A question of desert

I think the following is right:

The contemporary debate seems be tracking more toward offering something like rival accounts of what it takes to be responsible in the desert-entailing sense. There also appears to be a general sentiment that if incompatibilist theories worked otherwise, they would easily secure desert-entailing responsibility, whereas, even if compatibilist theories work, they still bear a special burden of explaining just why satisfying those compatibilist conditions is sufficient for the agent to deserve, say, blame and praise.

At least, the above is my impression. But apart from some comments on the Garden, (and a few remarks in personal correspondence), I'm not sure of any concrete instances which support the observation. That is, I think there is this general presumption that incompatibilist theories have at least this advantage: if they worked they'd secure desert-entailing MR. But this is a general impression I have - not something I could cite examples of.

So I thought I'd solicit The Garden, both to check to see if the above seems like an apt observation, and in the hopes of some pointers to better evidence of the above than my rough impressions.

Comments

[Tamler Sommers]

Matt, it seems to me that the substantive debate has to be about moral responsibility in the desert-entailing sense. Incompatibilists recognize that agents in a deterministic world could be responsible in other ways. The burden for compatibilists is to show that the particular conditions they lay out can ground DEMR. When philosophers who some call "compatibilists" don't try to do this (e.g. J.J.C Smart, Spinoza) they are arguing for skepticism, not compatibilism.

You write:

"I think there is this general presumption that incompatibilist theories have at least this advantage: if they worked they'd secure desert-entailing MR. But this is a general impression I have - not something I could cite examples of."

I guess this depends on what you mean by "worked." If you mean that the theories are AIMING to secure desert-entailing responsibility, then I'd agree. On the other hand, many think that even if the libertarian's indeterministic account of agency were correct, it would not help to secure DEMR. The most common criticism of Kane's theory, for example, is that his indeterministic account of SFAs cannot handle the problem of luck and so does not add anything DEMR-wise to rival compatibilist accounts. G. Strawson makes a similar criticism of agent-causal views like Clark's--arguing that the regress argument would apply equally well to an agent-causal self.

So I think libertarians share the compatibilist burden of showing how their theories could ground DEMR. Skeptics, of course, do not. That's my sense, anyway.

[Matt King]

That sounds right, too, Tamler. I should be more mindful to not forget about the skeptics.

Though notice that on G. Strawson's account, he, too, (as I recall) is silent on why his conditions, when satisfied, would entail desert. He seems to presume it in the same way the observation I posted about does.

[Sofia Jeppsson]

Matt, I agree too that something is lacking from Strawson's account. He thinks that if it was somehow possible to be the ultimate cause of oneself THEN it would "make sense" to claim that people could deserve Heaven or Hell. But he doesn't say anything about why that would suddenly make sense.

I think praise and blame can make sense whether the world is deterministic or not, but I don't think one can justify, say, torturing people because they've been bad. And I can't see any reason to change my mind on the latter topic, even if there would come up some brilliant new argument that proved that a theory like Kane's gave us both freedom and control in some ultimate sense, and empirical science then went on to show that this is how our brains work. And I don't know what it would even mean to say "okay, the claim that some people deserve to be tortured can still not be DEFENDED, but since this new philosophical argument and these new empirical findings came up it started to MAKE SENSE."

[Tamler Sommers]

Here's a charitable way to interpret Strawson. He doesn't think that being the ultimate cause of ourselves would make us deserve heaven or hell. He just uses the heaven/hell analogy to make sense of the concept of desert (because of its non-consequentialist nature, not because of its association with eternal torment or bliss). Nor does he think that being causa sui (which is impossible) would give us desert entailing MR. Rather, he thinks that (a) we take ourselves to be causa sui, and (b) this belief/phenomenology is tied to our belief in desert-entailing MR. Because of how the (a) and (b) are connected, he thinks that being causa sui is a necessary condition for DEMR. But I don't see him committed to the view that it is a sufficient condition...

[Cihan]

"That is, I think there is this general presumption that incompatibilist theories have at least this advantage: if they worked they'd secure desert-entailing MR."

I don't share that 'general' presumption. Even if 'incompatibilist' (libertarian?) theories worked, I don't see how they'd secure DEMR. (To me, indeterminism just looks like luck.)

Also, all libertarian theories are speculations at best as they don't have any/substantial empirical evidence for their account and religion inspired nonsense at worst (i.e. agent causation).

[Sofia Jeppsson]

Cihan: You wrote "Even if 'incompatibilist' (libertarian?) theories worked, I don't see how they'd secure DEMR. (To me, indeterminism just looks like luck.)"
I interpreted Matt's "if they worked" as "if it could be argued to everyone's satisfaction that the indeterminism in question isn't just luck" or something of the kind. Am I right Matt?

