of criminal law allow us to feel OK about punishing people from disadvantaged backgrounds. Originationist theories, according to Kaye, would make us more aware of mitigating social factors. But as Tom notes, given that we're ruling out libertarian theories of responsibility, originationist theories can make no distinctions whatsoever between people based on their social environment. Everyone deserves the exact same amount of punishment: none. This is as true of the privileged as it is of the disadvantaged. No cause is more or less mitigating than any other—at least from the perspective of desert.

So we're going to be originationists, we have to find some other justification for punishment besides desert like consequentialism. The gap in Kaye's argument, it seems to me, is that he doesn't give any reason to think that a consequentialist (or other non-desert based) approach to punishment would benefit people from disadvantaged backgrounds more than a compatibilist approach. It may be that a consequentialist approach would favor harsher punishments for disadvantages people because the temptation to commit crimes is greater. For all we know, a compatibilist approach—given that it does take backgrounds into account to some extent—may favor the socially disadvantaged over a system that disregards desert entirely. Nothing in Kaye’s article gives us reason to take sides on this (empirical) question one way or the other.

[tnadelhoffer]

Tamler,

I suspect Kaye might (or arguably should) prefer the kind of preventive regime developed recently by Slobogin that Kip mentioned earlier--which is elaborated upon fully in his very interesting book Minding Justice. Given this approach, more focus, energy, and resources would be directed to addressing the causes of crime and on rehabilitating offenders. This kind of approach to the problem of crime is well-suited for orginationists since it is based on empirically-informed predictions of violence rather than subjective moral intuitions concerning desert and retribution. As such, I think both you and Tom are incorrect in claiming that "given that we're ruling out libertarian theories of responsibility, originationist theories can make no distinctions whatsoever between people based on their social environment." If I am no longer interested in fixing blame but rather in preventing offenses and fixing those who offend, I must pay a great deal of attention to social background, etc. After all, I won't know how best to development a plan for rehabilitating and reintegrating an offender if I don't play close attention to his own unique circumstances. Now, you're both right that social background doesn't make a difference in terms of desert, but it does make a difference in terms of whether someone is an apt target for different rehabilitative measures. As such, I think that Kaye's best choice would be to forgo the punishment paradigm altogether and opt instead for a preventive regime. Moreover, I do think it is easier for the originationist to motivate this option than it is for the compatibilist.

Tom,

I take it a lot hangs on what we mean by "punishment" and "excuse." On some views, to punish someone just is to give them what they deserve. Similarly, to say that someone is excused is to say that they don't deserve to suffer for what they did. I assume that's not the kind of compatibilist moral responsibility you hand in mind. Instead, you are presumably interested in some kind of control condition that enables us to separate those who are apt targets for state intervention from those who are not. Is this correct? Setting that aside, I also wanted to comment on your following remarks:

"More accurately, I’d say they are (if he’s right in his accusation) corralling *causation* in service to restricting *causal responsibility* to individuals, after which most compatibilists go on to say that retribution is warranted, why I’m not sure. But not all compatibilists about moral responsibility (such as myself) go on to say this, and widening the scope of causal considerations, as the author advocates, can help to motivate consequentialism without posing any danger to MR compatibilism itself. You don’t have to be an originationist (incompatibilist) about MR, or suppose that hard social conditions count as excuses from being held (consequentially) responsible, to be interested in situational causation and be against retributive punishment."

First, I think you're right about how Kaye's view could be most helpfully construed. However, I find the rest of your remarks puzzling. At the end of the day, I have never understood why one wouldn't be a retributivist if one adopts a compatibilistic account of either free will or responsibility. If you believe in desert-based moral responsibility, then why wouldn't you go on to endorse retributivism? If, on the other hand, you believe in non-desert based responsibility, why call it moral responsibility rather than something else--e.g., answerability, accountability, etc.? What work is the "moral" doing absence some notion of desert?

[Tamler Sommers]

Thomas,

Right, I meant that originationist theories can make no such distinctions when it comes to mitigation and desert. So yes, it's true that:

"I won't know how best to development a plan for rehabilitating and reintegrating an offender if I don't play close attention to his own unique circumstances."

But you could substitute "punish and deter" for "rehabilitating and reintegrating" and this would be equally true. It's an empirical question whether (a) everyone would be better off if we took a rehabililative rather than a more punitive/deterrent approach to criminal justice and (b) whether the socially disadvantaged are better off with a non-desert based approach than a (compatibilist) desert based one. Given Kaye's claims about compatibilism's detrimental effect on the disadvantaged, it seems to me that he has to provide evidence that althernative theories of punishment would make things better (for the disadvantaged) rather than worse.

[tnadelhoffer]

Tamler,

Would the general success of Western Europe's move away from a criminal justice system that focused primarily on retribution and desert count as evidence? In general, I think it is safe to say that disadvantaged are less disadvantaged there than ours are here. Moreover, when the disadvantaged and mentally disordered violate the law there, they are treated far more humanely that they are here. Of course, not even the E.U. has a preventive regime in line with Slobogin's vision. But I find it hard to believe that the "empirical evidence" line is decisive on this front. After all, I can imagine the same argument could have been used in the past to prop up all sorts of otherwise sup-optimal social and political arrangements--e.g., slavery, oppression of women, etc. That being said, perhaps it would be helpful if you said a little bit more about what you would count as evidence. I, for one, think the move away from the so-called "rehabilitative ideal" in this country in the 1960's--which was curiously driven by folks on the left and the right, albeit for different reasons--has produced demonstrably disastrous results. So, there seems to be a pretty costly desert-based social experiment already in place. Does that count as evidence on this front? If not, why not?

[Sofia Jeppsson]

Tnadelhoffer wrote: "At the end of the day, I have never understood why one wouldn't be a retributivist if one adopts a compatibilistic account of either free will or responsibility. If you believe in desert-based moral responsibility, then why wouldn't you go on to endorse retributivism?"

This is how I see it. If we start by looking at everyday social interactions and friendships, there are two basic possible stances I can take towards other people as well as myself.
One would be like a trainer with animals, that you train to perform some behaviours and not perform others. We know from psychology that there are all kinds of training methods that are efficient, both reward based and punishment based. Of course we could be nice and choose mainly reward based methods, although it doesn't necessarily follows from viewing others this way that you focus on rewards.
The other would be like one moral agent interacting with other moral agents. That includes sometimes being proud, sometimes regretting things, sometimes being grateful towards others for what they've done, admiring them, or being angry with someone and going up and telling him what you think of his behaviour and why - and doing all this WITHOUT keeping a plan in the back of your head for how your reactions will work as training tools. Just doing it honestly.

Now I think the second attitude is the only one that can be really consistent. If I take up the first attitude, I can to a certain extent keep that attitude towards myself as well, but not completely. Some part of me will always be "the trainer", and act as a subject rather than an observed and trained object. Most of all, I simply think that the first kind of attitude might be appropriate in particular contexts, but to keep it all the time would simply not be desirable. It would be disrespectful and patronizing, not at all kinder to others and morally preferable.

Now a colleague once asked, when I tried to explain all this, if this view doesn't imply some kind of hidden utilitarianism, if it isn't the case that the "real" reason for taking up the second kind of attitude is that it has better consequences? No, I don't think so. If one take up the second kind of attitude without THINKING of it as a way to produce better consequences, then how could the consequences be what "really" justifies it? Sounds like some mysterious moral realism to me...

But I think it's mistaken to view the legal system as some kind of extension of normal social interactions. It COULD be an extension in a world where criminals reacted to punishments by going "oh, now I see, I totally deserve this, I caused other people to suffer so it's just and proper that I suffer myself". Just as people DO sometimes react to being blamed by a friend for inconsiderate behaviour by going "oh, now I see that I hurt you and was behaving badly, I deserve you being angry with me, but please forgive me". Well, of course people don't ALWAYS react like this to being blamed for inconsiderate behaviour either, but that's how it works ideally, and DO work time to time in praxis, when the two people involved manages to reach a kind of understanding, or when the person blamed has had more time to think about the whole thing.

I think people with retributivist views actually often imagine that harsh punishments for crimes can give rise to a similar understanding on part of the criminal, making him understand that he caused others to suffer by suffering himself, that the punishment will make the criminal finally internalize the legal rules. I think people often say things that suggest this kind of view, "well punishment is the only language these people understand", "they have to be made to understand what they've done" and the like. But it doesn't work this way, does it? So I think it's more to harsh retributivism than just moral intuitions, I think faulty empirical assumptions are often built in. Not necessarily a faulty empirical assumption regarding the deterrent value of harsh punishment (although I think that often plays a role too), but faulty empirical assumptions regarding how criminals will view their own desert of punishment.

I think that when making criminial laws and policies, that's a context where a more objective stance towards other people is appropriate, where attempts to use the first kind of attitude towards people simply won't work. It won't be a proper interaction of several moral agents taking responsibility for their own actions.

I think the closest one can REALLY come towards the first attitude is not throwing the criminals into terrible prisons, but rather the kind of psychotherapy programs carried out against sex offenders that aim at making the offender understand that he's actually caused another human being to suffer, and that HE did this, it wasn't the victim that caused everything to happen as it did. Such therapy, if successful, is probably what ACTUALLY comes closest to how retributivist people IMAGINE that terrible prison sentences work. Making people really understand that they've done something wrong, rather than just training them by rewards or punishments to avoid some behaviours and seek others.

