Free Will in Psychology Today

Yet another piece on free will by a psychologist can be found here.  This time, the author is Joachim Krueger--a social psychologist from Brown University.  In light of the gathering number of popular science pieces on free will, I thought this might be a good opportunity to get everyone to list the ones they know of in this thread.  I actually had someone contact me the other day when I posted the the link to the "free choice" study asking me if there was a rough and ready list for getting one's feet wet with respect to what psychologists are saying about free will (or lack thereof).  Since this is something that comes up here at the Garden, I figured this would be a perfect place to get the ball rolling.  Once the comment thread is flowering, I will compile it into a bibliography and post it here for others to use.  It might aslo be helpful if you state in your comments whether the psychologists take a pro, anti, or neutral stance on the existence of free will--which would make it easier for me to clasify the bibliography.  Since I know at least two Gardeners are already working on a paper about the contemporary interest amongst psychologists in free will, I assume they will take the lead here! :)

UPDATE:

The piece by Krueger is a response to an earlier post by Baumeister which can be found here.  Another response to Baumeister can be found here. 

Comments

[Eddy Nahmias]

Thanks Thomas. Yet again we see psychologists simply defining free will as being "a cause that is itself uncaused"; presenting determinism as, by definition, the opposite of free will; and conflating determinism with reductionistic materialism. I try to make some of these points in a comment posted there.

Hopefully, Manuel will post some of the sources he's found in our joint project. I'll try to as well (eventually).

[Cihan]

Baumeister writes:
"To a determinist, there are no counterfactuals. Nothing that didn't happen could possibly have happened. Everything that did happen was the only possible thing that could have happened at that point in time and space, given the causes."

Haha... I think psychologists ought to stay away from philosophy - what do you think? (Philosophers on the other hand are free to roam in psychology.)

On an unrelated note, I remember having discussions on GFP about how the phrase "had to happen" in a study by Nichols et. al. threw off folks' modal intuitions (and thus supposedly their compatibilist intuitions). After reading this, I say "pleaaaasee....."

[Eddy Nahmias]

Cihan, you say "on an unrelated note," but then it looks like you are saying Baumeister's view of determinism should call into question my point about the folk's modal intuitions. I don't get it. Baumeister is making my point about the problematic way Nichols and Knobe describe determinism, not refuting it. He is taking determinism to mean what I think is suggested in the Nichols and Knobe description of determinism--that everything that happens *has to happen*. It is easy to get people to understand determinism to mean a sort of fatalism where what happens must happen. What's hard is getting them to understand that determinism *does* allow for counterfactuals and possibilities, just not that things could be different *holding fixed* the past and laws. (I should add that Baumeister at least understands that determinism is not the same as reductionism and that he understands that compatibilism is not false by definition. I take his work on strength of will to be quite interesting and relevant to the free will debate.)

[Kip]

Eddy,

I have a question for you. I understand that you're distinguishing between:

A. Derminism->every state of the universe had to happen as it happened, including the first state

and

B. Determinism->every state of the universe had to happen as it happened, except for the first.

In other words, because the first state didn't have to happen, nothing that followed had to happen (as if the Big Bang was one big Agent Cause Libertarian).

You seem to cling onto this distinction pretty vigorously. Here is my question for you:

Do you agree that this distintion (between determinism A and B above) is completely irrelevant in deciding whether agents have free will?

In other words, the distinction between determinism A and B is nothing to hinge one's hopes for free will on.

It's not as if someone would say "Thank God, the first state of the universe could have been something different... therefore I can still have free will." It wouldn't be *in virtue* of determinism being determinism B (as opposed to A) that anyone had free will?

I think you would agree with this much, but I wanted to make sure. If I am right about the above, and you agree, then I don't understand your critique of Nichols and Knobe. The most they can be guilty of, I would think, is a slight lack of precision in their language, but not one that should affect the outcome of their study. But perhaps I am failing to fully understand your critique.

[Cihan]

Eddy,

I agree with you that Baumeister (and shame on him, he should know better than this if he is writing about it) and the folk have a mistaken notion of determinism.

Nonetheless, I believe that the distinction is so subtle and rarefied and what the folk think and the philosophers mean are so apart that a wording such as "had to happen" (as opposed to, say, "deterministically happened") will not make a difference. In fact, the folk will interpret "deterministically happened" and "had to happen" as essentially the same. Thus, I agree with Kip when he says

"The most they can be guilty of, I would think, is a slight lack of precision in their language, but not one that should affect the outcome of their study."

(Kip, by the way, you need to account for the contingency of natural laws as well - not just the past or the first event.)

With that said, I'm not so sure whether this difference between a contingent but deterministic event and a necessary event can do the work the compatibilist argues that it does. If we could debate the "fatalism" vs. "determinism" issue, I'd be grateful as I may be misunderstanding some of the things. (I remember that Kadri Vihvelin was a big fan of this distinction.)

[tnadelhoffer]

Kip and Cihan,

I am unconvinced that either of you have adequately addressed Eddy's worry. For starters, there is obviously a difference between fatalism and determinism (it's easy enough to show formally, but Eddy and Jason Turner have already taken the time to do this in print in response to Nichols and Knobe, so I won't rehearse their argument here). Moreover (and more importantly), it's clear that EVERYONE--philosophers, psychologists, and the folk--ought to think that fatalism is incompatible with free will. The whole debate is whether determinism is similarly incompatible.

As such, wording determinism in a way that either conflates determinism with fatalism outright or at least invites people to conflate the two is a bad methodological approach to adopt when there are other ways of wording it that don't make/invite this mistake. That point seems entirely obvious to me. So, even if the modal scope worries raised by the original wording of the Nichols and Knobe studies didn't affect the results (and I am not claiming that they did, only that it is quite possible that they did), it is certainly reasonable to point out that to the extent to which different language could have been used, it should have been used. Of course, you can dismiss this possibility out of hand by simply assuming without any supporting evidence whatsoever that the "had to happen" language didn't bias the results of their studies in the favor of incompatibilism, but if psychologists (and experimental philosophers) have taught us anything, it's that these kind of "slight imprecisions" in wording can and often do make a difference. It is precisely why we run manipulation checks to make sure that people understood the vignette in the intended way.

[Kip]

Thomas,

I was only asking whether Eddy agrees with me that, in principle, the difference between determinism A and B shouldn't matter. I agree with you that one aspirational goal should be to use the most accurate language as possible, and that slight differences in wording can make big differences in folk response.

Still, while we're discussing the issue, I'll ask you the same question:

Do you agree that the difference between determinism A and B (as described in my comment above) should make no difference as to whether free will exists, or is compatible with determinism?

[Kip]

Another of way of putting my point:

Compatibilist-powers are compatible with fatalism. In a fatalistic world, in which even the first state of the universe "had to happen", people could still be reasons-responsive, or act in accordance with their higher-level desires, etc. Because people can have these compatibilist powers in either fatalistic or non-fatalistic (but deterministic) worlds, this difference shouldn't affect an incompatibilist finding.

[Of course it might affect an incompatibilist finding because of subjects' inconsistency or irrationality. But the distinction wouldn't affect it in a way that compatibilists could rationally justify. Compatibilists don't want to hinge the existence of free will on the premise that "the first state of the universe didn't have to happen."]

[Eddy Nahmias]

1. I hope to have some results soon (with my student Dylan Murray) on whether the Nichols/Knobe description of determinism leads subjects to interpret it as fatalism (in the abstract condition)--e.g., that agents's decisions make no difference to what the agent ends up being caused to do. If so, then I'd think that is a significant problem for drawing any conclusions from their studies about folk intuitions about incompatibilism properly construed. (Kip, compatibilism is not really compatible with certain types of fatalism, like one that says you are fated to X no matter what you decide or try to do.)

2. Kip, I'm afraid your determinism B is still not accurate. Determinism is about the (logical) *relationship* between events in a universe (or system), not about the necessity of any one of those events (first or later, and Cihan is right that the laws are contingent). As I've said here before, I think the best way to see this is to forget about free will for a minute. Consider a dog trying to catch a frisbee. She can usually do it on this type of throw but this time she misses it and her owner says, "She could have caught it." Did he say something false in a deterministic universe? I find that hard to believe. Rather, in a deterministic universe, his claim must be interpreted in a way I think is actually most natural: "She would have caught it only if something had been slightly different just before she tried" (rather than "She could have caught it even even if *everything* had been *exactly* the same just before she tried"). The weird part is that this conditional reading, in a deterministic universe, then entails that other events would have had to be slightly different, perhaps going back to the beginning of universe, or the laws would have had to be slightly different (going Humean here helps). But indeterminism is kinda weird too (given exact same conditions and laws, more than one thing could happen--yeah (students wonder), but what makes the difference whether one thing happens or the other?? Quantum physics is weird!). My guess is the folk don't have deep intuitions about which of these is more or less weird, at least outside the free will arena (they are theory-neutral here). And my own theory is that the folk's intuitions *within* the free will arena do not commit them to an unconditional analysis of "could do otherwise" but that's trickier to defend.

[John Dell]

Here is an older article that addresses the issue of free will as a concept which can be studied scientifically:
http://psycnet.apa.org/?fa=main.doiLanding&doi=10.1037/0033-2909.108.1.19

Here is a book that I imagine many of you know of, but I thought I would link to it anyway just in case some do not:
http://books.google.com/books?id=6DrNDONGuAQC&dq=Free+Will+Psychology&lr=&source=gbs_summary_s&cad=0

Sorry about having to post the whole url, but I am not very tech savvy so I do not know how to do it another way.

