Pereboom on the Weirdness of Compatibilism

I want to write a post about an interesting point Pereboom makes about compatibilism.  It is a point I’ve tried to make, however inelegantly, several times before.  But Pereboom is an eloquent writer and does a better job than I could.  I first heard him give this argument at the symposium on Fischer’s work at Inland 2006.  Now he has published an article with largely the same arguments in Philosophical Books.  Here’s Pereboom:

“While this ‘legitimately calling to moral account’ notion may be a bona fide sense of moral responsibility, it is not the one at issue in the free will debate. For incompatibilists would not find our being morally responsible in this sense to be even prima facie incompatible with determinism. The notion that incompatibilists do claim to be incompatible with determinism is rather the one defined in terms of basic desert.”

Pereboom is writing in terms of desert and moral responsibility.  But, I think, the point also applies, and with more force, to talk about “free will.”  In short, Pereboom is pointing out the remarkable, but obvious, fact that compatibilists are committed to the following claim

WEIRD: The power or faculty of “free will” is something nobody would deny most humans have most of the time.

Now, suppose you were learning about the free will debate for the first time—as all of us must have at some point.  You learn about compatibilist conceptions of free will and you learn about incompatibilist conceptions.  Perhaps you’re sufficiently unprejudiced and trying to decide which one actually captures what “free will” means.  If you are like me—and, I imagine, most compatibilists—the weirdness of compatibilism counts against it.  Personally, I thought something like “well, the compatibilist thinks free will is something that most people have, most of the time, and that just can’t be right.”  But the question is: when I think “that just can’t be right”, am I correct?

Some points:

1. Compatibilism isn’t committed to WEIRD per se.  For example, I can imagine that a compatibilist is genuinely committed to requiring powers the existence of which is controversial.  The neurotic compatibilism of Nahmias is something like this (although, even then, the threats that concern Nahmias don’t seem so troubling, and he seems to share my optimism about this).  Similarly, a compatibilist might require “free will” to involve something that no human being has—perhaps something like an IQ of 2000.  But, in practice, most compatibilists are committed to this weird claim.  Whether we are discussing Fischer’s ownership and reasons-responsiveness, Frankfurt’s identification and hierarchy of desires, Watson’s concern with values, Dennett’s list of evolved faculties, these are all powers that nobody would deny most people have most of the time.

2. To appreciate this point, consider this analogy: people are arguing about whether ZERODOME exists (you can substitute the non-sensical word of your choice).  The existence of ZERODOME is very contentious.  People have been arguing for millennia.  People have been put to death because of the belief in ZERODOME; others have been spared the death penalty because people doubted whether ZERODOME exists.  One group says: ZERODOME is the Loch Ness monster.  This makes sense, because people have argued about whether the Loch Ness monster exists.  Sure, only one of the Loch Ness believers, or the Loch Ness doubters, can be right at the end of the day.  But at least they were arguing.  Another group, however, says ZERODOME is the city of London.  But this is strange.  It was extremely contentious that ZERODOME exists, yet it is not extremely contentious that London exists.  In fact, it’s not contentious at all.  So it would be very strange if ZERODOME and London were the same thing.  That analysis doesn’t seem to capture the controversial element of whatever ZERODOME was supposed to be.  Yet this is precisely what the London group says: ZERODOME is something that nobody would ever doubt exists.  So a newcomer, trying to decide whether ZERODOME means the Loch Ness monster or London, might feel inclined towards believing it means the Loch Ness monster.

3. It is important to note that you can’t say this about the other analyses of free will in the literature.  Consider first the libertarian conception of free will.  Unlike ownership and reasons-responsiveness, and unlike having five fingers on each hand, the question of whether we are unmoved movers (e.g agent causes) or whether spooky physics in our brains gives us free will is controversial.  Some people actually believe we have these powers and some people strongly doubt that we do.  The libertarian analysis of free will is not weird, in the same way that the compatibilist analysis is, because it is consistent with “free will” being controversial.  One can even make this point, with only slightly less force, about logically impossible conceptions of free will, because even their proponents, like Galen Strawson and Thomas Nagel, insist that—however impossible this kind of free will is—belief in it is incredibly hard to shake.  Fischer, Frankfurt, Watson, and Dennett say no such thing about their kind of free will—they never doubted, in their wildest dreams, that most humans, most of the time, have these weaker powers.  So, again, the other parties of the debate analyze free will in a way that is consistent with people doubting that free will exists.  Compatibilists do not—they make the weird claim that free will is something nobody would doubt that most people have most of the time.

I am not 100% sure that this argument, so formulated, is right.  Is it not possible, for example, that people might argue about whether something (“free will”) exists without knowing, and only later discovering, that it is identical to something that everyone agrees exists?  But I suspect that a better formulation would work.

The compatibilist can say: “Imagine that MYSTERYKEY is a key that opens the mystery door to the tomb.  People argue about whether MYSTERYKEY exists—it is unclear whether any key in existence will open that door.  But everyone agrees that KNOWNKEY exists.  KNOWNKEY is a big gold key that the mayor owns.  The mayor has never tried to open the mystery door with KNOWNKEY.  One day, he tried to open the door with KNOWNKEY and it opens.  People discover that MYSTERYKEY and KNOWNKEY are the same.  So, even though people doubted that MYSTERYKEY existed and agreed that KNOWNKEY existed, MYSTERYKEY and KNOWNKEY can still be the same.”

I suspect that a dogged incompatibilist can resist these sorts of counter-examples.  Pereboom’s point about the weirdness of compatibilism has a strong pull on my intuitions.  But much more needs to be said.  Perhaps the one who needs to say the most is the compatibilist.  Because compatibilism is weird, in a way that incompatibilism is not (as described above), it seems that compatibilists owe us an explanation for this weirdness, like the explanation the mayor can give for how MYSTERYKEY and KNOWNKEY are identical.  Although there is no shortage of compatibilist writings for me to look over, it’s not clear to me that compatibilists have provided such an explanation (and if they have, please feel free to remind me).  They seem to still owe us an explanation for how the existence of free will ended up not being contentious at all.

Comments

[Mark]

Kip,

Given that many (if not all) prominent compatibilists in the history of the debate have approached the subject from an ostensive perspective, why is WEIRD really that weird?

[Kip Werking]

Mark,

I'm not sure how you get from "approached the subject from an ostensive perspective" to "WEIRD isn't that weird"? Would you care to elaborate?

[Justin Coates]

Kip writes:

WEIRD: The power or faculty of “free will” is something nobody would deny most humans have most of the time.

I take it when you say "nobody" you mean "nobody" and not "no philosopher working on free will." I hope you mean the former because if you mean the latter, I'm not sure how that's supposed to be a problem for the compatibilist. After all, the fact that I (or other compatibilists) find it weird that some incompatibilists believe in agent causation and that others believe there's no such thing as moral responsibility isn't particularly telling against their position.

But if you meant the former, then there is empirical research that strongly suggests that almost all of the folk believe WEIRD to be true. Or, if they didn't believe WEIRD, almost all the folk do believe that normal adult humans make many decisions of their own free will. If this data is right (and I don't know any other way to interpret it), I'm not sure why being committed to WEIRD is a problem at all for the compatibilist. If anything, not embracing WEIRD would make your position much WEIRDer.

[Roman Altshuler]

Can't we account for the apparent WEIRDness by distinguishing between different notions of freedom? For example, take Frankfurt's distinction between (1) having a free will and (2) acting of one's free will. It looks like (2) is extremely common--it occurs in cases where individuals act without experiencing a conflict. But (1) is somewhat strange and mysterious and might, in fact, never occur. So nobody would doubt that most people are free most of the time, in the sense of (2). At the same time, many people can be very skeptical about freedom in the sense of (1). On this view, the existence of free will is at the same time largely agreed upon and contentious, depending on whether one is speaking about (1) or (2).

[john a]

Roman
I do not see how one can believe in 2 without believing in 1. 1 = I have x, 2 = I act from x, therefore x is present in both cases. How can one act 'of' x without x? I guess I do not understand exactly what the distinction is supposed to accomplish.

For all:

In another comment (on this blog or elsewhere, I cannot remember) I asked why it is the case that free-will appears to be an all or nothing situation where one believes that there is such a thing as free-will or does not believe that there is such a thing as free-will? My question is why can't free-will be a matter of degree so that some might have more of it then others, or none at all? I am thinking of free-will as a trait, like having hair. Some of us have a lot of hair, some less, and some not al all. Sometimes we do talk about 'losing free-will' such that we are referencing a fact that what someone once had, she now has less of, or not at all. This might apply to #2 in Roman's comment. No one ever answered my original comment, maybe someone here will shed some light on my question.

[Kip Werking]

Here are some responses:

John:

The problem with thinking of free will in terms of degrees is that most people, either historically, or even currently, do not seem to think of free will this way. But common usage seems to govern, at least in part, the meaning of terms. Most people do not seem too comfortable saying “he had more or less free will”, “this diminished his free will”, and so on. It seems much more natural to think of free will as something like a soul: you either have it or you don’t (perhaps it is a gift from God).

But I don’t think this discomfort is absolute. Empirical studies could try to measure just how comfortable people are with using “free will” in certain ways.

