Revising our Retributive Practices

Let me ask a question that I think brings together a lot of the issues in the previous thread:

Suppose many of us agree (roughly) with these three claims about the U.S. as well as other nations (which I think are consistent with Bruce Waller's points): (1) the criminal justice system punishes people in ways that are overly retributive (not to mention ineffectual for rehabilitation), (2) there is an unjust distribution of benefits and wealth that undermines equal opportunity, and (3) some of 1 and 2 derive from an overblown sense of free will--e.g., of people's ability to overcome internal and external constraints or limitations.

Practical question:  Are we more likely to improve the situation by (1) telling people that free will is an illusion, (2) telling people that free will has limitations and comes in degrees, or (3) neither.

I mention this point at the end of my review of Sam Harris' book, because he clearly think (1), while I'm more inclined to think (2), though the most likely answer may be (3).  Obviously, much depends on how you tell them what you tell them (and on  what exactly people already believe), but I'd like to hear what people think.  I will fill in my reasons along the way.

Comments

[Thomas Nadelhoffer]

Eddy,

Just a request for a quick clarification before I head out the door (with a promissory note to post a proper response later today):

When you say that "the criminal justice system punishes people in ways that are overly retributive..." what do you have in mind?

Any retributivist worth her salt will insist that to justifiably punish offenders is to punish them only in ways that are proportionate to their moral desert. So, on this view, it's not possible to be "overly retributivist." On this view, to say that a punishment is too harsh, is to say that it amounts to more punishment than the agent deserves--i.e., not a retributive punishment. A retributive punishment is one whereby the offenders gets only what he deserves and no more.

So, a retributivist can (and indeed should!) agree that our current system is too harsh. But that's not because it's overly retributive. Rather, it's because it isn't retributive enough! On this view, it is ultimately forward looking worries about deterrence, public safety, and social control that explain why we punish too harshly.

But this suggests that by "overly retributive" you shouldn't simply mean "overly harsh or punitive" since these two things clearly come apart. But then what do you have mind?

p.s. As you know, I am a non-retributivist, a free will skeptic, and someone who has suggested in print that disillusionment is the way forward! So, you know in general how I am going to answer your interesting and important question!

[Steve Petersen]

I'd be interested in any psych studies that investigate this question, since I think it's pretty urgent, for just the reasons you say.

Meanwhile I'm inclined to think that one thing philosophers can do is emphasize to undergraduates that standard compatibilism does not sit easily with retributive punishment or with libertarian (non-)distributive justice. I think too often the take-away for undergrads is that compatibilism lets them keep their intuitive notions of freedom without the magic of libertarianism.

[Eddy Nahmias]

Thomas, I think what I mean--but I may be misunderstanding the gist of your question--is that our criminal justice system (reflecting our society) tends to punish people more than they deserve. And that's because it tends to over-estimate the degree to which people in general (and many criminals in particular) possess the cognitive, emotional, and volitional capacities to make good choices and to control their behavior, and their opportunities to exercise those capacities. That is, in my view, we over-estimate how much free will people have and are able to exercise.

[Thomas Nadelhoffer]

Eddy,

That's what I thought you meant. I was just trying to make sure. But then you shouldn't call our current system "overly retributive" just because you think it's overly harsh since the retributivist can quite consistently agree with you that our system is overly harsh for precisely the same reasons you identify--namely, offenders deserve less punishment than they typically receive because they have less free will than is typically assumed. In short, the harshness of a instance of punishment doesn't tell us anything about the retributive nature of the punishment. Whether the punishment was retributive depends on whether it was designed to give the offender only what he deserves. If not, the punishment wasn't properly (or at least purely) retributive.

But setting that aside for now, I just want to point out that the main difference between us is that while you want to go the revisionist route I would prefer to go the eliminativist route when it comes to how to best to address the fact that (a) people are punished too harshly, and (b) people are assumed to have more control of their station in life than they do--which is related to (a). We agree that people who violate the law often have far less regulative control over their actions and emotions than legal decision makers and the public often assume. We also agree that the world would be better off if people had a more realistic picture of the nature and limits of human agency. But whereas you want to continue talking in terms of free will and moral desert, I would prefer to talk instead about the cognitive and emotional capacities we need in order to be accountable for what we do in a way that doesn't piggy back on our traditional notions of agency and responsibility--which I think are far too metaphysically and religiously freighted to be deployed in progressive ways. As we have discussed many times before, this is ultimately an empirical question--namely, whether revisionism or disillusionism is more likely to lead to the progressive changes we both agree are needed when it comes to both criminal responsibility and wealth inequality. I, for one, have my money on disillusionism--and partly for the reasons that Anders Kaye highlighted in his interesting piece on compatibilism as a tool for enshrining precisely the kinds of views you claim to be combating!

But before I say more, I will wait to see what others have to say!

[Kip]

Eddy,

1. I am happy to see you concede (or at least suggest) that people think they have more free will than they do. I rarely hear compatibilists say this, which makes me wonder how they account for the countless related biases known in the psychology literature. You're halfway to admitting that free will doesn't exist, and I look forward to that happy day.