[Tamler Sommers]

Thomas,

A couple quick things. First, I don't think it's fair to associate compatibilism with the current American approach to criminal justice, which is insane on a number of levels (e.g. the drug war) and almost certainly not effective at dishing out what compatibilists consider to be just-deserts. As Kaye notes, prisoners are suffering tremendously, getting raped etc.--and that should offend retributivists as much as anyone. Second, my sense was that the move away from rehabilative approaches was due in large part to the failure of these techniques to actually rehabilitate people. Moreover, with movies like A Clockwork Orange and One Flew Over the Cuckoo's next, you can see the general uneasiness about an approach that "cures" people of their anti-social tendencies. So again, empirical circumstances might force the consequentialist to abandon a rehabilitive approach in favor of a more punitive approach that is more effective at deterring criminal behavior in general.

Remember, I'm not denying that the consequentialist approach to punishment would lead to a more beneficial system for everyone. (By definition, it would actually.) The consequentialist approach also makes far more sense to me than one based on objective assignments of desert. All I'm questioning is whether the group that Kaye is concerned about would benefit from it.

I don't know enough about the legal system in Western Europe to respond to your comment on that. So let me ask a question: can you trace the relative well-being of the disadvantaged in Western Europe to the move away from desert-based punishments? (It might be that more general attitudes about justice are responsible for this.)

[tnadelhoffer]

Tamler,

I take it we need to distinguish between (a) the justifications we give for criminalizing certain behaviors--e.g., drug use, (b) the standards we use for determining whether a particular individual is legally liable for breaking a legal norm, and (c) how guilty offenders are treated once they have been subjected to state intervention. In the article, Kaye is more interested in (b) than in either (a) or (c), but I think the latter are related.

Keep in mind that his claim is that one way of understanding the motivation for adopting compatibilism is that it enables us to salvage desert and retribution from the clutches of scientific advancements in the behavioral sciences, genetics, neuroscience, etc. Indeed, I think that historically, he is right about this (at least post-Hobbes). First, you envision just how bleak a world without desert would be--unless, like P. Strawson you find such a world somewhat inconceivable--then you try to figure out how to make desert mesh with what scientists are learning about agency and causation. In this sense, some compatibilists adopt a strategy that is similar to the one adopted by libertarians (e.g., P.V.I)--namely, assume as a starting point that deep moral desert is possible, and then come up with an interpretation of science that allows you to keep that assumption in place. In the case of libertarians, it leads them to reject determinism. In the case of compatibilism, it leads them to try to figure out a way of adopting some fairly lean standard of "responsibility" that will nevertheless permit people to be blameworthy in a deterministic world. In both cases, the scientific data are viewed with an eye towards keeping desert in place.

Now, to get back to your recent comments, you also suggest that the move away from the rehabilitative ideal was driven by its failure to adequately address the crime problem. In some sense, you're correct. However, we must not overlook the fact that when we experimented with the rehabilitative model, we didn't concurrently take steps to address precisely the kinds of criminogenic sources of crime that Kayes is interested in--both rotten social background and non-psychotic mental disorders. Of course, if you leave some of the main sources of crime unaddressed and you just try to focus on rehabiliating criminals, it won't work. I take it that Europe's success on this front--and our corresponding failure--is driven by the fact that their citizens are provided (a) a better public education, and (b) and free health care, plus (c) there is far less socio-economic inequality. In short, they do a better job preventing career criminals in the first place. Moreover, when criminals are released, their rights are not eroded as they are in this country--e.g., felons often can't vote, take out student loans, etc. So, it wasn't the rehabilitative per se that failed in this country--it was a paired down and unrealistic version of it. That being said, specifying why it failed is not so easy. On the one hand, objections from the right were based on the fact that (a) it was ineffective, and (b) it failed to give people what they deserved. Objections from the left, on the other hand, were based less on its ineffectiveness and more on the sort of Clockwork Orange worries you mentioned.

Now, as for what might be called the "Clockwork Orange Worry," I have to admit I always find it less troublesome that others. For me, I am no less bothered by prison films that provide a fairly realistic portrayal of American penal institutions that I am by the fictitious dystopian films like Clockwork Orange. If you take someone as savage and disturbed as Alex as a starting point, I can't see how locking him up and throwing away the key forever with no treatment and no hope of redemption is any more humane or dignified that using modern science to try to rehabilitate him if possible--even if the methods are heavy handed. Operant conditioning is more dignified than gang rape--or so methinks...

[Sofia Jeppsson]

" In both cases, the scientific data are viewed with an eye towards keeping desert in place."

Yes, and if you're a hard determinist scientific data is viewed with an eye towards abolishing desert. What's the difference?

Empirical science cannot investigate the very question of whether blame can be justified, for example, since that's not an empirical question to start with. You won't get an answer until you add some philosophy to the empirics, BOTH compatibilists and incompatibilists have to do that to pull their conclusions.

[Tom Clark]

Thomas,

"Instead, you are presumably interested in some kind of control condition that enables us to separate those who are apt targets for state intervention from those who are not. Is this correct?"

Right.

"I have never understood why one wouldn't be a retributivist if one adopts a compatibilistic account of either free will or responsibility. If you believe in desert-based moral responsibility, then why wouldn't you go on to endorse retributivism? If, on the other hand, you believe in non-desert based responsibility, why call it moral responsibility rather than something else--e.g., answerability, accountability, etc.? What work is the "moral" doing absent some notion of desert?"

I guess I'm trying to make it possible for consequentialists to be realists about morality and MR, in that I think the moral domain need not be essentially bound up with desert and retribution. That is, moral attitudes and practices can exist without the assumption of non-consequentialist desert, in that people are still moral agents (they act within and are members of a moral community) and their behavior has a moral dimension (it impinges on the rights and well-being of others), which requires agent-directed responses in order for the behavior to be guided properly. Picking out these conditions is at least some of the work "moral" does in “moral responsibility,” and those conditions still hold under consequentialism. So I can be (and am) a compatibilist about MR (nothing contra-causal needed for being a moral agent or for being held responsible) and I can be (and am) a consequentialist (no conception of retributive desert needed to justify appropriate responses to moral infractions). All this seems perfectly coherent to me, which is why I question the standard bundling of MR, desert and retribution.

Mark,

I hope the above explains why there was no party.

[tnadelhoffer]

Sofia,

You make it sound like the hard determinist starts with the assumption that we are not free or responsible and then looks to see whether the scientific evidence can be made to fit this assumption. But I don't think that's right. The hard determinist looks at the evidence and says that given the nature of the universe and/or the nature and limitations of human cognition/agency, humans are not free and desert-based responsible. The way you framed the issue suggests that just as the compatibilist/libertarian assumes that moral desert exists the skeptic assumes it does not. But here again, I don't think that's right. The skeptic concludes that we are not free and responsible based on how the universe and/or the mind operates. This seems principally different than what's going on with some (though admittedly not all) compatibilists and libertarians. Of course, just as all compatibilists and libertarians are not created equally, the same can be said for skeptics. As such, we might want to distinguish conceptual skeptics--such as G. Strawson--from empirical skeptics--such as Baron d'Holbach. The former suggest that something about deep desert is incoherent regardless of whether the world is deterministic or indeterministic. The latter start with observations about the mechanistic nature of the universe and work backwards to deny free will and moral responsibility. But that is a story for another day...

[tnadelhoffer]

Tom,

Thanks. I agree with nearly everything you say. I think my main aversion to using "moral" to refer to the beliefs, behaviors, and practices you mention is that it just makes it more likely that the waters get muddied when arguing about free will and responsibility. You use "moral responsibility" to refer to both desert-based responsibility and non-desert based responsibility--which is obviously fine as far as it goes. I would just prefer that we simply talk about duties, obligations, liability, accountability, etc. when talking about non-desert-based responsibility rather than adding "moral" to the mix.

But at this point, our disagreement is merely (or at least mostly) terminological. If nothing else, your way of phrasing things admittedly has the added benefit of making the sort of skepticism you embrace seem less profound and far-reaching that it actually is! I suppose that makes it better for recruitment purposes :)

[Sofia Jeppsson]

Tnadelhoffer: What I meant to say was rather that the hard determinism starts out with a notion of free will or moral responsibility that is really far-out, or even incoherent, and then go "look, this doesn't fit into a scientific world picture", while a compatibilist start with more down-to-earth notions and then go "look, this does fit into a scientific world picture". Neither is more objective.

But as I wrote earlier, I didn't manage to open the file... so perhaps the article is about some particular brand of compatibilist who do things differently... Anyway, I think discussions betweeen free will sceptics and compatibilists USUALLY look like that - one camp start out with these really weird notions and the other camp start out with more down-to-earth ideas.

Then I think there are more dimensions to moral responsibility than just "desert-based" vs "consequentialist".

Take a concrete example. A couple of years ago I broke a promise to an acquantice, and made that person very sad and disappointed. My sister heard about this and blamed me. I then realized that I had behaved really badly, apologized to the person I had broken the promise to and decided never to do that again.

A pure consequentialist would say that the only reason the blame was of value in this situation, was that I was deterred from similar behaviour in the future. If my sister instead had heard about it, and promised to pay me 1000 Skr by the end of every year if I didn't break any more promises during said year, and the promise of the reward made me behave, this would have been even better on a consequentialist view, good behaviour without any initial discomfort on my part. Or suppose my sister could supervise me and give me a small electric shock everytime I broke a promise, and I behaved from fear of punishment - this might be pretty much on par with the actual scenario on a consequentialist view, at least if we (unrealistically perhaps, but not unusual in philosophical thought experiments) suppose that this didn't have any bad side-effects.
Well, I think that sometimes it can be better to just reward or scare people into behaving than letting them do as they please, but it is far better if people can actually take proper responsibility for their own actions. Since I think so, I don't think one can call my view on praise, blame and moral responsibility a consequentialist one.

