Blog Coordinator

« Books About Agency | Main | Books on Agency »

08/04/2014

Comments

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

Hi Marcus, really interesting thought experiment and questions. I hope to say more later, but for now three points:

1. Contextualism about free will has been discussed in articles by J Hawthorne, S Rieber, and W Sinnott-Armstrong (under label of contrastivism), perhaps others. I don't know if any of their presentations fit well with yours, but they're on the same track.

2. One feature you don't highlight from your thought experiment is that, presumably, we would experience choice and free will no different at all once we leave the matrix (not sure whether we'd experience it differently when we re-enter, but as you suggest, probably not for long). I think this suggests that whatever you've stipulated we add when we gain "libertarian FW" might not seem (or be) very important *relative to the deterministic law-governed* causation in the matrix. What does seem crucially different is that you've described the determinism in the matrix as complete manipulation by the creators, which is not the way the movie matrix is set up and which, I believe, introduces principled differences from determinism. So...

3. If we strip away the contentious claim that manipulation (complete control by the creators) is no different than determinism, it's not clear what we'd think when we come out of the matrix. I think we'd think what we should think if we accept the sort of view Chalmers presents in "Matrix as Metaphysics": "Huh, interesting, I didn't realize reality was made up of bits in a computer program [or was deterministic], but then again, I didn't know whether it was made up of atoms, or quarks, or strings either. In any case, I sure am glad that the creators made a nice consistent substructure that (for the most part) matches up nicely with (and subserves) my experiences and concepts of most things, such as friends, dogs, colors, joy, and even choices and space and time..." (And just think how much Neo and we would want to get back *into* the matrix if, when we got out, the 'reality' was a bizarre world we could not apply our prior concepts and experiences to!).

Sorry if I'm screwing with your thought experiment a bit. That's what happens when you bring in the Matrix, eh?

Hi Eddy: Thanks for your comment, and for drawing my attention to Hawthorne, Reiber, and Sinnott-Armstrong's work on this!

On (2), you write: "I think this suggests that whatever you've stipulated we add when we gain "libertarian FW" might not seem (or be) very important *relative to the deterministic law-governed* causation in the matrix."

I reply: it may or may not be important! It all depends on how we're thinking about morality and moral responsibility. For instance, suppose Kant is right and the Categorical Imperative applies to us if and only if we have libertarian free will (and, let's add, there's a way of making libertarianism consistent with *observed* determinism...hint, hint: my theory of Libertarian Compatibilism). If this is right, then the "ultimate" question--the addition of libertarian FW--matters a great deal, even relative to the deterministic reality of the simulation. The existence or non-existence of libertarian FW just might make all the difference in the world: it just might make the difference between the Categorical Imperative applying to us, or its not!

On (3), you write: "I think we'd think what we should think if we accept the sort of view Chalmers presents in "Matrix as Metaphysics": "Huh, interesting, I didn't realize reality was made up of bits in a computer program [or was deterministic], but then again, I didn't know whether it was made up of atoms, or quarks, or strings either. In any case, I sure am glad that the creators made a nice consistent substructure that (for the most part) matches up nicely with (and subserves) my experiences and concepts of most things, such as friends, dogs, colors, joy, and even choices and space and time..."

I reply: what grounds do you have for thinking that "we" would think that? Maybe we should do an experimental philosophy poll? ;) I rather suspect that, like most polls on these issues, some people would come out strongly compatibilist and others strongly incompatibilist. I, for example, don't find your reaction the least bit attractive. If *I* were confronted by these Matrix creators, I am quite certain that I would think to myself, "Well, then, nothing I ever did was really up to me. Every decision was made for me. I'm not responsible for anything I ever did." And, insofar as I suspect many people would have a similar reaction, doesn't this just show what I am suggesting: namely, that different people may well have quite different, contextually-sensitive concepts of FW?

Anyway, thanks again for your comment!

Hi Eddy: a quick follow-up. Suppose I change the story a bit so there was no manipulation involved. So, for example, suppose that instead of Simulation Creators we were freed from the simulation by Simulation *Finders*. That is, suppose when we were freed from the simulation, they told us, "To the best of our knowledge, your deterministic simulation just formed spontaneously. We've been living at this level of reality for eons, and every once in a whole we just find massive, deterministic simulations forming spontaneously. It's pretty cool, actually--and it seems your reality is one of those simulations!"

I still think that if we were released from one of these No-Manipulation Simulations, many--if not all--of us would say things like, "Free at last! For the first time in our lives, our choices are truly ours rather than the result of deterministic laws!"

Nice exchange! A few comments:

Can contextualism give you what you want: libertarian compatibilism? With contextualism you get: In ordinary contexts free will is compatible with determinism (since, perhaps, the relevant facts don't entail the future facts even if the facts do) but in philosophical contexts it is not. There's no context in which free will both is and is not not compatible with determinism. Maybe it's just me but this sounds like compatibilism since in at least some contexts free will is compatible with determinism. Incompatibilists want to take the more radical view: free will is never compatible with determinism.