[Tom Clark]

Thomas,

Regarding precise wording of deterministic scenarios taking full account of the proper placement of modal operators, see the Garden thread New Midwest Studies in Philosophy. The result of the discussion, more or less endorsed by Eddy, was the following deterministic scenario:

“It is impossible that the Ertan has the specific thoughts, desires, and plans he has, in the specific situation he's in, and then makes a different choice than he actually makes."

Re Eddy's point above, note that this scenario describes the *relationship* between events, not their individual necessity. Still, this statement seems equivalent to saying that the actual choice had to happen as it did *given* the specific states of affairs internal and external to the Ertan. Of course this determinism isn't fatalism, which is the idea that no matter what the Ertan decided to do, things would have turned out the same way.

Eddy,

I agree that the conditional reading of "She could have caught it" is the natural one (the only sensible one) in a deterministic universe. As to whether the folk have that reading in mind is another question, given that many of them don't believe we live in a deterministic universe. Many of them believe (I suspect, but it needs researching) that human beings (if not dogs) are loci of a strong kind of agent causation that makes them exceptions to determinism. They think of themselves as causa sui in some crucial respect such that "it's *possible* the Ertan has the specific thoughts, desires, and plans he has, in the specific situation he's in, and then makes a different choice than he actually makes."

[Kip]

Eddy,

I recognized that my description of determinism B could still leave something to be desired (although I think the gist was conveyed between Gardeners like us). But I'd still like to get a decisive answer to the question. So please let me revise the description of determinism B:

A. Derminism->every state of the universe had to happen as it happened, including the first state, in accordance with natural laws that determine the next state based on the previous.

and

B. Determinism->given that the first state of the universe was what it was, every subsequent state of the universe had to happen, in accordance with natural laws that determine the next state based on the previous. But the first state did not have to happen.

Now the question again:

Do you agree that this distinction (between determinism A and B above) is completely irrelevant in deciding whether agents have free will, and whether it is compatible with determinism?

[Brian Parks]

Eddy,

I'd like to echo Kip's concern here.

Assume our universe is Laplacian deterministic. I make some choice at t=now. Why would the question of whether the state of the universe "could have been" different at t=0 matter to the question of whether my choice was free, or the question of whether I am morally responsible for it?

[Paul Torek]

Kip,

While you're waiting for Eddy's answer, here's another. The key phrase of (A) is "every state of the universe had to happen", which sounds an awful lot like "every state is inevitable". That kind of absolute necessity threatens free will in a way that relative necessity does not. I think (A) could be called maximal fatalism.

The phrasing of (B) privileges the first state of the universe. But there's no particular reason to do that. How about the last state? How about the current state? The physics doesn't dictate any particular choice. This doesn't make phrasing (B) wrong, but the fact that it's optional throws a new light on it.

Living in a deterministic universe is a bit like being married. You both exert and receive influence. This has its benefits. Some of us consider it preferable to a radically individualistic mode of existence.

[Eddy Nahmias]

What Paul said.
And what I said earlier about admitting a weirdness to the conditionality going back to beginning of universe (unless you make it the laws that change or the fact that being a deterministic system is itself contingent). But remember indeterminism is weird too.
I'm off to Chicago APA (talking about scientific challenges to free will in a session with Al Mele), so will be silent for a while.

[Kip]

Paul and Eddy,

I appreciate your responses, but neither answered the question. Having read your responses, I still don't know if you would answer yes or no.

Paul wrote:

"While you're waiting for Eddy's answer, here's another. The key phrase of (A) is 'every state of the universe had to happen', which sounds an awful lot like 'every state is inevitable'. That kind of absolute necessity threatens free will in a way that relative necessity does not."

How does it threaten compatibilist free will? Suppose every state has to happen, including the first. What prevents you from being moderately reasons-responsive? What prevents you from satisfying the compatibilist mesh-theory of your choice? Nothing.

Paul goes on:

"I think (A) could be called maximal fatalism."

Let's call this True Fatalism (TF). What do we mean by TF? Here's a helpful example: suppose I throw a baseball through a glass window. Normally, the glass window shatters, because I threw the baseball. If I hadn't decided to throw the baseball, and therefore didn't throw it, the window wouldn't have broken. True Fatalism would be for the window to break no matter what I did. I decide not to throw, but I throw anyway, or I don't throw, and the window still breaks. Effects proceed, regardless of causes. True Fatalism is a strange thing.

Here's my point: Determinism A does not imply TF. In determistic A worlds, my throwing the baseball can still cause the window to break. There need not be any strange mismatches between cause and effect. In deterministic A worlds, people can still have the compatibilist powers of your choice.

In other words, let's distinguish between two distinctions:

First State of the World Distinction (FSWD): the difference between determinism A and B; and
True Fatalism or not True Fatalism Distinction (TFD): whether effects proceed, regardless of causes

My point is this: FSWD and TFD are orthogonal to each other. Both determinism A and B can accommodate compatibilist agents. And both determinism A and B can accommodate TF-style mismatches between cause and effect. FSWD and TFD don't necessarily have anything to do with each other.

But, while compatibilists seem genuinely worried about TFD, the criticism of Nichols and Knobe focuses on FSWD.

[Kip]

In further review, I see that Paul's "that kind of absolute necessity threatens free will in a way that relative necessity does not" implies "no" as an answer to my question. So please ignore the first two sentences of my previous comment.

[tnadelhoffer]

Kip,

You are spending a lot of time trying to flesh out from the armchair what I took to be a fairly straightforward empirical issue--namely, were the intuitions of participants in the Nichols and Knobe studies more incompatibilist than they otherwise would have been because the "had to happen" language makes it tempting to conflate determinism and fatalism. Perhaps this is why Eddy didn't answer your question. What he did say is that he would have precisely the kind of data you need to shed light on this issue in the coming months. That seems like the right response given the actual issue that was originally raised.

Setting that issue aside, your recent comment (where you state that TFD entails that effects proceed regardless of causes) seems confused to me. Keep in mind that determinism merely entails that, necessarily, given the actual previous state of the universe (Po) and the actual laws of nature (L), there is only one possible outcome (P). Fatalism, on the other hand, is the claim that it is impossible that the past, the laws, and the present state of affairs could have been otherwise.

I take it this is why Eddy made the following suggestion in his MWS piece:

"The problem is that determinism should not be described in a way that suggests that actual events, including human choices, could not happen in any other way. This is because determinism does not entail that nothing could happen otherwise—that all events are necessary. The fact that an event X is completely caused (or determined) by prior events does not entail that X has to happen (necessarily happens)" (Nahmias 2006, p.222).

Adam Feltz, Edward Cokely, and I talk about this as well in a paper we have forthcoming in Mind & Language. I plan to post a copy over at the experimental philosophy blog soon. Perhaps we can argue about this more then...

[Kip]

Thomas,

You're right that I'm spending a lot of time trying to figure this out. That's because the answers I've gotten seem evasive (didn't you object to evasive answers in another thread?).

There are some ambiguities here that are making conversation difficult:

1. What is fatalism? Determinism A implies a kind of fatalism. We can call this Apparent Fatalism. But it does not imply True Fatalism, which Eddy describes as "you are fated to X no matter what you decide or try to do." Compatibilist free will is compatible with Apparent Fatalism but not True Fatalism.

2. Are we concerned with the technically right answer, or with what the folk say the answer is? For now, I'm concerned with the technically right answer. This is why I'm working from the armchair instead of waiting for more data. Because I think the question I'm asking (when properly framed) is pretty clear cut, I think philosophers should be able to answer it without needing data.

With the above two clarifications in mind, let me ask again:

Do you (Thomas, Eddy, Paul) agree that determinism A can accomodate compatibilist agents, just as determinism B can? In other words, do you agree that whether the first state of the universe "had to happen" is irrelevant to whether fw exists?

I'm asking you, as philosophers; I'm not asking what the folk might say (for whatever reason).

I guess what is motivating me is that Eddy's comments hint or suggest that:

NOA: if determinism A is true, we can't have compatibilist free will.

NOA is false, as I've tried to explain at length. I'm just looking to get you guys to agree with me that it's false. This is a crucial point that can't be lightly dismissed.

Whether NOA is true or false has a big implication for the criticism of Nichols and Knobe (N+K). If NOA is true, then the criticism of N+K is *much* stronger. But, as I've tried to show, it's false.

[Paul Torek]

Kip,

I guess I am unclear what (A) implies or what might motivate it. Determinism as discussed in science cannot motivate it, I think. Perhaps some theological views could motivate it. Perhaps an Ontological Argument combined with a Best Of All Possible Worlds type approach.

There's a lot of tension between "had to happen" and "reasons-responsive". If I am reasons-responsive, this provides a straightforward sense (or so it seems to me) in which my action didn't have to happen. Avoidance of harm is a leading factor in the evolution of our big reasons-responsive brains; that doesn't fit in with inevitability.

