What are we fighting about?

Thanks to Thomas for this chance to discuss free will with all of you -- I'm looking forward to your contributions.

I’ve been arguing that the question at the core of the historical free will controversy is whether we have the kind of control required to be morally responsible for our actions in the basic desert sense. (For an agent to be morally responsible for an action in this sense is for it to be hers in such a way that she would deserve to be blamed if she understood that it was morally wrong, and she would deserve to be praised if she understood that it was morally exemplary, and the desert is basic because the agent would deserve to be blamed or praised just for the reason that she has performed the action, given an understanding of its moral statues, and not, for example, merely for consequentialist or contractualist reasons.)  But this hypothesis has been challenged recently by several of you, including Manuel Vargas (Book 2013) and Michael McKenna (Article 2013). One objection is that the control required for desert-involving, but not necessarily basic-desert involving responsibility is what lies at the core of the debate.

I want to resist this, for the following reason. Consider a conception of moral responsibility that’s as close to the basic desert notion as it can be, but actually rejects basic desert: We retain our unrevised ordinary moral responsibility practice, the one that includes basic desert suppositions and justifications, and this practice is morally justified solely on the ground that of all the competing practices possible for us it’s the one that does best at realizing the relevant good consequences. The underlying ethical theory might be thought of as practice-consequentialism, and the moral justification is exclusively forward-looking because the practice is justified solely by forward-looking reasons. Manuel may be open to this sort of position.

I say that this position isn’t one that’s at issue in the historical debate, because on this view justifications for blaming and praising are exclusively forward-looking, and the traditional incompatibilist has no quarrel with the compatibility of an action’s causal determination by factors beyond the agent’s control and forward-looking justifications for blaming and praising. And this is because incompatibilists haven’t seen any incompatibility between an action’s causal determination and the agent being appropriately subject to measures that aim at goods like moral formation, protection, and reconciliation. The incompatibilist might well want to resist such a consequentialist justification of our ordinary practice of holding morally responsible, or any relevantly similar ‘government house” consequentialist justification of the ordinary practice, and I’m one of those incompatibilists. But this would for ethical and/or empirical reasons that are independent of the threat to control (or the threat posed by luck) that results from causal determination or from indeterminacy in action.

So I conclude that we’ve been fighting about whether we have the kind of control required for basic-desert moral responsibility.  (Well, OK -- maybe this isn't quite the whole story – there’s another incompatibilist strand, one that Dana Nelkin addresses in her book (2011), that relies on a link between blameworthiness and obligation, and justifies a challenge from determinism by way of the ‘ought implies can’ principle. But I’ll leave that for another post.)

Agree or disagree?

Comments

[Robert Allen]

It certainly is not the "position at issue in the historical debabe," which was over the proposition that at the Final Judgment some people are going to be eternally rewarded while others are forever damned- because that is what they deserve. And they will be receiving their just deserts because they had complete control over the morally important decisions that they made in their lives. Medievalists such as Augustine, Anselm and Aquinas were certainly trying to philosophically justify the Catholic belief in Heaven and Hell. So if our discussion began with their work, then it's been about the sort of responsibility associated with being justly consigned to one of those places.

[Fritz Warfield]

Derk,
On your view of the history, is it a *stipulation* that there is such a thing as a "kind of control required to be morally responsible for our actions in the basic desert sense"? Or do you think there is a good argument for accepting this particular connection between freedom and moral responsibility? [I assume you mean "freedom" by "control"? - if not I don't know what "control" is].

This doesn't look to me to be the sort of thing one could properly stipulate.

Here's an alternative view of the history: the core of the free will controversy is whether we have free will. A separate question which different people answer differently is whether freedom is necessary for moral responsibility or any other way intimately connected wtih it.

[John Fischer]

Derk,

It seems to me that the rejection of a purely forward-looking account of moral responsibility would be compatible with also rejecting "desert" or "basic desert" as the relevant theoretical element, in favor of "fittingness" or "aptness" or something like that. So, I am wondering whether you think that desert is to be preferred to aptness or fittingness in the account of moral responsibility. More specifically, if someone contended that to be morally responsible for a bit of behavior is to be a fitting candidate for the reactive attitudes on the basis of that behavior, would that be problematic, on your way of thinking?

[Tamler Sommers]

John (and Derk),

Clarification question: can basic desert itself be defined in terms of fittingness or aptness? In other words, can we regard "deserve blame" and "fitting or apt candidate for resentment" as equivalent?

[James Laird]

Derk,

I agree with your conclusion that we’re “fighting about whether or not we have the kind of control required for basic-desert moral responsibility”. In order to end the fighting and resolve the issue, I believe there’s a series of questions that need to be debated in a certain sequence. In other words, the question at the core of the historical free will controversy may go deeper and result in several lower-level preliminary questions that must be answered first.

Here are the three preliminary questions as I see them: 1. How are we going to define the meaning of the word “individual”? 2. For some type of specific case, we’ll need to reach agreement that the “individual” is responsible for the forces that cause the agent’s body to move. 3. Finally, we’ll need to develop reasons why we believe the forces exerted by the “individual” are qualified to be called “free will” in nature (vs. predeterministic), thereby connecting the individual to the idea of being morally responsible.

If we could debate those three questions and come to some kind of consensus, we could then be in a position to argue intelligently about whether an individual has the control required for basic-desert moral responsibility. Until we lay that foundation, the debate is extremely difficult to hold at the top-level subject, since we all have different assumptions for the supporting ideas. In other words, we first need to stand on some kind of common ground. I’d enjoy debating all three of those questions.

[Nate]

John,

Some implementations of the fittingness view might involve (or even collapse into) a version of the desert-based view. This would be so if we thought e.g (i) that resentment essentially involves the thought that the resented agent deserves to be sanctioned / harmed and (ii) that resentment is fitting (if and) only if its constitutive thoughts are true. (Note that this toy view sees deserving harm as fundamental, whereas Derk seems to be seeing deserving blame / resentment as fundamental; I'm unsure what to say about this.) So if we had this view your hypothetical contention wouldn't be problematic (since fittingness is a matter of desert), except perhaps in the sense that framing the issue in terms of fittingness is less illuminating than framing it in terms of desert (because fittingness is less fundamental, i.e. blame is fitting, when it is, [partially] in virtue of the fact that the agent deserves to be harmed, and not the other way around).

The toy version of the fittingness view isn't inevitable and so perhaps there are conceptions of fittingness that do not essentially involve desert. Assuming there is an available concept of fittingness, I still don't see how it could be "problematic" to ask whether blame can be fitting under determinism since the answer doesn't seem trivial and since there is no wrong-kinds-of-reasons problem (as w/ the forward looking views).

