The truth about libertarian agency

In the comments to the discussion Dan started below about when it makes sense to go revisionist about a concept, he said the following thing:

Value-relevant considerations will be things like providing for a distinctive kind of human dignity, grounding certain deep kinds of moral assessment, generating a perspective more likely to encourage moral development, etc. So my reason for not becoming a revisionist here and now is that I don't yet think that the truth relevant considerations in favor of revisionism outweigh the value-relevant considerations in favor of retaining the folk concept..

This gets at something I've been puzzled about in a number of pieces of libertarian-friendly writings. To put it bluntly, I don't see how the fact that I might want libertarian agency, or that it might bestow on my a certain dignity on me that I might otherwise have, would give me any reason for thinking the theory is true. I mean, look, if I weren't happily married I would want Salma Hayek to woo me. But if I were not married to Steph (my wonderful actual wife), I don't see how the additional dignity I would accrue (relative to my unmarried state) from Salma wooing me, or my satisfaction at having her woo me would in any way give me reason to think that she really was wooing me.

Since this post is long-ish, I've stuffed the rest of it "below the fold" as seems to be the parlance in the blogosphere. So click below to keep reading.

We can spend all day painting pictures of things that we would like to be true, but it seems to me that the first question we have to ask when considering a theory (or at any rate, the non-revisionist part of it, if it is revisionist) is whether it is true, or whether the balance of reasons suggests that it is true. And even for this latter question, I don't see how value considerations (in the sense that Dan was carving them up) should play any role at all.

Of course, it is open to someone to argue on pragmatic grounds that even though libertarianism ain't true, we're all better of believing that it is true. And, of course, this is the sort of view that our own Saul Smilansky has (Saul, correct me if I've glossed this incorrectly). So maybe folks like Dan should opt for that view over views like mine. But even so, it seems to me that if you think that libertarianism ain't true, and that we are better off not realizing this, then unless you are involved in the false-consciousness of a PR campaign, you ought not identify as a libertarian. Because you aren't.

(There is a whole 'nuther blog post to be done on what it is, precisely, that libertarianism is supposed to exclusively get. Discussions tend to get pretty hand-wavy about these things, though of course there are important exceptions. Anyway, I'm certainly not persuaded of the moral development line that Dan suggests, at least not yet, and I think that John Fischer's discussion in "Free Wil and Moral Responsibility" (go to the papers blog at http://gfp.typepad.com/online_papers/, and scroll to the bottom) is pretty compelling- though I'd argue the point about indispensibility along revisionist lines, instead, because I think Kane may be right that we want independence. But of course, I (counterfactually) want Salma to woo me.)

Even if I really, really wanted it to be true that Salma was wooing me, I don't have any truth-relevant considerations to favor the belief that she is, in fact, wooing me. So, I might have some practical considerations that favor quickly checking whether she is wooing me but once I realize that there are no truth-relevant considerations to recommend the view that she might be wooing me, I should give up looking for proof of her wooing pretty quickly (Nope- no letters in the mailbox, no phone messages - hey, wait a minute now, I just realized I have no independent reason whatsoever to think she has any contact with me at all- so why am I spending so much time looking?). Similarly, unless libertarians have some reason, independent of our obviously strong desire to be free! free! free!, why should they suppose Salma is wooing them, I mean that we have libertarian free will? I mean, even if I grew up in a family that had a tradition of supposing that Salma was wooing all the members of the family, once I realized there was no independent evidence for this, it doesn't seem that I have much reason to keep believing in it, even if (strictly speaking) Salma wooing me wasn't metaphysically incompatible with other things I believed. Analogously, even if I grew up in a culture in which agent causation (or whatever your favorite form of libertarianism is) was believed by all of its members, once I realized there was no independent evidence for it, I should think I don't have much reason for continuing to believe in it.