I like the analogy between free will and epistemological skepticism but I have reservations about the Matrix example, and not just for the reasons that Eddy suggests. For one thing, the corresponding Matrix story for the free will case is inextricably tied with manipulation cases. Think of Pereboom's four case argument, which walks us from cases of overt manipulation to determinism cases. In this context, it is hard to see where the latter begins and the former ends. There is no use mining the intuitions of philosophers here since a great many of us will have already made up our minds and are avowed soft compatibilists, or hard compatibilists, or incompatibilists.

Of course, you don't need a Matrix world to tell your story. What you want is a world analogous to the Matrix world. But it depends on how the analogy works. In Matrix cases, knowledge is undermined because it’s ground – evidence, justification, truth – is undermined. What is the analogous model in the case of free will?

Hi Joe: Thanks for your comment!

What I think contextualism does is to get us to distinguish between different (though equally legitimate) types of questions.

The first set of questions are "ultimate" ones, or ones that concern fundamental issues in metaphysics, such as:

1. Is determinism true?
2. If it is true, can libertarianism (the ability to make choices *not* enacted by physical law) be rendered consistent with it?
3. If determinism is not true, can libertarianism play a fundamental role of altering behavior in (indeterministic) "gaps"?
4. Do we have any evidence in favor/against libertarianism?

My contextualism holds that if the answer to (1) is true, but the answers to (2)-(4) are all "no", we should say that we are not *ultimately* free or morally responsible at a ground-floor metaphysical level (but that we still may be free and responsible in a compatibilist manner).

On the other hand, I want to say that if the answers to these questions turn out to be different--if, in particular, we turn out to have some real evidence in favor of libertarianism--then we should say that we *are* ultimately free and responsible at a ground-floor metaphysical level, and that there would *still* be important compatibilist-level questions (about reason-responsiveness, deep selves, etc.).

As such, I am suggesting that compatibilism and incompatibilism are orthogonal to one another, and that in sense so are libertarianism and compatibilism. In each case, both could be true, both could be false. They are questions pertaining to different *levels* of reality (i.e. ground-floor metaphysics versus ordinary-everyday explanations of action), where these different levels are simply different contexts of explanation.

In terms of the Matrix-y example, my thought (though it may be incorrect!) was that it does something more than just mine old intuitions, but rather show, in a manner akin to thought-experiments on external-world skepticism, that our intuitions *shift* as we are moved from one context to the other, and then back again. This feature of the example seems to me to go beyond existing manipulation arguments (including Pereboom's four-case argument), since Pereboom is trying to mine the same (incompatibilist) intuitions with all four cases. This is not what I'm doing. I'm trying to mine the compatibilist intuitions, show that a shift in context lends itself to incompatibilist ones, and how a shift back in context leads again to the compatibilist ones.

Finally--and I probably should have given the revised case to begin with (but, with all due thanks to Eddy, I hadn't thought of it until he pressed me on it!)--the case I'm giving can be detached from any and every type of manipulation. Again, suppose, as I said in my 9:47AM comment, that instead of manipulating us into a determinstic setting, our Simulation Creators were instead Simulation Finders, who rescued us from a spontaneously self-generated deterministic simulation (a situation which is directly akin to the idea that our universe is "just here", rather than the manipulated design of a Creator!). In this case, even though all of the manipulations are absent, the shift in context--and shift in (many of our) reactions--still appears to exist.

Hi Marcus,

I'm happy with the idea that, if there are multiple FW concepts, it doesn't have to be that one is 'correct' and the others 'incorrect'. But how can they all matter the same? It sounds to me like you have your compatibilist and libertarian concepts (I'll accept for the sake of argument that you have identified these as distinct things) playing different roles; they both help us judge people's responsibility for their acts, but only the libertarian one lets us judge them ultimately responsible. Do both ordinary and ultimate responsibility matter? Should I, for instance, feel proud or ashamed of an act if I'm only ordinarily but not ultimately responsible for it? If I should, I can't see how ultimate responsibility and the libertarian FW concept matter; if I shouldn't, I can't see how ordinary responsibility and the compatibilist concept matter.

The way you describe what happens in your thought experiment when we are inserted back into the simulation, it sounds like we end up secretly accepting that no-one deserves blame for their actions, but either pretend otherwise or decide to fool ourselves into forgetting that. I can't read it as saying that the two concepts matter equally; it sounds like only the libertarian concept responds to anything of moral significance, whereas the compatibilist one is a politically useful substitute that responds to things that aren't of moral significance because the moral reality is too disturbing.