But non-consequentialists can still differ from each other on what justifies blaming me. One can think that it's justified because blaming each other is, in contexts like these, a natural part of treating each other as equals and participants in a moral community. Or one can think that it's justified in virtue of some mysterious metaphysical facts. These are pretty different views, but both of them differ from the consequentialist view on blaming others as simply a way to train people to behave.

A further difference within the non-consequentialist camp is between those who merely thinks it's fitting that one feels bad about oneself after having done something bad, and those who think that there's no limit to how much suffering one can be justified in inflicting on others in the name of desert.

There are all these dimensions that cannot be captured by simply seperating a "desert-based" notion from a "consequentialist" notion.

Personally I think moral responsibility exists, since people do take responsibility for actions with a moral worth (just as I did in the example above). It's just like language exists because people communicate, and societies exist because people live together. I'm aware that some people don't think that's "real" moral responsibility, because "real" moral responsibility is a sort of weird metaphysical lump that only sticks to people in possible worlds where human beings have libertarian brains, or it's something so incoherent that it can't exist in any world at all. But I see no point in using the phrase "moral responsibility" to refer to something so strange that one can't even explain what it is.

[Mark Smeltzer]

Tom,

I don't think MR entails retributivism either. So, I don't see why there couldn't have been a party.

IMO, moral responsibility is bound up with non-consequentialist desert (the only kind worthy of being called desert), but desert is not bound up with retributivism. For example, if a criminal is MR for a crime and thus deserves to be treated as a criminal, it might be the case that the criminal deserves to be rehabilitated, given a no-more-crime serum, or put to death. In other words, what the agent deserves is a function of ethics (and of course there are consequentialist approaches to ethical questions), but the fact that the agent deserves something is a function of moral responsibility.

[Sofia Jeppsson]

Besides, I don't understand why so many people seem to think that all free will problems are solved by assuming consequentialism. Consequentialism is the view that an action is right if it has better consequences than all available alternatives, right? If one thinks that determinism precludes true alternatives, then this assumption together with consequentialism means that anything anybody ever does is trivially morally right. Endorsing a policy of torturing people to death for petty crimes is just as morally right as endorsing more liberal policies.
So if one wants to combine determinism and consequentialism without these absurd implications, one must say that in a morally relevant sense there are still different options for agents under determinism. But if one has gone that far, it's hard to see why all other normative theories would be excluded.

[Kip]

I just want to say that:

1. I disagree with Tom Clark, to the extent that I think: if you're going to get rid of free will and retribution, it's pointless to keep moral responsibility. If people don't have free will, they're not morally responsible for their behavior, not really. Someone who does a bad act is sick, or victimized (by bad genes, or bad parents, something), rather than being responsible. Holding otherwise strikes me as pretty inconsistent, in spirit if not (technically) in principle.

2. Like Thomas N., I've always found Clockwork Orange scenarios must less disturbing than others. Not just for the reasons he cites (is it really worse than prison conditions today). But, more importantly, because I think the standard for determining your general position on free will and punishment shouldn't be based on the faulty, imperfect situation we have today.

That is, the fact that the Clockwork Orange scenario (in the movie) had problems doesn't mean that a modified scenario would have the same problems. We can (*I* can) imagine an ideal, modified rehabilitation scenario. And I think that is what we should aspire to, and I think that is the standard we should measure against. And, if we are aspiring for that world, I don't think it makes any sense to be saying people have free will, are morally responsible, and deserve punishment, and then going on to give them an anti-crime pill (or whatever it is). The whole scenario is, in general and in spirit, inconsistent with the compatibilist paradigm.

[Eddy Nahmias]

OK, I read the whole damn 63 pages and 257 notes; I hope you are happy, Thomas and Tamler. I’ll start with 2 ½ positive points before I rip into it.

1. I agree entirely that there is a danger in defending human free will so strongly that it makes people (and legal institutions) blind to the causal forces that should mitigate responsibility and punishment, including “hard social conditions,” and Kaye is right to emphasize this point. (I think it would be more apt to focus on libertarians here, though I wouldn’t want to suggest that philosopher libertarians need to buy into the kinda crap spouted here: http://www.detnews.com/article/20090830/OPINION03/908300308/1008/Individual-responsibility-explains-differences)

2. It is plausible that criminal theorists could *misread* compatibilist philosophers to conclude that such theories enshrine human free will in a way that makes them blind to legitimate mitigating circumstances. After all, too few compatibilists point out that compatibilism entails nothing about whether humans actually have free will or that there may be lots of things that threaten human FW other than determinism (including discoveries from the modern mind sciences). And compatibilists sometimes devote their efforts to analyzing *sufficient* conditions for free will without analyzing the extent to which real humans satisfy those conditions (but I blame this move largely on incompatibilists, since compatibilists do this so they can say, “Look, if an agent has all THIS (complete rational self-control, sanity, identification, etc.), isn’t that enough for FW even if determinism were true?”)

3. I like that legal theorists are reading and citing philosophers. I only wish they’d understand them better. Kaye clearly misunderstands a lot of the current debate, and if he’s presenting his target criminal theorists accurately (e.g., Moore, Morse, Fletcher, etc.), then they are getting compatibilism wrong in some very dangerous ways.

Now, some objections:

Most importantly, Kaye’s main point is that “originationists” (by which he means hard incompatibilists or “skeptics”; he entirely ignores libertarians) are more likely than compatibilists to “see,” “attend to,” and “inquire into” hard social conditions and their role in causing criminals to commit crimes. That claim seems backwards in two ways. First, why would skeptics pay particular attention to the details of the causal explanations for criminal behavior, including specific deprived upbringings or specific situational influences? Skeptics say we are all in the same boat—no one has free will, no one has MR, no one deserves (genuine) credit or blame for what they do—regardless of the specifics of one’s upbringing, because we are *all* equally unable be the ultimate source (originator) of our actions. So, pointing out the particulars of a criminal’s difficult background is superfluous if the point is to show that they do not deserve blame or retributive punishment. This is clearly the case if the order of explanation is: [argument for skepticism] --> [criminals don’t deserve punishment]. If the order of explanation is (roughly) the other way around, such that attending to the causal history of a criminal (e.g., Robert Harris) makes vivid (or generalizes into) an argument for skepticism, then it’s not being an originationist (or incompatibilist) that gets one to consider the relevance of difficult causal histories, and one not end up an originationist just because one does consider such histories.

Indeed, it is the compatibilist who is best situated to attend to the specific details of a deprived upbringing. Unlike the skeptic or libertarian who suggest an all-or-nothing approach to FW and desert, the compatibilist can (or at least should) consider FW a set of capacities that can be possessed and exercised to varying degrees, and hence can and should attend to the degrees to which humans, in general, and specific wrongdoers possess those capacities and have the opportunity to exercise them in specific situations (e.g., crimes). And understanding the details of genes, brains, situations, etc. will be crucial to understanding these degrees of freedom. Kaye argues that extant compatibilist accounts (e.g., Fischer, Wolf, Wallace) are not well situated to account for the role of hard social conditions in mitigating the degree to which criminals lack these capacities and opportunities.

But I think that’s a mistake. Kaye treats these accounts as if they say: “If agent A meets (to any degree) the conditions we suggest for free will (acting on moderately reasons-responsive mechanism, being sensitive and reactive to the True and the Good, etc.), then A is *fully* free and responsible.” But that is a highly uncharitable reading of these philosophers. Rather, they are best read as saying: “To the extent that agent A meets such conditions, to (roughly) *that* extent they are free and responsible.” And it is highly plausible to think that hard social conditions generally reduce the extent to which agents possess and have the opportunity to exercise the capacities these compatibilists discuss *and* that careful scientific investigation can help us understand when and why certain conditions (genes, upbringing, situations, etc.) undermine (compatibilist) free will. Kaye gets sidetracked by the fact that compatibilists get sidetracked by responding to incompatibilist arguments (e.g., dealing with manipulation arguments and nefarious neurosurgeons). Overall, the key section IIIA3 fails.

As Kaye points out, originationists can and do attend to the same sorts of mitigating factors compatibilists might consider, but he argues, they can go further to consider hard social conditions. He’s stuck in a dilemma. Any theory of punishment will presumably need to attend to different people’s different psychological capacities. Skeptics tend to agree that compatibilist criteria are picking up on something important about agency and types of freedom. If originationists use the same sort of criteria in their own punishment practices, then any problems they claim compatibilists have (e.g., dealing with cases of addition, unfortunate upbringings, scientific discoveries), the originationists will have too. But if the originationists go on to offer some new account of how to deal with these problems, then compatibilists can simply use that account. The only account the compatibilist can’t use is the one that says no one is MR because to be MR requires self-origination. But *that* account cannot help the originationist deal with the tricky cases since it does not allow for *any* distinctions between different people or different histories (see above). So, either the compatibilist can help herself to any insights an originationist offers to handle the subtleties of hard social conditions and scientific discoveries, or the originationist “insight” simply wipes out any useful distinctions between various agents and various facts about them (e.g., their histories) and hence does not help us “see” anything about why a particular criminal should be treated any differently than another.

I want a theory that helps us see why Bernie Madoff and Dick Cheney are more deserving of blame for the evils they’ve committed than Robert Harris or the homeless addict who commits armed robbery (which is not to say they should be punished more or less, since punishment depends on lots of other factors). Compatibilists can focus on the fact that the former two jerks had privileged upbringings, have the capacities for rational deliberation, know a lot, etc., whereas the latter two jerks almost surely—but this needs to be explored not assumed—have much more limited compatibilist capacities, likely because of their upbringing. Compatibilists can offer such a theory. Originationists can only glom onto a compatibilist theory to make distinctions between these men, while saying that ultimately they are all equally unfree and non-deserving of (genuine) blame or retributive punishment. (Indeed, they may also be suggesting that it is precisely by attending to the unfortunate Harris and homeless addict that we see why Madoff and Cheney are not really any more free and responsible.)