So, is there good independent reason for thinking libertarianism is true? Nearly everyone thinks that our introspective sense freedom isn't an especially good guide to the issues here. So we can't count on that. The conceptual link between freedom and responsibility is obviously controversial, and (speaking personally) I've always thought arguments that relied on the obviousness of responsibility simply seem to ignore the possibility that our being responsible was a mistaken judgement, or that if these judgements could not be mistaken (perhaps they aren't that sort of thing), that nothing about that judgement implies a commitment to libertarian metaphysics AND that that commitment alone is sufficient for the truth of libertarianism.

There is, of course, lots more to say about this. And yes, for anyone who likes this line, I argue for something like this in a forthcoming Midwest Studies paper, so don't go borrowing Salma without a big fat credit. And for anyone who thinks that this is painfully simple-minded, I swear to you that the paper is much more subtle than a blog entry- no, really! But really, libertarians, what am I missing?

Comments

[James Gibson]

I don't happen to be a libertarian, but from reading the literature and talking to enough of them, it seems pretty clear that their belief in LFW is not *merely* from their desire to have it. What, then, is that good independent reason for thinking LFW (of some variety) is true? Simply put: they take it to be an obvious fact about our experience, and to the degree that it is (or should be) our default position on the issue of freedom and moral responsibility. For example, after quoting Searle and Nagel at length Hasker says, "Now the question posed by the libertarian is: Why shouldn’t we take the descriptions offered by Nagel and Searle at face value, as accounts of the way the world really is?" (The Emergent Self, 83). Granted, Hasker doesn't think this establishes a decisive case for libertarianism, but he does say that in the absense of powerful arguments for determinism, this puts the ball in his court.

Hasker's line of thought is typical of a variety of contemporary contenders for libertarian freedom. That is, there appears to be a large consensus on behalf of them that the phenomenological experience of making free choices (whatever that amounts to) gives the libertarian *prima facie* evidence. But Manuel, you asked, "is there *good* independent reason for thinking libertarianism is true?" I've recently written a paper (currently in revision) where I argued this is *not* a good reason for being a libertarian over, say, a compatibilist. But nonetheless, it seems that the independent reason offered on their behalf is phenomenology.

[tnadelhoffer]

In a forthcoming paper (Journal of Consciousness Studies) entitled "The Phenomenology of Free Will," Eddy Nahmias, Stephen Morris, Jason Turner, and Thomas Nadelhoffer (me!), we a) discuss some of the relevant literature from psychology about the phenomenology of free will (there has unfortunately not been very much research), and b) present the preliminary results of our own attempt to get at the phenomenology of free will in an empirically informed way. We did not find that people's phenomenological experiences are as overtly libertarian as some libertarians assume. But even if libertarians best capture our experiences (which is an open question), that is no reason to assume that these experiences adequately represent reality--i.e. no reason to assume that libertarians claims about free will are true. Take, for instance, Velleman's suggestion that our phenomenological experiences of free will are much like our experiences of color. And just as it does not follow that the objects in the world really are colored just because we "see" them that way, it does not follow that we have libertarian free will even if we feel like we do. In this respect, I think Velleman's point hooks up to the one made by Vargas: Why assume that we have free will just because we want to have it? Similarly, why assume that we have it just because it feels like we do?

[James Gibson]

Thomas, all I think Velleman's point regarding the analogy of color proves is that having the feeling of freedom does not *entail* that we actually are free. I think what is really being questioned here is this: what degree of justification do our beliefs have from our phenomenological experiences, if any justification at all? So... if it appears to me that there is a cup on my desk, am I not justified in the absense of any defeaters in believing that there is a cup on my desk given the phenomenology? Likewise, if the phenomenology presents itself to me as if I have libertarian freedom, why wouldn't I be justified in believing this in the absense of defeaters (assuming the phenomenology *does* tip on the side of the libertarian)?

Btw. I saw a copy of your paper from Eddy. Nice job. Very interesting.

[Dan Speak]

This is good stuff. Manuel is getting chippy. The empiricists are out. Compatibilists are coming to the defense of libertarians. Very cool.