For what it's worth, as a true-believing compatibilist, my response to being extracted from the simulation and thinking back on some past misdeed (assuming that my beliefs, character, memory and identity survived the extraction) would be something like this: part of my mental life was devoted to bringing about that wrong; it was a part of my intentions and imaginings that the world would contain a little more evil; this wasn't some accident, or panicked impulse, or compulsion - I devoted myself to wrong over right for the duration of that choice; what a horrible thing to have been deliberately incorporated into my life! It may be that I don't know what guilt is - in which case, the rest of you are welcome to it - but I find it very hard to say that that response wouldn't amount to an entirely appropriate experience of guilt. However, it doesn't depend at all on what I was constituted by at the time - any kinds of particle, following any kind of law, or bits of information, or indivisible spiritual substance, or anything else. So, that guilt would be appropriate before and after I exited the simulation, and consequently, any FW concept that is sensitive to whether I am in a simulation or not is irrelevant to whether I ought to feel guilt or not. But I can't see that there's anything of greater significance than the appropriateness, or not, of that guilt (well, there are other attitudes/feelings/responses - behaviousrs towards myself and others - but they will be treated along much the same lines). So your libertarian FW concept doesn't play a role in what I take to be important, and there's nothing more important for it to add.

I'm sure you'll see the above as just more evidence that we are using different concepts. To an extent, that's fair enough (though I should point out that I don't know of another kind of contextualism-about-X in which there are different 'contexts' corresponding to people with different theories of X; it's not like epistemic contextualists think that, say, Bonjour and Dretske are using 'know' in different senses because of their competing theories about knowledge - that would be more like a dangerous sort of relativism). But the point of laying out what this particular true believer might think is more to demonstrate that I, for one, don't have any hole in my conceptual jigsaw for your libertarian FW concept, or for the ultimate responsibility that goes with it, to fill. If I were to take up an additional FW concept, it would be a spare wheel, with no new conceptual role. If I am to accept your contextualism, you'll have to show me that both of your proposed FW concepts play distinct and genuinely important roles. (It's worth noting that just trying to persuade me that my compatibilism is false doesn't go far enough, because I could well take up incompatibilism but think that only the libertarian concept captures anything of real importance.)

I think my questions for clarification are in some ways following up on some of Joe's comments. You ask whether you are right or confused. At this point I don't know but I do think your terminology is confused and I suspect you know this. So I ask:

1. Is there some advantage I'm not seeing to the odd terminology? There are fairly standard uses of "compatibilism" and "incompatibilism" in philosophy of action. It's formally trivial that a coherent position called "incompatibilist compatiblism" has rejected at least one of those standard uses (probably by rejecting presuppositions of the standard definitions).
Compare: so far as I am aware, contextualists about knowledge do not adopt labels such as "non-skeptical skepticism" or "anti-skeptical skepticism"... Probably because these labels would be thought to only distract from the discussion of skepticism.

2. Relatedly, when you write this:

"All of this suggests, I believe, that compatibilism and incompatibilism are both true. We have incompatibilist concepts of free will and moral responsibility which we would almost certainly deploy in certain contexts (i.e. when confronted by our Creators with the Deterministic Truth). At the same time, we also have plainly compatibilist concepts, which we would almost certainly deploy in other contexts (i.e. when going about everyday life)."

is what comes after the first sentence your (partial) unpacking of what you mean by the first sentence?

3. If the answer to #2 is "yes", how does that relation between the unpacking and the claim work? I don't see what the relation is between some people having various concepts of freedom (and determinism?) and both compatibilism and incompatibilism being true. Unless you are defining those terms non-standardly in which case I would suggest you explain those alternative definitions and how the issues and claims expressed using the alternative definitions related to the issues and claims expressed in the normal way.

4. Returning in a way to issue #1, though I see why Joe (in one of his comments) says that your position seems like a compatibilist position, I think I disagree. Where compatibilism about freedom and determinism is the view that freedom and determinism are compatible, I'm currently understanding your view to be the view that "compatibilism" as a thesis is defined in a defective way because the definition falsely presupposes there is one thing freedom when there is no such unique thing. [You seem to make the point one level up, so to speak, about there being multiple *concepts* of freedom but for now I'm trying to bracket the issue of whether these issues are about *concepts* (or the concept) of freedom or about, well, freedom].

Joe may well be thinking that you understand compatibilism as the thesis not that freedom is compatible with determinism but as something like "there is at least one freedom compatible with determinism" [and incompatibilism as "there is at least one freedom incompatible...].

**
Lastly, and more critically, on "libertarian freedom" as a "concept" of freedom rather than as a short hand for the thesis that freedom exists and is incompatible with determinism" I recommend Peter van Inwagen's discussion of this phrase in "How to think about the problem of free will"


Hi CJ: Thanks for the awesome comment. You're raising a lot of great issues!

First, on your question of whether the different senses of free will and moral responsibility "matter the same", the short answer is no: they matter very differently! One matters when it comes to an *ultimate* evaluation of our actions, and the other matters in a more proximate way. Allow me to explain, by virtue of a couple of analogies.

Consider the question of whether we can explain, say, why this plant photosynthesizes. In a proximate sense, we can explain it full well. We can explain how photosynthesis evolved, why it evolved, why plants give off CO2, etc. At the same time however, this is not a *full* explanation. A full explanation would have to explain why the laws of nature are what they are, as they play a role in the plant's evolution.