[tnadelhoffer]

What I keeping trying to explain to you is that your struggles to understand the difference between fatalism and determinism and whether either one is compatible with free will and moral responsibility is entirely orthogonal to the issue about the problematic (or non-problematic, as the case may be) use of the "had to happen" language in the Nichols and Knobe studies. So, if you simply drop the latter issue--which is thoroughly empirical--then the conversation won't be so difficult.

Keeping that in mind, I will now say a bit about the rest of your worries. For instance, you suggested that you don't know either what fatalism is (or at least you are unsure what Eddy and I mean by fatalism). But I just posted both what Eddy and I have said in print about what we take the difference between fatalism and determinism to be. The whole point is that there is no sense in which an agent could have done otherwise in a fatalistic universe. There is a sense in which an agent could have done otherwise in a deterministic universe (namely, the coulda/woulda conditional kind). Now, of course it is an open question whether this later kind of conditional could have done otherwise is enough to ground free will and moral responsibility (I happen to think that it is not), but it is nevertheless the case that there some minimal sense in which agents could have done otherwise in a deterministic universe but not in a fatalistic universe.

I can't be any less evasive than that :)

Having said that, I want to make sure you're clear about the worry Eddy and I have about Nichols and Knobe's studies:

They conclude that people are natural theoretical incompatibilists who may nevertheless give compatibilist-friendly answers provided enough negative affect is induced by the vignette. They are only entitled to this conclusion if the incompatibilist answers they get in the low affect cases are being driven by intuitions about determinism rather than fatalism. After all, everyone ought to be incompatibilist about fatalism and free will and moral responsibility since if fatalism is true, there is no sense whatsoever in which agents could ever have done other than they did. For instance, fatalism even rules out reasons responsiveness--which requires a minimal sort of conditional "coulda/woulda" done otherwise.

Does that help?

[Kip]

Paul,

By "had to happen" I mean only and just what Eddy, and the critics of N+K, mean when they say that determinism doesn't imply that the first state had to happen.

To help clarify, consider this:

In addition to all of the laws of nature, there is a meta-law (ML) that states: the first state of the universe will be, and can only be, X.

"Determinism" as we usually use the term is not committed to ML. For all we know, the Big Bang could have produced X, Y, Z, etc.

But suppose the Big Bang couldn't produce Y or Z. Suppose it had to be X. Suppose, in addition to determinism, that ML is true. Would that disturb your compatibilist intuitions? It shouldn't, right?

Compatibilists (generally) believe they have the power to do otherwise. They (generally) believe they can change or steer their future. But all of these beliefs are based on certain structures or mechanisms: reasons-responsive mechanisms, organized desires and meta-desires, etc. They are not based on whether the first state had to happen. Even if it did have to happen, even if every state of the universe had to happen, compatibilist free will could still exist. (And, therefore, if N+K's questions suggested that the first state had to happen, this would appear to be a harmless error.)

[Kip]

Thomas,

You say "I can't be any less evasive than that" and ask me if your generous post helps. Directly answering the following question will help tremendously.

Do you agree with the following:

NOA: if determinism A is true, we can't have compatibilist free will.

By determinism A, I only mean determinism in which the first state of the universe "had to happen". I do not mean True Fatalism (I distinguish between True and Apparent Fatalism above).

[tnadelhoffer]

I am a free will skeptic, Kip, as should be clear from anything I have ever posted on this blog. So, my answer is obviously no at first blush since you are asking me whether I think people can be free in a determined universe. Of course, if by "compatibilist free will" you simply mean "conscious control" or "reasons responsiveness" (i.e., if you're focusing on the "compatibilist" part rather than the "free will" part) then my answer is yes. I think there is a sense in which we could have both of these in a determined universe--whether it be of the A or the B variety. I just prefer not to call these kinds of capacities "free will." But that is a story for another day/post...

[Kip]

Thomas,

We're making some progress.

[BTW I know, and I'm glad, that you're a skeptic. That's why I used the qualifier "compatibilist" free will.]

Now you might be in a better position to appreciate my original (poorly articulated) point.

We agree that agents can still have compatibilist powers in deterministic A worlds. Deterministic A worlds are worlds in which, not only do the future states follow depending on the previous states, but the first state *had* to happen. So, we agree that the difference between determinism A and B *should* not (not *won't*) disturb any compatibilist's intuitions about free will.

My point is this: I read Eddy as saying that the difference between determinism A and B should make a difference. And, because it makes a difference, the N+K data is flawed.

Specifically, Eddy writes in footnote 8 of his critique:

"Determinism entails that BOX[(Po & L) ⊃ P]—i.e., necessarily, given the actual past state of affairs (Po) and the actual laws of nature (L), there is only one possible present state of affairs (P). But determinism does not entail (fatalism) that BOXP (or that BOXPo or BOXL)—i.e., that the actual state of affairs (or the actual past or laws) are necessary (could not be otherwise)."

Unless I am reading this passage (and Eddy and Paul's various other comments) incorrectly, Eddy is relying on precisely the sort of distinction that we just agreed *shouldn't* matter.

In other words, the defender of N+K should be able to read footnote 8 and say:

"So what? Suppose you are right, and the folk read our study to imply fatalism (as its logically defined in the footnote) instead of determinism (as its defined in the footnote). Nothing about this sort of fatalism prevents agents from having compatibilist free will. Agents in this sort of world, in which the first state of the universe (and the laws of nature), had to happen, could still be reasons-responsive. They could still have an organized hierarchy of desires and meta-desires. Nothing about fatalism, as you've defined it, prevents agents from having any of these compatibilist powers. So, even if we made an error, and allowed the folk to read fatalism into the question, instead of determinism, as we've defined it, this error *should* be harmless."

Do you agree?

[tnadelhoffer]

Kip,

I am in a rush, so this will have to be short. I wasn't claiming the "had to happen" language evokes what you are calling determinism A rather than determinism B. It's that it evokes fatalism! Look back at the way the modal operators are working here. Even if the original state of the universe "had to happen" as determinism A suggests--and I am still unclear what that would be like if no gods were involved--it is still not true that the laws are necessarily the way that they are. Think of David Lewis' analysis of free will and determinism here.

If you go on to assume that the laws are themselves necessarily the way they are in addition to determinism A, then you just get fatalism (or, at least, so it seems to me). And, of course, if fatalism were true, then everything that has ever happened had to happen by necessity full-stop. In this kind of universe, it is unclear what sense it would make to say that an agent was free and responsible. Or so me thinks.

Your own question about determinism A and determinism B is an interesting one in its own right (even if I still think your determinism A may just be fatalism in disguise for the reasons I mentioned above), but I am still unsure why it is more than tangentially related to the modal scope worry Eddy and I have raised about the Nichols and Knobe studies.

[Kip]

Thomas,

Suppose that, in addition to the first state, the laws were also necessary ("had to happen").

How does this change your previous statement, which is that agents can have compatibilist powers in such worlds?

For example, what about the first state and the laws being necessary prevents an agent from being reasons-responsive? Prevents an agent from having integrated desires and meta-desires? Prevents an agent having the compatibilist conditions of your choice?

Nothing, as far as I can tell.

"Fatalism" is scary word, sure. But, the way we've defined it here (and in Eddy's article), it seems to me *perfectly* consistent with agents still having compatibilist free will. And, because it is consistent with compatibilist free will, the fact that N+K's questions suggested fatalism, should be a mostly harmless error.

Here's another way to think about this point: compatibilists do not believe that compatibilist agents have free will *in virtue* of the first state being contingent, or *in virtue of* of the laws being contingent. Compatibilism doesn't care how the first state came to be the way it was, or how the laws came to be the way they were. Compatibilists tend to only look at the agent's powers and capacities (e.g. reasons-responsiveness), and these remain undisturbed in the fatalistic worlds we define above.

[tnadelhoffer]

Kip,

I fly home to Atlanta tomorrow and I still have a lot to do before then, so I shouldn't really be playing along anymore. But since I still think you are confusing things when it comes to the worries Eddy and I have about the Nichols and Knobe studies, I at least want to take one more pass at this.

It makes no difference whether you or I think that fatalism as we have defined it rules out free will and moral responsibility. The question when it comes to the N&K studies is whether the participants were incompatibilists about determinism and fw/mr rather than incompatibilists about fatalism and fw/mr.

You begged this off earlier by saying you wanted to know what I, qua philosopher, thought--not what N & K's participants thought (and why). Yet now you're back to trying to draw inferences from what I think about the issues under discussion to what I think about how these issues may have played out in the minds of the participants in N &K's studies. These two issues can be entirely decoupled. Having a commitment with respect to the underlying theoretical issues doesn't commit me to anything when it comes to how the folk understood (or misunderstood, as the case may be) the "had to happen" language in the N &K vignettes. All that matters to me is that given the way the vignettes were worded, we simply don't know whether their participants were fatalism incompatibilists or determinism incompatibilists (or both). I am unsure how to make this any clearer...

Setting that aside, I want to say one more thing about the underlying theoretical issue. You ask: "How does this change your previous statement, which is that agents can have compatibilist powers in such worlds?" It makes a difference because if everything that has ever happened was necessary full-stop (and not just conditionally, based on the past, the physical state, and the laws), then there is no sense in which things could have been otherwise (conditionally or not). I actually think Paul's way of fleshing this out earlier bears repeating. He said:

"There's a lot of tension between "had to happen" and "reasons-responsive". If I am reasons-responsive, this provides a straightforward sense (or so it seems to me) in which my action didn't have to happen. Avoidance of harm is a leading factor in the evolution of our big reasons-responsive brains; that doesn't fit in with inevitability."