Roman:

I’m very sympathetic to the sort of move you make. I do think people sometimes use the phrase “of one’s own free will” in a legal or conventional sense that is obviously not intended to raise any philosophical or metaphysical issues. It just means something like “a normal, rational human being without any uncontroversial legal excusing condition.”

Justin:

When I wrote “nobody”, I meant “virtually no reasonable human being.”

You raise an interesting point about the weirdness of incompatibilism. Here is a possible distinction: when the compatibilist says “your theory of agent causation of no-free-will doesn’t fit with my intuitions”, the agent causal theorist and no-free-will theoriest don’t share your intuitions. It’s weird, but only weird through the eyes of the compatibilist—or so I would argue.

The weirdness of compatibilism isn’t like this. The controversial nature of “free will” is something every party agrees on: agent causal theorist, no-free-will theorist—even the compatibilist. The agent causal theoriest doesn’t know that it’s weird to think of free will as being an agent cause—s/he thinks it’s perfectly natural. But even the compatibilist knows that, whatever free will was, people have been arguing about whether it exists for millennia.

Here’s another try: when I say “compatibilism is weird”, I mean “compatibilism seems to be obviously wrong, but I have difficulty formulating the argument.” I agree that weirdness, per se, shouldn’t invalidate a philosophical theory. I believe some pretty weird things. But my comments here are meant to inspire an argument about wrongness, and not just weirdness. Whether incompatibilists will formulate such an argument is another question (and will involve debates about examples like MYSTERYKEY).

Also, here I am not defending incompatibilism. It might be, as you suggest, that both compatibilism and incompatibilism are weird (perhaps in very different ways). Fair enough. Then we might have to choose between two weird theories and we would need further arguments to settle the matter.

Finally, you insist that virtually all folk believe WEIRD is true. Here are some points in response:

1. So? The folk believe lots of ridiculous things.
2. I doubt your empirical claim. You seem to be basing the claim “virtually all people believe in WEIRD” on the claim “virtually all people believe in free will.” But these are distinct questions. I think, even if most people believe in free will, many or most recognize that “free will” is something the existence of which is at least somewhat controversial. For example, the claim “most people believe in free will” can be divided into “the compatibilist portion of free willists believe that nobody ever doubted that we have compatibilist powers” and “the incompatibilist portion of free willists believe that nobody ever doubted that we have incompatibilist powers.” But, at least in the latter case, this isn’t obvious at all. A non-specialist, who thinks people are unmoved movers, might think “I happen to think free will involves being an unmoved mover but I could see how people would doubt that we have this extraordinary power.”
3. Even if your claim is largely true, in a quantitative sense, it is not true, in a qualitative sense. In another thread, I speculated that 3% of the folk would doubt the existence of free will. This number might be much higher in communities with large Lutheran or Calvinist populations. But this minority, however small, is extremely telling. If 3% of the population doubts the existence of free will, then free will is still very different than London—because *nobody* doubts the existence of London. It is this distinction, and this remaining controversy, on which the incompatibilist can rest hir weirdness argument.

[Manuel Vargas]

Hi Kip- interesting issue, though I'm not yet convinced that any "weirdness" is all that weird. Consider the weirdness of similar debates about personhood or morality. It doesn't seem (to me, anyway) dialectically salient that on one side of each of these issues there is a group of people who think personhood and morality consist of things that no in their right mind has ever doubted, and that there is another group of people who insist that these things also require something more, something that a number of reasonable people have doubted.

Indeed, I am inclined to think free will is hardly the only domain in which we are prone to strong intuitive convictions about what is needed for something that grossly exaggerates what is necessary for its existence or justification. Besides (perhaps) the mind, morality, personhood, and free will, perhaps love and marriage are also in that category. Not that anything I've said here should lead you to agree with me about the particular examples; I'm just making a point that the "weirdness" of this situation in the free will debates doesn't seem all that different from lots of other situations, whatever you think about how those other cases should sort out.

[Roman Altshuler]

John,

Let me rephrase my initial suggestion. As I understand Frankfurt's position, the distinction is something like this:
(1) having free will: I can actively choose which of my first-order desires will be effective, based on the evaluations of my second-order volitions. This ability, I think, is necessarily going to be a massive problem for compatibilism.
(2) acting of my free will: when I act, the first-order desire I act on happens to be the desire that is endorsed by my second-order volition.
Frankfurt wants to show that if (2) obtains, then my action was, for all intents and purposes, done from my free will, since the only reason I could need any extra kind of freedom would be to actually change my first-order desire; but if I already endorse that desire, then I would not need to change it, and so I would not need any extra freedom, like (1): my action is already as free as it could be.

I think that the account Frankfurt actually gives doesn't quite work, and suffers from serious internal inconsistencies. But there are other approaches in a similar vein (like John Fischer's distinction between regulative control and guidance control) that maybe work better: the idea is that we can come up with the conditions needed for responsibility (that the agent is satisfied with her action and agrees that she is responsible for it) and provide a reciprocal kind of freedom to go with it. In this way, we can have the ability to act of our free will while retaining serious doubts about any more ambitious and metaphysically troublesome libertarian kind of freedom.

Kip,

With this in mind, I am still very confused about why WEIRD is supposed to be more of a problem for the compatibilist than for the libertarian. The compatibilist seems to me to have a ready explanation for why "the folk" believe that we have (libertarian)free will: We do in fact have (2). Most people recognize this but, since they are not metaphysicians, they naturally extend this to think we also have (1). (No condescension is implied: like Strawson, one could say that it's actually the metaphysician, not the folk, who gets things backwards here.)

But the libertarian has the problem of explaining why, if we really do have (1), the issue might be controversial. So while WEIRD itself (that most people believe we have free will) is not a problem for him, the fact that there is any controversy about free will is a little weird; at least it seems to me to be as weird as the predicament you attribute to the compatibilist.

[Mark]

Kip,

Approaching the problem ostensively involves looking at paradigmatic cases and drawing conclusions from those cases. Given this, compatibilists tend to prefer cases that the maximal number of rational persons would identify with. So it seems that WIERD should be the goal, which makes me wonder whether it should be relabed as OBVIOUS.

[john a]

Kip
Thank you for your response. The reason I have a tendency (I am not sure what I actually believe about this matter) to think of free-will in degrees is because of people who suffer certain debilitating mental diseases and the fact that the law recognizes deminished capacity as an excusing condition. Often times I do think we talk about these people as not having the ability to make choices freely. If by having free-will we have as a condition that there must be relevant alternative options available to choose from then it seems under these conditions their choices are compromised and not free. Furthermore, if it is simply their 'minds' that are adversely affected, I do not see the relevance of refering to free-will in decision making. We could simply reference a person's reasoning ability, decision-making competence, etc, without the need for the concept of free-will.

I also like your idea of doing an experiment.

[john a]

Roman
Thank you for the clarification. It has been years since I read Frankfurt (or much on free-will for that matter) and obviously I need to do so again. Reading the threads on this blog as well as the work being done in experimental philosophy has rekindled my interest.

[Richard]

Kip,

There seems to be a simple explanation, on behalf of the compatibilist, for why there is dispute over the existence of "free will" (de dicto) even though no-one disputes the existence of reasons-responsiveness. It is because not everyone recognizes that "free will" is reasons-responsiveness.

Am I missing something?

[Kevin Moore]

Can the "weirdness" of compatibilism count as evidence against the "queerness" of objective moral values(Mackie)or vice versa?

[john a]

WEIRD: The power or faculty of “free will” is something nobody would deny most humans have most of the time.

This probably has absolutely no philosophical merit, but I was thinking about a consequence of accepting this claim. A person can 'deny' x without having any clear reason to do so (stubbornness). Because of the use of 'nobody' only one person is needed to deny that WEIRD (therefore compatibalism?) is true to make it false. So I will deny that it is true thereby making it false.

[Jonathan Farrell]

I’ve three points to make here. The first concerns terminology: we seem to be talking about soft determinists (and libertarians) not compatibilists (and incompatibilists). To be a compatibilist, I take it, is only to claim that it doesn’t follow from the fact that a universe is deterministic that there can be no free will in that universe. It is not to make any claims about whether we have free will. And likewise, (with the appropriate changes) for incompatibilism.

My second and third points are, unfortunately, much longer. In the second, I’ll explain why I don’t think that soft determinism is any weirder than libertarianism. And in the third, I’ll try to explain why the ‘weirdness’ of soft determinism (in the sense that Kip outlines above) isn’t problematic (or at least that if it is, it’s not for the reasons that Kip suggests).

Okay, my second point: WEIRD can be read in a couple of ways (as people have implied or pointed out already). One of the ways to disambiguate WEIRD is shorter than the other, so I’ll call them SWEIRD and LWEIRD respectively.

SWEIRD: nobody disagrees that most humans have free will most of the time.

As it stands, this is controversial (there are free will deniers), but it seems pretty clear to me that if we change ‘nobody’ to ‘almost nobody’, then this claim is true. So everybody should be committed to (the slightly weakened version of) SWEIRD.