Now, of course, we presently lack an ultimate explanation of this sort--and this concerns us (and should concern us). Physicists and philosophers have every reason to seek the ultimate explanation...and maybe we'll find it, maybe we won't. None of this, however, makes the proximal explanation less worth having. All it shows is that there are two types of issues: (1) the issue of explaining the proximal conditions of photosynthesis, and (2) the issue of providing an ultimate explanation. Both are important, just in different ways, and in different contexts.

Turn now to free will. You write: Do both ordinary and ultimate responsibility matter? Should I, for instance, feel proud or ashamed of an act if I'm only ordinarily but not ultimately responsible for it?"

I think this very way of asking the question posits a false dilemma--the very one I am trying to use my contextualist account to disarm.

To see how, consider again the epistemic case: do I know there are tables and chairs, or don't I? Following other contextualists, I think the answer is: I do and I don't. I do know there are tables and chairs relative-to-ordinary-everyday epistemic standards, but I don't know there are relative-to-skeptical epistemic standards.

By a similar token, when you ask, "Should I feel proud or ashamed of an act if I'm ordinarily but not ultimately responsible for", my answer is: (1) As far as ordinary life is concerned, you should feel ashamed, but (2) in an ultimate, philosophical sense, you should not be ashamed. And I think these are *precisely* the right things to say. Here's why.

So, I've done some smart and good things in my life, and some not-so-smart and some not-so-good things. If someone were to give me definitive evidence today that determinism holds (and I do not think such evidence exists--more on this in future posts), I would momentarily--and in my more philosophical moments--believe that, indeed, I'm not ultimately responsible for *any* of it. I would chalk up my life to fate, "thank the gods" for the smart and good things I've done, and curse them for the bad and dumb things. (No joke, this is what I would do!).

But, of course, although I could well hold these views--and really take them seriously in my more reflective moments--there's is also something that life would force upon me: I would have to start *living* again. And, as Hume points out, when we actually go about living, there are things we cannot really avoid: holding ourselves responsible for doing dumb and bad things, holding others responsible for the same, congratulating people for doing good things, etc.

Now, of course, maybe there are people who would be *so* overcome by the fact that we are not ultimately free or responsible for our actions (in a libertarian way) that they would stop living this way. By a similar token, there in fact were people--Pyrrhonian skeptics--who were so convinced by external-world skepticism that they stopped living like normal people (Pyrrho reportedly sat in a bathtub all day, I believe!). Most of us, though, would probably not respond this way. Moreover, we might even adduce grounds for *believing* in ordinary life that we ultimately responsible when we're not (see this paper on how disbelief in free will promotes aggression and reduces helpfulness: http://psp.sagepub.com/content/35/2/260.short ).

Now, to this you might say (as indeed you did say!): "The way you describe what happens in your thought experiment when we are inserted back into the simulation, it sounds like we end up secretly accepting that no-one deserves blame for their actions, but either pretend otherwise or decide to fool ourselves into forgetting that."

But this too, I think, is exactly right--and it's exactly right in the same way as it is in the case of external world skepticism! I'd wager that if you asked the average person--after watching the Matrix--or even the average philosopher whether we *really* know the external world exists, a lot of them would say, "no!" (There's a reason, after all, that the skeptical problem has stuck around so long! Viz. Appearance vs. reality and all that). But, for all that, most people will say--just like Hume said--that although we know this in the back of our heads (we can't prove the external world exists), so far as *ordinary life* is concerned, we know it exists.

Now, I say all of this in a sense that is very sympathetic to your worries. For most of my career, I was no fan of compatibilism (an enemy, really). I thought that libertarian free will is the only kind of free will worth wanting. But now I think this is a mistake. It is like saying infallible knowledge is the only kind worth wanting. Which I think is clearly wrong. Infallible knowledge is awesome...it just so happens that we have hardly any of it, and that if we want to keep using the term "knowledge" in ordinary life (and we do), fallible knowledge will just have to do.

By a similar token, libertarian free will is awesome. I hope we have (and think we well might!). But it's not the only kind of free will that's relevant, at least not to ordinary, everyday thinking, speaking, and holding responsible. For these reasons, I now think that although there is a sense in which compatiblist free will is "free will lite", it's still a free will lite worth thinking and theorizing about--just as fallible knowledge is "knowledge lite" worth thinking and theorizing about.

Anyway, I want to say couple of things about your compatiblist reactions to the case.