[Matt King]

Yes, that's right, Sophia. And we needn't limit ourselves to libertarian views, hard determinist or hard incompatibilist views are relevant, too.

The question is: given a theory's conditions on DEMR, were they satisfied, has the account specified why such agents would deserve blame and praise *in virtue of satisfying those conditions*?

I think this is an open question for the extant accounts I can think of - and my diagnosis of why we don't see this question addressed in the literature is the presumption above. But maybe I'm wrong - and I'd love to have examples either way.

[Mark Smeltzer]

Matt I agree that this is a central issue. I wrote a post related to this issue back at the end of July -- c.f. Desert and Responsibility.

In that post I suggest that the right way to approach the question of moral responsibility is to start by looking at the question of desert: what would make something deserve to be treated a certain way?

It is by pursuing this question that we are able to make progress on answering questions like, "how should this person be treated?"

Although I only discovered Zimmerman's work on this topic after I had already set off on a similar track, his approach of "taking luck seriously" and offering an account of counterfactual responsibility is also worth looking into.

One thing you'll find is that this question has largely been taken for granted for the past 5,000 years. Or at least it has been taken for granted with respect to the journals, books, etc., that anyone I've known has had access to (I would love discover that there is a large body of work on this topic that I am unfamiliar with but so far I have been left empty handed).

I am quite sure that there is a precedent of appealing to a widely shared common sense view of desert that allowed the debaters to skip over some of these vital details, and this has led to much confusion for those who do not share the same common sense starting points about desert.

Resolution? There needs to be more straight forward talk about what desert is.

[Cihan]

Sofia, if "if it could be argued to everyone's satisfaction", then any view would work. In this regard, compatibilism is no different.

I interpreted "if they worked" as "if the indeterminism was actual" -i.e. the human decision making process actually "worked" that way.

There was a post by Kipman Werking about how both hard determinists and libertarians agreed that 'free will was controversial' but compatibilists thought that less magical powers could grant it. That post may be illuminating:

http://gfp.typepad.com/the_garden_of_forking_pat/2007/04/pereboom_on_the.html

[Tamler Sommers]

I don't agree with the general sentiment expressed here if I understand it right. Nobody has been letting libertarians off the hook about the problem of luck. Hume discussed it in the Treatise and Enquiry and people have been harping on it ever since. I mentioned Kane already and the most common complaint about his account. Another example is Nagel in "Moral Luck" (1979) who writes:

"Once we see an aspect of what we or someone else does as something that happens, we lose our grip on the idea that it has been done and that we can judge the doer and not just the happening. This explains why the absence of determinism is no more hospitable to the concept of agency than is its presence--a point that has been noticed often."

Note the "noticed often"...

If anything, until recently, both compatibilists and libertarians have assumed that one of them must be right (i.e. one way or another we're DEMR)--and so refuting their opponents meant supporting their own position. But compatibilists were just as guilty of this maneuver as libertarians. Thanks to contemporary skeptics like Pereboom, Strawson, and Waller this doesn't happen too often now.

[Sofia Jeppsson]

"I don't agree with the general sentiment expressed here if I understand it right. Nobody has been letting libertarians off the hook about the problem of luck."

I don't think you understand it right, Tamler. Yes, libertarians have been widely critisized for providing theories that wouldn't give agents any more control over their actions than they have on compatibilist accounts. They have been critisized for just substituting randomness for some of the determination.

But suppose that some new libertarian theory comes up that explains how it would be possible to have ULTIMATE CONTROL over one's actions, so that one's actions were neither fully caused NOR random in any way. Okay, lots of people (including myself) have trouble imagine how a third alternative to "fully caused" and "random" could exist, if we look at things from an objective standpoint. But let's just assume it does. Let's just assume that this brilliant new theory comes up, or these brilliant new arguments to supplement some old theory, that shows it possible to be in ultimate control of one's actions in a way that precludes both randomness and full causation.
Suppose next that it is shown that real people have this amazing capacity.

The assumption that lots of people seem to share is that it would now be obvious that people could deserve this or that. But why would this be obvious? It's not like it could follow in a strict logical sense from the fact that we have this ultimate control of what we do. So how would it follow? Is there an argument anywhere?

[R. Clarke]

What's desert? It would be nice to have a good account of that. Many of the posters seem to think that something rather grand would be needed to ground desert. How come?

[Eddy Nahmias]

Good question, Randy. But here's one for you (and others who develop libertarian theories, especially agent causal ones): if *not* to secure the sort of free will required for desert, what is libertarian free will supposed to be required for (i.e., what is it supposed to get us beyond what compatibilist theories get us)? And if desert does not require something grand, does it require libertarian free will (agent causation)?

It seems to me like the only reason to favor libertarian theories over compatibilist ones is that the former secure something essential to our most essential practices and beliefs regarding moral responsibility. If not, why go for such a theory? If so, what essential thing is it that libertarian theories get you? And if it's supposed to be desert, then (as others have been asking), what is it about desert that suggests libertarian theories *might* (if they "worked") secure it?