[Michael McKenna]

Hi Derk, Great post! There are also lots of interesting comments so far. Let me just pick up on one thing John suggests: that rejection of a purely forward looking account would be compatible with rejecting basic desert. I take it John meant to suggest that it would be compatible in such a way as to preserve the link to the traditional free will debate. (I would add a nod to Fritz that there's work to do to establish this link, though I think it can be established.)

I think John is on to something. And I would just say more generally that there may be other modes of justifying our blaming practices (not just consequentialist) that leave no significant role to play for freedom of a sort that could give rise to serious philosophical worries. That, however, leaves other nonbasic-desert-based options open. What I have been thinking lately is something like this: Any account of the normative warrant for blaming is at least preliminarily in the business of the (or *a*) traditional free will debate so long as the normative warrant on offer leaves room for the idea that it is *at least in part* up to an agent, by her so acting, whether she is blameworthy. The idea is that somehow at least a part of what grounds her being to blame must be dependent on an exercise of her free agency. And once we have such a requirement--however it is normatively underwritten--we are in the game of asking what that 'up to the agent' condition comes to. Worries about determinism or indeterminism can now start to arise.

John suggested aptness or fittingness, and while I am not sure these will do, the idea might go something like this: our practices and the norms we embrace limit fittingness of blame to circumstances in which it was, at least in part, up to an agent whether she was to blame. Now we can begin to have a debate about how to understand ‘up to us’. Does it entail (or just amount to) being able to do otherwise? Does it require “ultimacy”? And so on…

Just a quick note on why I worry that mere fittingness or aptness will not do. Aptness seems to me merely to involve a claim that there exists some normative warrant. If it is apt to blame, then there exists some normative warrant for blaming. But we are given no content as to what that normative warrant is. (Basic desert, by contrast, has rich content, and so is one reason that it is satisfying to think ‘apt because deserved’.) My worries about fittingness are much the same. But think, for example, about Wallace’s view. He unpacks claims about the aptness of blame in terms of fairness, and is explicit in rejecting a full-throated desert thesis. Yet with just the notion of fairness, we can ask if it is fair whether a person ought to be blamed if it was not up to her whether she was to blame, or if she was not able to do other than she did. Isn’t this a way into the free will problem? And doesn’t the incompatibilist have a perfectly legitimate path to press her case when the issue is cast in terms of fairness rather than basic desert?

One minor remark in signing off, Derk: Given the general Zen-like nature of your position, and even more so your much beloved cool Zen-like demeanor, I am surprised to see you describing our debate in terms of fighting. I don’t think I could bring myself to fight you! You’re just too good of a guy.

[V. Alan White]

Well, I just lost an hour's work here with a sneeze that made me delete a prior post. Neil--is that the result of luck?

So to start over.

The history of the connection of FW to ultimate responsibility has (I think) definite connection to the history of criminal responsibility, especially with respect to 20th century American law.

The background of insanity law goes back to McNaughten (1843, English) where the absence of rational intent (knowing right from wrong) was the first wide-ranging precedent for establishing complete lack of guilt.

In the 20th century "irresistable impulse" was another, logically independent criterion of exoneration that gained legal standing. This was clearly tied to scientific advances that linked causation (behaviorism/OCD) to a lack of responsibility.

The Model Code (early 60s) put the latter into legal statute: one should not be held responsible if one had not the substantial capacity to conform one's acts to the law. This is specifically a FW condition that is logically separate from McNaughten, and introduces the counterfactual question of the ability of minds to choose the good when they in fact have chosen evil (Susan Wolf's famous point about asymmetry placed into law).

Since the vexed case of John Hinckley, various jurisdictions have reacted to such laws in different ways, including completely deleting the Model Code's empahasis on FW, though it remains important in certain states. (It is still important here in Wisconsin, where it played a major role in the Jeffrey Dahmer case. Dahmer was nearly exonerated on the question of his ability to confrom his behavior to the law--it was a split vote by the jury, where preponderance of guilt prevailed.)

But my point is about Fischer's semi-compatibilism, which I would argue fuses questions about reasonableness of minds with questions about dual-ability of choosing the good. Reasons-responsiveness combines both simply into one issue. So the separate traditions of reasonableness and choice are combined by semi-compatibilism into one eloquent answer as to what constitutes responsibility.

[John Fischer]

Hey Tamler,

Derk's getting paid the BIG BUCKS for doing this this month...

[Sofia Jeppsson]

I think the historical free will/moral responsibility-debate has always involved several issues, not just one. Robert Allen above mentioned Augustine - well, he was in part very concerned with arguing that human beings didn't have libertarian free will, because that would threaten God's omnipotence (but people could still deserve to go to Hell!). I think the debate, even thousands of years ago, had elements both of basic desert, whether we have it or not, what kind of free will we have, in what sense humans might have the ability to do otherwise and so on.

Anyway. I think it's pretty hard to pin down exactly what might be meant by "basic desert". As Derk defines it here, one can interpret P F Strawson as talking about basic desert, because blame on his account is justified when an agent did something wrong. One needs no forward-looking considerations in order to justify it. Sure, Strawson talks about blaming each other as being an important part of human interactions, but he doesn't seem to argue for the position that blame is justified because it promotes a further value which is "human interaction", the way consequentialists argue that blame is justified if it promotes the further value of a society of well-behaved people.

I think it is important to discuss whether "basic desert" is merely a normative-ethical concept, or if it's simultaneously a meta-ethical concept. That it's normative-ethical seems obvious. That we ought to blame people when blaming can serve to promote a well-behaved society, or that we ought to blame if and only if someone did wrong, are distinct normative-ethical positions, and it's the latter one which is the "basic desert" notion. But suppose one has the metaethical position that all moral concepts, right and wrong as well as moral responsibility, are created through some kind of implicit contract - stuff that rational agents could reasonably come to accept in order to be able to live together or something like that. This metaethical view is compatible with holding the normative-ethical view that one ought to blame people if and only if they did wrong, regardless of the consequences that come out of the blaming.

Now, would a person combining this normative-ethical and metaethical view believe in basic desert or not?

[Robert Allen]

The reports of Augustine's Compatibilism have been greatly exaggerated. Consider: "Therefore nothing is so completely in our power as the will itself, for it is ready at hand to act immediately as soon as we will." Sounds like Agent Causalism to me. He also asserts that the issuing of Commandments is inconsistent with us being unable to act differently than we do.

What is omnipotence? Must God be controling us to be almighty? If our wills, which He created, were inviolable, would He be any less powerful than He would be if our actions originated in forces beyond our control? I think not. But thanks for the citation, Sofia.

[Patrick Mayer]

I want to echo Sofia Jeppsson's claim that Strawson is making a meta-ethical point not a normative-ethical point. In Skepticism and Naturalism this is made very clear I think.