WEIRD is controversial, however, when we disambiguate it the second way:

LWEIRD: The power or faculty of “free will” is best understood as requiring ownership and reasons-responsiveness (or identification and a particular hierarchy of desires, etc.), and nobody would deny that most humans have these powers or faculties (i.e. that most humans have ownership and are reasons-responsive, etc.) most of the time.

Soft determinists are committed to LWEIRD (with the relevant analysis of free will slotted in). But LWEIRD is only controversial because its first conjunct is controversial.

Kip suggested (in 3.) that libertarians are not committed to anything that is both analogous to WEIRD and controversial, but I don’t think that’s so. As I suggested above, libertarians are as committed to SWEIRD as anybody else. But SWEIRD is not controversial. The libertarian analogue of LWEIRD is:

LWEIRD*: The power or faculty of “free will” is best understood as requiring us to be unmoved movers (or for there to be spooky stuff going on in our brains, etc.), and nobody would deny that most humans have these powers or faculties (i.e. are unmoved movers, etc.) most of the time.

LWEIRD* is controversial because the second conjunct is controversial. (Some might say that the first one is too, but even if they are mistaken, I take it that because the second conjunct is controversial, so is LWEIRD*.)

So I think that the situation is (roughly – one claim might be more controversial than the other) the same for soft determinists and for libertarians.

I could (I think) have said all of the above in terms of the de re/de dicto distinction that Richard mentions (and probably in far fewer words!). But I’m not confident I’d have got it right (for some reason I tend to see stars whenever someone mentions this distinction), so I didn’t.

Third point: that soft determinists are committed to LWEIRD is not a problem.

The argument put forward that the soft determinist’s commitment to LWEIRD is problematic seems to go like this:

(1) We don’t doubt (= it is not controversial) that we are reasons-responsive.
(2) We (some of us) do doubt (= it is controversial) that we have free will.
Therefore,
(3) Being reasons-responsive is not having free will.

Obviously, you can slot in your favourite compatibilist analysis of free will in the place of ‘reasons-responsive’ if you want.

This argument commits the masked man fallacy (and it looks as if this is the problem of the ZERODOME example, too). What we know (or doubt) about something depends on how it is described. I suspect that we all know whether Kirk Douglas was in Spartacus. I also suspect that at least some of us don’t know whether Issur Demsky was in Spartacus. But we would be making a mistake to conclude from this that Kirk Douglas is not Issur Demsky, because, in fact, they are the same person. Likewise, just because we know that we are reasons-responsive (or whatever) and we don’t know whether we have free will, we can’t conclude that having free will just is being reasons-responsive.

In other words, (1) and (2) aren’t relevant to whether or not (3) is true, so we have no reason to think that the weirdness of soft determinism is a problem for this position.

[Joe]

Richard has it right.

Everyone agrees that we have the kinds of powers and abilities that the compatibilist thinks that we have. The disagreement is over whether those powers and abilities constitute free will (the freedom relevant condition necessary for desert entailing moral responsibility). The compatibilist says that they do and the incompatibilist says that they don't. These are substantive claims.

[john a]

"I suspect that we all know whether Kirk Douglas was in Spartacus. I also suspect that at least some of us don’t know whether Issur Demsky was in Spartacus. But we would be making a mistake to conclude from this that Kirk Douglas is not Issur Demsky, because, in fact, they are the same person. Likewise, just because we know that we are reasons-responsive (or whatever) and we don’t know whether we have free will, we can’t conclude that having free will just is being reasons-responsive."

Jonathan
I am not sure that the last sentence is consistent with your example of KD = ID.

From "We know that we are r-r, but we do not know whether we have FW" we cannot conclude anything regarding r-r = FW. But if your example of KD = ID is the point you are making then we should be able to conclude that r-r = FW.

Or am I missing something?

Furthermore, I think you move to SWEIRD is not sound because SWEIRD is subject to the same problem as WEIRD. What is the range of 'almost nobody? such that if there is one person more beyond the set of 'almost'nobody' that thinks SWEIRD false then SWEIRD is false. By extending the set from an empty set (nobody which is problematic) to ? you have created the 'loyal opposition' and you then need to account for the epistemic status of their beliefs. After all, if Kip is correct and FW is something like the soul then there are ontological issues involved. One side of the dispute must be wrong even if we do not know which one.

There is also an issue underlying LWEIRD; does 'ownership and reason-responsiveness' extend to fetuses, children and to those that have mental disabilites such as Alzheimer's. They are, afterall, human beings. LWEIRD contains an empirical claim that needs to be tested. If there are human beings where applying the concepts of ownership and r-r does not pertain and this group is larger then that group to which they do pertain, then LWEIRD would be wrong. For LWEIRD to be acceptable then the empirical claim must be verified.

[Kip Werking]

Let me respond to some of the excellent comments here:

Manuel:

I agree about the similarity between “free will” and the other things you mention. In the Non-Reality of Free Will, Richard Double concludes that “free will” is ill defined but comes to this conclusion grudgingly—he does not seem to think other philosophical terms (like truth or duty) suffer from the same problems. But I suspect these other terms have all or most of the same semantic problems that “free will” does.

In fact, I favor a “semantic ambiguity plus cognitive biases” explanation for most philosophical problems. Either the terms are vague and so people argue over their meanings or certain biases dispose people to resist truths about terms (even if the terms are sufficiently well defined for there to be truths about them). Most philosophy problems, it seems to me, involve some mix of these two (and not always in the same proportion: some words are just vaguer than others).

The difference between you and me, perhaps, is that you seem more inclined to salvage such terms despite these problems. In contrast, I have a much stronger deflationary attitude.

I will split hairs, however, and say that the existence of free will is controversial in a way that mind or personhood is not. However small the percentage of people is who say “free will doesn’t exist”, it’s significantly bigger than the percentage who say “mind doesn’t exist” or “persons don’t exist.”

Roman:

This is only a problem for compatibilists who are not content to say “people act of their own free will”, but insist on also saying “people have free will.” You might be willing to make this concession, and so perhaps my (or Pereboom’s) argument is no problem for you. But I suspect most compatibilists will not follow you down that road.

Richard:

You wrote:

“It is because not everyone recognizes that "free will" is reasons-responsiveness.

Am I missing something?”

What is missing is this: an explanation for how “free will” is reasons-responsiveness (or the compatibilist conditions of your choice). Consider the MYSTERYKEY example. If the mayor says “MYSTERYKEY and KNOWNKEY are the same”, people can respond “sure, you say that, but how can you prove it.” At that point, the mayor can take KNOWNKEY, put it in the mystery door, and open the door--*proving* that MYSTERYKEY AND KNOWNKEY are the same.

Now consider the compatibilism debate: the compatibilist says “free will and reasons-responsiveness (or whatever) are the same.” Then, just as the people challenged the mayor to prove that MYSTERYKEY and KNOWNKEY are the same, incompatibilists such as myself can say “ok, compatibilist, if you think they are the same, *prove* it. The mayor can put MYSTERYKEY in the door and open it. What can you do?” And it seems to me that the compatibilist can do nothing, if only because MYSTERYKEY was well defined as any key which opens the mystery door, but “free will” is not well defined—there is no obvious task the satisfaction of which would prove that something is free will.

John:

By “nobody”, I mean something like “most reasonable people acting in good faith.” Sorry.

Jonathan,

I sympathize with your first point, but the term compatibilist, at least as it is used in these circles, tends to mean not just “free will and determinism are compatible” but also asserts that most people have free will most of the time. At least, whenever someone calls themselves a compatibilist, it is safe to assume that they make this additional claim—if only because I know of no compatibilist who denies the existence of free will (maybe Nahmias in one of his more neurotic moods).

I don’t think “soft determinist” wouldn’t help matters because, if I am not mistaken, if it differs from “compatibilist”, it is in asserting the truth of determinism, and not in asserting the general existence of free will. But I don’t know of any compatibilists who assert determinism.

I agree this terminology is sloppy. But it’s the way it is.

About your second point, there is one fatal problem for your argument: libertarians are not committed to LWEIRD*. Libertarians are not committed to LWEIRD* because they recognize that the existence of these powers (unmoved mover, spooky physics, etc.) is controversial. Compatibilists cannot make a similar claim. They are committed to WEIRD/LWEIRD.

About your third point, the distinction seems to be this: you have a story to tell about why Kirk Douglas and Issur Demsky are the same. Similarly, in the MYSTERYKEY example, the mayor has a story to tell about why MYSTERYKEY and KNOWNKEY are the same. But what story does the compatibilist have to tell about how “free will” and “ownership and reasons-responsiveness” (or whatever) are the same? There seems to be a gaping hole here, and it does not seem too demanding for the incompatibilist to ask the compatibilist to explain hirself.

For more on this third issue, I would point you to my response to Richard above.

[Jonathan Farrell]

John A,

Thanks for your questions.

First, the point of the KD = ID example is to show that, just because we know something about KD, but don’t know something about ID (or vice versa), we can’t assume that this means that KD is not identical to ID. But nor does it mean that we can assume that KD is identical to ID. We need more information to do this. We know KD was in Spartacus. We might not know whether Tony Curtis was in Spartacus (I’m sure he was in The Vikings, but Spartacus...?). But this tells us nothing about whether or not TC = KD.