It would, of course, be a mistake to insist that what I've called "value-relevant" considerations in favor of a theory make it more likely to be true. That's why they aren't "truth-relevant" considerations. So, Manuel's trouble seems to come from assuming that the only considerations that should play a role in theory preference are of the truth-relevant variety. But that surely isn't right. Suppose Manuel and I have both received strong but not conclusive evidence that my brother (missing in action for some time now)was actually killed in combat. It seems reasonable that I withhold the belief that he is dead while Manuel does not. Why? Because Manuel and I are differentially related to the value of my brother. He means more to me than Manuel. So, though my brother's value to me does not make it more likely that he is alive, it does give me a reason to withhold my belief that he's dead.

There's lots more to say in response to Manuel's post, but I'll save that for later. I will add, however, that the problem might be that Salma REALLY IS wooing ME. I hope this doesn't muddy the water.

[Andrei A. Buckareff]

Regarding Dan's case of learning that his brother *might* be alive, it seems that my relation in this case does not matter one bit when we are assessing the epistemic rationality of your belief. Only evidence for the truth of the proposition you take a doxastic attitude towards counts when making judgments about the epistemic status of your belief. The same would be true with theories. Insofar as a theory purports to describe some phenomenon, whether or not I would like it to be true should not count in its favor from the standpoint of assessing whether the theory ought to be accepted or not. But the reasons for theory-acceptance are reasons that any similarly situated person in your position who evaluates the evidence ought to similarly accept. If she accepts another theory that is not as well-supported by the evidence, then she is irrational.

Now, you may have non-theoretical reasons for accepting a theory (just like in the case of the news about your brother). Take libertarian free will as an example. I may find that if I accept that I exercise libertarian free agency some or most of the time, I will feel as if my life has great purpose. Is this a good theoretical reason for accepting libertarianism? No. It may be a prudential reason for accepting it. But I will still be theoretically irrational if I believe it.

So if Dan accepts that his brother is alive (or at least may not be dead) and forms a belief the strength of which is not proportional to the evidence he has for the truth of the proposition believed, then, assuming that he has the exact evidence that Manuel has, his belief is more weakly justified than Manuel's. I fail to see how value-relevant considerations count in favor of Dan's belief except from a non-epistemic standpoint. Similarly non-epistemic considerations should not count in favor of accepting one theory over another, especially when our goal is to try to describe reality.

I fail to see what real cash value non-truth relevant considerations can have in constructing theories. If we allow for other considerations to govern theory construction and acceptance, then it seems our standards for ourselves as philosophers are lower than what we expect from scientists. Of course, we may bicker about what will count as relevant truth-considerations.

[Rico Vitz]

The discussion seems to conflate as many as four different sets of norms:

(1) the norms that govern attributions of knowledge,
(2) the norms that govern attributions of rational (or justified, or virtuous) belief, which might not amount to knowledge,
(3) the norms that govern the acceptance of a scientific theory, and
(4) the norms that govern the preference of a scientific theory.

Off hand, it seems to me that (1) is SOLELY concerned with “truth-relevant” considerations, and (2) is PRINCIPALLY concerned with “truth-relevant” concerns. But what about (3) and (4)? (3) seems to be largely, perhaps solely, concerned with “value-relevant” considerations such as parsimony, elegance, and predictive power. So, to assess Dan's comment (and Andrei's critique), we need to answer a pertinent question: Which set of norms governs the debate about free-will and moral responsibility?

[Andrei A. Buckareff]

Rico's comments are helpful, but unpersuasive. It seems that all of the relevant norms are in the service of the knowledge goal. (I could say something about (1) being SOLELY concerned with truth relevant considerations and (2) being PRICIPALLY concerned with such considerations, etc., but I'll avoid that for the time being. Although I do have some thoughts about that, but this isn't an epistemology blog.)