But I wanted to ask a different question. Derk says:

"For an agent to be morally responsible for an action in this sense is for it to be hers in such a way that she would deserve to be blamed if she understood that it was morally wrong, and she would deserve to be praised if she understood that it was morally exemplary..."

And I wondered whether you would be amenable to a slight revision. So on a standard moral theory act types like 'killing children to achieve local racial purity' are wrong and particular tokens of this type are wrong because of their being tokens of this type. Consider someone who accepts a non-standard moral theory according to which actions of this type are not wrong, but rather obligatory. Let us further suppose that this non-standard moral theory is wrong. If this person kills a child to achieve local racial purity, and knows that he has killed a child to achieve local racial purity, but does not believe that this action is wrong, then he fails to meet the conditions you put on basic desert. But it isn't plausible that he does not deserve blame simply because he accepts an incorrect moral theory.

So how about the following revision:

For an agent to be morally responsible for an action in this sense is for it to be hers in such a way that she would deserve to be blamed if she understood that the action was the token of a type of action with property X, where property X is the property of being morally wrong according to every moral theory which might be true....

Or in other words, if moral wrongness and rightness are grounded in non-moral properties we only demand that the agent know that her action has the properties which are the grounds for the moral property.

[Randolph Clarke]

I seem to recall that Feinberg takes the fittingness of reactive attitudes to be what desert of them comes to, and I'd think that desert, as Feinberg thinks of it, counts as basic on Derk's view.

I have a worry about the idea that the fittingness of a reactive attitude is simply a matter of the truth of its cognitive component. It seems that there might be different and conflicting attitudes with the same cognitive component. For example, the thought that one has oneself done wrong might be an element in different attitudes -- guilt, one that I'd be likely to experience, and pride, one that a devil might experience. But I wouldn't say that the devil's response is fitting. Perhaps there's something to this view that I'm overlooking or misunderstanding.

[Derk Pereboom]

The remarks by Robert, Fritz, Michael, and Sofia make me want to rethink some of these broader issues. I’ve been assuming that the core issue in the free will debate is whether compatibilism or incompatibilism is true, but it might well not be right to think that there’s just one central issue, as Sofia proposes. And so even if it’s right to say that the main controversy between compatibilists and incompatibilists is about whether we have the control in action required for basic-desert moral responsibility, that doesn’t make this the one central issue in the free will debate. As recent work on blame has shown, there are quite a few different notions of blame on offer, and as Michael points out, for each there will be an account of the normative warrant for blaming, which raises the issue of the kind of control required for justifiably being blamed on that notion. Several competing notions of blame may be compatibilist-friendly, but they may differ on the kind of control required; for example, John’s reasons-responsiveness view may differ in this respect from Frankfurt’s position. Such a controversy is as central to the free will debate as any is.

Robert’s and Alan’s historical remarks about the free will debate are interesting and helpful.

On Fritz’s comment, I don’t distinguish between freedom and control in this context. Suppose we’re wondering if the main controversy between compatibilists and incompatibilists is about whether we have the control in action required for basic-desert moral responsibility. I wouldn’t want to stipulate this, but instead to say that it’s a hypothesis that makes best sense of what’s going on in this debate. But couldn’t it be instead, as Fritz proposes, that the core of the controversy is whether we have free will? That would be the most elegant hypothesis. Here’s the concern for this proposal that I’ve often heard expressed. As is the case for any term that’s been around in the language for a long time, ‘free will’ has a number of different meanings; in one sense, to have free will in doing A is to for both A and refraining from A to be open to the agent; perhaps in another sense, to have free will is to have the control in acting required for moral responsibility. But let’s restrict ourselves just to the first of these. In his discussion of the consequence argument in Chapter 5 of On Action, Carl Ginet considers several different possible accounts of what it is for an action to be open to an agent – Lewis’s account, the backtracking view, and his own incompatibilist conception. Arguably, each of these corresponds to a sense of an action’s being open to an agent. But in Ginet’s account, the final arbiter between Lewis’s and his own view is a manipulation argument , which suggests to me that for him the issue is settled by which notion of an action’s being open to an agent is appropriately linked to moral responsibility. But perhaps it’s possible, say, for the incompatibilist to rule out the local miracle and backtracking conceptions of free will as without an appeal to moral responsibility, and this would be illuminating.

On John’s suggestion that desert and basic desert be rejected in favor of fittingness or aptness, I’m at least initially sympathetic with Tamler’s challenge, and that perhaps ‘he deserves blame’ and ‘he’s a fitting candidate for blame’ are equivalent. But both Michael’s and Nate’s comments make me think that ‘aptness’ and ‘fittingness’ are more general notions than ‘desert,’ and thus to spell out one’s account in terms of ‘fittingness’ doesn’t imply that’s it’s not also an account that involves desert. James’s suggestions for questions that need to be answered to make further progress are valuable. At the same time, I’m not especially optimistic about achieving the kind of consensus he mentions.

I agree with Michael that basic desert is a richer and more specific notion than desert or aptness or fittingness. On Sofia’s comments on how to construe basic desert, in my initial post I spell out a view according to which we retain the ordinary basic-desert-involving practice, and doing so is justified on the ground that it’s the practice that results in the best consequences. There’s also a related contractualist view – James Lenman develops a position like this in his “Compatibilism and Contractualism” (2006). I’m thinking of the practice-consequentialism as a normative ethical theory, but not one that has a role in actual practice. But maybe, as Sofia proposes, there’s a related metaethical position in the wings.

Michael suggests that the there’s a tension between the Zen spirit and the use of the term ‘fighting’ in the title of the post. The tension is resolved here: http://www.sanfranciscosentinel.com/?p=60722. This may be worth some of those BIG BUCKS…

[Sofia Jeppsson]

Robert, I wrote a paper once on the argument between Augustine and Pelagius when I studied Religion at university, and from what I remember a lot of the stuff he wrote towards Pelagius (who was a libertarian) sounds definitely compatibilist. BUT this was years ago, so I won't press the point. :-)

[Sofia Jeppsson]

Patrick Meyer, I think this is a tricky point. On the one hand, I don't think accepting an incorrect normative-ethical theory should be an automatic excuse. Possibly most people who commit blameworthy acts (if we assume that Derk is wrong, and there are such acts in the first place) think they're justified in doing what they do. On the other hand, one might want to excuse people who just don't grasp moral concepts. On the third hand (?), maybe such people are already excused if we demand that the act must be wrong in order to be blameworthy and the other way around. Possibly acts that are performed by people who just don't understand morality at all can't have the property of being right or wrong, just, say, fortunate or unfortunate, desirable or undesirable.

[Patrick Mayer]

Sofia Jeppsson,
I think that there may be a good bit of controversial stuff being assumed in the case of the person who does not grasp moral concepts. On certain understandings of the content of our moral concepts I find the idea of a person who grasps the normal range of non-moral concepts but lacks any grasp of moral concepts very hard to conceive.