Regarding SWEIRD, it doesn’t matter how many people think that SWEIRD is true (of false), all that matters is that it is. And I think we have good reason to think that it is. (I think this responds to what you said, but I’m not sure – does it?)

I think you are right that the claim that SWEIRD makes is a vague one, but I don’t see why this is a problem. And I don’t see why the fact that there are ontological issues involved is relevant (though I accept that it’s true). SWEIRD makes no (interesting) claims about ontology, it only makes claims about what people think.

I also think you are right that it is an empirical matter whether most humans are reasons-responsive (or whatever), but, again, I don’t see why this is an issue. LWEIRD doesn’t say: most humans are r-r. It says: most people think that most humans are r-r. This is another empirical claim, but it’s not about whether we are r-r.

Kip,

First, a minor point regarding your response to Manuel: at least some forms of Buddhism deny that there are selves. I’m not sure whether all forms do, and whether we can take selves to be the same as persons, and nor am I clear what the consequence of this is for belief in free will, but given that there are a lot of Buddhists around there might be more people who deny the existence of persons than who deny the existence of free will.

Terminology: you are absolutely right to say that soft determinism entails holding that our universe (or at least the bits of it that are relevant to free will) is deterministic – I got this wrong. (My only excuse is that I was still seeing stars after having to think about the de dicto/de re distinction!)

But I’m not sure that I agree with you about the use of the term ‘compatibilist.’ I take compatibilists to be saying something like this. “It seems that we have free will. But it also looks as if (a) we might seem to be free in a deterministic universe and (b) there are good reasons to think that we can’t be free in a deterministic universe. On closer inspection, however, it turns out that can have free will in such a universe. Does this mean we have free will? No, but it does mean that the possibility that our universe might be deterministic gives us no reason to think that our seeming to be free is not a consequence of us being free. Of course there may be other reasons why our seeming to be free is not good evidence for our being free, but that’s neither here nor there as far as compatibilism goes.”

Maybe ‘compatibilism’ is sometimes used to refer to those who think FW and determinism are compatible theses, and also that we have FW, but I’m not sure that this is a good thing (and so I’m not sure that we should continue doing it). One reason why you (and I) may not know of any compatibilists who deny that we have FW would be because, under the stricter understanding of the term, whether or not we have free will is irrelevant to whether or not we should be compatibilists. (A silly example: I don’t know of any compatibilists who don’t like chocolate. But this is because not liking chocolate is not relevant to whether or not we should be compatibilists. My lack of knowledge shouldn’t lead me to conclude that to be a compatibilist is to have any particular opinion about chocolate.)

(Having said all of this, for the rest of this post, I’ll use compatibilism to mean ‘compatibilism plus free will.’)

LWEIRD* etc.: Are libertarians committed to LWEIRD*? You suggest not, and you may be right. The reason I discussed it was that it seems to be to be the closest analogue of LWEIRD. I was trying to show only that compatibilism is no weirder than libertarianism, but perhaps I was trying to be too charitable to the latter.

If you’re right to say that libertarians are not committed to LWEIRD*, then they certainly seem to be committed to:

LWEIRD**: The power or faculty of “free will” is best understood as requiring us to be unmoved movers (or for there to be spooky stuff going on in our brains, etc.), but some would deny that most humans have these powers or faculties (i.e. are unmoved movers, etc.) most of the time.

LWEIRD** is controversial because the first conjunct is controversial.

Further, as I suggested in my first post, libertarians are (or should be) committed to:

SWEIRD: (almost) nobody disagrees that most humans have free will most of the time.

But by holding both SWEIRD and LWEIRD**, I think we could argue that libertarianism is weirder than compatibilism. Both camps accept that most people think we have free will, and both provide a controversial analysis of free will. So far, so (a bit) weird. But then libertarians have to go further and admit that most people do not accept that humans have what (they say) is required for free will. I actually don’t see this as being particularly problematic, but it certainly seems that the libertarian is no better off than the compatibilist here.

You respond to my third point by asking ‘what story does the compatibilist have to tell about how “free will” and “ownership and reasons-responsiveness” (or whatever) are the same?’

First, can’t we ask exactly the same question about the libertarian: ‘what story do they have to tell about how ‘free will’ and ‘being an unmoved mover’ are the same?’ (As you say in your response to Richard, ‘there is no obvious task the satisfaction of which would prove that something is free will.’ If you are right, then clearly this is not just a problem for compatibilists.)

Presumably both will say something like this. “Look at the way we use the term ‘free will’ (and similar terms). Doesn’t it seem that this (reasons-responsiveness, being an unmoved mover, etc.) captures the important aspects of the kinds of things we’re pointing at when we use terms like ‘free will’?” And then we have to decide whether the proposed analysis does capture these important features.

What makes this difficult is that, although (as you note in your response to Richard) MYSTERYKEY is well-defined, ‘free will’ is not. This allows compatibilist analyses of the notion to be criticised by pointing out what the analysis does not include, but which free will seems to (according to the critic). And it allows incompatibilist analyses to be criticised for including what does not need to be included (according to the critic).

[Joe]

I haven't been following the details of this thread but it all seems like a non-issue.

Both the compatibilist and the incompatibilist are relatively good at coming up with necessary conditions for free will. Where they run bad is in getting from the necessary conditions to a set of sufficient ones.

In response to Richard's question, Kip writes: "What is missing is this: an explanation for how 'free will' is reasons-responsiveness (or the compatibilist conditions of your choice)."

But there is NO EXPLANATION of how indeterminism, or agent causation, or ultimacy fits into the full picture of free will. At most there is an argument that one of these conditions is necessary for free will. That is precisely what the compatibilist has. And the story about how we get from the necessary condition to a full-blooded understanding of free will is still missing from the incompatibilist/denialist theory. (This is precisely what we should learn from the problem of luck.)

Until the incompatibilist can meet Kip's challenges, I don't see why the compatibilist should worry. This is a non-issue dressed up like an issue.

[Kip Werking]

Joe and Jonathan,

Joe writes:

“Until the incompatibilist can meet Kip's challenges, I don't see why the compatibilist should worry. This is a non-issue dressed up like an issue.”

You shouldn’t worry if you think the only threat to compatibilism is the success of incompatibilism. But this might not be the only threat. It’s not very convincing to say “I can do no better than the opposing view, which I find invalid and unsupported.” Surely, you want to do better than that (presuming you have compatibilist sympathies). It might be the case, for example, that free will is poorly defined, and so neither compatibilism nor incompatibilism is well supported and ultimately successful.

Jonathan,

My thoughts on your points about “compatibilism” and terminology remain the same as in my last comment.

Your argument that libertarianism is weird (which it surely is in one obvious sense) made me think. The libertarian can say “it would be weird if I thought everyone shared my conception of free will, everyone believed in free will, and yet not everyone believed in being an unmoved mover or spooky physics, etc. But the disagreement might not be empirical. It might be semantic. It might be that many people think free will means what libertarians means, and many people think free will means what compatibilists mean, and in this case, it is definitely not weird that not everyone agrees that people are unmoved movers or have spooky physics in their brains.”

I think that argument would work. The problem is that the compatibilist can say the same thing to Pereboom: “It would be weird if everyone agreed that free will is incompatible with determinism and yet some people say that free will is something that isn’t incompatible with determinism, at all. But it was never the case that everyone agreed that free will is incompatible with determinism. Compatibilists have always existed (I think). So it is not weird at all that the compatibilist is arguing for a conception of free will that isn’t prima facie incompatible with determinism. They’re compatibilists! Why would they do that?”

And, in this case, what we are left with is two groups of people, using the term “free will” to describe two very different things, and seemingly without any way to prove that their definition is the ONE TRUE DEFINITION of free will.

Here’s another way to think about Pereboom’s argument: the incompatibilist can allege that “whatever free will is, it is something that reasonable people could argue about whether it exists.” “At least slightly controversial existence” is an *element* of the definition or concept of free will: if something wasn’t controversial in this way, people would say “well, that doesn’t look like free will then, does it?” So there must be some contest, some empirical investigation that could turn out “yeah, we have free will” or “no, it turns out we don’t have free will.”

Note that even a compatibilist can have a definition of free will that makes it at least somewhat controversial whether free will exists. If free will involved something like fourth order meta-desires, it would be controversial whether human beings generally have these. Fourth order meta-desires are entirely consistent with determinism. But no compatibilist, with the exception of perhaps Nahmias, seems to have such a definition for free will.

In part because even a compatibilist can agree, in principle, that free will’s existence is controversial, the incompatibilist might expect the compatibilist to agree with the incompatibilist about this. Pereboom might say: “listen, I understand that we disagree about what free will is. But we can agree on this much, right: free will is supposed to be something that reasonable people can doubt exists—people like Spinoza and Einstein and Russell. It doesn’t have to be so doubtful that it can’t exist—I’m not insisting here that free will doesn’t exist. But it has to at least be something that there could be a reasonable disagreement about, a reasonable fight, an empirical contest. You’ll give me that much, won’t you, compatibilist?”