2. Although I find the question "Are we more likely to improve the situation?" intriguing, I also think that it's seductively misleading. I'm interested in the truth behind the free will debate, not in what is best for society. It might be that the truth is very bad for society - some research suggests that telling people free will doesn't exist increases cheating.

What worries me, when compatibilists ask what is best for society, is the possibility that this question is influencing their answer to the free will question. By asking what is best for society, instead of asking what is true, you start to sound like a revisionist (about which I'm sure Vargas will correct me). Note, also, that everyone is pretty emotionally invested in what is best for society.

3. I agree that society tends to punish way too much, especially in the United States. But, unlike you, I am mostly motivated by concerns about constitutive luck, and not by concerns about flaws in human rationality/intelligence/willpower. It seems to me that even perfectly rational human beings, with any sort of personality, would still experience huge constitutive luck in receiving their personality/values/drives endowment. These are two very different kinds of concerns (about constitutive luck, and about weak rationality, respectively). Are you concerned at all about constitutive luck?

[Brent]

Thomas,

It's not that criminals have "less regulative control over their actions and emotions" than society assumes; NO ONE HAS ANY control over their actions and emotions. No compatiblist has ever successfully explained how someone has "MORE" regulative control over their actions and emotions than someone else. Unless you can somehow explain why, even though all people are ultimately shaped by the same laws of the universe (or the same randomness), they can be assessed to have different levels of control, then you must reject meaningful personal control altogether.

[Thomas Nadelhoffer]

Brent,

I have more control over my emotions than I did when I was a child. I also have more than some adults and less than other adults. Some adults are so impaired when it comes to regulating their emotions--e.g., people with intermittent explosive disorder--that they may not be properly held accountable for their actions. Does that mean that the people who have regulative control over their emotions are deeply deserving or have free will? No, I never said they were. After all, I am a skeptic when it comes to free will and desert.

But it seems positively bizarre that anyone would deny that certain kinds of creatures have a power or capacity to regulate their desires and emotions that other creatures lack. Moreover, regulating emotions isn't the only important factor here. For instance, both young dogs and young children can learn not to defecate and urinate in inappropriate places--e.g., the house or in their pants, respectively. But infants and incontinent dogs lack this kind of control. While I wouldn't call this free will, I see no reason not to call it control. You seem to assume that the only kind of control is some kind of global or ultimate control. I see no reason to make that assumption.

Just because I don't think people are ultimately in control--because luck swallows all, so to speak--it doesn't follow that there aren't forms of local control that are completely relevant to how I interact with pets, children, and other adults. When pets and people have local regulative control, certain attitudes and responses to their behaviors make sense. When they don't have control, these same attitudes and responses become misplaced or misguided. So, when my dogs were very young, it made sense to use some negative reinforcement when they pissed in the house. But when my 14 year old dog would accidentally piss in the house from time to time because she couldn't control her bladder as well as she used to, it would have been both cruel and absurd for me to give her any negative reinforcement. What she lacked late in life was a certain kind of accountability-facilitating control not free will--which she never had from the start!

In short, I see no reason why I can't consistently continue to talk about local forms of regulative control--and the attending forms of accountability that can go along with it--even once I have dispensed with the idea of free will and moral desert.

p.s. Use *x* instead of ALL CAPS if you find yourself having to place emphasis! It doesn't make it seem like you're cyber-screaming :)

[Brent]

Haha, sorry about the capps.

I agree there is such a thing as an empirically measurable faculty of impulse control. I also agree that this faculty should inform us about how to deal with all people, particularly anti social types and children/old people.

I was thrown off by the word "accountability." Why is that word necessary in a purely scientific account of behavior or of questions regarding behavioral engineering choices? Obviously I don't think any amount of this impulse control faculty would result in someone being accountable in any sense of that word I’m familiar with. Do you have a specialized definition for "accountable"? Perhaps: a fit target for the necessary evil of behavior fixing punishment?

[V. Alan White]

Great question and thread. The main problem as I see it is that there's reason to believe that two tendencies seem to be at work in the evolution of views of crime and punishment in the US. One is an increasing concern to convict criminals merely on the basis of establishing the fact of criminal action--that some defendant X commited crime Y. (I've said a lot about this in previous posts and won't repeat it here.) A consequence of that is less concern about the mental states of so-called criminals, including questions of L/mind-based compatibilist FW. But another concern is to punish harshly--I'd say in retributive terms of perceived harms determining punishment (most likely though from just an emotional sense of revenge). The fact that these two concerns present possible conceptual conflicts doesn't seem to faze either the public at large or legal scholars. (One bit of evidence for the latter: some states' rejection of the insanity plea have been found to be Consitutional, although that is hardly evidence of a thorough philosophical laundering of the mens rea issues at hand.) This is why I'm pessimistic about the potential for even sound philosophical discoveries to influence criminal justice practices: people are going to continue to believe that acts trump perpetrators' mind-sets based on the relative perceived outrageousness of those acts, and therefore acts should determine punishment, and some inchoate sense of universal justice demands something like an eye-for-an-eye treatment, all other things being equal.