For a simple example suppose you were some kind of expressivist who thought that sentences like 'X is wrong' was equivalent in meaning or perhaps had the same force as an expression like 'I do not like X'. Such an expressivist, if asked what the content of the concept of wrongness was might say that it is just the same as the content of the concept of liking. I cannot imagine a person who did not grasp the concept of liking.

But of course such simple minded expressivism is really just a strawman. I think the problem comes out though, when you consider some different varieties of moral realism. On some realist theories our moral theories come out as consisting mostly of claims with a priori justification. On such views understanding the concept of rightness or wrongness gets you a long way towards the general principles that govern which actions are right and wrong. I take it that Kant had a view like this. On such an account of the content of moral concepts it is easy to imagine someone who doesn't grasp the concept, its anyone who doesn't accept the correct moral theory.

But consider a theory on which a good many of the general principles of the theory have a posteriori justification. A theory which said, for example, that while 'Good' doesn't mean 'desire satisfaction' nonetheless Good = desire satisfaction (in the way that water = H20), the concept of goodness might be really thin. Suppose that the theory says that Good = desire satisfaction because the concept of the Good is just some kind of phenomenal concept (Goodness is the thing that makes me happy about my life when I reflect on it). I find it hard to imagine an otherwise competent person who lacked that concept. Of course this is an oversimplified theory, but the point I want to make is that it seems like on these kinds of theories we get moral concepts that are so low in content that it gets hard to imagine someone lacking them who was not also cognitively deficient in other areas.

So long story short I am not sure I can think adequately about the deeply morally incompetent who nonetheless can do things like identify children or act types like murder.

But of course if my uncertainty is unfounded, then it would be easy to revise my suggested revision by just attaching the rider that the person who knows the non-moral facts that ground the moral fact must also be morally competent.

[V. Alan White]

Derk, thanks for the kind remark. What I intended was to show that law has pretty recently made a specific place for something like dual-able FW that entails basic dessert because its lack is completely exonerating (and still is in some jurisdictions). As well it occurred to me while writing that Fischer's view captures both classic components of mens rea in his view, though of course he does not rule the counterfactual test of reasons-responsiveness as significant of the presence of FW.

So here's a thought experiment to expand on your question. Say that a scientist discovers an L-detector that detects in people the presence or absence of an indeterministic process that always and only attaches to ordinary capacities for reason, and thus effectively establishes the presence (or absence) of libertarian FW. (In other words when L is absent, effective reasoning capacity is too, so it maps the usual territory of the presence or absence of basic-dessert-bearing responsibility in libertarian terms.) Libertarianism is vindicated, retributivism is too, even though we continue as well forward-looking responsibility practices.

But evolution goes on and something happens: a genetic variation appears that no longer involves the presence of L--our detectors can definitely say that is true--and yet ordinary reasoning capacities are, quite astoundingly, unaffected in these cases. Furthermore, some of these mutants frequently (and reasonably) do good things, and some bad things. Based upon reason assessment, these mutants are outwardly indistinguishable from Ls, and furthermore make reports about their mental phenomenology that are distinguishable from what Ls report. Yet we train our detectors on these mutants and damn--they lack L.

Why on earth would we then treat these mutants any differently based on the sole fact that they lack L? They do the same good deeds and commit the same heinous crimes as Ls. Their reasons for doing such are (apparently) indistinguishable inwardly and outwardly from Ls save the one difference that the indeterminism is lacking in reasonable mutants. Should we withhold our retributive practices from them and punish them only in forward-looking ways? Why should we set up a two-tiered system of justice based on the fact that we can detect this lone difference between Ls and mutants, where, for instance, we justify capital punishment for at least some L-vicious murders but not for mutant-equally-vicious murders?

[James Laird]

Patrick,

It seems to me that when a person judges the morality of a certain action, they do so with respect to a reference. For example, if I want to eat another life in order for me to live, I generally don’t perceive of any morality issues with that action (depending upon which life I consume). In other words, I decide the morality of my action based upon what’s best for me. If one of my peers wishes to harm me and take my meager resources so that he may live (thereby effectively eating *me*), I tend to see morality issues with that, whereas he may not.

For cases where a person doesn’t want to spend lots of time analyzing the morality of a certain action, it’s easy for them to simply judge morality by “I like that” or “I don’t like that”, and effectively they create a quick reference by which to judge. (e.g., I like eating, but I don’t like to be eaten.)

If humans really want to discover/define “absolute morality”, then we’ll first need to develop the ability to judge what actions produce absolutely the greatest overall ascent of life (a difficult task to say the least).

[Sofia Jeppsson]

Derk, I think Lenman's position should be read as a metaethical one, but I can see how it can be interpreted as normative-ethical. I think P F Strawson's position is definitely most plausibly interpreted as a metaethical position, either a kind of expressivism or a kind of naturalism. On the expressivist interpretation compatibilism follows from the fact that we express certain attitudes even if the universe is deterministic, and on the naturalistic interpretation compatibilism follows from the natural fact that people who are capable of normal adult human relationships sometimes do things on purpose, and this is true regardless of determinism.

If basic desert is supposed to be a normative-ethical concept, compatible with various metaethical stances, perhaps you should include in your definition that "blameworthy" and "praiseworthy" can't be REDUCED to a statement involving what it's right or obligatory to do.
On consequentialism, "blameworthy" can be reduced to "it would be right to blame this person" and likewise with praise. I think that Christine Korsgaard's notion of moral responsibility might be seen as reductive in the same way, and thus not be a "basic desert" notion. She writes that praising and blaming should be guided by the virtues of respect and benevolence; sometimes not holding someone morally responsible is the benevolent thing to do, sometimes holding someone morally responsible is the respectful thing to do. This idea, just like consequentialism, seems to reduce praise and blame to what is right to do. On the other hand, with the basic desert notion, it's right to praise and blame people because they deserve to be praised and blamed - but it's NOT the case that people deserve to be praised and blamed because it's right to do so.

I think the difference between reductive/non-reductive ideas of moral responsibility might better capture what you're after than the distinction between backward-looking and forward-looking, since Korsgaard's notion doesn't seem to be exactly forward-looking. (But perhaps not backward-looking either... but a kind of here-and-now-moral-responsibility, perhaps.)

[Sofia Jeppsson]

Patrick Meyer, I realise that it's problematic to discuss "grasping moral concepts", since, as you point out, what does that mean? At earlier stages in my dissertation I tried to spell out sufficient conditions for being a moral agent and morally responsible, and I included understanding moral concepts (later on I settled for freedom-related conditions only). One of my supervisours said that if that's what it takes, nobody is a moral agent or responsible... since who really understands moral concepts?
Obviously I meant it in some simpler, pre-philosophical sense, but it's hard to spell out what that sense would be. Maybe everyone who has a bit of rational thinking going on has enough of a grasp to be a moral agent and morally responsible - if we assume for the sake of argument that general scepticism about these matters is false.