First, I think in one sense they are a *permissible* reaction. You can well conceptualize the case that way. The error, I think, is in supposing that everyone should have your (compatibilist) reaction. This takes us pretty far afield, but as I've explained elsewhere (see http://philosopherscocoon.typepad.com/blog/2014/06/analytic-philosophy-continental-philosophy-and-natural-philosophy.html and http://philosopherscocoon.typepad.com/blog/2014/06/misled-by-language.html ), I think it's a big error of contemporary analytic philosophy to suppose that we all share the same concepts. We may have different concepts, and there may be no fact of the matter which way of carving up the world is objectively correct. If so, so much the better for both of us! ;)

Second, I think that when you say that libertarian FW "adds nothing" for you, you may not be thinking of certain important connections that libertarian FW has to other things--specifically, meta-ethics and normative ethics. This is something of a pet-peeve of mine. When people say that free will and moral responsiblity are compatible with determinism, I want to say, "You can't answer that question until you pin down what morality is!" To see why I think this, consider Kant's argument for the Categorical Imperative. Kant thinks the Categorical Imperative applies to us if and only if we are noumenally free, and our actions not ultimately caused by the laws of physics. If Kant's right, then libertarian FW is pretty big deal, whether we like it or not. If we have it, the Categorical Imperative applies to us; if we don't, it doesn't. That's a pretty important thing!

Anyway, to close, I think you may be right. There may not be anything I can do to convince you that you have a whole in your conceptual jigsaw puzzle...well, unless I can convince you that we *do* have a crazy kind libertarian free will, and that it makes a real difference with respect to morality. But that, obviously, will take a lot of work (provided I can accomplish it at all!). On these counts, though, I guess we'll have to wait and see. I've got a lot of arguing-for-libertarianism forthcoming this month. We'll see what you think. :)

Hi Fritz: Thanks for your comment! Maybe I know that my terminology is confused, but perhaps I just don't know that I know it yet. ;) Anyway, thanks for the good questions, and let me try to address them.

You write: "1. Is there some advantage I'm not seeing to the odd terminology? There are fairly standard uses of "compatibilism" and "incompatibilism" in philosophy of action. It's formally trivial that a coherent position called "incompatibilist compatiblism" has rejected at least one of those standard uses (probably by rejecting presuppositions of the standard definitions)."

I reply: I'm not sure that there's any advantage to the odd terminology...well, beyond the obvious advantage that it highlights the fact that I think both positions are true! The real question is: how can think this, given that as you note, it's formally trivial that on traditional interpretations of the two positions, they are mutually exclusive? Your suggestion that I have to reject some presupposition(s) of the standard definitions--and this is exactly right.

Let's consider the standard definition of "free will" from the SEP entry on compatibilism. Here's the definition: "free will can be defined as the unique ability of persons to exercise control over their conduct in the fullest manner necessary for moral responsibility."

I'm suggesting that there are two fundamentally different types of moral responsibility: (i) metaphysically ultimate, ground level moral responsibility, and (ii) ordinary-everyday moral responsibility. (I think I've been pretty clear on how I'm understanding the distinction, though I'd be happy to try to continue clarifying if I haven't).

Anyway, if this is true, then the question, "Is free will compatible or incompatible with determinism?", is not yet well-formed. We have to disambiguate. There are two questions:

(I) Is metaphysically ultimate, ground-level moral responsibility compatible with determinism?

(II) Is moral responsibility as it is understood in ordinary-everyday life compatible with determinism?

Since there are two distinct questions, compatibilism and incompatibilism--as I'm understanding them--are no longer mutually exclusive.

It's at least conceptually possible for us to (A) have free will in an everyday sense (i.e. the ability to exercise control over our conduct in the fullest manner necessary for *ordinary*-moral responsibility) in a deterministic universe, while, at the very same time, (B) *not* have free will in at an ultimate level (viz. the ability to exercise control over our conduct in the fullest manner necessary for *ultimate*-moral responsibility). Which is just to say that compatibilism and incompatibilism could both be true simultaneously...relative to different contextual standards of metaphysical focus and explanation.

I think all this addresses your queries in (2) and (3) as well (as I've just unpacked how I'm non-standardly understanding compatibilism and incompatibilism). I hope I've done so adequately.

In terms of your fourth query, I'm not quite clear on it.

You write: "Returning in a way to issue #1, though I see why Joe (in one of his comments) says that your position seems like a compatibilist position, I think I disagree. Where compatibilism about freedom and determinism is the view that freedom and determinism are compatible, I'm currently understanding your view to be the view that "compatibilism" as a thesis is defined in a defective way because the definition falsely presupposes there is one thing freedom when there is no such unique thing."

Is your worry that my position isn't compatibilist because I've defined it in a non-standard ("defective") way? If this is the worry, what's defective about my construal? Compatibilism holds that free will is compatible with determinism. On my analysis, free will is defined as "the unique ability of persons to exercise control over their conduct in the fullest manner necessary for moral responsibility." My claim then is that it *is* defective to think that there is a single, unique way to understand the crucial clause here ("fullest manner necessary for moral responsibility"), as I think there are two equally legitimate ways to disambiguate that notion (again, moral responsibility at an ultimate, ground-floor level, and moral responsibility at an ordinary-everyday level). However, once the crucial clause is properly disambiguated, I am suggesting that compatibilism is true: namely, that ordinary-everyday moral responsibility is compatible with determinism. But perhaps I'm misunderstanding your query...

I'm also a bit unclear on your last point.