*If* elected politicians and public figures in the future came across as more concerned about *why* we punish people for what they do, maybe our profession could have hope to influence the discussion. But the fact that increasingly plutocracy determines political discourse, and the self-interest of plutocracy is best served by a deceived (and self-deceived) ignorant electorate, I am not sanguine about our discussions having any influence on how society behaves with respect to crime and justice. We are subservient in our pragmatic influence to states of affairs far beyond our reach as a profession.

Don't misread me. What we do is important--I think some of the most important work done in academia. And I certainly will continue to do it. I just don't think we have the means to make most people care about what we do, and I don't know how to change that.

[Thomas Nadelhoffer]

Brent,

Specific words are never necessary as far as it goes. But it would be quite tedious to constantly talk about "fit targets for the necessary evil of behavior fixing punishment." So, rather than pointlessly limiting myself to such a long-winded phrase, I opted for shorthand which I assumed wouldn't throw anyone off since I tried to make it fairly clear what I had in mind. In this case, answerable would have worked just as well as accountable (see Gary Watson's seminal work on the difference between responsibility, accountability, and answerability). Both accountable and answerable could carry their ordinary meaning and serve the roll it was supposed to play in my earlier argument--namely, highlight that certain creatures can be accountable or answerable for their behavior even in the absence of free will and moral desert.

[Tom Clark]

Eddy,

I really like your Wallerian hypothetical consensus about what needs fixing. To improve the situation we should definitely disillusion people about any overblown sense of freedom or control they might harbor, for instance any supernatural, spooky or logically impossible contra-causal capacities. That way they'll be more realistic, more practically effective, and more compassionate when it comes to human failings, which will carry over into criminal justice reform and policies to equalize opportunity. Since many (most?) folks are supernaturalist dualists, there's a lot of disillusioning to be done.

But of course we have to make sure that people don't make the mistake of thinking that without libertarian free will they lose control (Thomas's point to Brent above) and any basis for moral judgments. Telling the truth about human agency (what Kip recommends, contra Smilansky) is the way to go since it doesn't involve paternalism or dissimulation (plus it's true!), but it requires careful, repeated spelling out of the naturalistic replacement for soul control.

Given the psychological tendency for retributive blame (and Randian hero worship) focused on the individual to be attenuated by putting her in a wider causal context, I think that telling the naturalistic truth about agency is just what the US needs. But saying that free will is an illusion, without any qualification or explication, obviously isn't helpful.

[Kip]

As Tom points out, there are two competing fears here.

Fear #1: if we say that free will doesn't exist, then people might mistakenly believe that they lack powers (control, willpower, etc.) that they actually possess

Fear #2: if we say that free will does exist, then people might mistakenly believe that they possess powers that they actually lack (souls, origination/self-creation, quantum-split-world alternative possibilities, total freedom from genetic and environmental constraints).

Now I think it's clear that Fear #2 tends to motivate people like me, Tom Clark, Thomas N., Pereboom, Galen Strawson, etc. And I think it's also clear that Fear #1 tends to motivate compatibilists like Eddy and Dan Dennett etc.

Here's my point: I don't understand Fear #1. I don't understand it because I experience total, immediate, subject, undeniable access to compatibilist powers. I have total, immediate, executive access to my powers of control, and my powers of deliberation, and my experience of consciousness.

The argument here is really just like Descartes': I can deny anything, except I can't deny that I think. By the same logic, I can deny anything, but I can't deny that I have compatibilist powers. They're *obvious*. As I type these words in this blog comment, I am obviously controlling my actions. I want to type the word "word." So I deliberate about it. I type it. Then the word "word" appears in this blog comment. All of this fits perfectly within my understanding of my compatibilist powers. I experience it, immediately, undeniably, constantly, through every minute of the day.

NOTHING could make me doubt that I have these powers. So why, really, are compatibilists so concerned that people might lose faith in these powers? Can you lose faith that you're alive? That you're breathing? That you're seeing? Fear #1 just seems overblown - especially in contrast to Fear #2.

[Manuel "Revise This!" Vargas]

I'm too late to try to wade in here on the substantive issues, but this blog has taught me to not go on vacation for two weeks- it seems to lead everyone to start talking about awesome things. Or maybe the lesson is to be on vacation all the time?

Anyway, it sounds like it might be worth having a conference about these issues at some point. Would someone (who is not me) like organize?

[Allan McCay]

The conference sounds like a good idea!

I just wanted to introduce another option in response to Eddy's practical question. He said;

'Practical question: Are we more likely to improve the situation by (1) telling people that free will is an illusion, (2) telling people that free will has limitations and comes in degrees, or (3) neither.'

This is a pragmatic issue and I wonder if the answer might depend on the audience.