V Alan White, my intuitions are that it wouldn't make any difference.
I suspect that people sometimes have a hard time separating general non-retributive intuitions from incompatibilist intuitions. We are so often presented with thought experiments featuring agents that were determined or manipulated in some way, and then asked whether this person really deserves to be punished (or at least the latter question is often implicit in thought experiments about moral responsibility featuring, for instance, murder). In order to properly distinguish incompatibilist fron non-retributive intuitions we should make thought experiments like the one you just presented, where people with libertarian free will are compared to people who merely have compatibilist free will, and ask whether we really would like to treat them differently.

[Derk Pereboom]

Patrick, I think I’m happy with your suggested revision. I’m wondering if some of the issues for your proposal that you, Sofia, and James were discussing might be dealt with by adding a provision to the effect that the agent must have some level of understanding of morality, although it’s difficult to make this precise. And I need to think more about Sofia’s points about how the issues divide up between normative ethics and metaethics.

Alan, you say: “my point is about Fischer's semi-compatibilism, which I would argue fuses questions about reasonableness of minds with questions about dual-ability of choosing the good. Reasons-responsiveness combines both simply into one issue. So the separate traditions of reasonableness and choice are combined by semi-compatibilism into one eloquent answer as to what constitutes responsibility.” I can take this on board, provided that ‘responsibility’ is interpreted in a forward-looking sense, where the practice of holding responsible aims at goods such as protection, moral formation, and reconciliation. So I’d want to say that there’s a non-retributive view that can accept John’s eloquent answer.

In reference your thought experiment, you ask about the mutants who lack the indeterministic free will that the Ls have: “Why on earth would we then treat these mutants any differently based on the sole fact that they lack L? They do the same good deeds and commit the same heinous crimes as Ls. Their reasons for doing such are (apparently) indistinguishable inwardly and outwardly from Ls save the one difference that the indeterminism is lacking in reasonable mutants.” I would agree with you if the only difference between the L’s and the mutants is indeterminacy. But I would have the contrary intuition if the Ls had indeterministic agent-causal free will and the mutants lacked it.

I’m thinking that Randy’s concern about the idea that the fittingness of a reactive attitude is simply a matter of the truth of its cognitive component is persuasive.

[V. Alan White]

Derk, again thanks: your excellent questions have really pushed me to think harder (though maybe not hard enough; I suppose I could have called my mutants "FW Zombies" just for the cache'). The reason I required that the L-detector detected indeterminism only in people with associated reasoning power was to be inclusive of event- and agent-causal forms of libertarianism by correlation of some role of indeterminism with the exercise of reason. My thought is that typically libertarians exempt some people from basic desert by lacking FW with respect to at least some classes of actions (completely irrational actions over a wide range or specific OCD-types of unreasonable action that are partitioned from more reasonable ones that someone with OCD might otherwise engage in). So the verification that Ls always and only have indeterminism associated with reasonable capacities is compatible with event- or agent-causal accounts. The introduction of the mutation that destroys this correlation yet preserves a consistent description of some people with ordinary rational capacities forces an issue in a deliberately pragmatic way: we now have two slightly different types of people who behave indistinguishably with respect to moral action and rational capacity yet inhabit the same actual world. What this does from my perspective is eliminate the uncertainty of leveraging one possible world description of Ls against another possible world of mutants (or if you will FW Zombies). After all, we can easily pigeonhole our intuitions within possible worlds and claim precedence for one or another world, championing Ls or mutant-FW Zombies as reflecting some other more salient and no doubt abstract intuition. But my thought experiment uses mutation as a plausible means to place the contestants in one world, and thus more pragmatically and straightforwardly offers a challenge of how we are to regard Ls and mutants co-inhabiting one actual world doing good and bad stuff. Seems to me that if one wishes to avoid the single-world confrontation I offer then one would have to argue that there is an inherent conceptual way that Ls and mutants differ with respect to rationality and moral responsibility. I do not see that, at least without avoiding circularity; and thus the question resolves (for me) to a pragmatic moral one--why should we treat them differently? Now if we had previously treated Ls retributively because of the strict correlation of indeterminism with rationality, and yet we see later that rationality does not require indeterminism, maybe we should revise or eliminate retributive practices, or justify them in ways that doesn't refer to the indeterminism difference. My thought experiment tries to force the issue one way or another.

[Manuel "I'm a Lover, Not a Fighter" Vargas]

Sorry to arrive late to the party, but I'm wondering about the initial characterization of what we are supposedly fighting about.

Derk writes that the forward-looking justification of praise and blame view isn’t one that’s at issue in the historical debate, because on this view justifications for blaming and praising are exclusively forward-looking, and the traditional incompatibilist has no quarrel with [this]" because the real issue is the threat to control (or the threat posed by luck) that results from causal determination or from indeterminacy in action."

I take it that basic desert isn’t interesting unless it is supposed to be about a notion of desert that is required for our existing responsibility practices to be warranted, licensed, or morally adequate. And, I’ve also thought that at least Derk (and some others) are worried about varieties of control that are important for real world responsibility. But the link to actual practices better persist here, too. If we’re talking about notions of control that are distinct from the ones implicated in the actual practice of ascribing and holding people responsible, that seems like an uncharitable way to interpret many of the debates, at least in the responsibility-centric wing of the debate.

So, I just disagree that “the threat to control” is the basic issue. Instead, I think the integrity of our responsibility practices is the basic issue. Or, at least, that’s the issue I’m interested in, and that I’ve taken folks like Derk and others to be interested in. (I’ve argued in several places that there are distinct debates going on here; I don’t think everyone is interested in the forms of agency sufficient to sustain normatively adequate responsibility practices.)

As I see it, what should be threatening about forward-looking (and other normative) accounts—to hard incompatibilists, anyway—is precisely that they purport to provide an independent justification for responsibility, a justification that doesn’t rely on the notions of control threatened by determinism. Normative approaches (of this sort) are effectively attempting an end run around the hard incompatibilist’s worries, showing that *if* what we are worried about is justification of praise and blame, including ascriptions of desert and backwards looking judgments, then we don’t need the incompatibilist notion of control to do that.

Maybe incompatibilist powers would be sufficient for those practices, too. But, on this view, incompatibilists are (sometimes) mistakenly thinking that there are no other paths sufficient to supporting our practices as we find them. Of course, we might need incompatibilist pictures of agency for other purposes, but those are, well, different purposes.

It is a bit like us agreeing that we need to get across town, with one person pointing out that a bus is sufficient to do it, and the other person objecting that the empty gas tank on our car will keep us from getting there. No matter how empty that tank is, it is just irrelevant to the question of whether we can get across town once we've got public transit sufficient to get us there.