Well, this is going on way too long, so I will not try to respond here to the interesting points raised in previous comments or go on to fume about why I think Kaye is being unfair in his political commentary in the latter sections of the article.

[Tamler Sommers]

Thomas,

Here's a slightly more succinct way of putting my point. Kaye argues that (hard) originationism will illuminate the mitigating factors that lead to criminal behavior among the disadvantaged. As you note, consequentialist do have to take these factors into account in a way that (perhaps) compatiblists do not. The gap in the argument, it seems to me, is that he doesn't give any reason to believe that causal factors relating to a disadvantaged circumstances would mitigating rather than aggravating.

(Of course, it terms of desert, the disadvantaged are exonerated but so is everyone else. It's the claims about social factors being "mitigating" factors that bother me...)

[tnadelhoffer]

Eddy,

First, you claim that Kaye “clearly misunderstands” a lot of the current debate, but you don’t actually point out where you think he clearly goes wrong. I, for one, thought he did a pretty fair and thorough job given his target. So, I am curious to hear how and why you think he has misstated or misrepresented the debate as far as compatibilism is concerned—especially with respect to the different ways compatibilists try to ground responsibilities in various cognitive capacities. Of course, I am sure you're not a fan of his claim that compatibilist free will isn't *genuine* free will, but nearly every incompatibilist could tell you that :)

Second, you make the following claim:

“Skeptics say we are all in the same boat—no one has free will, no one has MR, no one deserves (genuine) credit or blame for what they do—regardless of the specifics of one’s upbringing, because we are *all* equally unable be the ultimate source (originator) of our actions. So, pointing out the particulars of a criminal’s difficult background is superfluous if the point is to show that they do not deserve blame or retributive punishment.”

That’s true, as far as it goes, but these circumstances will be relevant to the project of trying to prevent crime and rehabilitate offenders—which is an issue that I already mentioned earlier in this thread. In short, an orginiationist is forced to give up desert and focus instead on prevention, rehabilitation, and deterrence. And while some compatibilists *could* go this route as well, nothing about compatibilism forces them to opt for this approach rather than simply holding on to the retributivistic system we currently have in place. I take it that Kaye’s point is that so long as the law adopts a compatibilist concept of responsibility, people will not come to see the true criminogenic nature of poverty, child abuse, etc. So long as we are *fully* responsible for breaking the law provided that we are minimally reasons responsive—which is the current legal standard of liability!—this makes it less likely that people will appreciate the true relationship between poverty and crime. As a result, the sort of progressive changes that we would need to make at the societal level in order to actually deal with the problem of crime in a scientifically minded way will continue to be overlooked in our drive to give offenders what they deserve. It is in this sense, that compatibilism, unlike originationism, is not “politically innocent.” It still peddles in desert, thereby distracting us from focusing on the real problem of crime—i.e., it moralizes a problem that should be instead be medicalized.

Third, I find your “degrees of freedom” defense of compatibilism on this front to be unpersuasive. You state the following:

“But I think that’s a mistake. Kaye treats these accounts as if they say: “If agent A meets (to any degree) the conditions we suggest for free will (acting on moderately reasons-responsive mechanism, being sensitive and reactive to the True and the Good, etc.), then A is *fully* free and responsible.” But that is a highly uncharitable reading of these philosophers. Rather, they are best read as saying: “To the extent that agent A meets such conditions, to (roughly) *that* extent they are free and responsible.” And it is highly plausible to think that hard social conditions generally reduce the extent to which agents possess and have the opportunity to exercise the capacities these compatibilists discuss *and* that careful scientific investigation can help us understand when and why certain conditions (genes, upbringing, situations, etc.) undermine (compatibilist) free will.”

I think his case is stronger than you are letting on. Let’s take burglary as an example. On your degrees of freedom view, whether someone is fully free and fully responsible for burglary will depend on their “genes, brains, situations, etc.” That’s fine as far as it goes. But then who ends up being fully free and responsible for burglary on your view. For the sake of argument, let’s (safely?) assume that the overwhelming majority of burglars had rotten social backgrounds. Would it follow on your view that the overwhelming majority of burglars are not fully free and responsible? Would you set up the criminal law in such as way that a burglar’s sentence would depend on whether or not he had a rotten rather than a privileged background? Would you say that a burglar born of poverty *deserved* less punishment? Of course, you *could* say this. But I find it hard to believe that many of your compatibilist kin would agree. But I will leave it to them to make their voices heard. I, for one, think that if you actually look at the specific conditions for responsibility as they are defined by compatibilists—as Kaye has very carefully done—it is unclear why we should think that a rotten childhood matters unless is causes *extreme* cognitive and volitional impairments. Origionationists, on the other hand, have an easier time accommodating less severe kinds of criminogenic sources. Once you give up on the idea that the criminal law is administering desert, you can actually turn your attention to figuring out how to best prevent crime and rehabilitate offenders. And, as I have already mentioned both in this comment and earlier in this thread, this tasks require us to examine the very kinds of factors—“genes, brains, and situations”—that you misleadingly claim the originationist must ignore. In one sense, these things are indeed irrelevant from the originationist point of view. But in another more important sense, they will be the *only* things that matter when it comes to addressing the problem of crime.

Fourth, you try to foist the following dilemma upon Kaye:

“If originationists use the same sort of criteria in their own punishment practices, then any problems they claim compatibilists have (e.g., dealing with cases of addition, unfortunate upbringings, scientific discoveries), the originationists will have too. But if the originationists go on to offer some new account of how to deal with these problems, then compatibilists can simply use that account…So, either the compatibilist can help herself to any insights an originationist offers to handle the subtleties of hard social conditions and scientific discoveries, or the originationist “insight” simply wipes out any useful distinctions between various agents and various facts about them (e.g., their histories) and hence does not help us “see” anything about why a particular criminal should be treated any differently than another.”

This only works if you have successfully shown that compatibilism has a way of defining free will and responsibility in such a way that a rotten social background is mitigating or exculpating. As it stands, you have not provided us with such an account. Instead, you have simply asserted that, of course, compatibilists *should* care about these things. But Kaye has presented textual support of a number of compatibilists who explicitly reject the very claim you are making on behalf of compatibilists. As such, you would either need to provide examples of compatibilistically grounded notions of free will and responsibility that make room for socio-criminogenic sources of crime from the literature or you would need to flesh out your own “degrees of freedom” view in a way that showed that your view, unlike other versions of compatibilism, can accommodate these kinds of mitigating/exculpating factors. Simply stating that compatibilists should care doesn’t show that their views have the conceptual machinery needed for the task at hand. Until someone shows that Kaye is wrong on this front—by argumentation and not stipulation—I still find his claims about the political non-innocence of compatibilism compelling.

[Sofia Jeppsson]

Let's be clear here about the difference between "compatibilism IMPLIES using harsh and terrible punishments on criminals", and "compatibilism does NOT imply that we should ABOLISH the use of harsh and terrible punishments on criminals". If the former was true, that would be really problematic for the compatibilist. But it's not. It's the second one that is true. And that just means we have to go to ethics per se rather than free will theory to find arguments against harsh and terrible punishments. What's the problem?

It's the same thing for free will scepticism, really. The only thing scepticism implies is that one cannot justify harsh and terrible punishments by appealing to free will and desert. But nothing in free will scepticism per se implies that one must replace the current american system with something that is NICER to people. The current system may be terribly inefficient if the goal is to lower the crime rate. But psychology and the social sciences has taught us that there are many EFFICIENT ways to change and control behaviour as well, and not all the efficient ways are nice. To argue for a NICER system, not just a more efficient one, one would still need to bring in ethics.

In my own country, from the thirties to the seventies, people could be sterilized against their will if medical expertise believed they carried bad genes. During the later decades of this policy a lot of the focus lay on preventing crime, by sterilizing people one believed to have a hereditary tendency to criminal behaviour. Now the empirical basis for these decisions was probably pretty bad, but with better genetic knowledge we might in the future be able to carry out such a program with greater success. Just a concrete example to show that to arrive at a humane policy on how to treat criminals, one needs something else than free will theory.

[Sofia Jeppsson]

Tnadelhoffer, you wrote: "As such, you would either need to provide examples of compatibilistically grounded notions of free will and responsibility that make room for socio-criminogenic sources of crime from the literature or you would need to flesh out your own “degrees of freedom” view in a way that showed that your view, unlike other versions of compatibilism, can accommodate these kinds of mitigating/exculpating factors."

I'm not quite sure that I follow here. I'd dare say that most people, or at least lots and lots of people, assume that a rich kid with a loving family who one day steal a jacket from a shop, citing "excitement" as his reason, is MORE culpable than a poor kid who's neglected at home and don't have the money to buy fancy clothes. Are you saying that this view is incoherent? I mean, assuming that libertarianism can't be made to work, this view must be incoherent if it can't be accommodated within a compatibilist framework, since according to a free will scepticist they both have zero culpability.

[tnadelhoffer]

Sofia,

First, I am unsure you're right about the way intuitions track here. If you used food instead of a fancy jacket, you might be right. But given the example you chose, I suspect many people would find them equally liable for legal sanction. Of course, that is an empirical claim. So, since we don't have the data at hand, I am going to set it aside for now. Instead, I want to point out that I did not make any claim about coherence or lack thereof. What I did say is that if Eddy and other compatibilists want to claim that they can accomodate socio-criminogenic factors as mitigating or exculpating, they need to explain how the conceptual machinery of their views accommodate this. As it stands, it looks like in order for these factors to mitigate or exculpate, they have to cause very extreme cognitive or volitional impairments--so extreme, that the mitigating/exculpating factors will be very rare. I am not claiming that compatibilists cannot accomodate these factors, I am simply suggesting that as it stands none of them have (at least not in the this comment thread).