If determinism A or determinism B is true, then it is (at least) arguably false that everything is inevitable full-stop. If, on the other hand, fatalism is true, there is only the inevitable. It is hard to see how even a Frankfurtian compatibilist can squeeze fw or mr into a world governed by that kind of inevitability (since I am not a compatibilist, it's not my job, thankfully). To see why, try to develop a coherent Frankfurt case while at the same time maintaining that everything that happens is entirely inevitable. In short, there is no room for possibility of any kind in a fatalistic world. Indeed, modally speaking, you might even think the truth of fatalism rules out the possibility of possibility (and hence of possible worlds). There is only the given. It is inevitable and unchangeable with no logical space for different unfoldings. If you would like to try to show that we can be free and responsible in such a world, have at it. I, for one, think it is a project that is doomed (in the non-fatalistic sense!) to fail. But I am well-aware that those more able-minded that me may think otherwise...

OK, that's it for now. I need to be grading and not blogging. I am sure these issues will come up again in a future thread. I look forward to arguing with you again when that time comes. For now, I am bowing out.

[Cihan]

Kip,

You may be right about meta-desires - so I don't know, maybe Harry Frankfurt would be a fatalism compatibilist, too. I don't know.

However, some compatibilists depend on counterfactuals when defining their control conditions. For instance, if I'm not mistaken, Fischer asks whether the mechanism that lead to action was reasons-responsive and suggests a test that involves counterfactuals to determine whether a particular mechanism was reasons-responsive.

More traditional compatibilists ask, for instance, whether the agent could do otherwise if she had wanted otherwise. In a fatalistic world, she couldn't do otherwise even if she had wanted otherwise - just ignore the question whether she could have wanted otherwise in a deterministic world.

With that said, here is a worry I have. Imagine two worlds that involve exactly the same events. One is deterministic and the other is fatalistic. It seems counterintuitive to me that in one world, agents are responsible and in the other they aren't.

[Kip]

Thomas,

I understand you can't play along any more. Since you've written such a long response, I'll try to make a last effort to address your concerns.

Let me summarize your view, as I've understood it so far.

You agreed with me that the first state "having to happen" wouldn't disturb compatibilism.

But, you add: the laws of nature didn't have to happen. And if they did have to happen, then fatalism would be true, and so we cannot have even compatibilist free will. In other words, you grant my point about the past, but you draw the line at the laws of nature---that would be going too far into fatalist territory (which you see as per se ruling out even compatibilist free will).

Why? You write:

"It makes a difference because if everything that has ever happened was necessary full-stop (and not just conditionally, based on the past, the physical state, and the laws), then there is no sense in which things could have been otherwise (conditionally or not)."

To see why I disagree, consider your typical Frankfurt style agent, with integrated desires and meta-desires. Let's call him Joe. Begin by considering Joe in a deterministic B world. We can imagine that Joe wants to drink a beer, and wants to want to drink a beer, and so goes to the fridge and gets a beer.

Now make a little change: Joe lives in a determistic A world instead of a deterministic B world. Does this make a difference? As you already agreed above, it doesn't. When Joe wants the beer, he goes to get one, etc., etc.

Now make another little change: Joe lives in a deterministic A world in which the laws of nature "had to happen." Back at the Big Bang, the laws of nature (including the laws of thermodynamics, the law of gravity, etc., with all of their particular values), had to be just the way they are. They couldn't have been different.

Now ask yourself: has anything changed about Joe's free will? Has anything changed about Joe at all? No. Nothing has changed. Joe has the exact same powers he had before.

So, it seems to me that whether the laws had to happen is irrelevant too. I understand that you disagree, and have to get to your work, but I wanted to explain my view. And I hope the Joe example gets my point across.

How does all of this relate to N+K? I take the criticism to be the following:

1. Define fatalism as the conjunction of the following: A. determinism, B. the laws of nature had to happen, and C. the past had to happen. Let's call this fatalism*.
2. Because N+K use the "had to happen" language, they imply fatalism* and not just determinism
3. Thus, their responses really test fatalism*-compatibilism, and not free will compatibilism.
4. Free will is incompatible with fatalism*, as even compatibilists agree
5. But free will is compatible with mere determinism: A without B and C.
6. So the fact that people report that free will is incompatible with fatalism* is not a mark against compatibilism, because the folk might be rationally justified in still asserting free will compatibilism (that is, fatalism*-incompatibilism doesn't imply free will incompatibilism)

I've been attacking 4-6. What I've been trying to show, and what the Joe example tries to show, is that the difference between fatalism* and determinism is irrelevant to compatibilist powers. Joe still has the same powers in fatalistic* worlds, as in mere deterministic world. If Joe is a Watsonian agent, his action might reflect his values; if he is a Frankfurtian agent, his actions may reflect his organized desires and meta-desires; if he a Fischerian agent, he might own his moderately reasons-responsive mechanism. Nothing about fatalism* (as defined above, and as discussed in Eddy's paper) strips these compatibilist agents of their powers. Joe still wants to get the beer, his still wants to want to get the beer, he still gets up to get it, gets it and drinks it. Nothing changes.

And because nothing changes, 4-6 above are not quite right.

But maybe I've made a fundamental misunderstanding somewhere...

[Kip]

Cihan,

I don't know if Frankfurt would be a fatalist compatibilist. Maybe the possibility never occurred to him. I like to think that, if compatibilists realized that their compatibilist powers were consistent with fatalism (at least the type described above), they would abandon their compatibilism. But I won't hold my breath.

As for Fischer: he defines reasons-responsive in terms of counter-factuals. But I don't think he commits himself to these counter-factuals having been actually possible. Rather, he seems to consider *conceptual* counter-factuals: what would Joe have done, if something had been slightly different, nevermind that nothing could have been slightly different. I could be completely wrong about that (Fischer is the better judge of what Fischer's view actually is).

Here's another way to think of Fischer's view in this context: take Joe's RR mechanism in a deterministic B world. Define it in terms of the genuinely possible counter-factuals in that world. Then remove the mechanism from that world, and insert it into Joe' in a fatalistic* world (just like the previous world, but now the first state and laws of nature "had to happen").

Do we have any principled reason for thinking that Joe can be morally responsible for his actions, but Joe' can't? I certainly hope not. Joe' has the exact same mechanism as Joe, and has the exact same powers as Joe had. Whether the past and laws of nature could have been different is completely irrelevant.

You conclude:

"With that said, here is a worry I have. Imagine two worlds that involve exactly the same events. One is deterministic and the other is fatalistic. It seems counterintuitive to me that in one world, agents are responsible and in the other they aren't."

Exactly. I agree 100%. Fatalism* doesn't matter. The critics of N+K imply that it matters. And to that extent, and only that extent, I think their criticism is not as strong as otherwise supposed.

[Cihan]

Kip,

You write that,

Rather, he seems to consider *conceptual* counter-factuals: what would Joe have done, if something had been slightly different, nevermind that nothing could have been slightly different.

I think fatalism is such a strong condition that it affects what you call *conceptual* counter-factuals (I'm not so sure how these differ from common-garden-variety counterfactuals) as well. In a fatalistic world, if the event A is "fated" to occur, "if something had been slightly different", A would still occur - period. Fatalism is that strong a condition. Think of Oedipus. Even if everything in Oedipus' life were different, he would still marry his mother and kill his father. I think the claim is that it's not conceptually possible that Oedipus does not marry his mother and kill his father - call this ~O.

Of course, one natural question is that if it's not conceptually possible that ~0, how can we imagine ~O? I don't know. So maybe I'm wrong. Maybe you are imagining some other character that's close to O.

Let me ask you. What are the notions of possibility at play? I think there are two notions of possibility (and I think you think there are three) - conceptual possibility and physical possibility.

If the first event of Big Bang "had to happen", it would have to be true across all possible worlds - hence all possible worlds would have to have the first event of Big Bang, which just seems weird.

[tnadelhoffer]

Since (a) it is 2:15am and I am finally done grading (hooray for me), and (b)I still have a bit of unspent energy, I wanted to briefly elaborate upon Cihan's remark that:

"If the first event of Big Bang "had to happen", it would have to be true across all possible worlds - hence all possible worlds would have to have the first event of Big Bang, which just seems weird."

Right. And it becomes even more bizarre when you add that whatever laws that govern this first event are necessarily the way they are as well. Why would that be especially weird (keep in mind that the mere fact that I can talk this way seems to assume fatalism is FALSE)? Because then there cannot be POSSIBLE worlds at all. There can only be this actual world with its necessary beginning and its necessary laws. As such, all "possible" worlds collapse into precisely the world we inhabit. In this world--the only possible world by definition!--nothing could have been different even in principle (i.e., all is inevitable). As such, we cannot make counter-factual claims about what agents would have done had things been different since it is necessarily the case that nothing could have been different. To suggest otherwise is like suggesting that it could be false that "A therefore A."

p.s. Kip, you still appear to be misunderstanding my worry about the N&K studies. Whether Eddy feels the same way is for him to decide! The issue is simply whether the number of participants who had so-called incompatibilist intuitions would have been less had the "had to happen" language been dropped for alternate ways of cashing out determinism that don't invite the conflation of determinism with fatalism. Of course, this, too, assumes the falsity of fatalism--but that is an ironic aside! OK, now I really am bowing out. Until the next post/round!