Regarding (3) and (4), parsimony, elegance, and predictive power still seem to be truth-relevant considerations. I should say that taken together they are truth-relevant considerations. Taken separately, they may not count for much. But, at any rate, the final goal of (2)-(4) is to acquire some knowledge of the world. So the norms governing (2)-(4) are all in the service of of the norms governing (1), viz., knowledge. (Many hold that the epistemic goal is truth--so it seems that insofar as the relevant norms in (2)-(4)are epistemic, they are in the service of (1).) If you think differently about these matters, you owe us a story about the goal of inquiry.

This seems relevant for the metaphysics of action and free will because (I hope) we are trying to describe reality. This puts our efforts on a continuum with scientists. I know that sounds platitudinous these days. But if the goal of scientific inquiry is to acquire knowledge of the world, then our goal is similar. This then puts any norms governing inquiry in the service of the truth/knowledge goal. Things should not be different when thinking about free will.

[Rico Vitz]

Andrei, I think it would be fun for you and I to get together to chat about belief and acceptance, but (as you rightly note) this is not the blog for that. So, let me return to Dan's point about theory preference.

I take his point to be about the legitimacy (rationality, etc.) of pursuing a research project in defense of a particular position -- in Dan's case, a research project in defense of libertarian freewill. As I understand Dan's reply to Manuel, there is a question about what counts as 'sufficient evidence' for (or against) the truth of a position, and what the participants in the debate count as 'sufficient evidence' for the truth of a position seems to be influenced by 'value-relevant' considerations. (Did I get that right, Dan?)

So, Dan's point, as I understand it, seems to suggest that the participants in the freewill debate, in Andrei's words, "owe us a story about the goal of inquiry" -- well, at least about the goal of inquiries concerning questions of freewill and moral responsibility.

[Dan Speak]

Rico has me right (if you know what I mean).

Is "Describing reality" the only goal of the philosophical enterprise? Maybe this is too big a question for this context, but I'm not sure what to think on this one.

Taken neat, reality would seem to be what IS. But philosophy is also interested in what could be, what couldn't be, what should be, what shouldn't be, what it would be rational to believe about each of these, etc. Now, I suppose it's possible to expand our concept of reality to include all of these interests. But then our pursuit of knowledge of reality will include a lot more stuff than we might have supposed when the continuum with science was invoked.

But I don't want this discussion to get too far off the interesting question Manuel initially raised about the role "what we want" can play in what we have reason to believe (in particular, regarding free will). If it wasn't 1:15am I'd have something really compelling to say about it right now.

[Hilary Bok]

Hi; nice blog. -- As I read your original post, it was about our reasons for adopting or abandoning a concept. But the concerns you raise here are chiefly about adopting or abandoning a belief. The two seem to me quite different. I agree that the fact that I want something to be true is normally no reason to think that it is true. The question whether or not some concept is correctly applied to a given object is a question about what is true of that object, and so in answering it we should be governed not by our desires, but by the evidence. But when we ask whether or not we should use (or: interest ourselves in) some concept, why isn't the fact that we want to use it, or that we find it of interest, a perfectly good reason to use it? If I want to define a predicate tat is correctly applied only to those objects that are either tin cans or George W. Bush, is there something I am getting wrong? Or some other reason to suppose that if I find this concept useful for some reason, I should not use it?

This line of thought does not allow us to define away the problem of freedom of the will, or any other problem. Suppose I want a concept of freedom that has two properties: (a) it allows me to argue that beings that are free in this sense have dignity, responsibility, etc., and (b) it allows me to show that normal human beings are free. It might be that there is no such concept, in which case my freedom to use whatever concepts I find useful will get me nowhere. But if we can find a concept that does whatever it is that we want a concept of freedom to do, and that is correctly applied to human beings, I don't see why we shouldn't use it.

[Rob]

The phenomenology of (what it pleases us to call) "free will" is one thing; the a priori case for its impossibility is another. The latter, it should need no saying, is entirely insensitive to the apparent deliverances of the former.