That being said, I am not sure you are framing your own claims about coherence (or lack thereof) very clearly either. Of course, the skeptic cannot accommodate folk intuitions about retribution/culpability--since the skeptic reject this framework. The skeptic would nevertheless suggest that (a) we ought to adopt social programs that make it less likely that poor people are led to break the law, and (b) we ought to treat offenders on an individualized basis such that people with different "genes, brains, and situations" will get the particular attention they need to eventually be reintegrated with society. Now, it is of course true that a compatibilist can opt to focus on these as well--but then the compatibilist ends up with a strange commitment to a legal system that, according to her own account of moral responsibility, does not give people what they deserve. That is a tension that does not arise for the skeptic. I also believe it is that tension that Kaye has correctly picked up on...

[Eddy Nahmias]

Take a view like Fischer and Ravizza's that identifies MR with guidance control and analyzes guidance control in terms of capacities to recognize reasons and respond to reasons. I'd have to look to see if they explicitly talk in terms of degrees, but in any case, such an account is clearly well-suited to say that different agents possess the capacities for guidance control to varying degrees and have the opportunity to exercise these capacities in specific situations to varying degrees. If so, then empirical data can inform us about the degree to which a particular agent (e.g., criminal) possesses the relevant capacities and had the opportunity to exercise them in her situation. An agent will then be less *deserving* of retributive blame to (roughly) the extent she lacks these capacities and opportunities. (I say "roughly" because other factors may also influence degrees of desert; and I'm not suggesting F&R or anyone else has a fully worked out account of how to analyze the capacities and opportunities or the degrees to which they are possessed or exercised--this is hard stuff.)

Crucially, this degree of desert will not simply entail lesser punishment(s), since other issues will come into play here (e.g., deterrence)--this seems to be one of Sofia's points. But it should suggest different reactions to the agent and likely different *kinds* of punishment (or telishment or whatever), not to mention rehabilitation where possible (and it should suggest we react to the agent with diminished guidance control differently than someone like Madoff or Cheney who possess the relevant capacities to a greater degree; don't you feel something is lost if we can't blame those bastards in a way we don't want to blame those with diminished capacities and/or miserable upbringings??).

There is every reason to think that certain "genes, brains, and situations" will cause/constitute lesser degrees of guidance control. In general, it will not be the mere *existence* of a bad upbringing that will mitigate desert, but the causal outcomes of that bad upbringing--that is, in the presumably very common cases where such bad social conditions will break down agents' abilities to be receptive and responsive to reasons.

If philosopher-compatibilists have not drawn enough attention to these auspicious implications of their accounts, that is too bad... but fodder for my work! If legal theorist-compatibilists have denied or contradicted these implications, then they may be doing so for the political reasons Kaye suggests and should be taken to task (I still find it strange that libertarians do not come into the picture here). But I see no textual evidence that the *philosophers* have advanced the problematic conclusions Kaye suggests. As I said, he misreads them when he says their view commits them to saying: "If agent A possesses *any* degree of the compatibilist capacities, then A is *fully* deserving of blame/retribution." Thomas, you seem to be making the same mistake when you say, "in order for these factors to mitigate or exculpate, they have to cause very extreme cognitive or volitional impairments." Indeed, "mitigate" is a degree term; compatibilists can offer detailed accounts of mitigation. Originationists have no use for mitigation. Everyone is entirely exculpated; there are no mitigating factors.

[Anders Kaye]

I thought I’d (very tentatively) introduce myself into this discussion at this point. (I’m the author of the piece.) As you can probably imagine, I’ve got mixed feelings about doing so, and I’m afraid the most likely outcome is that I will let down those who have defended some of the ideas in the piece, while confirming the critics’ doubts. I do want to say at the outset that I think Thomas, Kip, and Tom (despite his significant reservations about the piece) have done a really wonderful job of presenting and polishing the core claims in the piece – better than I would have done – and that those critical of the piece have done an equally wonderful job of spotlighting its deepest fault lines (while generously putting aside its many more not-quite-as-serious problems).

As one further preliminary, I’d also like to segregate my piece from the work of other legal theorists writing about responsibility, determinism, compatibilism, and free will. To the extent that my discussion suggests those other theorists make unfortunate mistakes, it is vastly more likely that I have misinterpreted them than that they have actually made those mistakes.

[Finally, I want to apologize now for the length of this comment; now that I move it from Word to the web, I see that I should probably have done this one bite at a time.]

Now, about the piece. You all have raised so many interesting concerns about the piece that I don’t know where to start. It might be helpful (not so much to convince you of anything, but just to clarify where the piece is coming from) to say something about one premise of the political analysis in the piece. The premise has to do with what lay people in our culture think about determinism. My view is that most lay people are not convinced of determinism, and therefore think that origination is possible. To complicate things further, however, I also believe that most lay people are open to case-specific causal explanations of particular acts, and that when convinced of such explanations, they conclude that those acts are not original. In legal theory the view I’m attributing to lay people is sometimes called (usually pejoratively) partial determinism. Privately, I think of it as “show-me determinism,” to highlight the way it combines suspiciousness about determinism with receptivity to causal explanation. I think that many lay people in fact have these sorts of “show-me determinist” inclinations, and that they are cultivated and encouraged by significant cultural phenomena.

It is against this backdrop that compatibilism looks most like (as Kip suggests) a sort of super-bacteria evolution from libertarianism. For the show-me determinist, compatibilism is an account of responsibility that better resists the incremental disintegration of responsibility in a time of incrementally increasing causal explanation (by which I mean, a time in which the menu of intellectually accessible causal explanations that can be fit closely to particular acts is perceived to be expanding, in part due to the expansion and popularization of the behavioral sciences). And it is against this backdrop that the choice between originationist and compatibilist criminal justice institutions has the most immediate political bite.

With that clarified, it may be easier to answer some of the questions Gardners have raised in this discussion. I’ll try to address three; but please forgive me if I misconstrue or misunderstand those questions.

First, I hope it makes more sense now that I don’t think criminal justice institutions that adopt the originationist perspective will immediately have to give up punishment, or shift to a consequentialist perspective on punishment. While an originationist-determinist might be faced with those choices, an originationist-show-me-determinist criminal justice system would not. Instead, such a system will assume most acts are original and responsible, but will provide opportunities to argue that particular acts are not original and thus not responsible. Unless and until criminal justice actors come to believe that “all has been shown,” the system can and will continue to issue punishments on originationist and retributive grounds (though the frequency of such punishments may gradually diminish as the universe of case-specific causal explanations expands).

Second, I hope it is also clear why I do not think that determinism renders the originationist criminal law immediately unable to distinguish between conduct traceable to hard social conditions and other conduct. It is true that an originationist-determinist would have to concede that all conduct is unoriginal, and thus could not rely upon the origination criteria to distinguish one group from the other. But, again, I take our contemporary criminal justice institutions to have show-me-determinist inclinations rather than determinist ones. If this is correct, then the originationist criminal law will, at least for the time being, identify some acts as original and others as not-original (at least until “everything has been shown”). There will not be an immediate blanket determination that “nothing is original,” and no immediate need to lump both groups of conduct under such a blanket.

A third concern (if I read the comments correctly) is that originationism does not actually provide better tools for attending to the influence of hard social conditions than compatibilism. Indeed, as several people have said, their sense is that compatibilists tend to be more sensitive to social condition arguments than libertarians. (The same observation was made to me before publication as well). I do not have a quick or sound-bite response to this concern; but, as readers of the paper know, I dedicate a very long chunk of the paper to arguing that compatibilism has a noteworthy blind spot here. Let me mention two pieces of that argument:

(1) One piece of the argument is that originationists can attend to everything the compatibilist attends to, plus something more (the origination piece). That is, the originationist can take the position that responsibility requires everything that Fischer and Martin (or Wallace, or Wolf, or …) say it does – and one thing more (origination). By making origination a prerequisite to responsibility, the originationist opens an additional line of inquiry that the compatibilist has foresworn, without giving up any of the inquiries the compatibilist has embraced. And, as it happens, as an added “bonus,” this additional line of inquiry is especially well-suited to “hear” claims about hard social conditions, in that many such claims are (in practice) formulated in causal terms, perhaps because they are so often inspired by behavioral science research expressed in such terms. In this sense, an originationist evaluation of responsibility can (at least in theory) take account of everything a compatibilist evaluation can, plus a little more; and the “little more” happens to be especially suited to taking account of hard social conditions.

(2) A second piece of the argument, more fully developed than the first, is that the compatibilist accounts of responsibility that we actually have right now are fairly resistant to claims of non-responsibility grounded in hard social conditions. I do not mean that these theories are hard-hearted or disdainful of such claims; on the contrary, the theorists’ whose work I discuss seem to show a strong predisposition to compassion regarding such claims. Nevertheless, as I attempt to show in some detail in the piece, their accounts of responsibility do not actually make much room for such claims. That is, they are structured in a way that would render most realistic versions of such claims irrelevant to evaluations of responsibility. Of course, I am painfully aware that I may be horribly wrong in this reading of their work. But I am also so addictively interested in and curious about it that I would be grateful for arguments to the contrary. What have I overlooked in Fischer & Ravizza’s handling of hard social conditions? In Wallace’s? Etc.?