And, what I think you’ve made me realize is: no. The compatibilist doesn’t have to give the incompatibilist this much. And the reason the compatibilist doesn’t have to do this is because the compatibilist recognizes that the contest was not just empirical. It was also semantic. Compatibilists and incompatibilists have been fighting over what the term “free will” means, and not just about whether some specific faculties or powers exist. Because there were fighting over more than whether some specific faculties or powers exist, there doesn’t have to be a genuine controversy about whether any such powers exist. The semantic dispute is enough.

Note, also, however, that it’s a little silly to fight over what words mean. If enough people say “free will” means X and other people say “free will” means Y, then perhaps both parties need to just recognize that there are two senses of free will, or that free will is vague or ill defined. There could very well be no ONE TRUE DEFINITION of free will, looming out there in the ether, waiting to be discovered.

[Joe]

Kip,

You can't get from the claim that 'free will' is poorly defined to the thesis of denialism.

I still say: If NO ONE can have it, I don't want it.

"No one can be both here and there!" Is this a slight on my powers and abilities?

Perhaps you disagree. Then I say: "Get over it. You're not Superman and neither am I. None of us is the ultimate source of anything but each of us plays a role."

[Kip Werking]

Joe,

You seem to think that I'm vigorously defending denialism here.

You say "You can't get from the claim that 'free will' is poorly defined to the thesis of denialism." I say "You can't get from the claim that 'free will' is poorly defined to the thesis of compatibilism."

I'll make you a deal: you give up compatibilism and I'll give up denialism. I still think that I don't have certain powers or faculties, like being causa sui, to which I once thought "free will" referred. But, if you insist, I'll call that free will* and say that "free will" itself is pretty vague.

My point remains: compatibilism remains unsupported.

Fair enough?

[Joe]

Kip,

Just a few clarifications.

My “you can’t get there from here” point was in reference to your comment: “It might be the case, for example, that free will is poorly defined, and so neither compatibilism nor incompatibilism is well supported and ultimately successful.” Maybe this isn’t denialism but whatever it is, the inference to this conclusion is faulty.

Nor do I think that 'free will' is vague, at least not for the reason that you note. Saying that we haven't yet come to an agreement about what 'free will' means is different from saying that we've figured it out and its vague -- as if the expression 'free will' were like the term 'bald.'

Think of the term ‘personhood.’ There is a lot of disagreement about the application of this term but it doesn’t follow from that that the term is poorly defined or vague or that nothing is a person or that none of the theories of personhood are ultimately successful.

Second, I wouldn't say that compatibilism is unsupported. Surely compatibilists have provided some support for compatibilism. I admit that they have not provided conclusive support or, perhaps, even convincing support. Which is only to say that we have not yet solved the problem of free will.

Personally, I’m not sure how one can provide support for compatibilism other than to show the flaws in various arguments for incompatibilism. That's where I think the action is.

[Mark]

Kip,

In the spirit of trampling ambiguity, what does "unsupported" mean in the phrase, "compatibilism remains unsupported"?

[Kip Werking]

Mark,

By "unsupported" I mean something like:

"The compatibilist says that 'free will' means certain powers and faculties that nobody denies people have. But compatibilists are the only ones who say this. Libertarians say free will means something else. And impossibilists say it means something else as well. So it is not sufficient to assert that any one analysis of 'free will' is the ONE TRUE DEFINITION of free will, because each party in the debate can do this, and there does not seem to be any way to settle the question.

In order to support any view, the defenders of that view would need to do more than just assert that their analysis is correct. They would need to tell a story about why it is correct, about why "free will" and the the powers or faculties they discuss, mean the same thing. They need to do something like the mayor does, when he explains that MYSTERYKEY and KNOWNKEY are the same.

Here's one way they could do that, which would satisfy me: they could produce a survey and ask thousands of people what "free will" means and then list the options, in simple and friendly language, that compatibilists, libertarians, and impossibilists suggest. If all of the people say "oh, no, of course free will doesn't mean unmoved movers, spooky physics, or something impossible, of course it just means these specific powers that everyone agrees exist" then I'll become a compatibilist.

Of course, this probably won't happen. If you've ever asked the average Joe in the street what "free will" means, many or most of them have never heard of the term or, if they have, have only the vaguest idea of what it means. The dictionary definition is vague enough to admit of compatibilist, libertarian, and impossibilist definitions. This is why I suspect free will is vague.

Joe,

You wrote:

"Maybe this isn’t denialism but whatever it is, the inference to this conclusion is faulty."

How is the conclusion faulty? It seems sound to me.

"Nor do I think that 'free will' is vague, at least not for the reason that you note. Saying that we haven't yet come to an agreement about what 'free will' means is different from saying that we've figured it out and its vague -- as if the expression 'free will' were like the term 'bald.'"

I'm not 100% sure free will is vague in a way that is fatal to arguments about its existence or compatibility with determinism. I'll reserve judgment until more empirical work is done to show how comfortable are with certain usages of the term "free will".

But I do find it telling that people have been arguing for millennia about what "free will" means and yet they cannot reach agreement. Today, some of the smartest people I know continue to argue about it and yet cannot reach agreement. My hypothesis about this, that the term is vague, at least helps explain why people continue to disagree. But, if you don't care for my hypothesis, then it seems quite mysterious why people disagree, and have disagreed for so long, about what free will means.

"Think of the term ‘personhood.’ There is a lot of disagreement about the application of this term but it doesn’t follow from that that the term is poorly defined or vague or that nothing is a person or that none of the theories of personhood are ultimately successful."

Mele makes this same point against Double at the end of Autonomous Agents. But there is a subtle yet crucial difference between "free will" and "personhood". Most people subscribe to the principle "whatever personhood is, it is something that people, in the conventional sense, have, because persons obviously exist." The same cannot be said about free will: enough libertarians hinge the existence of free will upon shaky empirical foundations (unmoved movers or spooky physics) and enough skeptics insist that free will is something no human being possesses, that one cannot say "whatever free will is, it is something that people, in the conventional sense, have, because free will obviously exist." Free will *doesn't* obviously exist.

"Personally, I’m not sure how one can provide support for compatibilism other than to show the flaws in various arguments for incompatibilism. That's where I think the action is."

As I said in my previous comment, killing incompatibilism isn't sufficient to support compatibilism. Compatibilism is not the default view; it doesn't win because something else loses. It must win on its own merits. For example, if free will is vague (and I understand that you're not sympathetic to my arguments on this point), then neither compatibilism nor incompatibilism wins. You said that my "inference to this conclusion" is "faulty", but I don't see how it is.

[Mark]

Kip,

So you're willing to rest your metaphyics on the results of a popularity contest?

[Jonathan Farrell]

Kip,

Thanks for your comments. I think we kind of agree on the point we’ve been discussing (or at least we agree more than we seemed to at the beginning of your post!).

I’m not sure that the problem is a semantic one, though. It seems to me that it’s more to do with the concept of free will than with the use of the term ‘free will.’ I don’t think that there is one true definition of free will. I think that what we refer to when we say ‘free will,’ is a whole ragbag of faculties/powers/etc. This is why I’m not sure how helpful surveying the folk is. (I suspect that after lots of careful surveying has been done, we’ll find that, depending on the particular situation or how it is described, people (sometimes the same people!) sometimes have compatibilist, sometimes libertarian, and sometimes have no-free-will intuitions. This seems to be true of me, at least.)

I would suggest that one way to move forward would be to think about what the significant features of free will are (to put it in Kane-ian terms) or which features are worth wanting (to put it in Dennettian terms). Then we can see under which conditions (if at all) we can have these features. If we can have few or none of the important aspects, then perhaps we should conclude that we cannot have free will. But if we can have many or all, then perhaps we should conclude that we can have free will: just because salamanders don’t live in fires doesn’t mean that there’s no such thing as salamanders: http://en.wikipedia.org/wiki/Salamander#Mythology.

Everyone,

On supporting compatibilism: I agree that compatibilists have to say something positive, but I think this is what compatibilists do (they offer alternative analyses of free will to those put forward by libertarians, and argue that we can have the faculties that these analyses require). Also, we might think that showing incompatibilism to be false does suffice to show compatibilism to be true, if we take these views to be only concerned with whether the existence of free will in a deterministic universe is possible or not: by showing P false, we show not-P to be true.

[Kip Werking]

Mark,

You wrote:

“So you're willing to rest your metaphysics on the results of a popularity contest?”

Not at all. I’m willing to rest my semantics on the results of a popularity contest.

If 99% of people say that there is a teacup on the side of Mars, that doesn’t prove that there is a teacup on the side of Mars. But if 99% of people, in some world, say that “the word ‘teacup’ means a red planet fourth from the sun” then “teacup” means the same things as “Mars”. People can be wrong about empirical claims in a way they cannot be wrong about semantic claims, because they are the authorities on the truth of those claims.

This is an important truth about language, which I’ve mentioned earlier at the GFP. Others, who know much more about philosophy than I do, have informed me that common language is not the ultimate arbiter of a term’s meaning. They cite examples about mass, whales, and salamanders, etc. I’m not even convinced that these examples show what they’re proponents think they show. But even if they did, all of them are “rigid designators” in the sense of referring to something in the physical world that we can reliably inspect. If I want to learn something about whales, I can go out into the world and inspect whales. So we understand how people might be wrong about whales: they refer to whales as the things they can reliably locate in the ocean; they might wrongly believe whales are fish; then they go inspect whales and discover whales are mammals. But free will is not like this. You cannot reliably go out and inspect free will, whatever it is. It is not even clear that it exists.