[James Laird]

Manuel,

I like your transportation analogy, but there may be a rub… I’m thinking that the public transportation is fictional, unless the “individual” actually exerts forces that aren’t predeterministic in nature. Here’s another way of looking at it: we can model the interaction of two ocean waves running into one another on the surface of the ocean as having interaction at the “wave level”, but the truth is: all of the interaction of the two ocean waves is controlled solely by the four fundamental forces of physics (4FFOP), and modeling the interaction from the top-down is simply a mirror image of modeling the interaction from the bottom up – there isn’t any control truly exerted from the “wave level” (i.e., top-down causation) that humans can perceive of.

That gets us closer to the heart of the matter: if the actions of an agent are controlled *solely* by the 4FFOP in a bottom-up predeterministic manner, then there isn’t any reason to believe that the agent could have done otherwise, and therefore the agent isn’t morally responsible in the “strong sense”. Yes, the agent is morally responsible in the “weak sense” (i.e., it’s true that the agent’s physical body performed the action and therefore his body is responsible), but is that really what we’re interested in? Aren’t we searching for something more? To say that a bowling ball is responsible for knocking down the pins doesn’t really get us closer to our search for “right and wrong”.

In order for arguments regarding moral responsibility to hold any water, I’m thinking that we *must* show that something more than predeterminism exists. We need to show that new life emerges, and by doing so, we’ll be able to fundamentally connect moral responsibility to the actions of the “individual”. I’m glad you made it to the party!

[Kelly McCormick]

It looks like there are at least two interesting questions at issue here. First, the one that Derk identifies in the initial post: whether we have the kind of control required to be morally responsible for our actions in the basic desert sense. Second, the one that Manuel identifies: whether there is a form of agency sufficient to sustain normatively adequate responsibility practices.

These questions are conceptually distinct and both look like candidates for what it is we are interested in when it comes to responsibility (what it is we’re fighting about). To borrow from Manuel’s transportation analogy, let’s say that providing an answer to each of these questions represents a place we might be trying to get. According to Derk it’s A, where we end up with an account of the circumstances under which agents are responsible for their actions in the basic desert sense. According to the Manuel it’s B, where we end up with an account of a form of agency sufficient to sustain/warrant normatively adequate responsibility practices.

It’s important to note that A and B could end up being the same place. It’s possible that the best account of when agents are responsible in the basic desert sense is also an account of agency sufficient to sustain normatively adequate responsibility practices and vice versa. In fact, this sounds ideal. But it’s also possible (and perhaps more likely) that A and B are on opposite sides of town. Finally, it’s possible that getting to A requires a trip through B (or, again, vice versa). So, here’s a third question: are we more interested in getting to A, or to B, and does getting to one require a detour through the other?

I’m inclined to say that both of these questions are central to what we’re interested in, and that pursuing one without attention to the other leads to problems on both sides. As Manuel points out, an account of responsibility-in-the-basic-desert-sense that fails to link up with our existing responsibility practices, while theoretically interesting, fails to provide any guidance about the justification of those practices (or practical questions about whether or not they should be sustained, revised, or jettisoned) if it does not also address whether or not there might be other forms of agency sufficient to sustain and warrant them. But purely consequentialist accounts that attempt to provide an independent justification for our responsibility practices face serious shortcomings too. They appeal to facts about what is independently valuable to justify our continued participation in these practices, rather than facts about the conditions under which agents deserve praise and blame. A consequentialist account therefore leaves responsibility held hostage to independent axiological claims and our assessment of whether or not the relevant value can be promoted or sustained without our responsibility practices. And those inclined to take the responsibility-in-the-basic-desert sense to be central to our interests are likely to think that this kind of view is missing the point in some way.

The main point here is just that I think it’s helpful to get clear about these two distinct questions. I’m not sure either of them can make, on their own, a claim to being the bedrock issue we’re arguing about, and thinking about the relationship between them could potentially be pretty fruitful.

[Robert Allen]

Manuel,

It is not just getting across town with which I'm concerned. I want to get there on my own, instead of relying on someone else. I would be disappointed were I not in control of the vehicle in which I was riding (not to mention without XM radio). My very self-respect depends upon the belief that I have to depend upon myself for transportation, that Someone has empowered me so that I am completely in charge of how I move about this earth, making me solely responsible for where I end up and how I got there.

Having said all that I should add that I actually detest driving a car and much prefer using public transportation, like the kind you guys have in the great city of SF. That way I can read a book or listen to Philosophy Talk on my iPod. "Leave the driving to us," as Greyhound used to say.

[Saul Smilansky]

Manuel - I am confused. Are you now in agreement with Derk that all our free will-related reactions and practices make sense only in a forward looking, consequentialist way? So that (assuming the absence of libertarian free will), revisionary compatibilism amounts to hard determinism cupled with consequentialism?

As I argued in my own posts, I do not agree that HD can, from its perspective, end up with a version of consequentialism, and still be true to itself as a moral position; e.g. be just. But setting that aside, do you understand your position as something like "HD-consequentialism- but no reason not to call it compatibilism"? Or, to put it another way, do you see your dispute with Derek as basically semantic, i.e. his mistake is only that he doesn't see that actually he is a compatibilist?

And if that is what you believe, how can you say that "incompatibilists are (sometimes) mistakenly thinking that there are no other paths sufficient to supporting our practices AS WE FIND THEM". Derek clearly doesn't seek to vindicate our current reactions and practices. And surely a consequentialist forward looking revision of compatibilism will mean something very different than traditional compatibilism, which is happy with blame and punishment based upon backward looking evaluations.

So please clarify.

[Derk Pereboom]

I like Kelly’s way of separating the issues raised in Manuel’s post – (1) whether we have the kind of control required to be morally responsible in the basic desert sense, and (2) whether there is a form of agency -- or, alternatively and more broadly, a justification – sufficient to sustain normatively adequate responsibility practices. But as Kelly indicates, these issues are related. Suppose we want to retain our practices as we find them, which involve appealing to basic desert justifications for how we treat others. If we lack the kind of control required to be morally responsible in the basic desert sense, the burden falls on the kinds of resources to which consequentialists or contractualists appeal to justify these practices. I’ve questioned whether those consequentialist and contractualist groundings are up to the task, and so I’ve tried to show that we might be able to do without the part of our practice that involves appeal to basic desert justifications.

To address Saul’s point, one can imagine a free will skeptic maintaining that the part of our practice that involves appeal to basic desert justifications can be vindicated on consequentialist or contractualist grounds, but as Kelly suggests, it can’t be assumed that these strategies will be successful. A further worry about this sort of view – expressed by Robert in his post – and maybe also by James – is that the conception of ourselves and others we would then assume in our practices would be importantly false or inaccurate. A related concern -- raised by Bruce Waller (1990: 130-135) – which I use to motivate the last two chapters of Living without Free Will (156), is that on such a proposal we would be treating people as if they basically deserved to be blamed or punished, while they would not really basically deserve to be blamed or punished, and this would seem unfair.