(A separate concern would be that it is inappropriate for me to generalize from the compatibilist accounts we have today to something intrinsic in compatibilism itself. I do try to address this in the paper, and would be happy to talk more about it here too.)

Finally, all that being said, I also want to be forthright about my personal perspective. As some of you have speculated, my personal views (at the moment, always subject to revision) include the views that (1) origination is a prerequisite for responsibility, and (2) origination is impossible. If I came to believe that my fellow citizens, my culture, and my criminal justice institutions shared this set of views, I would have to reevaluate my claims about the political function of compatibilism. (After all, its political function is inevitably dependent on the cultural context.) I also think it is fair to ask a person with my views how he would approach punishment. On this issue, I think Tomas, Kip, and Tom have given very appealing answers. I, myself, would probably lean toward the view (always subject to revision) that the practice of punishment cannot be justified, and that (morally speaking) crime prevention should be accomplished in other ways. But, to be clear, the argument in the paper reflects my belief that most people do not share my views about determinism – that most people are, instead, show-me-determinists -- and that the political functions of views like originationism and compatibilism should be evaluated against the backdrop of what people in our political culture actually believe, rather than what I personally believe or what I would urge others to believe.

I’ll stop there, with apologies for being so long-winded, and for nevertheless failing to address so many of the other very tough questions Gardners have raised in this thread!

[Kip]

Eddy,

Let me address a couple of your points (although I think Thomas N. already did a good job):

First, why would skeptics pay particular attention to the details of the causal explanations for criminal behavior, including specific deprived upbringings or specific situational influences?

Because skeptics follow the regress back farther. That's why the Basic Argument is in the form of a regress (an infinite regress). Why did John Doe do X? Because he wanted to. Why did he want to? Because he had genes that made him want to? Why did he have those genes? Because the universe had a certain structure at the beginning of time, leading to John's having that DNA today. And so on.

Compatibilists don't follow that regress back as far as anti-realists do. At some point (it varies dramatically between compatibilists), the compatibilist just stops caring. Why did John want to do X? It doesn't matter. He does. That's what Frankfurt meant when he wrote:

A manipulator may succeed, through his interventions, in providing a person not merely with particular feelings and thoughts but with a new character. That person is then morally responsible for the choices and the conduct to which having this character leads. We are inevitably fashioned and sustained, after all, by circumstances over which we have no control. The causes to which we are subject may also change us radically, without thereby bringing it about that we are not morally responsible agents. It is irrelevant whether those causes are operating by virtue of the natural forces that shape our environment or whether they operate through the deliberate manipulative designs of other human agents (2002, p. 27).

Frankfurt is talking about manipulators there. But the point applies to compatibilism in general: after a certain set of conditions are met, compatibilists stop caring how agents came to be who they are. They become blind to the past, after a certain depth.

Anders's point is that (1) this blindness includes blindness to hard social backgrounds and (2) the blindness to hard social backgrounds is ethically undesirable/incorrect.

And it is highly plausible to think that hard social conditions generally reduce the extent to which agents possess and have the opportunity to exercise the capacities these compatibilists discuss *and* that careful scientific investigation can help us understand when and why certain conditions (genes, upbringing, situations, etc.) undermine (compatibilist) free will.

I disagree. It is entirely possible (perhaps even likely) that both: (1) hard social backgrounds do not significantly disturb any of the conditions for compatibilist free will that compatibilists' generally endorse and (2) hard social backgrounds leads to very unfortunate things happening to these criminals and their victims.

Those two points can mutually coexist. That's a major point of the article. If you can't see how the points can coexist, of course you will find the article unpersuasive. But I think that just represents a failure of imagination.

I think there is a general parallel between the "cosmic designer" scenario (that I am always talking about) and the hard social background that Anders is talking about. As incompatibilists love to point out, a designer can design an agent to (1) satisfy the compatibilist criteria of your choice while also (2) leading the agent to commit a crime, and generally cause and experience tremendous suffering. Similarly, (1) a hard social background can leave a person's compatibilist free will entirely intact while also (2) leading the agent to commit a crime, and generally cause and experience tremendous suffering.

Your passage above suggests to me, Eddy, that you believe the crimes and suffering can't happen without the agent lacking (full) compatibilist free will. That's just not true.

Further, and as Anders's article goes on to explain: we wouldn't expect it to be true. Compatibilism is largely directed to making the world safe for blame (and praise!). To the extent that you think compatibilism can accommodate Anders's concerns about blaming people with hard social backgrounds, you're also undermining that (tacit) purpose of the compatibilist project. Even if you think compatibilists can and should reduce the blame given to criminals with hard backgrounds (and I wonder how far you would follow through with that), I think it's unlikely that most compatibilists would follow you, precisely because they would then be undermining the whole point of compatibilism (making the world safe for blame).

My only last point is that your previous comment repeatedly says that compatibilists can do this, or compatibilists can do that. The question is not can they, but do they. Anders is attacking compatibilism in practice, not in principle. And, to the extent that you think compatibilists are guilty of the errors he notes but can modify their views to accommodate his criticisms (which is already a huge concession), Anders gives fairly compelling reasons to think that they won't do so, as explained above. To be precise: compatibilists won't modify their views in a way that significantly diminishes or undermines the amount of blame in the world.

[Justin Coates]

*I haven't read all the comments (I've been grading), so if this has come up, please forgive me and disregard this comment*

Let's distinguish between two things: first, when an agent is not morally responsible for A-ing, and second, when it is inappropriate to hold an agent morally responsible for A-ing. Let's suppose that an agent's failing to be morally responsible entails that it is inappropriate to hold her responsible. So originationists have a good explanation of why certain histories make punishment inappropriate; namely, the agent in question isn't responsible.

But look, many times it is inappropriate to hold an agent morally responsible even if she is responsible. Why do we feel less indignant towards Robert Harris when we discover his history? Well, Watson explains the shift in our attitudes by appealing to our ignorance of how those conditions might have impaired his current agential capacities. And furthermore, we realize the pervasiveness of luck--we too could have become monsters if we had been raised in such a situation. As a result, our ignorance of Harris' agential capacities, as well as our recognition that we might very well lack the standing to judge Harris, should cause us to temper our reactions, to adjust our punishment. Notice however, what makes it inappropriate to hold Harris responsible in this case isn't that he isn't responsible--for all we know he is morally responsible. Rather, ignorance on our parts makes it inappropriate to hold responsible, and that is perfectly consistent with compatibilism.

And Kip,
You say Kaye's objecting to compatibilism in practice. Well, then the compatibilists y'all know are boneheads (but I doubt this, so I expect it's more likely that since this topic rarely comes up, they don't explicitly address Kaye's worries). Moreover, he needs to address Watson (I saw he cites the relevant article, but there is no substantive discussion) and a forthcoming paper by Tognazzini and Fischer, who do an excellent job of showing that it may be inappropriate to hold agents responsible (by engaging in retributive practices of punishment) even if they are morally responsible. And as long as that's a genuine distinction, compatibilists have a perfectly coherent explanation of why deviant social histories can make retributive punishment inappropriate without making the originationist mistake.

To sum up: it's hardly a knock on compatibilism that "the practice of punishment cannot be justified" in some or all cases in which individuals come from deviant environments. Why? Because showing these instances of punishment to be unjustified isn't the same as showing that the agents in question aren't morally responsible.

I also found it humorous that in my limited experience, almost all the "originationists" that I know (and I know quite a few) are quite conservative politically, while almost of the compatibilists I know are very liberal, and in general, are much more sensitive to these issues than my originationist friends. Of course, this doesn't answer Kaye's conceptual objections, but if my experience generalizes, then it might assauge some of his concerns about us compatibilists. Now, time to go watch some FoxNews (sic).

[Mark Smeltzer]

Anders,

In the second sentence of the second paragraph of the paper, you write, "Since there is no desert without responsibility, those who say we have a retributivist criminal law must also say something about how the law determines responsibility."

If you could be persuaded that desert logically precedes responsibilty (which is the reverse of what you claim), do you think that accepting this reordering would have a significant impact on your overall view?

That is actually the view I hold, and if you've read your Zimmerman (e.g. Zimmerman, M., 2002, “Taking Luck Seriously”), you'll have a rough idea of where I am heading with this...

[Mark Smeltzer]

Kip,

Surely compatibilists can draw a defensible, valid distinction between looking at things from the objective view of compatibilist theory, and the practical view that a legal system must take.

From the objective view of compatibilist theory, we (might) conclude that there are cases when hard social conditions (HSCs) do not exculpate and that there are cases when HSCs do exculpate (in the very piece you quoted, Frankfurt said that a manipulator can succeed in giving an agent a new character but not that the manipulator will *always* succeed in doing so -- when the manipulator fails, the agent is exculpated), and that it will be hard to tell the difference between the two kinds of cases in practice.

From the practical view of a legal system, compatibilists would have strong reason to suggest caution and careful review of historical factors in a given case (because the objective theory tells us that there are some cases that do exculpate and some that do not, and that it is hard to tell the difference). As a result, this cautionary approach is aimed at dissecting the historical factors in a given case (including HSCs) and determining whether they are exculpating or not.

Kip, you cannot actually believe that compatibilists are as simple minded as you portray them. This distinction is so obviously necessary that it sincerely pains me to have to defend it (again).

[Sofia Jeppsson]

Kip, you wrote: "First, why would skeptics pay particular attention to the details of the causal explanations for criminal behavior, including specific deprived upbringings or specific situational influences?

Because skeptics follow the regress back farther."