So if legislators, judges and the voting public overestimate the control that offenders have then perhaps it makes sense to to tell them that offender's control is more limited than they think.

If on the other hand offenders underestimate the control that they have perhaps it makes sense to tell them that they have more of it than they think.

[Eddy Nahmias]

Perhaps I'm too used to asking folk to give a forced-choice answer, but I'm still waiting to hear more direct responses to the initial question, now restated in Kip's (via Tom's) helpful dichotomy: which fear is more significant?

"Fear #1: if we say that free will doesn't exist, then people might mistakenly believe that they lack powers (control, willpower, etc.) that they actually possess.

Fear #2: if we say that free will does exist, then people might mistakenly believe that they possess powers that they actually lack (souls, origination/self-creation, quantum-split-world alternative possibilities, total freedom from genetic and environmental constraints)."

We could have a conference on such questions, of course (and if we do then Manuel is definitely organizing it), but until then, what say you?

Meanwhile:
1. Kip, I'm surprised to hear you talk that way about your Moorean knowledge of compatibilist capacities, since one might try (and some have tried) to use such a move to defend libertarian free will too. In both cases, I think the move is problematic. And many "willusionists" certainly think you are deluding yourself about the compatibilist capacities too. And yes, I'm most interested in the truth too, but this sociological/psychological questions is interesting and important, and its connection to the truth about FW and MR is not (obviously) non-existent.

2. Allan (and perhaps Alan), I really like the idea that the answer to the practical questions might be audience-dependent. For a similar sort of idea, check out today's David Brooks' timely column on this idea: http://www.nytimes.com/2012/08/03/opinion/brooks-the-credit-illusion.html?_r=1&ref=davidbrooks

[Brent]

I think it would be advantageous to map this case to an older, already settled one.

The big fact that the public has yet to accept: There is no God.

Fear #1: if we say that Got doesn't exist, then people might mistakenly believe that there is no morality, even though there are logical/biological bases for being pro social.

Fear #2: if we say that God does exist, then people might mistakenly believe in harmful superstitions (discrimination, anti-science behavior...)."

Now I think we can all agree that in this case the greater fear is obviously # 2. I propose that the only difference between this case and the free will case is that the god case is further along, or at a more advance stage. That is, it has had a longer time to be soaked in by the general public. We've been hearing from public atheists/agonistics for centuries. Somewhat rarer, is public free will/ MR skeptics.

Once free will skepticism gets more exposure, perhaps the reality that number 2 is the greater evil, will be more obvious.

[Joe Campbell]

Brent: It isn't that easy. We can play the game with "knowledge," too.

Fear #1: If we say that skepticism (about knowledge and justification) is true, then people might mistakenly believe that all views and opinions are on a par and that there is no more reason to take broken glass for a headache than there is to take aspirin.

Fear #2: If we say that we have knowledge, then people might mistaken it for infallible certainty and this will lead to arrogance, dogmaticism, and intolerance.

Only here the question isn't so much what should we do as what did we do. We junked the infallibilist conception of "knowledge" and replaced it with a fallibilist conception. We realized that we were wrong about the nature of knowledge and altered our understanding of the concept to fit our actual human capacities.

That's why I'm a compatibilist. That's why I advocate retaining our link with history and altering our conception of "free will" to better match our contemporary understanding of ourselves.

When I read discussions like this I think it is no wonder that the folk don't care what philosophers think because for the life of me I can't see any substance behind this debate. It is a mere verbal dispute. Do we all (or most of us) think that humans have powers and capacities relevant to action and control that other creatures lack? Yes. Do we all (or most of us) think that were you to consider the state of the world a few seconds prior to the performance of any action that (even if determinism is false) the action is if not inevitable at least close enough to being inevitable that the difference doesn't make any difference at all? Yes. So what do we disagree about? Nothing other than which set of powers and capacities upon which to hang the label "free will." If that is the case, it is hard to blame the folk for thinking that philosophers are wasting their time.

I'm not trying to be offensive! I'm as much of an armchair philosopher as anyone. I've been working on a book review and reading a lot of Hume, so perhaps I've become temporarily perverted (though I've been thinking this way for some time now).

Still, when we focus on the practical issues -- as Hume suggested we philosophers should do -- it is interesting that there is near universal agreement. We all seem to agree that reform is needed. I can't help but think that there is some way that we might be better able to pool our resources and worked toward that end.

With that in mind, although I'm not able to organize a conference on this topic just yet (the next INPC will be on neurophilosophy in July 2013; write me off list if interested) I would be willing to edit a volume on this topic. Anyone interested in contributing to that venture, contact me off-list and I'll see what I can do.

[Mark Young]

Fear #1? Fear #2? It seems to me that these "fears" assume that people are going to be *convinced* by what you tell them. You have no such guarantee. What you should be looking at is what are the pragmatic results of *arguing* that people have no free will/have a weaker kind of free will than they heretofore believed.

And I think that people who argue that we have no free will should be afraid of these possible results:

(a) People dismiss you as an idiot/philosophy as useless.
(b) People behave worse.
(c) Both (a) and (b) -- sometimes in the very same person.

