Blog Coordinator

« JOB: postdoc in philosophy (TU Delft, The Netherlands) for the "Enhancing Responsibility" project | Main | Philosophers Fight Back »

08/16/2011

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

I don't see how this line of reasoning doesn't end up sawing away the branch that it sits on.

Either "experience is (generally) veridical" or it is not. If it is, then we have conclusive reason to believe that at least some experiences of free will are veridical.

If experience is, as you seem to say, NOT veridical, then the entire edifice of empirical science is almost certainly invalidated, based as it is on sensory experience.

We need a much more subtle way of talking about the non-veridicality of introspection than to simply call into question our powers of observation.

Right, Nick. We need to be finer-grained. Whether there is some principled way to divide up sources of evidence into those that are reliable and those that are not, I'm not sure. But it is clear that (a) the senses are pretty reliable under most conditions and (b) the content of the experiences at issue - the experience of acting, deciding, choosing, and so on - are not acquired through these channels. Some philosophers think there is an inner sense. If there is, then we might need either to get finer grained still, or deny that the inner sense has the kind of reliability that the other senses have. I'm claiming that libertarians can't show that the content of the experience of freedom is acquired by a reliable channel. I am saying that they have done nothing to show that it is not acquired by channels that we have good reason to think are unreliable, and that there is a prima facie case for thinking that acquisition is via an unreliable channel.

Nick raises an interesting challenge, but I think it can be met. I think a distinction can be drawn between the role of experience in science and the appeal to experience by libertarians in a way that generally supports Neil's point without damning the whole of empirical science.

An individual relying only on sense experience (unaided by measuring sticks or other devices) might be unreliable as a detector of object size, but nonetheless, operating in consort with other investigators and aided by theory and instrumentation, object sizes can be reliably ascertained. This overcoming of individual sensory unreliability is achievable in part because there are multiple converging lines of evidence for the phenomenon at hand (as well as a theory that can guide in the calibration of instruments and the correction of individual divergences).

The above sort of thing is what keeps empirical science objective, not some reliable connection of the form: sense-seems that P therefore P. But the above sort of thing doesn't seem to be what the libertarians are appealing to in the quotes that Neil presents. They seem instead from making a move from it introspectively seeming that P to P, a move that the stuff Schwitzgebel collects helps cast into doubt. Indeed, one might argue that there's nothing else that they *could* appeal to, nothing sufficiently analogous to the multiple lines of converging evidence that grounds the ultimate reliability of our size judgments.

Great post Neil! I think a couple of clarifications are needed before we can assess the force of your worries. The first is that there are different possible avenues of support for premise 2. One can appeal to introspective evidence, but one can also base it within broader epistemological framework. One might, for example, adopt phenomenal conservativism or some broadly Reidian common-sense approach. So the case for two does not stand or fall with the reliability of introspection. The second clarification concerns the content of the belief one is attempting to justify. Premise 2 makes an assertion about free will that is theory neutral. That is, it does not claim that we satisfy either libertarianism or compatibilism, but simply that we have free will. I would suggest that the content of this belief (“that we have free will”) be interpreted as something like, “that we possess the kind and degree of control to warrant praise and blame towards others and ourselves for our choices and actions” (this is obviously rough but I just wanted to give a flavor for what I was thinking). When interpreted this way libertarians and (many) compatibilists will endorse it. We begin with the obviousness of free will, it is part of the data, and then we seek to analyze it. We gain support that the universe is indeterministic only if we have support for incompatibilism.

There are two steps. The first is to assess what evidence we have for premise 2 and this project is non-partisan (at least among non-skeptics). The second step is to assess which theory offers the best analysis of free will. It is only when both steps are combined that we might gain evidence for indeterminism. Note that nothing about this process suggests that science is irrelevant. Suppose we come to think we have good evidence for 2 and incompatibilism. Science might still defeat this evidence. Certainly the evidence we get for 1 and 2 will be defeasible and science is one source of possible defeat. But are scientific theories not also defeasible?