So these sorts of examples do not work. But I have not seen any other kind of example showing how the meaning of “free will” is not a popularity contest. Perhaps some day a sophisticated philosopher like Manuel Vargas will show me how it isn’t. Until that day, it seems to me that I should rest my semantics on the result of a popularity context.

Jonathan,

When you say:

“I think that what we refer to when we say ‘free will,’ is a whole ragbag of faculties/powers/etc. This is why I’m not sure how helpful surveying the folk is. (I suspect that after lots of careful surveying has been done, we’ll find that, depending on the particular situation or how it is described, people (sometimes the same people!) sometimes have compatibilist, sometimes libertarian, and sometimes have no-free-will intuitions. This seems to be true of me, at least.)”

This seems, to me, to support the idea that free will is vague and not that it refers to any one thing. Perhaps you think that people think that free will is some big, complex, variant thing but that everyone is consistent about the details. But I doubt this. If this were the case, philosophers would agree that free will is this big, complex, variant thing. But they are not. And the folk are not either.

About Dennett: I find Pereboom’s critique of Dennett’s move here, in the intro to his LWFW, completely convincing.

Third, you write:

“On supporting compatibilism: I agree that compatibilists have to say something positive, but I think this is what compatibilists do (they offer alternative analyses of free will to those put forward by libertarians, and argue that we can have the faculties that these analyses require).”

1. It is irrelevant that they “argue that we can have the faculties that these analyses require.” No-free-will theories offer own their concepts of free will but do not argue that we can have these faculties or powers. But that fact, by itself, does not count against the no-free-will theorist’s view.
2. This leaves only the fact that the compatibilist “offer[s] alternative analyses of free will to those put forward by libertarians.” But every party of the debate does this. The question isn’t just whether they put forward their own analysis; the question is whether they can show that their analysis trumps other analyses. And there seems to be no way to settle the matter.

[Mark]

Kip,

Assuming we could once and for all settle the semantic debate with a most exquisitely crafted survey, what profit would that yield?

Were the results to favor "free will", "free will*", "free will**", or even "free will***", what would that tell us about the conditions that are required to satisfy the moral practices that people everywhere, from all time, have continued to cherish and zealously defend?

It would seem that you are sprinting down a path toward a rather lackluster haunt, even as ghost towns go.

[Kip Werking]

Mark,

It is a purely descriptive project. It need not solve or settle any normative questions whatsoever. So it might not tell us "about the conditions that are required to satisfy the moral practices that people everywhere, from all time, have continued to cherish and zealously defend?"

But it would tell us one thing: what free will is. And from there, we could probably decide whether it actually exists and whether it is compatible with determinism. And if we do that, even if we don't answer any normative questions, we solve one of the largest and oldest problems in philosophy, something that others have not been able to do for millennia. That may not be much, but it is something.

[Mark]

Kip,

I don't think anyone here, or in the history of the debate, has ever been very interested in the project you're describing... So, what's the point?

Even the experimental philosophy published to date has been related to specific empirical conjectures that fit into the inductive grounds for certain theories (e.g. Smilansky's claim that we would be worse off if the falsity of free will was broadly accepted).

[Joe]

Kip,

You write: "As I said in my previous comment, killing incompatibilism isn't sufficient to support compatibilism."

I never said that showing that an argument for incompatibilism was unsound, or was inconclusive, is sufficient to support compatibilism. I can't speak for others but my aim is not to prove that compatibilism is true, or prove that we have free will.

You write: "Compatibilism is not the default view; it doesn't win because something else loses. It must win on its own merits."

Here is my view: I believe that we have free will. I have no conclusive argument for this belief. I wonder whether the belief is, as some think, irrational. Thus, I study the arguments for free will skepticism. I note that they happen to presuppose incompatibilism (I just finished a paper on this if you'd like to see it.) Thus, I look at the various arguments for incompatibilism and so far I have concluded that they are left wanting.

What can we conclude from this? The argument for free will skepticism is not as good as you seem to think that it is. There is no good reason to deny that we have free will. That is good enough for me. I'm not trying to prove that we have free will, or to prove that compatibilism is true and I have no idea of what either proof would look like.

You write: "... if free will is vague (and I understand that you're not sympathetic to my arguments on this point), then neither compatibilism nor incompatibilism wins. You said that my 'inference to this conclusion' is 'faulty', but I don't see how it is."

I’m still not sure what you mean when you say that ‘free will’ is vague.

If 'free will' is vague like ‘bald’ is vague, then there is good reason to think that something like contextualism is true, and if that is true then likely so is compatibilism. Thus, compatibilism wins; there is no draw.

If “‘free will’ is vague” means that there is no single concept shared by all people, then the compatibilist wins, for one of these concepts is the compatibilist concept. It is indeed possible for someone to have free will even if determinism is true, at least in the compatibilist sense. The incompatibilist denies this. To deny this is to privilege some other sense as the real meaning of ‘free will’ and thus deny that the term is ‘vague’ in the way that you say that it is. (‘Vague’ is not the right word here. Maybe the right word is ‘ambiguous’ but I’m not sure.)

I don't think that your new conclusion – which I confess that I don’t fully understand – follows from the evidence that you have provided, as I've tried to explain previously and just now.

[Kip Werking]

Mark,

You wrote: “I don't think anyone here, or in the history of the debate, has ever been very interested in the project you're describing... So, what's the point?”

If I agreed with your premise (that nobody here, or in history, has been very interested in this project), then I might agree with your conclusion (that it’s pointless). But I think the project I am describing—determining whether free will exists and is compatible with determinism—has quite obviously interested philosophers both here at the Garden and throughout history. It seems to me quite uncharitable, and bizarre, to suggest otherwise.

Perhaps you are referring to the particular method in which I am engaging in this project: by seeking out the common usage of “free will” to determine whether it prefers compatibilism, libertarianism, and/or no-free-willism. I agree that not many philosophers have engaged in this particular method. But this just proves that my method, so described, is innovative—not that it is pointless.

You write: “[e]ven the experimental philosophy published to date has been related to specific empirical conjectures that fit into the inductive grounds for certain theories[.]” But much recent experimental philosophy has been directed at determining whether people use the term “free will” (or at least the term “moral responsibility”) in a compatibilist or incompatibilist sense. I’m thinking of the research by people like Nichols, Knobe, Nahmias, and Feltz, and their colleagues.

Joe,

You wrote:

“I never said that showing that an argument for incompatibilism was unsound, or was inconclusive, is sufficient to support compatibilism. I can't speak for others but my aim is not to prove that compatibilism is true, or prove that we have free will.”

Perhaps you did not say it was sufficient, but you did say (what I immediately quoted before discussing this point earlier):

“Personally, I’m not sure how one can provide support for compatibilism other than to show the flaws in various arguments for incompatibilism. That's where I think the action is.”

And I think this weaker principle is still false and for the same reasons: undermining one view does not support, and certainly is not sufficient to establish, the other view, because both views might be wrong (because, for example, free will might be vague/ambiguous).

You wrote:

“[The failure of no-free-will arguments] is good enough for me.”

Let’s be clear here. When you say it is good enough for you, do you mean that you remain unconvinced by no-free-will theories? Or do you make the stronger claim that free will exists? You claim that you believe in free will but cannot prove it. This comes dangerously close to saying that you believe something for no good reason, or without sufficient reasons, or at least without reasons you can demonstrate to others.

I believe in the laptop I am typing—but I also have good reasons for believing in it, the least of which is not that its existence remains uncontroversial. You don’t seem to have the same good reasons for believing in free will, if only because there is no consensus on what “free will” even means. So how can you be so confident that free will exists? I can understand that, if no-free-will arguments do not convince you, you should refrain from adopting them. But I do not understand taking the defeat of no-free-will arguments to imply the existence of free will (to be “good enough for me”). Perhaps, for example, there are good no-free-will arguments that you have not discovered, or that nobody has discovered. Alternatively, perhaps “free will” is ambiguous in the way I have been describing.

The possibility that free will is ambiguous motivated my remarks above (and not in the sense that bald is vague, because the existence of baldness is uncontroversial). This is a possibility you address explicitly and reject. You wrote:

“If “‘free will’ is vague” means that there is no single concept shared by all people, then the compatibilist wins, for one of these concepts is the compatibilist concept. It is indeed possible for someone to have free will even if determinism is true, at least in the compatibilist sense. The incompatibilist denies this. To deny this is to privilege some other sense as the real meaning of ‘free will’ and thus deny that the term is ‘vague’ in the way that you say that it is. (‘Vague’ is not the right word here. Maybe the right word is ‘ambiguous’ but I’m not sure.)”

Let’s follow your logic here:

1. No single concept of free will shared by all people -> one of these concepts is the compatibilist concept -> compatibilism is true
2. Incompatibilism denies that free will is compatible with determinism -> incompatibilism privileges some other sense as the real meaning of “free will” -> incompatibilism denies that “free will” is ambiguous

1 is just a non-sequitur. Unless my eyes are horribly jaundiced, it does not even begin to be a sound argument for compatibilism. Sure, the first two premises are true. But the conclusion “compatibilism is true” doesn’t begin to follow from them. Why would you think otherwise?