You write: "Lastly, and more critically, on "libertarian freedom" as a "concept" of freedom rather than as a short hand for the thesis that freedom exists and is incompatible with determinism" I recommend Peter van Inwagen's discussion of this phrase in "How to think about the problem of free will."

Are you suggesting that "libertarian freedom" is just shorthand for "freedom exists and is incompatible with determinism"? (I'm a bit unclear if this is what you're suggesting, as the sentence in which you express it seems to be missing a few words, making it difficult to parse its content). Anyway, if that's the suggestion, I think that's a bad definition of libertarianism--one that should be replaced with the more general definition of libertarianism as "the brute ability to self-cause action independently of any physical or psychophysical law".

At the very least, what I argue in my research is that there's a natural way of construing "libertarianism" (i.e. in the manner just described) such that, surprisingly, it can be reconciled with the appearance of determinism (at least the *appearance* of determinism from our frame of reference).

Anyway, thanks again for your comment. If I've misunderstood anything, just let me know and I'll try again!

For Fritz (and Marcus), what's wrong with defining free will as the abilities to control action in such a way that one can deserve blame OR as the ability to do otherwise (as PvI says we should) OR whatever, and then define *libertarian free will* as whatever features of free will so defined are necessarily (and uncontroversially) incompatible with determinism--such as the *unconditional* ability to do otherwise holding fixed everything including laws or, depending on one's theory, agent causation--and define *compatibilist* free will as whatever features of free will (so defined) are clearly (and uncontroversially) compatible with determinism--such as the ability to act on desires one identifies with or to act in accord with reasons, etc.?

And then incompatibilism is the thesis that "lib FW" is necessary for FW (and perhaps MR), and compatibilism is the thesis that "compat FW" is sufficient for FW (and with some epistemic stuff, MR). Having written it out, I'm not sure exactly what we get by doing it this way, but I think that's what people usually mean by "lib FW" and "compat FW".

Hi Eddy: Thanks for your comment. Unfortunately, I'm a bit unclear on it.

You ask, "What's wrong with defining free will...[such that] *libertarian free will* [is] whatever features of free will so defined are necessarily (and uncontroversially) incompatible with determinism...?"

I reply: what's wrong with it is that, on my own Libertarian Compatibilist theory of free will, the unconditional ability to do otherwise holding everything fixed is actually *compatible* with determinism...at least from the standpoint of one particular reference-frame. (I'll say more on this as I forward throughout the month, but the reference frame stuff is a crucial move I make. What appears deterministic from the standpoint of one reference frame, I contend, may *not* be deterministic from the standpoint of a different reference frame).

Anyway, I get your final point that people want to define incompatibilism as the thesis that "lib FW" is necessary for FW and MR, and compatibilism the thesis that only "compat FW" is necessary for FW and MR...but this is the very way of defining things that I'm endeavoring to challenge.

I'm suggesting that the crucial clause of both definitions ("is necessary for FW and MR") is fundamentally ambiguous between two different *types* of FW and MR, and that once we disambiguate adequately, we see that:

(1) Lib FW is necessary for ultimate FW and MR, but not necessary for ordinary-everyday FW and MR.

(2) Compat FW is *not* necessary for ultimate FW and MR, but *necessary* for ordinary-everyday FW and MR.

In other words, I am alleging that the traditional ways of parsing "incompatibilism" and "compatibilism" (i.e. the ways you mention) are completely orthogonal to the real issues at hand, and so ought to be abandoned in favor of my alternative of parsing them (a way that tracks the relevant phenomena better).

Hope this helps--and I hope I'm not misunderstanding you!

Thanks for the (impressively long) reply Marcus.

I think this quote of yours gets to the core of my worry: "... when you ask, "Should I feel proud or ashamed of an act if I'm ordinarily but not ultimately responsible for", my answer is: (1) As far as ordinary life is concerned, you should feel ashamed, but (2) in an ultimate, philosophical sense, you should not be ashamed." I can accept that 'should' is sometimes context sensitive (just as in your explanation example, both the term 'can' and the set of admissible explanations being quantified over are plausibly context sensitive). But I have a lot of trouble accepting that it could be permissible, in any sense, to feel shame for some act if in the ultimate sense of 'should' I should not feel ashamed. If you say you should do X, and in another sense that you should not do X, you're going to fail to treat at least one of those senses as if it really means 'should', because you can't accord with both in what you actually do.

(Notice that contextualists about knowledge aren't doing anything like this. Knowledge is very relevant to our obligations, but the contextualist doesn't have to postulate multiple senses of 'should'. For instance, say there are two zoo inspectors inspecting the zebras at different zoos. The first goes to a zoo where painted donkeys have recently been passed off as zebras. The second goes to one where this is unheard of. A contextualist won't have to say "The first inspector should, on one sense of 'should', only pass the zebras when he knows they are zebras, but on another sense, may pass the zebras even without quite knowing they are zebras." They also won't have to say "The first inspector should, on one sense of 'should', check the so-called zebras for paint before passing them, but on another sense, is not obliged to." They can just say that both inspectors should pass the zebras only when they know they are zebras (with the sense of 'know', but no other term, varying), and the first inspector should check them for paint.)