We have experimental evidence of (b) -- and I imagine many of you have lots of anecdotal evidence for (a).

What would be the corresponding fears for arguing that free will is weaker than many think?

[Tom Clark]

Joe,

Beyond the "mere verbal dispute" about "which set of powers and capacities upon which to hang the label free will" there is substantive disagreement between compatibilists and skeptics about moral responsibility and desert. Compatibilists such as Dennett defend free will (that is, define it as being compatible with naturalism and recommend that we continue to say we have it) partly in order to justify desert-entailing MR and its associated practices, such as retributive punishment. Skeptics and revolutionaries such as Waller (who agrees that we have compatibilist control capacities, and even calls this free will) say MR and retribution should be abandoned since they aren’t compatible with naturalism. So there isn't "near universal agreement" on practical issues since there isn't agreement about MR and desert.

It's great if we mostly agree here about the need for criminal justice reform and equalizing opportunity, but the way I see it, compatibilist apologists for MR are regrettably and mistakenly keeping alive the desert-based rationales for harsh punishment and social inequality.

[Joe Campbell]

Tom,

Thanks. Two related points.

First, I don't want to dwell on the meta-issue but it isn't clear to me that the verbal dispute (if there is one) doesn't play out again at the level of "moral responsibility," as a debate between "social-regulation" theorists (and pragmatic considerations) vs. desert-based theorists.

Second, the punishment debate and the moral responsibility debate are logically unrelated. One can combine ANY theory of punishment with ANY theory of moral responsibility IF what we care about wrt "punishment" is actual behavior (mistreatment, for instance). One can be a skeptic about moral responsibility and still advocate incarceration, water board procedures, whatever. All that has to be the case is that the story is told in terms of "treatment" not "punishment."

In other words, just because we're not blaming or punishing anyone, it doesn't follow that we aren't mistreating them.

[Kip]

Eddy,

"Kip, I'm surprised to hear you talk that way about your Moorean knowledge of compatibilist capacities, since one might try (and some have tried) to use such a move to defend libertarian free will too. In both cases, I think the move is problematic."

It just seems obvious to me that I have the ability to generally control my actions in view of my goals and desires. Like typing this blog comment right now. Although libertarians might say that they have internal access to proof that the world is indeterministic, or that they have radical freedom from environmental and genetic constraints, I think they are overreaching here. Libertarians say lots of crazy things. The form of the argument ("I have internal proof of X") shouldn't obscure the fact that I and libertarians are arguing for different X's, and I think their X makes the argument far more problematic for them. Notably, I think libertarians would agree with me that I have internal access to uncontroverial compatibilist powers. We just disagree about whether we have internal access to further crazy libertarian powers.

"And many 'willusionists' certainly think you are deluding yourself about the compatibilist capacities too."

I don't know about many. You call many scholars "willusionists" who are inclined to incompatibilism, and might readily agree with me that we possess uncontroversial compatibilist powers. Greene and Cohen, Dawkins, perhaps also Hawkins and Einstein. Most of the remaining "willusionists" base their view on the unimportance of consciousness, or a lag effect to consciousness. This remaining literature is not without controversy. Moreover, I don't see these results as very threatening to free will - I generally take ownership of the unconscious processes that cause my decisions, and rarely do I feel surprised or disturbed by them. Perhaps I have an unconscious desire to eat a steak, which is the primary cause of me doing so, and my consciousness plays little role - so what? I like steak, I wanted to eat the steak, I ate the steak, I felt neither surprised nor disturbed by the result. All of this seems consistent with at least a bare minimum of compatibilist powers that I feel I undeniably have. These "willusionists" act as if we routinely experience random and bizarre outcomes from unconscious desires that aren't integrated with our conscious desires - as if we find ourselves routinely trying to eat rubber tires and have sex with lamps, and then look in horror at what we've done. That's not how our brains or consciousness work, and the threat to free will from this research is overstated. In my view, the real threat to free will is constitutive luck, to which compatibilists seem blind. But that's an argument for another day.

[Tom Clark]

Joe,

I don’t see that there’s much debate about the meaning of moral responsibility. Neil Levy defines it this way at the start of his interview at Philosophy Bites: “Well, as I use moral responsibility, it’s the property that makes agents appropriate targets of blame and praise and mabye even punishments and perhaps benefits. An agent is morally responsible if they performed an action and they deserve some kind of treatment on that basis, not on the basis of consequentialist considerations, not because it’s good for society or good for anybody else, but because they deserve it.” His host, Nigel Warburton, calls this “a fairly standard view of moral responsibility.” Tamler also defines it this way – as essentially involving non-consequentialist desert – at the start of Relative Justice.

But even if there’s another strong contender out there for the meaning of MR, *this* meaning is routinely used to justify imposing suffering, deprivation and death on offenders beyond what’s necessary for (or even inconsistent with) deterrence, public safety, restitution, and rehabilitation. So if no one is MR in this sense, *this* justification for harsh treatment evaporates, which is why the debate about MR has direct relevance for criminal justice.