Perhaps this is non-scientific, but what is unscientific about it? Compare: We have free will. If we have free will, then reasons are appropriately involved in the causal etiology of free choices. Would you also say this is unscientific? If so, then how does it differ from libertarianism? If not, then why do so many philosophers think only libertarians founders on the scientific front?

I also have a suspicion that much of what drives issues about determinism and indeterminism in the science is partly driven by philosophical assumptions. Many assume that the strength of a theory is a function of the degree of uncertainty it has concerning predictions: the less uncertainty the better. This assumption directly clashes with a libertarian conception of the universe under which we can be assured that even our best theories will have a degree of residual uncertainty. So it’s unclear to me that a defender of libertarian ought to just sit back and submit to what the physicists say since part of what might be driving the physicist theory are controversial philosophical assumptions, some of which libertarians may reject. The idea that there is some clear line between physics and philosophy concerning the issue of indeterminism, such that philosophers who make claims about indeterminism clearly transgress it strike me as a fiction.

Finally, I am curious what body of brain science libertarians are ignoring? I am not an expert but I know of a decent bit of evidence concerning the workings of the brain that is quite consonant with libertarian friendly prediction that the brain is indeterministic.

Neil,

I would rearrange your argument slightly.

1) Physics shows the universe to be indeterministic (at least at the level of small particles, with the macrocosmos being "adequately determined" since it averages over large numbers of particles).

2) The standard argument aqainst free will says that we cannot be responsible for our actions, if they involve randomness.

3) As you note, Campbell, Lehrer, Searle, and many others observe - we think we have genuinely open alternative possibilities for action.

Since William James in the 1880's based his "two-stage model" of free will on Darwinian evolution, which involves objective chance generating alternative possibilities (mutated genes), we can say that our thoughts appear to be free for similar reasons. They just "present themselves" as James said, they pop into our heads, apparently at random.

Now this randomness does in no way make our willed actions themselves random, if our actions result from a careful evaluation and selection in the second stage, based on our reasons, motives, feelings, and desires, which "adequately determine" our will.

In this "two-stage model," our will is compatible with the determinism of our actions by our reasons, etc. It should satisfy compatibilists and determinists in that regard?

But it also should satisfy libertarians, since those actions are in no way pre-determined by the past and the laws of nature from the moments just before our brain/mind begins to think about and generate genuinely indeterministic thoughts and alternative possibilities. Right?

I am glad that you mention Bob Kane. He and I have agreed that we can integrate his "Self-Forming Actions" into my two-stage model.

It goes like this. Normally, the two-stage model generates multiple alternative possibilities in the first stage, which are narrowed down in the second stage, to a single "self-determined" action.

I like to say that it is our thoughts that are free, our actions are willed.

The two-stage model for "free will" separates "free" from "will."

But there are times when we cannot narrow down our alternatives, when two or more options remain. Kane calls this a case of a "torn decision." Each alternative has excellent reasons. Kane introduces indeterminism at this moment, but he says the decision was not caused by the randomness per se, but by the agent's efforts in support of the final decision, whichever that may be. As a result, Kane has cleverly invalidated the argument that we are not responsible when randomness is involved in a decision.

Note the subtle difference between Kane and myself. My two-stage model uses indeterminism up front to explain not only freedom but creativity, the generation of genuinely new ideas in the universe.

Kane's injects indeterminism at the end of the process to resolve torn decisions. His use of indeterminacy is just as valid as mine. We both provide a break in the causal chain of predeterminism.

So Kane and I have agreed to offer a combined model of libertarian free will, one that we feel is the most practical and plausible current model, one fully consistent with a limited quantum indeterminacy.