It's concerning for me that you might end up saying the equivalent of "you're going to blame people sometimes anyway, and this is socially acceptable in some circumstances, so I'll call this acceptance 'moral responsibility' and I'll call the conditions on it 'free will' and then that makes compatibilism true". The reason it's concerning is that this is what too many incompatibilists and lay-people seem to think compatibilists do anyway - that the meanings of 'free will', 'moral responsibility' and so on commit us to incompatibilism, but we invent second meanings for the terms so we can keep talking as if we had the real thing. If you are going to call this kind of thing 'compatibilism', you're going to need a new word for people who think that 'free will' refers to one thing and that thing can co-exist with determinism.

I hope the above doesn't come across as unfriendly.It just feels a bit like if I was part of an entirely harmless protest march and some guy pops up next to me and yells "Come on, comrades, let's bust some windows!" I don't want people to start believing all the worst stereotypes about compatibilists!

Back on topic, I think Manuel Vargas says something along the lines that, if you count concepts in a liberal way, you can say there are as many FW concepts as you like, even one for each person witha theory on FW, but there's only room for one conceptual role in the area. You could take a bunch of people with a shared (confused) concept of FW, but different theories about it, and characterise them as having different FW concepts, but then they'll just be arguing over whose concept best fits the role. (I'm sure someone will correct me if Vargas never said anything like this, or if I've adapted it too carelessly.) I suppose my objection to your contextualism is that only one FW concept can actually play (for example - the role will have other facets) the role of guiding my behaviour towards other people. It sounds like you take your behaviour to be guided by the 'ordinary' concept. In that case, isn't the 'ultimate' concept just empty? You don't have to be a full-blown pragmatist to think that, at some point, if you keep saying that we lack free will in an ultimate sense, but you otherwise don't act any differently from someone who disagrees, that sense doesn't have an awful lot to it.

All this said, your point about Kantian ethics is relevant here if it would provide a distinctive role for the 'ultimate' concept. I mostly agree that talk about moral responsibility needs to be closely tied to the rest of moral philosophy. But I don't understand how you can think that the CI applies to us only if determinism is false, and also think Hume-style that all ordinary talk about MR not only would but also should go unchanged were determinism true.

Lastly, I do indeed think that everyone should have my reaction to the case. That's because they should all share my compatibilism, down to the last detail. How else should I think? How could I really believe anything at all without thinking everyone should believe the same? Of course, I also have reasons for thinking as I do (which is the same thing as saying I have reasons I think should make anyone else agree). (I don't know if you want to get into this, but I doubt your worries about vagueness will make much difference in the end. We're not just butting our pre-existing concepts against each other, we're trying to find one to satisfy shared interests, which is very different from debating how many grains make a heap.)


Marcus,

Nice post! I look forward to reading about your theory as the month goes on (time permitting).

I did have a question for you, you say:

"Although we would always know in the back of our minds that the Creators are responsible for everything we do, we would almost certainly go back to our old ways of speaking and thinking. We would very likely distinguish between "free" choices and "unfree" ones, and treat people as morally responsible for some choices (telling lies, committing premeditated murder) but not others."

But why think this is true? Scientific evidence often shifts the way that we think about the world and our and views about those that we encounter within it. For instance, psychological studies about psychopaths tend to change the views that many have about them regarding their blameworthiness and overt blame. The same seems to be true of addiction and those afflicted by the disease. In these cases our knowledge of the psychological condition of the agent in question changes our attitudes toward them. Why wouldn't this be the case once we understood that the creator is responsible for everything we do, and not they we are? What could do the work between “free” choices and “unfree” ones? And why would a libertarian be forced to accept the revisionist proposal?

Further thoughts – why think that we would carry on as usual? Some of our attitudes, moral concepts and appropriate emotional responses seem predicated on a libertarian notion of free will that is unavailable to us if we acknowledge that we are not morally responsible agents. I could offer a bunch of examples (forgiveness, moral right, wrong, obligation, etc.) but consider an apology for a moment. Can you sincerely apologize for doing something (wrong) to another if it was true that what you did wasn’t up to you? I don’t think that you can. You can say “I apologize” and you can feel bad that it happened, and even wish that it didn’t happen, but a sincere apology seems to be something more robust. An apology in the scenario you describe would seem empty or generic. It would be like giving a sincere apology for accidentally stepping on a person’s foot even though you knew it was an accident. Or like me apologizing for something someone else did. Think about it, if I apologized to a first nations person for the atrocities that happened to them in our nation’s history, it seems less sincere (or a different sort of thing altogether) than if I apologize for something I did of my own free will. But the former sentiment would be the only one was available in your world. The point of all of his is to show that there is at least some reason to be skeptical that we would go on as business as usual. Peter Strawson makes a similar point when detailing what he called ‘the objective stance’ and Stephen Darwall speaks of a similar stance when he talks of the ‘second-personal standpoint’(?).