[Manuel]

Hi Hilary-

Thanks for joining in! The concept/belief issue is a tricky one. Assuming I've understood things correctly, I think the answer to the questions you raise turns on the nature of particular commitments generated by or associated with concepts such as freedom, responsibility, etc. The background picture we were assuming is that there are concepts of freedom and responsibility, and that they likely have libertarian commitments. For example, we might have a concept of dragon, and it may well be the case that the concept does some work for us (e.g., in stories), but the commitments of the concept are such that few of us would really want to be in the business of defending a theory that claimed that there really were dragons in any substantial sense. In particular, I take it that the commitments of being fire-breathing, flying, and capable of speech are especially problematic. With respect to the discussion Dan started, if we think that specifically libertarian commitments to concepts of responsibility and freedom are exceedingly difficult to square with a plausible picture of the world, the issue becomes the point at which we should opt for revision of our commitments (whether through elimination of the concepts that imply those commitments — eliminativism — or by changing our commitments, as I argue). That's where the norms for belief become relevant- i.e, whether we should believe or accept libertarian commitments.

I agree with you that there are good reasons for holding on to our concepts of freedom and responsibility. I even think we have good reasons to hold on to these things even if the particular commitments implicit in commonsense thinking about these concepts turn out to be, as Dan put it, screwed up. But that’s why I think of my view as moderately revisionist— it says commonsense thinking is screwed up enough that we should change it, but the concepts with those commitments are sufficiently functional, useful, and capable of playing justifiable roles in our conceptual and practical economies that we should retain them, once they are sheared of the problematic stuff. There are lots of things to debate about this- including whether we still have the same concept when we shear off the commitments that many people think of as central to the concept, and what accurately characterizes these concepts (I take it that this is one way to construe the compatibilism/incompatibilism debate) . I try to say more about some of these things in a forthcoming paper, “The Revisionists’s Guide to Responsibility.” If you are interested, it will soon be up on the papers blog.

[Manuel]

Many comments ago on this thread Dan said I was “getting chippy.” Well, that phrase ain’t much used in my linguistic community so I had to it up. Here’s what I found:

chip•py
—n.,
—pl. -pies.
1. Also,chip'pie. Slang.
a. a promiscuous woman.
b. a prostitute.
2. See chipping sparrow.

chip•py
—n.,
—pl. -pies.
a chipmunk.

chip•py
—adj., -pi•er, -pi•est.
1. Ice Hockey.using or characterized by aggressive, rough play or commission of fouls: a chippy player; a chippy second period.
2. Canadian.irritable; ill-tempered.

chippy
—n.,
—pl. -pies. Brit. Informal.
1. a carpenter.
2. a store selling fish and chips.

First, I was beloved but now I know what Dan really thinks: I’m either a prostitute, ill-tempered, or a carpenter! But I know I’m not a carpenter . . . .

[Hilary Bok]

I guess I just think that we should figure out why exactly having freedom matters to us, and then try to see whether there is any account of freedom that does those things that matter, and that in addition makes it the case that we are in fact free. Whether it's the normal concept of freedom or not seems less important to me than whether it gives us what we want. As before, the question whether we are free in whatever sense we're investigating would presumably be determined on the basis of evidence etc., but the question which concept of freedom to use would not be. Or so it seems to me.

[Rob]

Why should it be expected that any supposed alternative concept of freedom will be any less vulnerable to the a priori case against ultimate moral responsibility than is our ordinary, unreflective experience of ourselves as agents? Isn't any such alternative at bottom merely a deferment of the corrosive course of thought inexorably demanded by the a priori case?

For a quick and easy rehearsal of the case, see this link: http://www.rep.routledge.com/article/V014SECT3

[John Fischer]

Speaking of the truth about libertarian agency, I will be teaching a graduate seminar on Randolph Clarke's recent OUP book this fall at UCR! All "Gardeners" are welcome! Additionally, Randy will visit the seminar to discuss the issues.