[Brian Parks]

Kip,

"and what the Joe example tries to show, is that the difference between fatalism* and determinism is irrelevant to compatibilist powers."

It seems that the point that Thomas and Eddy are trying to make is that the *folk* believe that there is a free-will-relevant difference between fatalism and determinism. In other words, if you prime fatalism, you get no free will, but if you prime determinism, you get yes free will.

Personally, I would find that to be extremely puzzling. I don't know of anyone who actually thinks that my freedom or my responsibility for my actions hinges on whether or not this universe "could have had" different dynamical laws or constants, or on whether the universe could have been in a slightly different state at the big bang. If that is in fact what the folk believe, then philosophers should not be going to the folk for guidance on this issue.

(Actually, philosophers shouldn't be going to the folk for guidance on *any* issue. The whole concept of using interviews with people off the street as an argumentive crutch in favor a specific philosophical position is just ridiculous IMHO. Of course, if the point of experimental philosophy is to learn more about this species of ours, rather than to provide support for a specific philosophical claim, then that's a different story. But I do get the impression that compatibilist experimental philosophers desperately *want* the folk to be compatibilist so that they can use this fact as an argument in favor of compatibilism--again, ridiculous.)

Finally, a quick question for all:

Suppose our universe is deterministic, and not fatalistic, such that the statement "the laws of nature could have been otherwise" is true. How does reality manifest the truth of this statement? Where exactly does its could-have-been-otherwiseness lie? If it were to suddenly lose its could-have-been-otherwiseness, what exactly would that change entail?

It seems to me that there are events. The notion that those events "could have been otherwise" is a projection of human thought, not something that reality actually bothers with.

[Kip]

Responses for all:

Cihan wrote:

"Of course, one natural question is that if it's not conceptually possible that ~0, how can we imagine ~O? I don't know. So maybe I'm wrong."

Right. We can imagine it happening. That's all we need to make Fischer's system work. Another reason why I think Fischer would agree: as I've argued at length, agents in fatalistic* worlds don't have any more power than in deterministic B worlds. So I like to think he would reach the correct answer (in my view). But he remains the better judge.

"Let me ask you. What are the notions of possibility at play? I think there are two notions of possibility (and I think you think there are three) - conceptual possibility and physical possibility."

I agree. In a fatalistic* world, there are conceptual possibilities but no physical possibilities (other than the actual one).

"If the first event of Big Bang "had to happen", it would have to be true across all possible worlds - hence all possible worlds would have to have the first event of Big Bang, which just seems weird."

It is a weird idea. Thomas goes on at length about how weird it is. But:

1. I didn't bring up the idea, the critics of N+K did. Read footnote 8 of Eddy's critique (above). The footnote defines fatalism in precisely the weird way I've been discussing. If the critics are allowed to raise this weird idea, and say that it rules out compatibilist free will, defenders of N+K should be allowed to point out that: no, actually it doesn't.
2. My point is that agents in fatalistic* worlds can still exercise compatibilist powers. The fact that fatalism* is a *weird* idea isn't a sufficient rebuttal to that argument. I would just say: "Yes, it's weird, but those weird worlds can still have compatibilist agents."

Thomas wrote:

"The issue is simply whether the number of participants who had so-called incompatibilist intuitions would have been less had the "had to happen" language been dropped for alternate ways of cashing out determinism that don't invite the conflation of determinism with fatalism."

I understand that this is one (simple) characterization of the issue. I've tried to spell out the issue in more detail, by recasting the critics' argument in 1-6 above. I think that, when my comments are read in light of 1-6, their relevance becomes more apparent. If you think that 1-6 mis-characterize the critics' argument, I'd be very interested to know why (then I might agree that my point is irrelevant). But so far I have no reason to think I've misunderstood their argument.

Brian wrote:

"It seems that the point that Thomas and Eddy are trying to make is that the *folk* believe that there is a free-will-relevant difference between fatalism and determinism. In other words, if you prime fatalism, you get no free will, but if you prime determinism, you get yes free will."

That is not the only point they're making. They also assert (in general, if not in the four corners of their published work) that the folk are *justified* in believing that fatalism* conflicts with compatibilist free will. To see this, just look at Thomas's post above. He refuses to concede that fatalistic* worlds could accommodate compatibilist agents. I read Eddy's comments (and his published critique) the same way. They insist: not only would the folk likely believe that fatalism* makes a difference for compatibilism, but they are right to do so.

"Personally, I would find that to be extremely puzzling."

It is extremely puzzling! That's why I've written more comments in this thread, in the last 24 hours, than in perhaps all other threads for the last 3 months combined. If compatibilists themselves can't/won't recognize that traditional compatibilist powers are consistent with fatalism* (as described in Eddy's paper), this is a big deal.

"Actually, philosophers shouldn't be going to the folk for guidance on *any* issue."

In general, I agree with you. But I think there is a huge exception: when we are trying to define a term. In my view, terms are defined, ultimately, by common usage. At least, common usage has an important role to play in defining terms. Thus, it's appropriate to poll the folk to determine what common usage actually is. The question of compatibilism is just the question of whose definition of free will is right. The folk can help us answer that (I suspect that the answer we'll get is that there is no consensus about how "free will" is defined).

[tnadelhoffer]

Brian,

IMHO, it is ridiculous to tilt at strawmen when you could simply read the articles the compatibilist experimental philosophers have written (god, this is a broken record). Therein you would find very careful attempts to motivate the relevance of the data for first order philosophical theorizing. Of course, you may not agree with the arguments, but addressing them would be more useful than thumbing your nose at work you obviously haven't read. FWIW, I am NOT a compatibilist experimental philosopher (I am a skeptic) but I do think compatibilism comes naturally to the folk and I also think this is relevant to the free will debate. If you disagree, look at the arguments we have put forward and address them head on.

Kip,

Now that Brian has lured me back into the mix, I might as well take one more stab at your line of reasoning. For starters, it makes no difference whether Eddy or I are justified in believing that fatalism rules out compatibilist free will and moral responsibility when it comes to our objection to the conclusion N & K draw from their data set. Nor does it make a difference whether fatalism really does rule out free will and moral responsibility. That you keep suggesting it is relevant demonstrates that you have, in fact, misunderstood our worry (or at least my worry). Presumably, since it's my worry, I am in a decent position to know whether you seem to be understanding it correctly. Now, perhaps there are other worries you have captured quite well with your six step argument--but mine is not one of them (presumably, Eddy agrees--we'll have to wait and see).

All that matters as far as my worry is concerned is that the "had to happen" language used in their scenarios may have led people to think of fatalism rather than determinism. Let's call this the Fatalism not Determinism Worry. Given that this worry is legitimate--and I have been at pains to argue that it is (that's the whole modal scope concern mentioned above)--it raises a second worry. Let's call this the Fatalism Incompatibilist yet Determinism Compatibilist Worry. According to this worry, it is possible--indeed, I believe likely--that some of the the participants who were led to think of fatalism rather than determinism happen to agree with Eddy and I that fatalism rules out free will and moral responsibility even though they nevertheless would think that determinism (properly described) is compatible with free will and moral responsibility. Whether they're right about the purported incompatibility of fatalism and determinism is irrelevant to the question of whether they believe in both fatalism incomatibilism and determinism compatibilism. For participants who have this mixed view, you cannot draw conclusions about whether they are determinism incompatibilists based on their responses to the "had to happen" language since you may simply be getting their responses to the fatalism incompatibilism issue.

And since (a) there is no way of knowing whether some of N & K's participants have this mixed view (it would require more studies), and (b) it is quite possible that some of them do, you cannot draw any firm conclusions about folk determinism incompatibilism based on the data set of N & K. Just because you don't agree with Eddy and I that fatalism should make a difference, it doesn't follow that it doesn't make a difference in the minds of the participants.

That is as clear as I can make my worry about their data. If there are other worries--which there surely are--they are for others to motivate and defend. I am only committed to this one--and it is one you have not yet adequately addressed. Indeed, I am confident the only way you could address it would involve more studies. It is not the kind of claim that can be dismissed a priori.

[Brian Parks]

"That is not the only point they're making. They also assert (in general, if not in the four corners of their published work) that the folk are *justified* in believing that fatalism* conflicts with compatibilist free will."

I think that's Eddy's claim, but I don't think that's Thomas's claim.

I think Thomas's claim is more experimental psychological: that the N&K results would be significantly different if fatalism were not primed in the phrasing. This claim implies that the folk see a free-will-moral-responsibility-relevant difference between fatalism and determinism, which further implies that the folk believe that my free will and my moral responsibility hinge on whether the laws of physics "could have been different" (!?) Obviously not a flattering portrait of the folk.

As for the question of the meaning of "free will", I take a different view.
Free will is a term. The fact that it is used in different senses does not have to be a problem as long as the user is clear about the sense in which she is using it.

To someone working in chemistry, the term H means enthalpy. To someone working in physics, the term H means Hamiltonian operator. Notice that chemists and physicists get along just fine. So why do compatibilist philosophers need to fight with social psychologists? Compatibilist philosophers know more or less what social psychologists mean when they use the term "free will." Why can't things be left at that?