Please see chapter 24 on Robert Kane in my book, and chapter 28 on a new more comprehensive compatibilism - comprehensive because it is compatible with both determinism and indeterminism, as well as with biological evolution. PDFs of these chapters are online here.

http://www.informationphilosopher.com/books/scandal/

Chris, that's an interesting response. I'm no friend of compatibilism either, but the partners in crime move is still a powerful one since I meant to claim that libertarianism is especially extravagant here. I'm not convinced that I can't show that though.  First, we need to distinguish different libertarian claims regarding  2. One possible claim is the common ground one you make: that we have free will. But many libertarians make a different one: that we experience alternative possibilities. That claim is not common ground. Second, I'm not sure the appeal to common sense is in any better shape ( given the psychological evidence) then the appeal to experience. Finally, some of the considerations you mention about how evidence assessment is holistic seem to support me. The move from 2 - understood as a common ground claim - to appropriate indeterminism is such a bizarre one (" hang on,  can I really do physics in the armchair?") that it ought to make the person tempted by it to rethink their commitment to 1. The move to the idea that reasons are causes is not similarly bizarre. 

I didn't mean to endorse the criticism of libertarian neuroscience I mention, though I doubt that the evidence available is compatible with anything really libertarian friendly.

Another great post Neil.

Certainly the easy conceivability of logically possible alternatives of action that are close-world relevant to our situations aids the libertarian here, but not in ways ultimately that finally and reliably back that view. I walked out of the classroom left; I easily could have gone right. So it seems. Nothing about my going left appears to dilute my sense of counterfactual ability that I could have gone right, despite the fact that I only have evidence about left-going-ability in that instance, but at least there were no opportunity barriers to right-going. I'm unconcerned about other logical possibilities such as walking through the opposite wall because they are not close-world relevant to either my abilities or opportunities, since I've no evidence in my past I'm a "4-D Man" and walls generally seem to stop any previously drunken attempts to prove that I was.

But the fact I have gone right on occasions in the past seems to show that I could have even though I didn't in this case. Right?

Except all of this is consistent with a being who (i) is aware of not just what I do, but what I have done, and (ii) is mentally equipped well enough to compare similar situations in terms of matches and differences and (iii) yet accomplishes everything in (i) and (ii) through deterministic mind/brain processes. But if the fact that (iii) is true is not something available to me in carrying through (i) and (ii) stuff (how could it be introspectively?), then my ongoing functioning in terms of (i) and (ii) might well yield error about my consideration of (iii) as a possible fact: I take my (i) and (ii) generation of close-world logical possibilities to constitute sufficent grounds for the denial of (iii). Essentially I take the complex contents of my mind as significant for how it works (or doesn't work). That is to confuse logical and epistemic possibility with real possibility. And I do not see how introspection as in (i) and (ii) can possibly show that this account (i)-(iii) is wrong. So the libertarian starts out with the resources of (i) and (ii) to deny (iii), but they finally can't do that.

C. A. Campbell at least had a strategy here: there is introspective phenomenological evidence and external scientific evidence, and they are evidentially exclusive. All the biological evidence about the deterministic brain we can muster does nothing to undermine our inner sense that we can choose. Of course, this seems (to me) to bring metaphysical dualism to the rescue by positing equivalent dual realms of epistemology. And that offends my entia non sunt etc. sense.

Hi Alan,

My argument kicks in when a libertarian accepts your parsimony claims. They can push back against it by arguing that there is no external scientific evidence that the brain is deterministic (as Mark Balaguer argues). As I said to Chris, I doubt the available evidence is really helpful, because there is no reason to think, say, that the brain is more a locus of indeterminism than are (say) computers, or that the indeterminism is located and constrained appropriately, but there is also no evidence that the kind of process Kane posits *doesn't* occur. So the libertarian can say that they are not multiplying entities unnecessarily. My claim is only that there are prima facie reasons to think that the experiential evidence many cite is unlikely to be reliable (by the way, these kinds of claims are certainly still made. I attended a series of seminars in London a couple of years ago run by Tom Pink, in which he claimed repeatedly that one could be a compatibilist only by denying the unshakeable experience of libertarian freedom). Of course there may be other reasons to think that (2) is true that do not turn on experience. You might think that deliberation requires commitment to (2), or you may have theological reasons to accept it. My sense is that the strategy cited is a common one, but I may be wrong.