This is no answer. A sceptic has a GENERAL regress argument that appeals with equal force to all people. Therefore it's completely superfluos to pay attention to facts that DIFFER between people. So a free will sceptic could say about a person from bad social background commiting a crime, that "he was determined to do that, by the laws of nature and the past, and the past includes a bad social background". The point is, we could pick another criminal from a great social background, and the free will sceptic would have to say exactly the same thing about him: "He was determined to do that, by the laws of nature and the past, and the past includes a great social background". The free will sceptic doesn't have any tools within his free will theory for differing between these people. They're on par. Both were equally determined, despite one person having a lousy background and the other one a great background.

Tnadelhoffer: I guess I read too much into your text, when I asked if you claimed that it's incoherent to think that some people are more culpable and others less so. And never mind the jacket example... I think most people would find the poor person from a bad environment less culpable, but first of all I don't have any empirical data to back this up, second folk intuitions may differ between (for example) USA and Sweden, third there may be pragmatic reasons for judging them alike in a legal system even IF most people agree that they differ in culpability. So never mind this example. But my main point wasn't that it's a failure of free will scepticism that it cannot accommodate for common folk intuitions, since it's not aiming to do that to start with. I just wanted to know if you think there's actual incoherence in claiming a difference in culpability between different people, and the only coherent options are either that they're both non-culpable or that they're both equally culpable as long as none of them have a serious mental defect. But apparently you didn't want to make as strong a claim as that, you were just wondering whether there are any compatibilist theories that can easily accomodate culpability in degrees.

On that, I think Eddy is right in claiming that it's pretty easy to work in mitigating circumstances into a compatibilist framework. One way is to look at things like reason-responsiveness. A maximally reason-responsiveness person would only change his mind on something if new relevant and well-supported reasons were brought to him, and would always have a rational explanation for his action. A maximally continent person would always do what he judged best. Nobody completely lives up to this ideal, but some people may fall short of this more drastically than others, without being mentally handicapped. Some people may more easily be swept away by sudden mood swings and therefore act incontinently. They may have a harder time recognizing what's a relevant reason for what, be less reflective etc. If you look at what for example Alfred Mele writes about compatibilist requirements for free will, he expressively says they come in degrees.

Besides this, it's plausible to praise people more for managing to do something which is really hard, and blaming them more for failing to do something easy. Imagine two agents who have the same general level of reason-responsiveness, continence and the like. Now place one of them in circumstances where there's a lot of pressure on him to become a criminal (lots of criminal friends, peer pressure), and place the other one in circumstances where it's piece of cake not to become a criminal. It might be that it's possible for both of them to stay away from criminal life (meaning, say, that there are close possible worlds where the agents don't differ in their personality traits and general psychological make-up but still stay away from crime). But for the first agent it would have required a huge determination and strength of will, while for the second agent it wouldn't have required anything like that. There's really no problem for a compatibilist to say that in this context the second agent would be more blameworthy for criminal behaviour than the first agent, simply because one is more blameworthy for failing an easy task than a hard task.

And finally I want to repeat a point I made earlier. You seem to think that a free will scepticist is forced on pain of inconsistency to not just be against a desert-based punishment system, but also to be against all kinds of cruel treatments. But one can be cruel in a non-desert-entailing way... For example, trick people into believing they're having their appendix removed and then remove their ovaries instead since they're carrying criminal-tendency-genes. Whether one is a free will sceptic or not, one needs to bring in ethics per se in order to guard against a cruel criminal policy.

[Sofia Jeppsson]

I earlier asked if there's any empirics on how views on free will might or might not correlate with views on retributive punishment in the legal system - and I was perhaps unclear, but I meant among PHILOSOPHERS, who have explicit standpoints in free will theory, not lay-people that can be primed this way or that. That would be pretty relevant for this discussion, wouldn't it? At least for debate between Kip and Justin... (I'd frankly be surprised if it turned out that compatibilist philosophers to a great extent favour retributive punishments.)

[Cihan]

Hello everyone,

I'd like to chime in with my legal issues to this legal topic. Tomorrow, I have a mandatory court appearance for a speeding ticket. I'm representing myself.

As I have no other defense, I will be doing the no-free-will defense :D. Right now I'm reviewing the basic argument by Strawson. I will also refer to this paper. I will also quote from Clarence Darrow.

Hey, I'm glad all that philosophy paid off somehow.

More on this as it develops...

[Eddy Nahmias]

Anders, thanks for joining us. And thanks for the paper. Despite my criticisms, I found it impressive in terms of its depth and breath of research and its ambitious argument. And, as my comments suggest, I think it is valuable in making compatibilists consider how better to analyze the capacities and opportunities relevant to freedom and desert, and in such a way that they can explain, as I think they should, why certain conditions mitigate freedom and desert. This will include paying close attention to empirical information about how certain hard social conditions tend to produce agents who possess the relevant capacities to diminished degrees and who have diminished opportunities to exercise them.

But this leads back to my central question for you (and Thomas): Can you explain why the presence of hard social conditions (HSCs), *in and of itself*, should exculpate (or even mitigate) desert? Putting aside weird thought experiments, we can all agree that HSCs are (extremely?) likely to lead to diminished CCs (compatibilist capacities, leaving opportunities aside for now). This makes it hard to pinpoint whether any intuitions we have about diminished MR (moral responsibility) and reactive attitudes are picking up on the HSCs or the diminished CCs or both, which is related to the epistemic point Justin raises.

When I try, then, to test my intuitions (e.g., by weird thought experiment), I find no intuitions (or reasons) to mitigate MR if an agent has HSCs but nonetheless has fully functional CCs, whereas I find it intuitive to mitigate MR if an agent does not have HSCs but does have diminished CCs (e.g., brain damage)--and I have the strong intuition that I want to make big distinctions between criminals who have both HSCs and diminished CCs and criminals who have neither (e.g. Madoff). I find the thought experiments implausible because I think the connection between HSCs and diminished CCs is hard to break, but I find myself thinking that MR should track (roughly) the degree to which the CCs are diminished, not the degree to which the HSCs were hard. So, how do y'all explain your intuitions or conclusions that HSCs alone should exculpate--and can you do so in a way that does not rely on a more general argument for universal exculpation (i.e., hard incompatibilism)?

This leads to another way of putting the question, going back to my concern in first post about the order of explanation. If one was not already convinced that origination is required for MR but also impossible (i.e., hard incompatibilism), why would one think that HSCs per se should mitigate desert?

[Tom Clark]

"So, how do y'all explain your intuitions or conclusions that HSCs alone should exculpate--and can you do so in a way that does not rely on a more general argument for universal exculpation (i.e., hard incompatibilism)?"

Let's assume the HSCs explain why the person with (let us assume) normal CCs has criminogenic values, criminal behavioral propensities, and a norms-transgressing character. Without the HSCs, she wouldn't have engaged in crime. Thus the crime isn't solely the result of the exercise of her CCs, in which case responsibility for it is distributed outside the agent, lessening her culpability. This is, I suspect, one reason why juries sometimes opt for lesser penalties when hearing about the causal role of HSCs in determining behavior. This is an example perhaps of Anders' "show-me determinism."

This is all in the context of desert-based MR and justice of course. If you stop thinking in terms of desert, and simply think about causes, then HSCs stop being mitigating excuses (there is no desert to be mitigated) and instead are simply causal contributors to crime needing to be addressed, just as the agent needs to be addressed by means of non-retributive interventions that take into account the status of her control capacities. What’s not to like?

[Anders Kaye]

Eddy,

You've presented so many really hard questions that I think I can only make a very preliminary start. Let me give first cut answers to two of your questions, one having to do with degrees of responsibility, the other having to do with the significance of HSCs-in-themselves for desert. (I’ll do it in two separate comments.)

1) Degrees of responsibility. You've suggested that compatibilists are interested in degrees of responsibility, and that this opens the door to compatibilist consideration of HSCs. I agree that at least some of the compatibilists I discuss anticipate that responsiblity will be measured on a continuum, and that at least one (Wallace) explicitly states that HSCs may be relevant to where a person falls on that continuum. (I say something very cursory about this in footnote 125 in the piece.) But I think much more would need to be said here.

For one thing, as Kip and Thomas have anticipated, I don't think HSCs would turn out to be relevant to responsibility – not even to degrees of responsibility -- on the compatibilist accounts of responsibility I discuss. My skepticism here rests in part on an empirical premise: that HSCs generally do not diminish the capacities these compatibilists consider relevant to responsibility. My view is that people exposed to HSCs almost always emerge with about the same cognitive and executive capacities as other people, that they have about the same degree of moral sensitivity, and so on. (As I say in the piece, see pp. 397-98, I take the criminogenic effect of HSCs to lie in their influence on desires and attitudes, rather than capacities. I don’t know if this strikes others as a peculiar empirical premise; FWIW, it comes (for me) from various theories of antisocial conduct, and (more anecdotally) from a potpourri of ethnographies, autobiographies, from my impressions of my clients when I was a public defender, and my personal experiences with the role of HSCs in my life and the lives of others I have known. Of course, I realize that many if not all of these sources may be suspect.)

I have other reasons for being skeptical here, too, though I have not thought these reasons through. One source of skepticism for me is that the discussion of degree responsibility in the compatibilists I discuss is exceedingly sparse – a few hundred words in hundreds of pages (I think – but correct me if I’m wrong). For me, this raises two concerns: (1) are these authors really invested in degrees of responsibility? (If not, that might support my claims about the “motivations” for compatibilism.);(2) perhaps it is actually quite difficult to operationalize the degrees of responsibility idea in the framework of these compatibilist theories (i.e., more difficult than it sounds). (In this vein, I would be really interested to see someone sketch out a compatibilist approach to degrees of responsibility. What are the landmarks on the continuum? What is the significance of falling near these landmarks? How does the continuum translate into criminal justice adjudication?) But, as I said, I haven’t really thought these concerns through, so it won't shock me if they dissolve under scrutiny.