The dispute about "free will" does not have to be a difficult dispute. There is a simple recipe for resolving it. Just precisely and rigorously define what you mean by the term, and then assess whether its referent actually describes human cognitive structures.

So Eddy can say: "I define the term free will as follows. An individual has free will to the extent that her behavior is reasons-responsive. Some individuals have free will." Agreed.

A social psychologist can say: "I define the term free will as follows. An individual has free will to the extent that she can trump the physical processes that occur in her brain. No individual has free will." Agreed.

We stop there. Where is the disagreement?

The disagreement is actually normative, though it's hidden in a poor choice of words. What the two parties really disagree about is on the *kind* of free will that would be necessary to render an agent morally responsible, i.e., the *kind* of free will that would have to exist in order for some person to really, truly, ultimately deserve praise or blame for some fact.

My take on this dispute is in line with Tamler's. Reality does not bother to resolve it. The question it asks is not well-defined.

Reality does not have norms, it has facts. So to ask what norm reality imposes in this or that situation is to ask gibberish.

Norms are entirely a consequence of the way that our minds construct reality. They exist only in the feelings, emotions, sensitivities, and so on that we experience when we make contact with the world. They do not exist in mind-independent entities.

There is no such thing as the property of "badness", "or goodness", or "justifiedness", or "deservingness", or "should-be-punished-ness", or "should-be-rewardedness", or whatever.

Does anyone disagree? Does anyone honestly think that those terms describe actual features of reality?

[Kip]

Thomas,

I understand your argument (in this thread) as being:

1. The folk might be fatalism*-incompatibilists while also being determinism-compatibilists. Let's call this asymmetry.
2. They might be this way, regardless of whether they are *right* or *justified* to be this way. That is, the asymmetry could be rational or irrational.
3. Eddy and I's conclusion is limited to the possibility of asymmetry. We're silent about whether the asymmetry is rational.

And, so, you go on: please, Kip, stop implying that the rationality of the assymmetry is important to the critique of N+K. It's not. We're merely noting that there could be an asymmetry.

So there are two issues:

A. Whether the asymmetry is rational/correct.
B. If it is irrational/incorrect, does this weaken Eddy and your critique of N+K.

I'm interested in both A and B. A alone is interesting enough to merit thorough discussion, as I think you would agree. So, even if you are right about B, we could continue discussing these issues in the context of A.

You insist, though, that the rationality of the asymmetry is irrelevant to the critique. Having reviewed the relevant passages in detail, I agree with you: the rationality of asymmetry is not so important. What is important, however, is the intuitiveness of the asymmetry.

The following passage shows that Eddy is relying on the intuitiveness of the asymmetry:

"Fatalism, coercion, reductionism and epiphenomenalism are each threatening—at least intuitively—to free will and responsibility, but it would take an impressive argument to show that determinism entails any one of them. The reason each of these views are threatening is not because the past and laws are sufficient for our behavior but because they suggest that our behavior is caused by forces that bypass our conscious mental life—or at best, by forces that manipulate our conscious mental life.12"

But, as I've tried to argue, the asymmetry between fatalism* and determinism is *not* intuitive. And, as the comments in this thread show, there are reasons for thinking that it isn't intuitive.

For example, Brian Parks wrote (February 19, 2009 at 11:28 PM), "Personally, I would find that to be extremely puzzling."

Similarly, Cihan wrote (February 19, 2009 at 08:57 PM): "It seems counterintuitive to me that in one world, agents are responsible and in the other they aren't."

So it seems as if: the only reason for you and Eddy to expect the folk to answer differently, is the alleged intuitiveness of the asymmetry. That is the one leg your expectation stands on. If we can knock out that leg, it is not clear what the expectation stands on.

That is, if we can show that the asymmetry is *not* intuitive (as myself, Brian and Cihan have expressed above), then I don't see why we would expect the folk to answer differently. We could just as well expect them to give the same incompatibilist-friendly answers they always gave.

[A caveat: I am only talking about expectations here. I agree that the *best* way to settle the question is to collect more data. I don't think that can prevent us from making expectations, though.]

I hope this explains my reasoning regarding the N+K studies.

[Brian Parks]

Well I'm glad I managed to lure you back in Thomas ;-)

"FWIW, I am NOT a compatibilist experimental philosopher (I am a skeptic)"

And I never said you were.

"But I do think compatibilism comes naturally to the folk and I also think this is relevant to the free will debate."

I've read most of Eddy's papers on EP and free will. (BTW, yours are not available because the URL links on your Dickinson page are broken.)

One of the ways in which Eddy argues for the relevance of EP to the free will debate is in terms of the standard incompatibilist argument. He points out that this argument rests on the claim that "most people are intuitively incompatibilist", or in less careful phrasing, that "incompatibilism is more intuitive." Obviously, if EP were to show that most people are intuitively compatibilist, then that result would be relevant to a key premise in the standard incompatibilist argument, and therefore relevant to the free will debate in general. Point taken.

Now, does the relevance of EP to the free will debate extend beyond the sense discussed above? If so, how?

"According to this worry, it is possible--indeed, I believe likely--that some of the the participants who were led to think of fatalism rather than determinism happen to agree with Eddy and I that fatalism rules out free will and moral responsibility even though they nevertheless would think that determinism (properly described) is compatible with free will and moral responsibility."

So in other words, they believe that my free will and moral responsibility hinge on whether or not the laws of physics could have been different?

As in: If F=ma were the macroscopic law necessarily, then my free will and MR would be impossible. But because F=ma is only the macroscopic law contingently, that is, because the different macroscopic law F=mv could have taken its place, therefore my free will and MR are possible.

Seriously, Thomas. If the folk actually believe that, then props to you and Eddy for the diagnosis. I wouldn't know what to say, other than what I said in the previous post... "Who cares?"

[Cihan]

Kip,

I think there are subtle modal metaphysical issues here that are interesting in their own right.

I agree. In a fatalistic* world, there are conceptual possibilities but no physical possibilities (other than the actual one).

I would have to disagree. In a fatalistic* world, there are no *conceptual possibilities* (I'm still vague on what you mean by *conceptual possibilities*) and no physical possibilities. Everything that happens happens necessarily - or so I think.

In a deterministic world, there are "conceptual possibilities" but no physical possibilities - at least, this is what I think.

Let me ask you something else. Is it *conceptually* possible that there can be zombies who are physically, behaviorally identical to us but nonetheless lack consciousness? Is it *metaphysically possible* (I'm just probing for a third notion of possibility)? Is it physically possible?

(I answer no to all these questions.) My point with Oedipus was that no, you can't imagine Oedipus not marrying his mother and not killing his father - just as in the case of trying to imagine philosophical zombies. You can think you are but you really aren't.

I haven't gotten to all the other comments - I will write some more if and when I plow through them.

Brian, Daniel Dennett in a recent lecture, argued as follows: Each of your molecule isn't alive but you are. Or no single molecule of you is conscious/perceiving/feeling but you are. The idea is that free will might be something like that. Are all these other properties (aliveness, perception, feeling/emotion) just fictions projected by the mind too?

[tnadelhoffer]

Brian,

Sorry for the snarkiness, but when you use words like "desperate" and "ridiculous" to describe the work by people like Eddy--who is painstaikinglg careful both about the kinds of claims he makes about folk intuitions and the ways he motivated these claims--it suggests that you either haven't actually read his work or you have read it and have opted nevertheless to be particularly uncharitable. Moreover, nothing raises the hackles of an experimental philosopher more than the claim that we are trying to make the following move: The folk intuitively think that x, therefore x. This is a move that Eddy and I explciitly deny we are making. So, objecting to our work on the grounds that we are making this move is misplaced and unmotivated. Now perhaps you have someone else in mind. I'd be curious to know who it is...

[Brian Parks]

Cihan,

You've posed some really good questions. Let me give a long-winded elaboration of my view, and then I'll address your questions ;-)

I think we all agree that human behavior has causes. We can describe these causes in loose mental terms or in tight physical terms.

In loose mental terms, when you make a choice, you choose based on the psychological forces (PFs) present in your mind--your thoughts, feelings, emotions, beliefs, desires, sensitivities, motivations, and so on.

Put bluntly, if I could control all of your PFs, I could get you to do whatever I wanted you to do. You would become my toy.

The only real obstacles to my control over you would be the random features of your behavior. But those features are not relevant to the matter at hand, because they are random, and therefore not even you have control over them.

(On a partisan side note, notice that nature already controls all of your PFs. So you are already effectively nature's toy. To put the point in more philosophically attractive terms, you are a part of nature. Your behavior is ultimately an expression of *its* causality.)

Of course, we can translate these loose mental terms into more concrete neuroscientific terms. You choose what you choose because of the neurobiological processes that occur in your brain. You choose to pursue your love interest because the image of her face, when processed by your CNS, activates a powerful dopaminergic reward circuit in your VTA... you get the picture. If I were to shut down that circuit with a dopamine antagonist, you would not pursue your love interest. She would become a big "bla, whatever, who cares..." (this has been confirmed in testing btw)

Now, let's supposed that we've identified all of the causes of human behavior. We know what aspects of human behavior are determined by causes, and what aspects, if any, are purely random.

What more needs to be said? What else does a truth-seeker need to look for? I contend nothing.