I agree Neil that libertarians don't argue from a position of strength--as you know I think that libertarians mount a rear-guard "lam" retreat strategy at best. No current brain-science favors indeterminism (that I'm aware of at least, but I'm open to being schooled on that; Penrose-style speculation should not count). A lot of brain-science is put in terms of mechanism, even if much of that is in terms of probabilistic potentials, which I take it is entirely compatible with an underlying determinism of the neural net. The introspective/phenomenological data is completely inconclusive with respect to incompatibilism (as I argued above). So what's to love about libertarianism except that it dovetails with moral/religious world-views that require it as a necessary condition of their satisfaction?

But as you say, Kane, van Inwagen, etc. get major cred from me for their insights and even more their honesty about the possible shortcomings of their positions.

Consider this slightly altered version of a John Searle claim: “Reflect very carefully on the character of the experiences you have as you roll a die. You will sense the possibility of alternative outcomes built into these experiences…. that the die could land in any of six ways right here and now, that is, all other conditions remaining the same. This, I submit, is the source of our own unshakable conviction of indeterminism” (1984, p. 95). Or this alteration of C.A. Campbell: “Everyone must make the introspective experiment for himself: but I may perhaps venture to report … that I cannot help believing that it remains open, quite absolutely, which of two genuinely open possibilities occur when I flip the coin” (1951, p. 463).

Can someone please tell me what it is about the phenomenology of choice that provides any more evidence (or reason to believe) that the world (or the brain) is indeterministic than the phenomenology of rolling a die or flipping a coin provides for indeterminism? I entirely agree that there are differences between the cases regarding our experience of being active, or even ‘agent causal’ in some sense, but I am at a loss to see how our phenomenology has any content that suggests indeterminism (i.e., alternative possibilities holding fixed everything). And if we think about die rolls and coin flips a bit, we are likely to think that different outcomes are not possible given identical conditions—rather, some small differences, unknown to us, will account for different outcomes. So, once we “go theoretical,” it seems determinism is the more commonsensical response for dice and coins. Other than an incompatibilist argument, what would motivate us to adopt an indeterminist theory for humans based on the phenomenology?

Unrelated point: it seems incompatibilists attracted by manipulation arguments can’t put too much stock in the “phenomenological evidence” since those arguments require that our phenomenology would be the exact same if we were manipulated (or determined) to choose what we choose.

Eddy,

About the manipulation point: A libertarian who is fond of appealing to the phenomenology of action in defense of free will can (and almost certainly will) adopt phenomenal conservativism. According to this sort of libertarian, since we have no reason to think that we are being manipulated then we should accept that our experience of acting freely (I'm assuming for the sake of argument that we have that experience) is vertical.

Bob,

What is the mechanism distinct to "selection" that distinguishes it from a purely random process? what logical argument do you use to support the conclusion that "self-determination" does not entail the completion of a vicious, infinite regress?

Neil,

"What reason do we have for believing that (2) is true?" — indeed. More longwindedly, given that libertarians have failed to articulate the mechanism by which the "libertarian description of the phenomenology of action" can be distinguished from mere randomness, in what sense are we justified in supposing that libertarianism is logically coherent, much less "veridical"?

I wonder what it would be like if our experience of acting freely were "vertical." In any event, in my post above I meant veridical.

Ipad or iphone, Justin? Autocorrect is (causally) responsible for some of my howlers. Wish I could blame them all on it.

Anthony,

In my two-stage model, the first "free" stage is marked by significant randomness, which Bob Kane and I agree is traceable (in part) to the same quantum indeterminacy that drives biological evolution (cosmic rays damaging DNA, etc.).

In the second "will" stage, the alternative possibilities generated in the free stage are evaluated (according to character and values, for reasons, motives, feelings, desires, etc.) by a process that is "adequately determined."

Adequate determinism is the same determinism that leads to the Newtonian laws of motion. It is the result of averaging over vast numbers of microscopic particles, so as to produce all the regular Laws of Nature (which are completely consistent with the microscopic indeterminacy of quantum physics).