[Anders Kaye]

(2) “Can you explain why the presence of hard social conditions (HSCs), *in and of itself*, should exculpate (or even mitigate) desert?”

I’ll give two answers, both of which I fear will only be frustrating. Apologies if they are.

The first phase of my answer is to flag that there may be a sense in which this question is extraneous to my project (the political analysis). I have proposed that the choice between a compatibilist criminal law and an originationist criminal law has political significance, by which I mean primarily it has ramifications for the allocation of power and resources in society. In support of this claim, I have characterized compatibilism as less sensitive to HSCs then originationism, and argued that (as a result) choices between compatibilism and originationism can have a significant influence on the way people think about crime, its causes, and its solutions. (Primarily by encouraging people to see crime as driven by individual choices, rather than social and environmental phenomena.) This argument does depend on various assumptions about how people would “apply” compatibilism or originationism. For example, I have assumed that people are “show-me” determinists, so that adopting originationism would not transform them immediately into hard incompatibilists. Likewise, as set out in the paper, I have assumed that people will take HSC information as suggesting plausible causal connections between particular experiences of HSC and subsequent wrongdoing, so that they will see HSCs (in themselves) as relevant to originationist responsibility evaluations. But none of these claims depends on the further claim that people are “right” to be “show-me determinists” or to think that HSCs provide causal explanations for wrongdoings. So long as people *do” think this way, the choice between compatibilism and originationism should have the political ramifications I attribute to it, and this should be true even if people are in fact wrong to think the way I say they do.

All that being said, I understand that you may be asking me to put the political analysis aside and talk about the deeper question regarding the relevance of HSCs-in-themselves to desert. If so, I’d be glad to talk about that too. I should begin by reiterating that –as an originationist and a determinist -- I am (at the moment) a skeptic about responsibility. In that sense, I don’t think HSCs-in-themselves are relevant to desert – because I am skeptical about desert itself.

That being said, I do think that reflection on HSCs-in-themselves can be an important driver of originationist intuitions. I say this because I think reflection on particularistic accounts of HSC effects (on desires, attitudes, etc., apart from any effects on capacities) brings to life a number of concerns and attitudes that are, themselves, conducive to originationism. For example, I think such reflection can supercharge otherwise-dormant concerns about fairness and moral luck; I think it can summon anxieties about moral tunnel vision (“are we excluding morally relevant information when we sever actors from the environments that construct them in our responsibility evaluations?”); I think it can catalyze attitudes like compassion, empathy, and identification, and that these attitudes can motivate originationism; and I think it can bring to life the political implications of blaming actors shaped by HSCs in a way that may ultimately influence intuitions about the conditions necessary for responsibility attribution. (FWIW, by the way, I do not think that reflection on HSCs is unique in its ability to bring about all these results; on the contrary, I think reflection on particularistic accounts of hard family history can have similar impact, as can reflection on particularistic accounts grounded in situationist psychology, psychodynamic theory, etc. And I think that reflection on multiple such accounts from diverse sources can have a cumulative effect on our intuitions about responsibility.) In saying all this, by the way, I want to acknowledge work – like the Watson piece Justin mentions – that suggests that these sorts of reflections will *not* motivate originationism. The Watson piece is one of my very favorite pieces in any field or genre, but I am not (at the moment) convinced that concerns about moral luck and moral tunnel vision and attitudes like empathy and identification do not have the capacity to motivate originationism.

Apologies, again, if these are frustrating answers or appear evasive; and happy to elaborate further if anyone is interested.

[Anders Kaye]

Also, apologies to Mark (who posed me a question last night) and everyone else who I've yet to reply to -- I keep wildly underestimating how long it will take me to answer questions, and running out of time before the next class / faculty meeting / trip to the playground ...

[Kip]

Justin:

Well, then the compatibilists y'all know are boneheads (but I doubt this, so I expect it's more likely that since this topic rarely comes up, they don't explicitly address Kaye's worries).

I agree that their theories don't address his worries (that's the point of his article).

Tognazzini and Fischer, who do an excellent job of showing that it may be inappropriate to hold agents responsible (by engaging in retributive practices of punishment) even if they are morally responsible. And as long as that's a genuine distinction, compatibilists have a perfectly coherent explanation of why deviant social histories can make retributive punishment inappropriate without making the originationist mistake.

Some points:

1. I hardly think it's fair to fault Anders for not addressing every single theory of compatibilism (such as those you mention in your comment). His article is remarkably thorough, but it's impossible for him to address everything.
2. If there are, as you suggest, compatibilists who believe that we should believe most criminals are morally responsible without punishing them, this is certainly not a very well known view. It's also not clear that other compatibilists would follow that same path.

Mark:

Compatibilists will look at whether agent X has compatibilist requirement A,B,C, etc. These are requirements such as:

1. The agent would have done ~X if the agent had wanted to ~X;
2. The agent's higher order desires were aligned with the agent's lower order desires and choices;
3. The agent's values were aligned with the agent's choices;
4. The agent's had an evolved structure to respond to moral reasons (along with Dennett's other critera);
5. The agent owned its own moderately reasons-responsive mechanism.

Etc.

None of those inquiries necessarily look into Hard Social Backgrounds (HSB). And we wouldn't really expect them too, either, because HSB don't necessarily (or even likely) undermine those capacities. So I agree with Anders that compatibilists don't generally look into hard social backgrounds.

Your claim to the contrary seems to underemphasize the extent to which compatibilist theories (such as those I listed above) really don't care about hard social backgrounds.

Sofia

This is no answer. A sceptic has a GENERAL regress argument that appeals with equal force to all people. Therefore it's completely superfluos to pay attention to facts that DIFFER between people. So a free will sceptic could say about a person from bad social background commiting a crime, that "he was determined to do that, by the laws of nature and the past, and the past includes a bad social background". The point is, we could pick another criminal from a great social background, and the free will sceptic would have to say exactly the same thing about him: "He was determined to do that, by the laws of nature and the past, and the past includes a great social background". The free will sceptic doesn't have any tools within his free will theory for differing between these people. They're on par. Both were equally determined, despite one person having a lousy background and the other one a great background.

A Galen Strawson-style anti-realist wouldn't need to look into the background, I agree.

But Anders claim is not limited to GS-style anti-realism. Anders is contrasting originationism (e.g. incompatibilists) with compatibilism. Not all incompatibilists are GS-style anti-realists. Some, like Pereboom, believe that agent-causal free will is possible but most unlikely. A person like Pereboom wouldn't ignore looking into a person's background, and say (from the armchair) that the person doesn't have free will, in the face of contrary evidence. Similarly, most incompatibilists are presumably not GS-style anti-realists. Most are probably libertarians by default, who would have their hopes of libertarianism dashed by the slow encroachment of empirical science. And those people would be very interested to see how the evidence (about hard social backgrounds, etc.) turn out.

[Justin Coates]

Kip,

But look at why compatibilist theories don't address his point. It's not because the theories don't have the conceptual resources; it's because they're attempting to give a theory of when an agent is morally responsible. But it's a different thing completely to give a theory of when it's appropriate to hold someone morally responsible (e.g., by punishing them). So I don't understand why it's a felicitous criticism to say, "look we shouldn't punish these folks because of their deviant histories"--traditionally, compatibilists have not been trying to answer the question of when it's appropriate to punish agents; rather, they've been busy answering the question of whether the agents in question are responsible.

Furthermore, Kaye cites Watson's article, so he's at least aware that it offers a compatibilist explanation for why we adjust our attitudes in light of history that doesn't appeal to originationist considerations. Of course he doesn't have to address every article any compatibilist has ever written, but Gary's article is (indisputably) the best article with respect to these issues, and it's a fairly well known article, so I would've appreciated more.

Lastly, you suggest that it's an anomaly for compatibilists to distinguish between an agent's responsibility and the appropriateness of punishing that agent. Watson, Fischer, Wallace, Scanlon, etc., all seem to make this distinction, and I suspect they are not the only ones. In fact, it's compatibilists--not incompatibilists--who are working on the cutting edge of these issues (e.g., those previously mentioned, McKenna, Heironymi, Smith, Tognazzini, and I could keep going...). So yeah, maybe some compatibilists wouldn't follow this path, but I don't know any. At the very least, it's not a secret little community of unknown philosophers.

[Mark Smeltzer]

Kip,

Let's start with #5: according to Fischer, that is a necessary condition for moral responsibility. In practice, how would we know *for sure* that an agent has an MRR mechanism and that it was active in the actual sequence? We can't! We can only set forth a conclusion that is supported with the best empirical evidence we can find by taking a hard look at the agent and the agent's background. Fischer has a whole chapter at the end of his book Responsibility and Control on this is relevant to determining how to deal with HSCs in practice.

Regarding #2 and Frankfurt's hierarchical theory, it is possible that HSCs in practice may eradicate an agent's 2nd order desires, turning the agent into a wanton. Wantons are not free or responsible for anything. In cases involving HSCs, we would need to take a hard look at the agent and the agent's background in order to offer evidence to support the conclusion that the agent is not a wanton.

Regarding #1, even classical compatibilists have an out: if agent is unable to do otherwise due to HSCs in its background (if the HSCs render the counterfactuals false), then the agent is not responsible.

I could go on, but really this stuff should be obvious to anyone who's given compatibilist accounts a fair shake. Compatibilists may disagree with each other on the finer points, but the broad strokes are in alignment. I'm literally aghast that there is confusion here.