On the thought experiment as specified in Alan’s last post, I now find myself unsure as how to react, because if the L-agents had event-causal libertarian agency, which is epistemically possible from the perspective for those assessing the thought experiment, I would opt for eliminating the retributive practices for the L-agents and the mutants; but if Ls had agent-causal libertarian agency, which is also epistemically possible, I might opt for retaining the retributive practices for them but not for the mutants, even when I keep in mind that this is a single-world confrontation - (well, not really, since I think that retributivism can be successfully challenged on other grounds). But maybe I’m being obstinate in ways that people who are less settled in their views might not be.

[Mark Young]

Saul,

Manuel wrote "I think the integrity of our [existing] responsibility practices is the basic issue." He also wrote that "basic desert isn’t interesting unless it is supposed to be about a notion of desert that is required for our existing responsibility practices to be warranted, licensed, or morally adequate." He said "we don’t *need* the incompatibilist notion of control" (emphasis added) to achieve our goal "*if* what we are worried about is justification of praise and blame, including ascriptions of desert and backwards looking judgments" (emphasis in the original).

IOW, he's *not* saying "all our free will-related reactions and practices make sense only in a forward looking, consequentialist way".

If basic desert is the correct theory of how our current system of responsibility judgements works, and our current system of responsibility judgements have the good consequences that these theorists think they do, then those consequences justify our actions and it's irrelevant whether we have the control incompatibilists think we need.

If basic desert is *not* the correct theory of how our current system of responsibility judgements work, then either it's the basis for a *better* system or it's irrelevant. If it's a better system, then it'd be interesting to know *why* it's better....

[V. Alan White]

Thanks for your very thoughtful remarks yet again Derk.

The thought experiment came to me in something of an instant, but as I mull it over I think it has some real legs--because your comments have given me a lot more to think about.

The scenario I describe uses a logically possible device--the L-detector--to correlate indeterminism with a verifiable instantiation of reasoning capacity. That (in Humean terms) suffices to establish as you say sufficient epistemic grounds for believing some form of libertarianism is at work in the world, whether event- or agent-causal in type. I intended the narrative to do just that--encompass descriptively any form of L-type agent. Then, in terms suitable for naturalistic libtertarians like Kane or Chisholm I introduce a change that yields seeming agents (mutants/L-zombies) who are indistinguishable from the Ls on rational grounds, but lack the indeterminism as detectable by the L-device. The ploy of this occurring in one world forces the question of how to treat them as the same or not, at least on pragmatic grounds.

Now it occurs to me that the thought-experiment also forces a meta-issue: why should metaphysical considerations such as the differences between Ls and L-zombies matter more than pragmatic considerations as faces someone in the world I describe? What possible axiological grounds married to metaphysics are sufficient to dismiss pragmatic considerations of treating Ls and L-zombies the same? I just wanted to push the pragmatism point with the original narrative--now you make me think it has more critical application to basic questions beyond that.

[Paul Torek]

V. Alan,

Lazarus (http://www.pcworld.com/article/233371/lazarus.html) for Chrome, Firefox, or Safari browsers will save you that lost hour next time. It probably justifies the (minimal) effort to switch browsers, too.

Sorry, all, for the non-philosophical interruption, but maybe we'll get more philosophy and fewer lost hours.

[Robert Allen]

V. Alan,

We are going to turn "pragmatic" over something as exceedingly important as moral responsibility? Judgments leading to reward and punishment surely deserve greater credence than (say) the tactical decisions of a football coach. By your own admission the L-zombies are ultimately being caused to do things by forces independent of their own wills: nature, nurture or both. The fact that they inhabit the same world as those exercising true self-control is no more morally relevant here than in cases where 2 criminal defendants both appear guilty but only one has really committed a crime. Better yet, normal adolescents are strikingly similar to adults in lots of ways pertinent to reasoning and even manage for the most part to exhibit self-control, yet they are morally and legally judged by different standards. Like cases must be treated alike, but between the L and LZs you have all the moral difference in the world: true freedom versus determination by outside forces.

I for one hope that you remain "obstinate," Derk, as you are defending a vitally important truth.

[Paul Torek]

Derk,

Sofia already raised the point about contractualism that I want to discuss, but let's start with your definition of basic desert. What is "contractualism" in the phrase "not, for example, merely for ... contractualist reasons"? What makes it *mere*? Are there any ethical philosophers who have been called contractualists (or cognate terms), whose normative reasons in favor or against various social policies* would count or not count as "mere" contractualism in the relevant sense? Can we draw a cutoff in a space encompassing Gauthier, Binmore, Scanlon, and Habermas, among others, with "mere" reasons on one side and, I guess, "loftier" ones on the other?

And why complicate the definition of basic desert in this way?

*Note, the same sorts of reasons - that is, "mere" reasons or "lofty" reasons - could be applied to many different areas, only some of which relate to praise or blame or other particularly salient issues to the Flickers community. So even if a given philosopher says little about our favorite subjects, we can still discuss whether the kinds of reasons she likes to deploy would be "mere contractualism" in the relevant sense, if applied to the issue of praise and blame.

[Manuel "Lover, not Fighter" Vargas]

Saul: I think there are two issues here. One has to do with the range of options for those resisting hard incompatibilism. The general point here is that any independent justification of the practices is going to present a kind of challenge to hard incompatibilism. The second issue has to do with the particulars of my view. I'm inclined towards something like a rule consequentialist picture—allowing that the justification of the practices will largely derive from the forward looking benefits, while maintaining that the structure of the practices themselves have backwards-looking elements. The details of the view can be found in chapter 6 in BBB (available everywhere soon!).

[Manuel "Somos Guerreros" Vargas]

A quick thought on the exchange with Kelly and Derk: I do think that there are distinct issues here, and unsurprisingly I think the more important has to do with whether and how much of our practices we can justify on grounds independent of what incompatibilists invoke.

I suspect the nub of the issue between at least Derk and myself turns on (1) whether and how much our practices turn on basic desert and (2) whether and how much of practices can be captured on consequentialist/contractualist grounds.

I'm not convinced that much of our practices in fact turn on basic desert, and I think almost everything we want justified or normatively adequate can be gotten on broadly consequentialist/contractualist grounds.

I take it that Derk thinks a rather larger swath of our practices depend on basic desert, and that he's rather skeptical that non-basic normative approaches can recapture or re-ground those claims.