The determinism in my evaluation process is essentially the same as that wanted by compatibilists and determinists. Our actions are willed and self-determined.

But because we have access to quantum indeterminacy in our thoughts, our actions are not pre-determined by the "fixed past and laws of nature" at moments just before our considerations began.

Our thoughts are free. Our actions are willed.

Calling the will "free" is arguably to make it random. This is a conceptual error that has led to much confusion, as first pointed out by John Locke, who said

"I think the question is not proper, whether the will be free, but whether a man be free." (Essay Concerning Human Understanding, Book II, Chapter XXI, Of Power, s.21)

Neil,

Something very like your original argument is in fact maintained by two prominent mathematical physicists at Princeton.

John Conway (creator of the famous Game of Life and cellular automata) and his colleague Simon Kochen, who argued for something like Bell's Theorem years before Bell, formulated what they call the Free Will Theorem.

I treat it in chapter 15 (Physics of Free Will) of my book and on my Information Philosopher website here.
http://www.informationphilosopher.com/freedom/free_will_theorem.html
____________

Conway and Kochen assume three axioms, which they call "SPIN", "TWIN" and "FIN". The spin and twin axioms can be established by entanglement experiments. Fin is a consequence of relativity theory.

1. SPIN: Measuring the square of the component of spin of certain elementary particles of spin one, taken in three orthogonal directions, results in a permutation of (1,1,0).

2. TWIN: It is possible to "entangle" two elementary particles, and separate them by a significant distance, so that they give the same answers to corresponding questions. The squared spin results are the same if measured in parallel directions. If the first experimenter A (on Earth) performs a triple experiment for the frame (x, y, z), producing the result x → j, y → k, z → l while the second experimenter B (on Mars, at least 5 light minutes away) measures a single spin in direction w, then if w is one of x, y, z, its result is that w → j, k, or l, respectively.

3. FIN: There is a finite upper bound to the speed with which information can be effectively transmitted. Conway and Kochen say this is a consequence of "effective causality."

[But the collapse of the quantum mechanical probability amplitude wave function is instantaneous and not so limited. ]

Conway and Kochen's formal statement of the Free Will Theorem is then

If the choice of directions in which to perform spin 1 experiments is not a function of the information accessible to the experimenters, then the responses of the particles are equally not functions of the information accessible to them.

They then say:

Why do we call this result the Free Will theorem? It is usually tacitly assumed that experimenters have sufficient free will to choose the settings of their apparatus in a way that is not determined by past history. We make this assumption explicit precisely because our theorem deduces from it the more surprising fact that the particles’ responses are also not determined by past history.

Thus the theorem asserts that if experimenters have a certain property, then spin 1 particles have exactly the same property. Since this property for experimenters is an instance of what is usually called “free will,” we find it appropriate to use the same term also for particles.
___________

Conway and Kochen thus say human actions are not pre-determined by events just before their current decision process (van Inwagen and Fischer's "fixed past" and laws of nature), which is the same result that Bob Kane and I argue for in my two-stage model, augmented by his "Self-Forming Actions."

Note that "not pre-determined" does not mean completely random and undetermined, since the "will" stage of my model is adequately determined.

It is only our thoughts that are "free." Our actions are willed.

I have found that the idea of a "free choice" of experimenters goes back before Conway and Kochen to the 1935 Einstein, Podolsky, Rosen paradox and Niels Bohr's reaction to it.
http://www.informationphilosopher.com/solutions/experiments/EPR/

Cheers,

Bob

Bob, I think you're misreading my argument. As far as I can see nothing in Conway and Kochen turns on the phenomenology of choice.

Bob,

So, according to your account, "access to" quantum indeterminacy is indistinguishable from the outcomes of a random generator? why then use the word "free" when clearly the "will" is not "free" but is the result of random occurrences? also, why use the word "self-determination" when the concept is clearly self-contradictory, and therefore meaningless?

Neil,

Let me clarify how exactly free choice is the essential part of Conway and Kochen's Free Will Theorem, by framing it in terms of your original three steps.