The matter is complete. It's done, it's over, we've explained everything that there is to explain. There is no need for any more debate.

The questions,

"But should I feel guilty for my past behavior, given the fact that I am not ultimately responsible for its causes?"

"But should I feel proud of my achievements, given the fact that I could not have not achieved them?"

"But should we punish the criminal, given that we would have committed the same crime that he commited had we been thrust into his internal and external circumstances?"

...are not answerable questions. They are examples of projections, examples of the mind putting its own emotions, feelings, norms, values, and so on onto a reality that does not have them, trying to extract things from that reality that are not there in the reality.

Think for a moment. Suppose we have two deterministic worlds A and B in identical physical states. In A, moral responsibility exists. In B, moral responsibility does not exist. How does A manifest the existence of moral responsibility? What does it do *differently* than B to make it true that "Joe is morally responsible for his actions"? ...???

If you ponder this question, you will see that the distinction is empty. The difference between world A and world B is not an actual difference in the two worlds. All it can be is a difference in how our minds perceive each world. In A, we perceive the world as colored by notions of praise, blame, desert, responsibiilty, and so on--"whose fault is all this?." In B we put such talk aside as nonsense.

So, in my view, this whole debate is essentially misguided. It's like asking "What color is the apple?" But the apple doesn't have a color! Your mind adds the color.

"Are all these other properties (aliveness, perception, feeling/emotion) just fictions projected by the mind too?"

Perceptions, feelings and emotions exist, and they have intrinsic qualities. I do not dispute that.

As for aliveness, that's a trickier issue.

I would describe myself as a conceptual relativist in the tradition of Putnam.

You can take a partition of the universe in some spacetime, and you can give it a name, say "Cihan", and a set of well-defined spatio-temporal boundaries. If you precisely define the meaning of "alive", then you can say "Cihan is alive." Reality will take your side (or not take your side) on that claim.

But the mereology that gave me "Cihan" and "the rest of the universe" is not a privileged mereology. I can carve reality out into "Cihan" and "the rest of the universe", but I can also carve reality out in other ways. No way of carving out reality is THE way that reality actually exists, just like no inertial frame of reference in physics is THE frame in which things happen.

[Paul Torek]

Kip,

For what it's worth, I agree with Cihan: In a fatalistic* world, there are no conceptual possibilities. Which wreaks havoc on the counterfactuals of reasons-responsiveness.

Brian,

Chalk me down as one who believes that "justification" describes actual features of reality. Unlike most moral skeptics and relativists, I'm an externalisthttp://michaeldowney.net/moral-internalism-vs-moral-externalism/. I also think that color terms actually refer. I guess I'm just a metaphysical slut :)

[Kip]

Paul,

My general claim is that agents in fatalistic* worlds can have just as much freedom and control as agents in merely determistic (deterministic A) worlds. I stand by that 100%.

I agree that there are subtle issues about how to analyze counterfactuals with respect to fatalistic* worlds. But I don't think this defeats my general claim above.

For example, in my [February 19, 2009 at 09:36 PM] post, I considered defining a RR mechanism in deterministic A world, and then inserting the mechanism into an agent in a fatalistic* world. Because the mechanism is exactly the same, the fatalistic* agent has the same powers and capacities. The fact that the first state and the laws "had to happen" doesn't limit the agents powers or freedom at all.

This is all because compatibilist agents have their freedom in virtue of their structure (e.g. RR mechanisms) and not in virtue of any contingency of the laws of nature or first state of the universe. To suggest otherwise is bizarre (as seen in Brian Parks's explicit example: "As in: If F=ma were the macroscopic law necessarily, then my free will and MR would be impossible. But because F=ma is only the macroscopic law contingently, that is, because the different macroscopic law F=mv could have taken its place, therefore my free will and MR are possible.").

I'd be curious to know what you, as a compatibilist, think of this line of argument?

[Paul Torek]

Kip,

I think you're playing too fast and loose with possible worlds. If fatalism* is true at any world, then that is the only possible world and this world is that world. On the other side of the coin, if fatalism* is false (as I believe) then it's necessarily false. So, as I see it, your [February 19, 2009 at 09:36 PM] thought experiment is caught in a contradiction.

I suppose I should have complained earlier, but I hadn't thought things through very far.

[Tom Clark]

The way I see it, there are two sorts of necessity in play. One is deterministic relative necessity in which the actual past and laws necessitate (determine) the actual outcome, expressed in language that Eddy approves of in his paper:

"I recognize that Nichols and Knobe’s scenarios do not make exactly this modal error: when describing Universe A, they write, “given the past, each decision has to happen the way that it does” (I’ve [boldfaced] the crucial phrase)."

Eddy boldfaces the phrase because for him it's crucial to avoid the implication that a decision has to happen the way it does, *period*.

This is the other global, non-relative sort of necessity, in which no event or law could have been other than it was. This is what Eddy defines in his footnote #8 as fatalism, and what in this thread Kip has called fatalism*.

The question (or one question anyway) is whether compatibilist agent freedoms and powers of the sort that underwrite ascriptions of MR are undercut by global necessity but not relative necessity. Kip's point (I think) is that global necessity leaves untouched the same *completely (let us stipulate) deterministic relations* between events in agents and the world, such as reasons responsiveness, guidance control, etc. that exist under a strictly relative necessity *given the past*, which of course is what we're always given in the real world. So why, Kip wonders, does Eddy think that these powers disappear under fatalism*?

I agree with Kip. Why does e.g., reasons-responsiveness, which (let's stipulate) is deterministic, depend for its efficacy (and therefore on its capacity to underwrite MR) on a certain type of global contingency attached to the past or laws or anything else? It seems to me that Eddy's sort of fatalism (Kip’s fatalism*) leaves compatibilist powers untouched. What would undermine them is the sort of fatalism that says: necessarily, agent X will do Y *no matter what* X does (e.g., considering various courses of action, or not) antecedent to Y.

[Brian Boeninger]

Tom says:
"It seems to me that Eddy's sort of fatalism (Kip’s fatalism*) leaves compatibilist powers untouched. What would undermine them is the sort of fatalism that says: necessarily, agent X will do Y *no matter what* X does (e.g., considering various courses of action, or not) antecedent to Y."

Kip - I wonder if Eddy and Thomas couldn't reply to your worry (about their take on the N+K study) as follows:

It's possible that the folk understand the phrase "has to happen" in a way that suggests fatalism**, where fatalism** is something like the thesis Tom describes above (agent X will do Y "no matter what" X does, or tries to do). Notice that fatalism** is neither equivalent to, nor is entailed by, your fatalism*.

It doesn't seem relevant to the question "what do the folk think about FW, MR, and determinism?" whether understanding the phrase "has to happen" (even "given the past") as implying fatalism** is a reasonable, justified, or rational interpretation on the part of the folk.

Nor is it relevant to that question whether there are good arguments for the conclusion that fatalism* threatens compatibilist powers, or whether positing a relevant difference between determinism and fatalism* is intuitive (perhaps the folk haven't thought through their intuitions very carefully on these matters).

Nor is it relevant to that question whether fatalism**, as a thesis, is even coherent (indeed - the fact, if it is a fact, that the folk perceive the thesis to be incoherent might go some way toward explaining why they take fatalism** to be incompatible with FW/MR).

So it seems that Eddy and Thomas could, if they wanted to, insist that the folk (mis)understood the N+K prompt in a way that implies fatalism** (though I grant that Eddy's footnote does not seem to describe fatalism**), and if so, this would cast doubt on whether the data accurately reflect the folks' beliefs about compatibilism.

Is there some reason to think that no portion of the folk in the N+K study understood the prompt as suggesting something like fatalism**, rather than either determinism or fatalism*? (I take it as fairly obvious that the fact that the folk perhaps should not have understood the prompt in manner X does not at all imply that they did not understood it in manner X.)

[Kip]

Tom,

You're 100% right. I could kill to have put the idea as eloquently as you have.

Two additional points:

1. For all we know, we live in a fatalistic* world. Presumably, compatibilists do not want to hinge their freedom upon this possibility, just as they do not want to hinge it upon the possibility of mere determinism (e.g. Fischer).

2. It may be that Eddy meant that the "had to" language suggested another kind of fatalism, a more disturbing kind. For example, regardless of whether I throw the brick through the window, the window always breaks. Call this True Fatalism. In that case, the explication of fatalism* in footnote 8 is an arguably minor error (an error of presentation, not substance).

But I think it's a telling one.

Suppose Eddy meant True Fatalism and not fatalism*. It's less clear that the "had to" language suggests True Fatalism than fatalism*. In other words, True Fatalism is even more of a jump, from the explicit text of the question, than fatalism* is.

True Fatalism is a very strange thing. We would want a pretty good reason for expecting the folk to read True Fatalism into the question, when none of the facts seem to otherwise suggest it. It's not clear that the mere use of "had to happen" suggests fatalism*, much less True Fatalism.

Of course, maybe Eddy and Thomas are right: maybe the folk read "had to happen" to mean True Fatalism, maybe they are compatibilists about determinism and incompatibilists about True Fatalism, and so maybe N+K is flawed for suggesting otherwise. Maybe. The data will show.