I agree that “just because we're not blaming or punishing anyone, it doesn't follow that we aren't mistreating them.” Absent MR, unless there’s another justification for harsh treatment, for instance on consequentialist grounds, we would call it mistreatment. That we’re all concerned here not to mistreat people shows that any rationale for harsh treatment has to pass the humanitarian tests of minimizing suffering (use the least punitive means possible) and preserving human rights.

[Brent]

Retributive Justice = unnecessary evil, in that it is both unfair and ineffective
Consequentialist Justice (e.g. quarantine justice) = necessary evil, in that it is unfair, but effective.

The only time society should be granted the grave power of perpetrating necessary evil (quarantine justice), is when that society is actively trying to avoid the circumstances that called for the necessary evil in the first place. If society perpetrates necessary evil passively, that is without looking for radical solutions to avoid the need for that evil, then the once “necessary” evil slides into the realm of the unnecessary.

First, it goes without saying, our current justice system dispenses unnecessary evil.

But even our proposed quarantine system can become an unnecessary evil itself. If the theoretical society that uses quarantine as a lesser evil and last resort, fails to take every step possible to cure and prevent the mental conditions that produce dangerous behavior in the first place, then they can no longer use the excuse of ‘necessary’ evil.

My point is that no use of force by a given justice system is ever justified. *unless* the society that gives it power is committed to radical behavioral engineering and preemptive intervention. In other words, it is not justifiable to use even a liberal quarantine system if you allow the general public to keep the status quo belief in libertarian free will (because libertarians would never feel the need to attempt radical or scientific solutions to behavior).

This leads to the subsequent, more radical conclusion: a society cannot have a defendable justice system *and* believe in libertarian free will at the same time. Those two are incompatible for the above reasons.

This would all make more sense with a longer explanation.

[Paul Torek]

Steve P,

It depends what you mean by "retributive punishment" - see Mark Young's comment here: http://agencyandresponsibility.typepad.com/flickers-of-freedom/2012/01/degrees-of-desert.html.

Kip,

Think Libet for the sort of "willusionist" who doubts your Moorean certainties. Or better yet, lest you think only obscure neuroscientists are contesting the point: think Sigmund Freud.

Eddy,

You numbered both the statements in the hypothetical consensus, and the solution options. Let's letter the former. As for solutions:

(3) followed at a long distance by (2). We should point out that the large plurality of convicts are there for drug crimes that either primarily harm themselves, or do harm others, but not nearly as much as the drug laws increase the society-wide harms associated with drug use. Neither users nor dealers should be treated thus, regardless of free will issues.

Another important prong of solution (3) should be emphasizing fact (B) of your consensus, unequal opportunities. On just about anyone's views of distributive justice, the fact that no human created the land and its resources, and the fact that unequal access to same explains a lot of the unequal results we see, are relevant facts.

[Kip]

Paul,

I understand Libet (and similar others) as arguing (based on evidence) that there is a lag effect between decisions and consciousness of the choice. In other words, we experience conscious choice at time T1, but brain/other measurements slightly before T1 predict those choices. In that case, I think:

A. no one could reasonably say that they have internal access to proof that Libet's wrong. I don't make that claim, and I don't know anyone who does. Everyone has internal access to their consciousness. But they have no internal access to unconscious forces that predict those conscious experiences. So nobody has, or can reasonably claim to have, internal access to proof that there is no such lag effect.

In other words, when I say that I have internal access to undeniable compatibilist powers, I am *not* talking about powers that Libet claims to undermine. I'm talking about powers that Libet would agree that we all possess. Whether Libet undermines some forms of compatibilism is an open question. I'm just arguing that I have undeniable access to a basic minimum of compatibilist powers that Libet doesn't even pretend to undermine.

B. Libet and others are talking about consciousness. But I rarely if ever hear compatibilists talk about consciousness. Frankfurt talks about integrated desires, Watson talks about integrated values, and Fischer talks about reasons-responsive mechanisms, etc. But it seems to me that each of these systems could apply to creature who lack consciousness, at least hard consciousness/qualia. This alone makes me skeptical that Libet et al. have undermined compatibilist free will. So I don't worry about them too much. As I said before, my worry is constitutive luck.

Kip

[Joe Campbell]

It is far from clear what Libet's argument is. I thought that before reading Kip's post but it is all the more true after reading it.

It is clear that "we experience conscious choice at time T1, but brain/other measurements slightly before T1 predict those choices." Let's call this "the empirical given" since it is the empirical data that we're trying to interpret.

You could be saying that since there were events prior to conscious choice about which the agent had no choice, and these determined the actions that the agent performed, so the agent had no choice about his action. This is just the consequence argument.

But I'm told that Libet's argument is different. Perhaps this is a kind of bypass argument (Eddy). The empirical empirical given suggests that the agent plays no role in the action, for all the work is being done by independent neurological processes.