2 is entirely true (note that I was more of an incompatibilist when I started this thread than I am now).

Kip

[Mark]

Kip,

The project you've clearly defined for us is not even philosophical in the strict sense; it's more befitting the label of cultural anthropology. I would be most suspicious of any school that would authorize that as a topic suitable for a doctoral thesis in philosophy.

The major problem is that your descriptive-only project cannot tell us anything about the real free will project, that everyone else is concerned with, which is bent on investigating the conditions that ground our most cherished moral practices. These substantive questions cannot be resolved by survey.

Regarding recent articles relating to free will from the experimental camp, initially most were investigating emperical claims made in philosophical arguments, but recently some have been in response to previous studies. To the degree that these projects shift over into purely anthropological domains and lose their ties to the philosophical debate, they too will lose their philosophical relevance.

[Kip Werking]

Mark,

1. If you don’t want to call my project “philosophy” in the “strict” sense, I am happy to jettison that label. As far as I am concerned, philosophy’s record for solving problems is nothing to boast about.

2. You wrote:

"The major problem is that your descriptive-only project cannot tell us anything about the real free will project, that everyone else is concerned with, which is bent on investigating the conditions that ground our most cherished moral practices."

This claim is based upon the premise that the free will problem, as it interests philosophers today and throughout history, is strictly tied to concerns about the conditions that ground our moral practices. But I doubt that "free will" is necessarily tied to our moral practices in this way. As Saul Smilansky wrote:

“To believe that control matters because it is a condition for moral responsibility and for the concomitant moral (and legal) notions does help us understand what the free will problem is about. But seeing that problem as following only from concern about moral responsibility can be misleading: even a person who has no interest in morality can easily and reasonably become concerned about the free will problem.” (Free Will and Respect for Persons).

Similarly, Robert Kane lists free will as related to, and important to, the following things: creativity and novelty; autonomy or self-creation; desert for one’s accomplishments; moral responsibility; reactive attitudes; individuality and uniqueness; dignity; life hopes and an open future; and love and friendship (SoFW, chapter 6).

Some of these are clearly moral or normative in nature, some of them have moral or normative connotations, and some of them have nothing, per se, to do with morality. Consider the following statements:

“You are not the genuine author of that painting, because God designed a clockwork universe, and your life story, so that you would paint it. God is the real author.” Response: “No, that’s impossible because I have free will.”

“You’re very predictable. I know what you are going to do before you do it. If I knew all of the information about your body and brain, and I had a powerful enough computer, I could predict everything about your life.” Response: “No you can’t, because I have free will.”

“Your entire life is just the result of heredity and environment. The two of them combined, in a very complex way that we cannot fully describe, to produce who you are.” “No, I am more than that, because I have free will, and contributed to my own self creation.”

I could produce many more examples like these. The point is: all of these examples make sense. We could imagine how people use the term “free will” in this way. Furthermore, we understand how people might be genuinely concerned about creativity, novelty, autonomy, self-creation, etc. But none of these things necessarily influences our moral practices. To think of free will as just involving the conditions that ground our moral practices is, as Smilansky notes, to adopt an unduly narrow of what the free will problem involves.

[Mark]

Kip,

So you concede the point that your project is moot?

By "our most cherished moral practices" I intend to refer to anything and everything that has to do with how it is that we, as agents, impregnate reality with value and meaning -- if we do so at all (that's part of the philosophical debate, right?). That perspective is sufficiently broad enough to capture all of the examples you mentioned, and many more to boot. Put a different label on it if you must, but please accept my invitation rejoin the conversation.

[Kip Werking]

Mark,

No, I don't concede that at all. The only thing I conceded in my previous comment was that my project (as you describe it) might not be philosophy in the strict sense. It does not follow from this that the project is moot (dictionary.com: "of little or no practical value or meaning; purely academic").

I understand the conversation you are talking about. But it is not the only conversation academics can have or the only conversation philosophers can have. If you don't find it particularly interesting or compelling, fine. Many areas of philosophy bore me to death. But it does not follow, from your lack of interest, that the project is worthless or meaningless.

[Cihan Baran]

Kip,

I don't see why the common usage of the phrase "free will" should be binding in any sense.

A scientist may say, "according to the vernacular, a whale is a fish. However, my investigations conclude that whale really doesn't meet the necessary conditions for being a fish. The whale, more properly, is a mammal."

Analogously, the philosopher may say, "According to the vernacular, free will requires the ability to X. However, my investigations conclude that free will really requires the ability to Y. Hence, even though we do not have the ability to X, we do have the ability to Y and hence we have free will."

Or perhaps in a way you'll find better, the philosopher may say, "According to the vernacular, free will *is* the ability to X. My investigation concludes that we do not have such an ability. However, we do have the ability to Y and this is sufficient for moral responsibility, creativity...etc."

My point is that I don't have any real interest in free will if it does does not relate to moral responsibility, creativity..etc. I have no substantive philosophical commitments/interest for whatever free will is, in the unlikely case it does not relate to all those wonderful things Kane lists.

My real interest is whether we have moral responsibility, autonomy, etc... And if we don't have free will, so be it.

[Joe]

Kip,

I only have a few more weeks left in my sabbatical, so I need to stop posting after this. It's been fun, though, and I promise to continue reading.

You wrote: "I believe in the laptop I am typing—but I also have good reasons for believing in it, the least of which is not that its existence remains uncontroversial."

I think that your "good reasons" are not so good. Your evidence is inconclusive, at best. I didn't say that I don't have any reason to believe that I have free will. I said I don't have a proof. Apparently, your standards are much weaker than mine. If the above constitutes proof, then I can prove that I have free will!

For one thing, if the fact that "the existence of something remains uncontroversial" is proof that it exists, then I have a proof that free will exists! I just have to get you away from contemporary analytic philosophers and into the real world first. If you had gone into the basement of the Garden Lounge in Moscow, where they play pool, you'd be hard-pressed to find someone who thought that the existence of free will is controversial. If you were here in New Jersey with me, spending time with my family and some old friends, you'd find it difficult to find anyone who wouldn't laugh at the suggestion that we are not morally responsible for our actions.

I grant that not everyone believes that we have free will but clearly MOST people do. And, depending on which circles you hang around in, you are likely to be surrounded by people who think that the existence of free will is uncontroversial. Unless you happen to hang out with philosophers and neurobiologists -- but even in that crowd the belief in free will would register a high overall confidence rating. I'm sure you'll join me in noting that none of this means a lot.

And try bringing the belief that you know that there is a laptop in the group of philosophers and neurobiologists and see how it stacks up, especially once you start talking about your reasons for belief. (I'm not comparing the belief in free will with the belief in external objects, although I might in the future. Right now the comparison is between claims about free will and claims about knowledge.)

In short, I think that the argument for epistemological skepticism is much better than you think that it is and I think that the argument for free will skepticism (which I take to be a metaphysical view) is much worse than you think it is.

You've added some new wrinkles to your theory but I'm still not sure that I understand it. In my Canadian Journal of Philosophy paper, I put for the two-'can' view, which suggested that the the word 'can' is ambiguous. If we're really talking about a genuine ambiguity with respect to the expression 'free will,' then I have a hard time not seeing that as a victory for compatibilism.

My guess is that this is not quite what you mean. You think that there is no single concept of free will shared by all. According to one concept, it might turn that compatibilism is true and according to another concept, it might turn out that incompatibilism is true.

I still think that this is more of a victory for compatibilism than incompatibilism. Certainly, one who adopts the theory might see it as a draw. But I bet you would find few incompatibilists who would adopt your theory. The incompatibilist believes something stronger, e.g., it is impossible for one to have free will in a determined world. That stronger view seems to be wrong if your theory is correct.

This is enough for me. But, hey, maybe I'm just easy to please!

[Kip Werking]

Cihan,

I would just refer to my post in this thread from Apr 30, 2007 4:01:09 PM.

Joe,

I feel as if you’ve jumped on one rhetorical point I made and largely ignored the others I did make. That point is the distinction between my laptop and “free will”. I wrote:

“I believe in the laptop I am typing—but I also have good reasons for believing in it, the least of which is not that its existence remains uncontroversial.”

Note that I said this reason is not the least of my reasons—not that it is the most of my reasons. I agree entirely that “everyone believes in X” is not a particularly good reason for believing in X. If everyone believed in God, I still wouldn’t believe in God. But, at least for practical purposes, it is a useful heuristic: many or most of the things I should believe in (my laptop, London, the sun, etc.) are things that most people believe in and many or most of the things I should not believe in (unicorns, Leprechauns, Santa Claus) are things that most people do not believe in. So I grant your point that this is not a particularly good reason to believe in my laptop.

I would have elaborated on my larger point if I had thought you would jump on my wording here. But since you have, let me explain myself:

My laptop is a rigid designator in the sense that, if I have questions about what my laptop is, I can go inspect my laptop and find out. For example, I might believe my laptop is a Mac, but discover that it is a PC. But I cannot do this about free will.