What I've tried to show is: based on the apparent error in footnote 8 (regarding True Fatalism versus fatalism*), the critique is not as strong as it would otherwise be. And, regardless of the relevance of the N+K study, it is important for compatibilists to acknowledge that their powers are consistent, not just with determinism, but also with fatalism*---a non-trivial definition of "fatalism" (indeed, the definition of Eddy's choice). As I wrote above, for all we know, we currently live in just such a fatalistic* world.

[Kip]

Brain B.,

Thanks for your helpful comment. My previous comment addresses many of your points.

Let me try to directly answer a few:

1. Yes, Eddy and Thomas may have meant True Fatalism (what you call fatalism**). Of course, as you note, footnote 8 doesn't seem to suggest True Fatalism, and that alone is worth pointing out.

2. You ask for some reasons why the folk wouldn't have read N+K to mean True Fatalism. There is no decisive proof of that. Only future experiments will show. I offer the following only as slight suggestions.

This is the description of determinism in N+K:

"Imagine a universe (Universe A) in which everything that happens is completely caused by whatever happened before it. This is true from the very beginning of the universe, so what happened in the beginning of the universe caused what happened next, and so on right up until the present. For example one day John decided to have French Fries at lunch. Like everything else, this decision was completely caused by what happened before it. So, if everything in this universe was exactly the same up until John made his decision, then it had to happen that John would decide to have French Fries."

There are some things to note about this stimulus:

First. It suggests that an *ordinary* action happens, and not that anything strange happens. John decides to have the French Fries, and then he does. The stimulus doesn't state that anything bizarre happens, like John eating the french fries, even if he chooses not to eat them.

Second. The stimulus says that "what happened in the beginning of the universe caused what happened next, and so on right up until the present." This suggests a chain of events without gaps. State 5 depends on state 4, which depends on state 3, and so on.

In contrast, True Fatalism suggests that there are gaps: State 5 happens *regardless* of state 4. State 4 could be anything, state 5 doesn't care. Whether I throw the brick or don't throw the brick, the window still breaks. That's a bizarre, and additional, constraint, that doesn't seem to be suggested by the stimulus.

Further, consider two claims:

A. The folk misread N+K to mean True Fatalism.
B. The folk misread N+K to mean fatalism*.

To me, A is more of a reach than B. A seems to be a reach, for at least the reasons suggested above. In contrast, fatalism* is prefectly familiar: for all we know we live in a fatalistic* world.

Last, it's not clear to me that the defender of N+K has the burden of explaining why the folk wouldn't have read True Fatalism into the stimulus. Rather, the first question seems to be: why would they? The mere fact that the question says the events had to happen is not very convincing (at least to me). But I have to grant, of course, that this is possible. [Indeed, I genuinely believe that all sorts of subtle and irrelevant changes can dramatically alter folk responses.]

[Al Mele]

Thomas asked for references. Chris Zarpentine, my research assistant, kindly went through the bibliography for my new book, *Effective Intentions* (to be published next month by Oxford UP) and put together the following collection of science references (free will & closely related matters) from 2000 on.

Best wishes,
Al Mele

Baer, J. 2008. “Free Will Requires Determinism.” In J. Baer, J. Kaufman, and R. Baumeister, eds. Are We Free? Psychology and Free Will. New York: Oxford University Press.
Bandura, A. 2008. “Reconstrual of ‘Free Will’ from the Agentic Perspective of Social Cognitive Theory.” In J. Baer, J. Kaufman, and R. Baumeister, eds. Are We Free? Psychology and Free Will. New York: Oxford University Press.
Banks, W. 2006. “Does Consciousness Cause Misbehavior?” In S. Pockett, W. Banks, and S. Gallagher, eds. Does Consciousness Cause Behavior? An Investigation of the Nature of Volition. Cambridge, Mass.: MIT Press.
Bargh, J., and M. Ferguson. 2000. “Beyond Behaviorism: On the Automaticity of Higher Mental Processes.” Psychological Bulletin 126: 925–45.
Baumeister, R. 2008. “Free Will, Consciousness, and Cultural Animals.” In J. Baer, J. Kaufman, and R. Baumeister, eds. Are We Free? Psychology and Free Will. New York: Oxford University Press.
Brass, M,, and P. Haggard. 2007. “To Do or Not to Do: The Neural Signature of Self-Control.” Journal of Neuroscience 27: 9141–45.
Caldara, R., M. P. Deiber, C. Andrey, C. Michel, et al. 2004. “Actual and Mental Motor Preparation and Execution: A Spatiotemporal ERP Study.” Experimental Brain Research 159: 389–99.
Dweck, C., and D. Molden. 2008. “Self-Theories: The Construction of Free Will.” In J. Baer, J. Kaufman, and R. Baumeister, eds. Are We Free? Psychology and Free Will. New York: Oxford University Press.
Ehrsson, H. H., S. Geyer, and E. Naito. 2003. “Imagery of Voluntary Movement of Fingers, Toes, and Tongue Activates Corresponding Body-Part-Specific Motor Representations.” Journal of Neurophysiology 90: 3304–16.
Fisher, C. M. 2001. “If There Were No Free Will.” Medical Hypotheses 56: 364–66.
Haggard, P. 2005. “Conscious Intention and Motor Cognition.” Trends in Cognitive Sciences 9: 290–95.
———. 2006. “Conscious Intention and the Sense of Agency.” In N. Sebanz and W. Prinz, eds., Disorders of Volition. Cambridge, Mass.: MIT Press.
Haggard, P., and S. Clark. 2003. “Intentional Action: Conscious Experience and Neural Prediction.” Consciousness and Cognition 12: 695–707.
Haggard, P., and B. Libet. 2001. “Conscious Intention and Brain Activity.” Journal of Consciousness Studies 8: 47–63.
Hallett, M. 2007. “Volitional Control of Movement: The Physiology of Free Will.” Clinical Neurophysiology 118: 1179–92.
Howard, G. 2008. “Whose Will? How Free?” In J. Baer, J. Kaufman, and R. Baumeister, eds. Are We Free? Psychology and Free Will. New York: Oxford University Press.
Jankelowitz, S., and J. Colebatch. 2002. “Movement Related Potentials Associated with Self-Paced, Cued and Imagined Arm Movements.” Experimental Brain Research 147: 98–107.
Kilner, J., C. Vargas, S. Duval, S. Blakemore, and A. Sirigu. 2004. “Motor Activation Prior to Observation of a Predicted Movement.” Nature Neuroscience 7: 1299–301.
Lau, H., R. Rogers, and R. Passingham. 2007. “Manipulating the Experienced Onset of Intention after Action Execution.” Journal of Cognitive Neuroscience 19: 81–90.
Lau, H., R. Rogers, N. Ramnani, and R. Passingham. 2004. “Willed Action and Attention to the Selection of Action.” Neuroimage 21: 1407–14.
Libet, B. 2001. “Consciousness, Free Action and the Brain.” Journal of Consciousness Studies 8: 59–65.
———. 2002. “The Timing of Mental Events: Libet’s Experimental Findings and Their Implications.” Consciousness and Cognition 11: 291–99.
———. 2004. Mind Time. Cambridge, Mass.: Harvard University Press.
Marcel, A. 2003. “The Sense of Agency: Awareness and Ownership of Action.” In J. Roessler and N. Eilan, eds., Agency and Self-Awareness. Oxford: Clarendon Press.
Passingham, R., and H. Lau. 2006. “Free Choice and the Human Brain.” In S. Pockett, W. Banks, and S. Gallagher, eds., Does Consciousness Cause Behavior? An Investigation of the Nature of Volition. Cambridge, Mass.: MIT Press.
Pockett, S. 2006. “The Neuroscience of Movement.” In S. Pockett, W. Banks, and S. Gallagher, eds., Does Consciousness Cause Behavior? An Investigation of the Nature of Volition. Cambridge, Mass.: MIT Press.
Prinz, W. 2003. “How Do We Know about Our Own Actions?” In S. Maasen, W. Prinz, and G. Roth, eds., Voluntary Action. Oxford: Oxford University Press.
Ramachandran, V. 2004. A Brief Tour of Human Consciousness. New York: Pi Press.
Roediger, H., M. Goode, and F. Zaromb. 2008. “Free Will and the Control of Action.” In J. Baer, J. Kaufman, and R. Baumeister, eds., Are We Free? Psychology and Free Will. New York: Oxford University Press.
Shariff, A., J. Schooler, and K. Vohs. 2008. “The Hazards of Claiming to Have Solved the Hard Problem of Free Will.” In J. Baer, J. Kaufman, and R. Baumeister, eds., Are We Free? Psychology and Free Will. New York: Oxford University Press.
van de Grind, W. 2002. “Physical, Neural, and Mental Timing.” Consciousness and Cognition 11: 241–64.
Vohs, K., and J. Schooler. 2008. “The Value of Believing in Free Will: Encouraging a Belief in Determinism Increases Cheating.” Psychological Science 19: 49–54.
Wegner, D. 2002. The Illusion of Conscious Will. Cambridge, Mass.: MIT Press.
———. 2004a. “Frequently Asked Questions about Conscious Will.” Behavioral and Brain Sciences 27: 679–88.
———. 2004b. “Précis of The Illusion of Conscious Will.” Behavioral and Brain Sciences 27: 649–59.
———. 2008. “Self Is Magic.” In J. Baer, J. Kaufman, and R. Baumeister, eds., Are We Free? Psychology and Free Will. New York: Oxford University Press.