1) If the experimenter, in a Bell Theorem test, is free to choose ("free will") which kind of measurement to make, then indeterminism must exit in the universe.

2) The experimenter is free to choose.

3) Therefore, the fundamental particles of the universe are indeterministic.

The experimenter would not be free if all events in the universe were pre-determined from the beginning of time, a view consistent with the Laplacian universe and with an omniscient God as creator. In both these cases the information in the universe is a conserved quantity. All times, past and present contain this information, which may exist outside of time and space in some noumenal sense.

But true indeterminism (objective chance) threatens responsibility according to the standard argument against free will - if determined, not free, if not determined, not responsible.

The main new idea in my two-stage model is to limit the indeterminism to the first stage of a process that creates alternative possibilities for action, a process that adds new information to the universe that simply did not exist before we, as co-creators of the universe, added it.

Then the second stage has a limited determinism. Actions are determinations according to the agent's reasons, motives, values, desires, etc. These are all the kinds of determination that traditional compatibilists and determinists said were needed for responsibility.

For centuries, the main attack on libertarian free will was against a will that was not determined in any way by the agent's character. It was unlimited metaphysical freedom.

The second stage of my model is thus compatibilist. My free will is not metaphysical, but biophysical. Organisms evolved in the presence of indeterministic chaos, which is still there in all our cells, including our neurons.

My two-stage model is compatible with a limited determinism in the second stage, a limited indeterminism in the first stage, and with biological evolution.

I ask you and other Flickerers to consider the model carefully. I am hopeful that it will be discussed as a possible explanation for free will in future philosophy courses, one that should not be summarily dismissed for logical or linguistic analysis reasons.

Please see http://www.informationphilosopher.com/books/scandal/

Back to your original observation that your argument is no way to do physics, I agree completely. Conway and Kochen should argue from the evidence of quantum physics to the existence of free will, not the other way around.

Cheers,

Bob

Bob, this still has nothing to do with my argument m, which concerned the PHENOMENOLOGY of free choice!

Anthony,

Let me try to explain why I see "self-determination" as a meaningful and appropriate term to describe my proposed two-stage model of free will.

1) If freedom from a completely deterministic causal chain of events since the beginning of the universe is to exist, somewhere and sometime there must exist indeterministic events.

2) If such indeterministic events were the direct cause of our actions - if as you say our actions were "the result of random occurrences," I would agree with you that "self-determination" would be self-contradictory.

3) But in my model the randomness just generates options for us to choose from. Many other options are our regular habits with no randomness, but the essence of freedom is these occasional creative new thoughts that are not pre-determined by our "fixed past and laws of nature."

4) If I used quantum randomness to generate a number of possible options for you, your decision to choose one of them would not be random if you carefully evaluated them and selected the one that is best according to your character and values, right? But then your choice would also not have been pre-determined from the beginning of time, can you agree?

5) Following R. E. Hobart*, and common language usage, I call this kind of limited but adequate determinism a "determination." But how can I attach it to the agent's "self?"

6) The issue of who deserves responsibility has been discussed by Daniel Dennett and Robert Kane. Dennett's Principle of Default Responsibility (initially, but mistakenly attributed to Al Mele, see Freedom Evolves, p. 281) is concisely described by Mele (Free Will and Luck, p. 177) as "If no one else is responsible for your being in state A, you are."

7) Since the origin of new ideas is often in our own minds, and since the determination process is also in there, I feel comfortable calling my free will model by the traditional term of self-determination.

8) To say the choice is adequately determined is to say the choice itself is neither random nor pre-determined, but self-determined.

9) In Robert Kane's Self-Forming Actions, the two-stage model alone cannot narrow down the options to one best choice, so multiple options remain in what Kane calls a "torn decision." Kane has shown that the decision can be indeterministic and yet the agent be fully responsible since there are equally good reasons, motives, etc., behind all the options. Thus Kane disproved the old idea that if a decision is made at random, the agent cannot be responsible.