Whenever we are trying to decide questions about some term, (i) first we need to decide what the term means and then, having this understanding, (ii) we need to resolve questions about whatever it is.

I agree that, if we successfully completed the first project (i), we would still have thorny issues to resolve about our epistemics. Suppose, for the sake of argument, that we do resolve the meaning of “free will” and it means exactly what Robert Kane thinks it means: it involves, for example, making Self-Forming Actions such that agents can be Ultimately Responsible for their actions. Suppose you agree about this, and I agree about this, and everyone in the world agrees about this. Then, as you rightly point out, we are still left with the problem of (ii), even though we’ve solved (i).

Suppose, for example, that in this world, sophisticated neuroscientists inspect human brains while they make decisions. They find that there is spooky physics in the brain such that people simultaneously want to accomplish two different things, during important moral decisions during their lives, and that whether the person will decide one way or the other is ultimately indeterministic—just as Kane said it would be (or could be). In fact, we have all of the evidence in front of us for the existence of free will (as everyone agrees that Kanean free will just is free will), just like I have all of the evidence in front of me that my laptop exists. We can see it, we can touch it, and everybody agrees that it exists. Still, you remind me, the skeptic can say: “but we might all be brains in vats. Your laptop might not exist. Free will might not exist. A Cartesian demon might be totally deceiving your senses.”

In this way, regarding the second project (ii), I agree that free will and my laptop are equally vulnerable. [Note, however, that if you believe in the existence of free will, you are not helping your position by pressing this skeptical argument.]

But, in my previous point, I was not trying to distinguish between free will and my laptop with respect to the second project (ii)—I was trying to distinguish between them with respect to the first project (i). To raise epistemic worries, such as those involving the Matrix and Cartesian demons, is to get ahead of ourselves. This is why: before we can even decide whether free will exists in this empirical sense, we must decide what the term “free will” even means.

And this is where “free will” and “my laptop” are very different. If we put your skeptical worries to the side, I can ask questions about my laptop to almost arbitrary precision and ultimately obtain their answers (or at least their “answers” in my brain-in-a-vat world). If someone says “is my laptop a Mac or a PC?” I can answer them. If they ask me whether it is black or white, I can answer them. I can tell them how many keys it has, how fast the processor is, how much RAM it has, how many ports it has, how big the screen is, and so on. The number of questions for which I can give definite answers in this sense is practically endless.

But, if we put your skeptical worries to the side, I cannot even do this much with respect to “free will.” If someone asks me “is free will compatible with determinism?” I cannot point to anything and say “ah, it turns out it is compatible with determinism” or “ah, it turns out it is not compatible with determinism.” I cannot do this because I would not know what to point to. I could do this, for example, if we all agreed that free will is what Robert Kane says it is (or what Daniel Dennett says it is, or what Derk Pereboom says it is, etc.). But we don’t all agree with Kane. We all *disagree* with each other.

The number of questions for which I can provide definite answers about “free will” is drastically curtailed by this reality. In fact, there is very little I can say about free will with any authority. This is not because a Cartesian demon might be deceiving me; it is because I am not even sure what “free will” means—I’m not even sure what the demon would deceive me about. This is in stark contrast to my laptop. And this is the crucial difference to which I was referring in my last comment.

Finally, you write:

“I still think that this is more of a victory for compatibilism than incompatibilism. Certainly, one who adopts the theory might see it as a draw. But I bet you would find few incompatibilists who would adopt your theory. The incompatibilist believes something stronger, e.g., it is impossible for one to have free will in a determined world. That stronger view seems to be wrong if your theory is correct.”

You’re right: someone like myself would see it as a “draw”. But even “draw” implies that they both do equally well. I would think of it more as a draw in the sense that no party can win. Not only do I think that someone who adopts this view “might” think that no party can win—I think such a person *must* think that.

You seem to disagree. But perhaps you are just, as you say, easy to please.

[Mark]

Kip,

Given your descriptive project, at the very best you could report something like, "At the time the study was administered, 75% of the people believe free will means X, 15% believe it means Y, and 10% believe it means Z". From these results one could draw hypotheses such as: theory X is far more popular than theories Y and Z for cultural reasons C, theory X came to be the most popular for historical reasons H, a disposition toward theory X seems to be embedded in our DNA for survival reasons S, etc. These hypotheses would surely are all very interesting and worthy of further investigation... for scientists.

Granting whatever value these findings may have in other circles, they remain philosophically moot. So, if you concede that your project is philosophically moot, I have to wonder why you see fit to spend so much time discussing it on a philosophy blog.

[Kip Werking]

Mark,

I disagree that the findings you mention would be philosophically moot---for reasons I've already explained. I also think you have an unduly narrow view of this blog's purpose.

[tnadelhoffer]

Mark,

First, you seem to draw a sharp line between philosophy and science. Would you care to explain how you distinguish one from the other?

Second, even if Kip were to concede that his descriptive project is merely anthropological in nature, why would it follow that it is "philosophically moot"? After all, there are a number of research programs in the sciences that are anything but philosophically moot! You seem to assume something like: If a research program RP is scientific, then it is philosophically moot--which is transparently false.

Third, when you say "our most cherished moral practices," who is the "we"? You? High SES Western graduate students? Americans? People in general? Philosophers? Just in making such a claim, you run the risk of "inventing your psychology and anthropology from scratch" to suit your argumentative ends. For now, I am curious to hear how you are going to define "we" in a way that wouldn't itself call for the kind of work done by experimental philosophers.

[Mark]

Kip,

It seems you're conflating "philosophically moot" with "academically moot". I doubt whether it is a good use to my time to pursue this further.

Tom,

First, science surely has its place and its value is unquestionable. As an empirical enterprise, science is very good at coming up with theoretical models in response to inquires that can be expressed in terms of an empirically falsifiable statement. However, it is obvious that philosophy logically precedes science: the philosophical grounds for the scientific method logically precede science, the mathematical models (and the related philosophy of mathematics) that ground scientific conjecture also logically precede science, etc., etc. I'm sure this is all old hat.

Second, many premises in philosophical arguments are subject to empirical falsification. Science can play an obvious role here. The problem I raised is that the specific project Kip has in mind seems inconsequential to any of the free will related arguments that I'm familiar with (viz., "what do people mean by 'free will'?"), which constitute the mainstay of the historical and contemporary dialogue, and I see little reason to suspect that they would become relevant at a later date.

If a philosopher is concerned with the conditions to secure autonomy, creativity, moral responsibility, self-expression, etc., what relevance would it be whether any of the possible hypotheses I mentioned earlier were not proven false by succeeding studies and were thus elevated to the status of scientific theories?

Science could obviously add value to the philosopher's tool set. If theory X demands that we have capacity C in order to secure creativity, science could (presumably) investigate whether we have capacity C. If we further stipulate that we obviously possess creativity, we could thus invalidate theory X on scientific grounds.

Third, by "we" I simply mean to refer to the mainstay of the historical and contemporary conversational context surrounding free will.

[Mark]

Joe,

Surely there are many types of beliefs that do not require an "argument" to considered rational and warranted. There must be other means to produce warranted beliefs than arguments, since arguments depend on premises, which must themselves be warranted beliefs. The process has to get started somewhere, right?

Depending on which theory of knowledge you subscribe to -- whether an internalist or externalist account, there must be faculities that produce beliefs that have at least some warrant; regardless if they are warranted individually (foundationalism) or jointly (coherentism).

Regarding Kip's belief in his laptop, we can say that his faculity of sight is a source of basic warrant -- as are his faculties of touch and hearing. Since his belief was produced by those faculties, his belief in his laptop is warranted.

Regarding our belief in free will, we can say possession of an agential perspective is the source of the warrant for belief in free will. Since our belief in free will was produced by that faculty (which is most likely a meta-faculty or second-order faculty), our belief in free will is warranted.

Why go in search of an argument when it is unnecessary, if not impossible to produce?

Kip,

Just thought of another way of illustrating my point...

Take Smilanksy for example. He attacks "deep" free will from both fronts: he rejects both compatiblism and liberatianism. Regardless which is the dominant view amongst the public, he claims that the public view is wrong (since there is no "deep" free will), and he futher claims that we ought to uphold the illusion of "deep" free will, in whatever sense the public is comfortable with, for forward-looking reasons. Now, sure, he makes some claims in passing about which view he suspects is the dominant public view, but it doesn't really matter in relation to his actual project.

That's just one example. Can you name any philosophers whose projects would be in jeopardy if your study produced evidence one way or the other?

The best I could think of is imagining the world some time from now where people are very concerned with marketing philosophical ideas about free will and they would like to know how best to structure their ploys to insure that the most number of people will "buy in" to the view they are selling. So, if the compatibilists of the future want help selling compatiblism, they might want to know what the dominant view is so they can aim their appeals at the most relevant concerns. Then again... I doubt the good old shotgun approach that we're all used to would really hurt much either, but I suppose the results of your study could save some time and energy.

Mark,

I just want to make it clear that I am not a skeptic. My point was a comparitive one: the argument for free will skepticism is no better than the argument for epistemological skepticism. I think that there are adequate responses to each. However, I don't favor a purely externalist response, like the one that you offer, to the problem of epistemological skepticism.

I agree that "there are many types of beliefs