I just don't see how, if we press up against the later argument, that we don't end up with the consequence argument. Why deny that the agent has control? Ultimately the story is going to be that the causes of the action were beyond the agent's control and thus so was the action.

I wonder if you (Kip) disagree with this much.

[Eddy Nahmias]

Joe, it may be that we bump against the consequence argument either way, but I think there is a huge difference between two different arguments, call them Bypassing and Throughpassing:

Throughpassing looks like the standard worry of the CA:
T1. There are sufficient conditions P (over which you had no control or responsibility) for your conscious choice to do X [implicit: but your conscious choice to do X, with all the compatibilist control conditions intact, is the (proximal) cause of your doing X].
T2. If you are not responsible for (have FW regarding) a set of sufficient conditions P for your choice X, then you are not responsible for (or free regarding) X.
T3. So, you lack MR and FW regarding X (generalizes to all actions).

Bypassing could be presented in lots of ways but here's one that looks like it has the form of CA:
B1. There are sufficient conditions P (over which you had no control or responsibility) for your conscious choice to do X *and* for your doing X, such that your conscious choice to do X is not a causal factor in your doing X.
B2. If you are not responsible for (have FW regarding) a set of sufficient conditions P for your doing X, then you are not responsible for (of free regarding) X.
B3. So, you lack MR and FW regarding X (generalizes to all actions).

What's the difference? The difference is that I think T2 and B2 are false (as do you, right? and perhaps all compatibilists need to reject Transfer principle like this since I think they underlie all traditional incompatibilist arguments). I'm fine with T1 (and I think it makes no difference if we throw in some indeterminism here, since there will still be conditions P, over which you have no control, that are as causally sufficient as possible for your choosing X).

But I think B1 is troubling and I think it is the main thing that troubles ordinary people (and I think that something like it is what many people implicitly read into most incompatibilist arguments and that explains why they *appear* to have incompatibilist intuitions).

B1 leads to a conclusion like B3 not because of a premise like B2, but because of a premise like B2*: If your conscious choices (or compatibilist capacities that, at some stage, require conscious processing) have no (relevant) causal impact on your doing X, then you lack MR and FW regarding X.

(Note that the conscious processing need not be directly prior to X, which is one of many reasons Libet's and related results do not support generalizing from B1 to all actions.)

B1 is troubling precisely because it undermines most compatibilist views of what free will requires and *not* because it just offers a specific incarnation of the sufficient conditions discussed in the CA-style arguments above. One might argue that compatibilists do not require consciousness to play a causal role, but I think that view is mistaken (and highly counter ordinary views)--Neil?

More to say about all this, but I'll leave it at this for now.

[Joe Campbell]

Thanks Eddy. This is very helpful.

A key issue is the reason one gives for supporting:

B1. There are sufficient conditions P (over which you had no control or responsibility) for your conscious choice to do X *and* for your doing X, such that your conscious choice to do X is not a causal factor in your doing X.

B1 is like T1 except for the ending: "your conscious choice to do X is not a causal factor in your doing X." But why think your conscious choice is not a causal factor in your doing X, given the empirical evidence? Note that the empirical evidence does not SHOW that B1 is true; clearly B1 is an interpretation of the data.

One reason might be that your conscious choice was predetermined. That leads back to the first argument and ultimately the consequence argument.

But as I noted before, Harris gives a different argument. We might call it the "No-Self Argument" (Neil noted that others have discussed it, likely you as well). One might argue that there is no self to play the causal role needed to refute B1.

Even in this last case, I wonder.

Thanks!

[Kip]

Joe, I have no disagreement with your earlier comment, as far as I can tell.

Eddy, I didn't quite understand your distinction. But I think that's more my fault, since you are more familiar with Libet and related work.

Let me say that I think Libet is (unknowingly?) hitting on several possible concerns:

1. One concern is origination. We worry that our decisions / actions originate with us. More specifically, we tend to think that our decisions originate with our conscious choice. In other words, before I decide at time T1 to do do X, there simply is no fact of the matter about whether I will do X. It is impossible to know, at least with certainty, because the decision originates with me, and it has not been originated yet (or so the incompatibilist story goes).

To me this principle about origination is a highly intuitive feature of free will. I think it motivates most incompatibilism, although incompatibilists usually do not speak in terms of it. In short hard, the principle says that free will makes prediction impossible, even in principle. Libet tried to show that (at least some) prediction was possible, at least before conscious choice, and that's where the concern is.

2. Note a related concern here about identity: do we identify with out unconscious processes? If we don't identify with them, then their influence or dominance is obviously problematic.

But if we do identify with them, then it is less clear why they are problematic. If we don't identify, then there is no problematic prediction - I made the choice at time T0 when I made the choice unconsciously, the conscious choice just followed as an unproblematic matter of course. In this case, the concern reduces to one about epiphenomenalism about concsciousness.