10) As to the critical distinction between quantum indeterminacy and a pseudo-random generator, which Dan Dennett thinks is all that is needed for his two-stage Valerian model (and for biological evolution as an algorithmic process), please see the material I developed last Fall when Dennett invited me to participate in his graduate seminar on free will at Tufts.
http://www.informationphilosopher.com/solutions/philosophers/dennett/seminar/

In the final edition of my book**, Kane and I have combined our libertarian models of free will. We argue that this model meets many, if not all, of the requirements of compatibilists and determinists - except, of course logical and strict physical determinism of the Laplacian and theological kinds.
_____

* Many philosophers misquote the title of R. E. Hobart's classic 1934 paper in Mind as "Free Will as Involving Determinism and Inconceivable Without It." The word Hobart used is "Determination." Hobart, following William James in my opinion, does not deny the existence of chance in the universe.

** Anthony, if you (or any other FoF participant) would like a review copy of my book, send me your address - bobdoyle@informationphilosopher.com. And/or see the PDF at http://www.informationphilosopher.com/books/scandal/

Neil,

My apologies. Thanks for the strong emphasis.

On reviewing your post, I see that I paid little attention to your remarks on phenomenology per se, but focus instead on the main argument. You asked,

"why would anyone ever endorse the following argument?

(1) If we have free will, the universe is (appropriately) indeterministic.

(2) We have free will.

(3) The universe is (appropriately) undeterministic.

Whatever else one might think of this argument, surely – surely – this is a really bad way to do physics. Why would anyone think we can uncover the causal structure of the universe in this kind of way?

But maybe we can. Obviously, the heavy lifting is being done by premise (2) (reject that premise and you’re not a libertarian). We’re supposing that incompatibilism is true, so if premise (2) is also true, the conclusion follows. This would be a prima facie odd result, I think: it would remain odd that we can physics in this kind of way. But the argument seems sound."

I agree with you that this is no way to do physics, but found what seems to be this very argument made by prominent mathematical physicists. Just thought you might be interested in that.

But also, it seems a poor way to do philosophy to ask how people feel about their freedom.

As to phenomenology specifically, you said,

"I don’t much care what the phenomenology of action is (in this context), because I doubt very much that careful attention to this phenomenology can bear on premise (2). We have no reason to think that the content of our phenomenology can give us evidence about the causal structure of the universe, because we have no reason to think that the phenomenology is veridical."

I completely agree, and thus think that phenomonelogy per se can contribute very little. As other commentators said, individual sense experiences may or may not be veridical, but science and careful philosophy have developed very sound ways to decide between them.

Eddy Nahmias asked

"Can someone please tell me what it is about the phenomenology of choice that provides any more evidence (or reason to believe) that the world (or the brain) is indeterministic than the phenomenology of rolling a die or flipping a coin provides for indeterminism?"

The clear answer seems to be no more evidence. If we want more evidence, we must seek it elsewhere, which is what Bob Kane and I have been doing recently.

Eddy goes on,

"I am at a loss to see how our phenomenology has any content that suggests indeterminism (i.e., alternative possibilities holding fixed everything). And if we think about die rolls and coin flips a bit, we are likely to think that different outcomes are not possible given identical conditions — rather, some small differences, unknown to us, will account for different outcomes."

In my opinion, this goes to the heart of the free will problem. What is the source of those small differences, and can their randomness in any way invalidate our responsibility if we carefully evaluate our options and select the best one based on reasons, motives, feelings, etc., all supported by and adequately determined by our character and values?

Whether we are always fully conscious of all our reasons at the moment of choice, and whether we often confabulate new reasons to suit the situation (as Daniel Wegner has shown), and whether some semi-automatic decisions may be relegated to the subconscious, such as swinging our tennis racquets or flicking our wrists in a Libet experiment, these all may in the end be less important than showing that the origins of both true and false reasons are in our heads, that they are not pre-determined, and and thus are free acts of self-determination?