Notice, though, that Derk's position has to show two things: both that basic desert is a widespread feature of our practices and that those features can't be re-anchored in a consequentialist/contractualist story. In contrast, I'd be happy if either it turns out that basic desert really isn't much (or at all) implicated in our practices OR if those places where it is, we can hold on to those practices, attitudes, judgments for reasons disconnected from the basic desert convictions.

Naturally, the devil is in the details on all sides . . .

[Robert Allen]

No Manuel there are no devilish details here at all; the matter is quite straightforward. If I lack the ability to do otherwise or am not the sole cause of my choices, then you can talk all you want about how it might be helpful to me and others if I were nevertheless held responsible for them, the cold hard truth is they are not my responsibility. Go ahead sever the tie between judgments and BDCs, re-anchor them in pragmatic considerations, all you will succeed in doing is creating the illusion of responsibilty. Do it out of the best intentions and you will still end up with the same brave new world.

[Manuel "Cumbaya" Vargas]

Robert: I do agree that the position you have expressed is one that any credible non-libertarian account will have to address. Indeed, it is a view I may well have endorsed at one point.

At this point, though, I'm so pathologically lapsed from "the hard truth" that I doubt blog discussions will change my mind. However, in a conciliatory spirit, I hereby pledge to take seriously the possibility that I'm being dogmatic and/or culturally parochial about how I think about free will and responsibility.

[Bruce Waller]

I have received a message from God, and She offers us a very attractive bargain: If we will inflict punishment on Manuel Vargas, then She will guarantee that for a full year there will be no crime, no sickness, no pestilence, and global warming will be suspended. She doesn't require waterboarding or the rack -- just a couple of days of solitary confinement on bread and water will suffice. I don't know why She selected Manuel -- perhaps due to his profound moral virtue, or maybe due to his marvelous characterization of the issues between Derk and Kelly -- but anyway, that's the deal. I think we should take it: I'm sorry that Manuel will suffer punishment, since I greatly admire his insightful philosophical work together with his vigorous yet insouciant writing style; but the positive consequences are overwhelming. But even as we punish Manuel, it is important that we keep firmly in mind that he is not morally responsible, and that he does not justly deserve punishment. There may be very good pragmatic reasons to punish Manuel, but they are distinct from the important question of whether he justly deserves punishment.

[Manuel "Threadjacker" Vargas]

So that explains the past 6 months of my life!

I agree something needs to be said about desert, and that there is something dissatisfying about most accounts of "purely pragmatic" justifications.

(That's why I try to say something different, having to do with the role that praise and blame play in our self-governance. But maybe that doesn't work?)

The point here, though, is that we have more options for understanding the basis of desert than just basic desert (as Derk defines it), or useful effects for other people (as simple-minded versions of consequentialism suggest).

Have I killed this thread yet by over-posting? I'll stop now, I swear.

[Saul Smilansky]

Apologies for taking a while to respond, couldn't write earlier. The core notion of desert is responsibility-based and backward-looking, i.e. deontological. So I think that it is very problematic to continue to talk about desert if our view is only consequentialist and forward-looking. I at any rate would then be inclined to speak about "desert", "blame", "justice" etc, not the genuine articles.

I think that it is very interesting that we have this "consequentialist turn" in the debate, with both hard determinists and compatibilists going that way. I've said before why I think this won't work for HD. I disagree, Manuel, that this IS a strength for compatibilism. Going consequentialist leaves the central moral and personal concerns which lead to the worry about free will unaddressed. So either we conclude that the free will problem itself really doesn't matter and live happily ever after as consequentialists, or it is a weakness for compatibilism if it retreats to consequentialism. Indeed, it is not clear that it continues to exist.

[Mark Young]

I wonder if there's some mismatch in the meaning of "justifies" that's at issue here. I agree that jailing Manuel would be unjust -- and so the good consequences don't "show the justice" of the action -- but I think it would be the right thing to do, anyway -- the good consequences give us "adequate reason" to so act. Perhaps those who argue that consequentialist reasons can ground our current MR practices have only the latter sense in mind, whereas those who object have the former?

I'd also like to note that our current MR practices *are* backward looking, and so even if these people are correct, they're NOT making our practices "merely forward looking". There's a difference between justifying a system and justifying moves within the system.

Finally -- you're not over-posting, Manuel. Please don't stop.

[Derk Pereboom]

We discussed the two-level view at the Arizona conference over the weekend, and one point I was arguing is this: if the aim of a responsibility practice, as Manuel puts it, is to build better beings, there will be various candidates in the running, and the fate of each can be determined only empirically. We philosophers can make a case that some practice is in in fact in the running, and that’s what I try to do for the practice that dispenses with basic desert justifications, and others have tried to do for practices that don’t. But the mere fact that our current practice involves basic desert justifications, or for that matter non-basic desert justifications, doesn’t count in its favor, and so wouldn’t give it an initial edge.

Echoing a point of Bruce’s, suppose a practice that involves basic desert justification does win out, and that determinism is shown to be true. Some of the folk might still wonder whether agents are really morally responsible in the basic desert sense, and they could come up with manipulation arguments to support their concerns. And these concerns would at least be coherent.

[Saul Smilansky]

Derek - good, this helps to see the issues and where we disagree. Two points: first, viz. your first paragraph, it seems to me problematic if we see the issue only in terms of "building better people" (I am not saying that you are doing this). Surely central moral concerns are treating human beings with respect, and being just. And it is these concerns that bring up the worry over the free will problem in the most acute way. If we focus on BBP we are in danger of losing track why we cared about the FW problem in the first place.

Second, you write "But the mere fact that our current practice involves basic desert justifications, or for that matter non-basic desert justifications, doesn’t count in its favor, and so wouldn’t give it an initial edge". Well, I think that while it might be that some radical revision can be both viable and morally best, it is advisable to start from a conservative assumption. Compare a proposal for a radical revision in family life. Perhaps one wants to propose the idea of doing without fidelity in marriage, something like an "open marriage". Surely the fact that typically people ARE envious if their spouse sleeps with someone else is a good starting point for the discussion, rather than working on the assumption that we can "build better marriages" irrespective of the constraints of human nature. Similarly, it seems to me a good idea to start from social and psychological facts about blame, resentment, compunction etc, and then see how much, if any, room for radical revision there is. Likewise for our moral intuitions (e.g. that justice is inherently related to desert in certain contexts): we might be convinced otherwise, but best to start from conservative assumptions.

[Derk Pereboom]

Saul - I agree with you that well-attested facts about human nature are highly relevant for determining whether it would be best to make the proposed changes. But this would involve a defensible conservative predisposition, by contrast with maintaining that the mere fact that our practice currently has certain features is a reason to retain those features. At the same time, it's important to be cautious of too strong a conservative predisposition. Much of the social and political progress we've made over the past three centuries -- for example in government, in punishment practices, and in the ways children are raised and taught -- is due to conservative predispositions being overridden, despite the expected temporary instability produced by these kinds of changes.