[Tamler Sommers]

Sofia, I see what you're saying--but something is still bugging me. Maybe it's this. The problem of luck is one that compatibilists have to face as well. As I see it, when we say that an action is the result of luck, we don't mean that the action was random. We mean that it is the result of factors outside of our control. (Certainly moral luck applies to deterministic accounts of actions.) So the idea that a brand new libertarian theory might solve the problem of luck and therefore ground desert seems equivalent to the idea of a brand new compatibilist theory solving the problem of luck. And in that case, both sides would be presuming that these brand new (perhaps unimaginable) accounts could ground desert.

As for Randy's question (and mark's comment), well--that seems like the dark secret of this whole debate. We're not sure. I tried to lay some possibilities here:

http://gfp.typepad.com/the_garden_of_forking_pat/2007/06/what_do_we_mean.html#comments

[Tom Clark]

Eddy writes:

"It seems to me like the only reason to favor libertarian theories over compatibilist ones is that the former secure something essential to our most essential practices and beliefs regarding moral responsibility. If not, why go for such a theory? If so, what essential thing is it that libertarian theories get you? And if it's supposed to be desert, then (as others have been asking), what is it about desert that suggests libertarian theories *might* (if they "worked") secure it?"

To the extent that our MR practices are deontological/retributive, libertarian theories seem to provide the strong sort of originative agency that intuitively backs up such practices. The essential thing libertarian theories give us is that it was the agent's doing only, therefore all credit and blame goes to her. She essentially and ultimately *deserves* them whether or not it results in good consequences. Intuitively, agency grounds desert. Compatibilist theories can't confer such ultimate desert since they admit the causal contribution of non-agent factors. This raises the question of on what basis compatibilist accounts can support retributivist MR practices.

[Fritz]

Randy can answer for himself. But some of us attracted to libertarian theories are not trying to get anything. We're trying to figure out the nature of free action and what's required for it.

By this I mean at least the following: as we explore these issues about free action we are open to the possibility that there's nothing especially valuable about being a free agent. We're also open to there being nothing morally important about being free.
(I hasten to add that we're open to freedom being valuabe and being importantly tied to morality too.)

Issues in philosophy can be interesting and important, after all, even if they aren't about morality.

[Eddy Nahmias]

Fritz, right, my first question got submerged under the rest of the questions that focus on MR and desert. That question was: if *not* to secure the sort of free will required for desert, what is libertarian free will supposed to be required for (i.e., what is it supposed to get us beyond what compatibilist theories get us)?

In light of your comment, here's a related version of the question: If (part of) our theorizing about free action (and its requirements) is not driven by (or importantly tied to) issues related to MR and desert (or other normative issues), then what sorts of issues should we use to guide us toward the target? Are we interested in figuring out the type of freedom involved in our phenomenology of choice/deliberation/action, in the metaphysics of action, in the psychological study of agency, or what? (I'm with you, by the way, but I'd like to hear what libertarians think.)

When van Inwagen says that everyone should define "free will" in terms of the ability to choose in more than one way (in certain situations), that's fine, but we still need to know what sort of ability to choose we are talking about, and that seems to get cashed out in terms of the MR/desert debates.

[Fritz]

Eddy,

You'd like to hear "what libertarians think"; I'm one. But I don't think we all think the same thing so if you mean you want to hear what other libertarians think then I'm with you.

Additionally, I think everyone agrees that lots of the inquiry is "driven by" (that is, motivated by) thoughts about its likely connections with some moral issues. But we take it to be an open question what the ties, if any, turn out to be at the end of our inquiry.

You ask what we're interested in: phenomenology (not so much), metaphysics of action (yes!), psychology of agency (yes, and we take it to be constrained by the metaphysics).

Lastly, and this continues to be a deep divide between what's most common in the Garden discussions and what's most common in at least some other venues: the "kinds of freedom" talk....do you really mean "kinds of freedom" or do mean "different accounts of freedom"? Separate and apparently complicated topic, and it's come up before; so it's fair enough if there's not much new to say about it here and now.

Lastly, "we still need to know what sort of ability to choose we are talking about, and that seems to get cashed out in terms of the MR/desert debates."
Some, including van Inwagen on some occasions, may well try to explicate "ability to choose" via some connection to the MR debates. But at first glance that looks like mistake, by their own lights even, since they all agree that sometimes the "ability to choose" is present but no moral issue is. (This issue can get more complicated, but I doubt it would since under pressure I doubt anyone wants their final explication of "ability to choose" to be in moral terms.

[isatimage icons]

I think, that you are not right. I can defend the position. Write to me in PM, we will communicate.

P.S. Please review our icons for Windows 8