Consider the example again: did my unconscious desire to eat a rubber tire lead to me eat a rubber tire, which my consciousness discovered in horror? Or did my unconscious desire to eat a steak lead me to eat a steak, which I did, and my consciousness discovered it, and said "way to go, unconscious!" This is the classic distinction between integrated and non-integrated desires/values, and it applies regardless of whether the desires/values are consciously felt. The distinction is not teased out in Libet's work, as far as I can tell.

3. A final concern is the concern about epiphenomenalism, as I alluded to above. But epiphenomenalism does not worry me at all. As far as I can tell, epiphenominalism is true, and poses no separate threat to free will. Hell, as far as I know, there is a 2 year lag between my actions and my conscious perceptions of those actions - I don't care (my conscious experience is exactly the same either way). This is especially true if our unconscious desires are integrated with our conscious desires, and they generally are. Epiphenomenalism is also unrelated constitutive luck, which I regard as the genuine threat to free will.

It bears repeating that compatibilists and incompatibilists alike rarely if ever characterize free will in terms of consciousness. People accuse arch-compatibilist Dennet of not even having consciousness, which doesn't stop him from believing in free will. In this way, Libet's work and the traditional free will literature are talking past each other.

4. One last point: Joe and Eddy's comments are talking about causal chains. Consider the chain A->B->C

In a deterministic world, A causes C. But that does not mean that B doesn't also cause C. They both "cause" C in some broad sense. Thus, the fact that A causes C does not imply "that your conscious choice to do X is not a causal factor", as Joe characterizes his opponents. The spirit of compatibilism says: "I'm ok being the middle link B in a long causal chain. It doesn't bother me that A causes C, as long as I also cause C as an intermediary."

Incompatibilists completely disagree with that. This goes back to origination: if the choice originates with me, then it originates at B (where I'm located as intermediary), and it's impossible for the choice to originate at A. What especially worries incompatibilists is that A dominates the entire discussion: once you know A (and the laws of physics), B and C fall like dominoes. Thus, the fact that B/humans also "cause" C provides little comfort. It is a superfluous kind of causation, and we don't like to feel superfluous.

[Joe Campbell]

Kip,

Comment on item (4).

This is the whole point of the example I presented in Origination and Moral Responsibility. It CANNOT be the case that folks take seriously the view that an agent must always be the causal origin of those actions over which he is morally responsible. If one is in the midst of a prolonged disagreement, it is undeniable that the history of the disagreement plays some CAUSAL role in your actions. That is undeniable. No one in their right mind could think that the CAUSE of the action originates with the agent if what they mean by this is that there are NO other events that play a causal role.

Maybe there is a modal element here: The action could have been otherwise, even given the past; so past causes play no determinate role. But that the past plays no CAUSAL role in our free actions? Who would think that?

[Kip]

Joe,

I agree that taking the Origination Principle (or the No-Prediction Principle) to its extreme will result in absurdities. Clearly, when an agent does X, we expect that the agent does X based on what happened before X. The past and present have a certain integration, in that way. Hume made this point a long time ago.

The problem, from the incompatibilist perspective, is that the integration in a determinist world is that the past perfectly predicts the present. You could go back before the agent was born, put in enough data into a powerful-enough computer, and predict the agent's entire life. This, in a nutshell, in Mele's Zygote Argument (and so many other incompatibilist arguments, without the manipulation).

This is the difference between:

A. "the agent's action X makes sense in light of what happened before X"
B. "what happened before X perfectly predicts, and provides an exhaustive account, of the agent Xing, even before the agent X's, or even before the agent is born"

What disturbs incompatibilists like me is B. Moreover, it seems that incompatibilists can find B incompatible with free will, without also finding A incompatible. In other words, it seems conceptually possible that B could be false, while A could still be true. And that would seem to be enough to answer your concerns (at least on the surface).

Another way to put this: you write that "it is undeniable that the history of the disagreement plays some CAUSAL role in your actions." True. The question for the incompatibilist is: is this causal role sufficient? By sufficient, I mean: does the causal role dominate the explanation, so that everything else is superfluous, if you are trying to explain or predict the future? Something can play a causal role without being a sufficient cause. It has always seemed to me that if the past was a sufficient cause for my life story, even before I was born, then I can't really have free will - I will never deviate from the life story that is effectively predestined by that past.

[Neil Levy]

Eddy rightly points out that I believe that compatibilism requires that consciousness plays a role in action selection. Let me just say this is a view that is utterly dependent on the science of consciousness. It might be, as Eddy says, highly counterintuitve and contrary to folk views, but until someone produces evidence that the folk tend to read work on thalamocortical reentry loops with their morning coffee, I will not think this is relevant. However, I am not committed to Eddy's B1. Unlike other people worried about the role of consciousness, I am not concerned about consciousness being epiphenomenal. Were I concerned with that question, the view would be falsifiable from the armchair, and folk intuitions would matter directly: we could build theories about whether we could be responsible despite the fact that consciousness does not do anything. Since the view is instead built on claims about what the causal role of consciousness actually is, the view is not so easily falsified (obviously theoretical work will remain relevant, but not in quite the same way).