Neil,

Sorry for the tardy responses. “Partners in crime”? I was thinking more of “Partners in the pursuit of truth” :) But a couple of non-half-baked thoughts (or at least attempts thereof). First, I don’t see why appeal to an experience of alternative possibilities could not be advocated by a compatibilist. I am sure Joe or Kadri would take issue with your claim. But I take the point to be more broadly that sometimes libertarians make appeal to experience to directly show that we have a degree of control over our choices that is incompatible with determinism. I am (at least) sympathetic to your critique of these libertarians. But I would hasten to add that this objection to libertarianism is not an objection to libertarianism per se, but merely an objection to a particular argument for libertarianism, and thus you have not given us reason to think libertarianism is unscientific, but only a particular brand.

You claimed, “Finally, some of the considerations you mention about how evidence assessment is holistic seem to support me. The move from 2 - understood as a common ground claim - to appropriate indeterminism is such a bizarre one (" hang on, can I really do physics in the armchair?") that it ought to make the person tempted by it to rethink their commitment to 1. The move to the idea that reasons are causes is not similarly bizarre.” I find this claim incredibly revealing. I am no scientist so perhaps I am simply not appreciating the differences between physics and neurobiology (or whatever the level or kind of science is that is relevant for testing whether reasons-states are causes of action), but as far as I can tell there ought not to be a difference between the libertarian and compatibilist claims. Both are empirical claims and both are grounded partly in philosophical theorizing and partly in experience, or seemings, or common-sense. I think you (and I suspect the whole host of others who agree with you) owe libertarians an argument/explanation (at least if this claim is used as an attempt to ground an interesting asymmetry) and not merely an appeal to experience of the bizarreness of the libertarian claim. Why should one armchair science be allowed but not the other?

I think we need to be very careful in what we take the libertarian to be doing when he claims that the universe is indeterministic. First, it is not armchair science since libertarians’ claims are based in the empirical world (e.g. experience, seemings, etc.). Second, libertarians aren’t “doing” physics even when they make claims about the existence of indeterminism. Physics is conducted via specific methods and libertarians are just not doing this kind of work. Perhaps you just think that this exasperates the problem for libertarians, but it is not so clear to me. Libertarians are making claims about (i) how physics needs to turn out if it is to be compatible with our view of ourselves as free and morally responsible beings and (ii) that in the absence of reason to think otherwise we are justified in believing that physics will turn out this way. Saying that “the libertarian thinks he can do physics from the armchair” is at best unhelpful and at worst false. This is not meant to show that libertarians do not face something like the objection you’re running, I just want to make sure your objection is actually something they face.

Chris, thanks. Several things: I am claiming that those libertarians who appeal to this argument are doing physics from the armchair in that they are making claims about what the structure of the physical world must be, given their experience. Why aren't I evenhanded? Because compatibilists (of the non-Hobartian sort) are not doing physics from the armchair. Do compatibilists engage in similarly suspect armchair theorizing, and shouldn't they be called out for it? Sure: I've done some of that elsewhere. They don't make that mistake, but they make plenty of others (one example: Dennett, and following him several others, say that luck evens out across the course of a life. The evidence suggests this is false).

Your claim (1) is an incompatibilist claim, which I'm granting here. Making (1) is not, I agree, doing physics from the armchair: it is saying how physics must turn out if we are in fact free. But claim (2) is doing physics from the armchair. What else can it be? As you reconstruct it, it is saying that we are justified in thinking that the universe is (appropriatetely) indeterministic, on the basis of an experience (okay, maybe it is not doing physics from the armchair. Maybe its physics in the jacuzzi: I don't know what libertarians get up to).

I work these days in a neuroscience institute, but I continue to do free will as a largely a priori debate. I think that's fine. We're not going to solve these questions in the fMRI scanner. However we need to be aware of the range of empirical literature, because sometimes we make claims without seeing that they have empirical presuppositions, without imagining they could possibly be false, or using mechanisms (like the appeal to how things appear to us) which are known to be unreliable. When philosophers do these things - and again, I agree that the problem is not limited to one side or the other - they should expect and welcome being called out.

The comments to this entry are closed.

Books about Agency


3QD Prize 2014: Marcus Arvan