Anyway, I have rambled on long enough. Thanks again for the post and the nice discussion that has followed.

Hi CJ: Thanks for your reply!

You write: "I can accept that 'should' is sometimes context sensitive (just as in your explanation example, both the term 'can' and the set of admissible explanations being quantified over are plausibly context sensitive). But I have a lot of trouble accepting that it could be permissible, in any sense, to feel shame for some act if in the ultimate sense of 'should' I should not feel ashamed. If you say you should do X, and in another sense that you should not do X, you're going to fail to treat at least one of those senses as if it really means 'should', because you can't accord with both in what you actually do."

I reply: I don't think it's hard to see how it could be permissible to feel shame in any sense if in an ultimate sense of 'should' you should not be ashamed. Let me try to explain why.

I think Hume implicitly puts his finger on the relevant issue in the passage I quote from him on external-world skepticism. Hume says that when he considers the question from a philosophical, reflective perspective, he is drawn to conclude that we have no knowledge of the external world. It is only when his "instinct" (or passions) intervene that he is drawn to say that we *have* knowledge of the external world.

Now, you say that contextualists about knowledge doesn't have to posit multiple senses of "should." I don't think this is correct. Arguments for external world skepticism arguably imply skepticism about epistemic justification, and justification is a normative notion (viz. one should aim to have justified rather than unjustified beliefs).

With this point in mind, return to Hume's analysis. Hume is saying something like this:

(1) From a detached, reflective, purely epistemic point-of-view, one should not believe we have knowledge of the external world.

(2) From an impassioned, instinctual, instrumental point-of-view, one *should* believe we have knowledge of the external world.

Hume seemingly is, in other words, positing multiple sense of 'should.' And, I would say, contextualists generally do the same, at least implicitly. After all, when contextualists say that knowledge-ascriptions depend on context, they often say that contexts are given by our interests.

But this is beside the point. Contextualists don't need to say this kind of stuff for me to make this kind of point in the case of free will--so let's move forward with it.

I want to say something similar to the claims that I have just attributed (rightly or wrongly) to Hume. Specifically, I want to say that when I'm talking about "ultimate free will and moral responsibility", I am talking merely about these things when considered from a metaphysical perspective, as in:

(U-FW) From an emotionally detached, purely metaphysical perspective, one has free will and should hold oneself morally responsible for one's actions just in case one has libertarian free will.

When I then turn to "ordinary everyday free will and moral responsibility", I want to say:

(OE-FW) From an emotionally engaged, ordinary-everyday perspective, one has free will and should hold oneself morally responsible for one's actions just in case one has [fill-in-the-blank with your favorite compatibilist theory].

I think (OE-FW) is true even though (U-FW) is true because, well, what one should emotionally do--as a flesh and blood human being, given one's real-life interests--is not always given by what one should do in a purely reflective, emotionally detached sense.

What we should emotionally do is given by our instincts and flesh-and-blood interests. Consider, for instance, the case of moral luck. Suppose consequentialism is false, deontology is true, and that while I do nothing wrong, I accidentally run over my wife in the car. From an ultimate, metaphysical perspective, I have nothing to be (morally) ashamed of. But, even though this is true from a ground-floor metaphysical level, there may be many ordinary-everyday reasons--reasons grounded in my desires and interests--why I should feel morally ashamed. (Perhaps feeling morally ashamed is just "what I need" in order to make sense of the world).

Anyway, when I think of free will, both senses of "should" here seem entirely plausible to me. If determinism is true, then in my more reflective, emotionally disengaged moments, I should say to myself, "You know, nothing I do is really my fault. I don't deserve credit for the good things I do, nor do I deserve blame for the bad I do."

That being said, when I leave my reflective moments are reenter *life*, there may well be many reasons--grounded in my emotions and interests--why I should hold myself and others morally accountable. It might be the case, for instance, that it is only by blaming myself for the wrongs I do that my behavior improves--something which may be the case even if determinism is true. Conversely, even though I shouldn't blame myself in a deep reflective sense, if I don't blame myself, there may be many practical costs (e.g. people who don't believe they have free will show increased aggression and less helpfulness, https://www.google.com/webhp?sourceid=chrome-instant&ion=1&espv=2&ie=UTF-8#q=people%20who%20don't%20believe%20in%20free%20will%20aggression).

Now, you might say--as your later remarks sort of indicate--that there is something deeply infelicitous about all of this. And of course there is! It is to say that compatibilist free will and moral responsibility are, in a manner, ultimate-FW/MR-effacing (viz. we should emotionally believe we are free and responsible even though we're not REALLY, ULTIMATELY free or responsible). And I would agree with you that we should feel uneasy about it. Indeed, that's part of the point of the analysis! The analysis holds that we should be pulled in two opposite directions--that in one deep sense we should believe we're not responsible, but in another sense believe that we are responsible.

For a long time in my life, I rejected compatibilism for precisely the reasons you're giving. But I've come to reconsider and revise my view for the reasons I give in the post (and indeed, for the reasons I have just given in the comment). We human beings are not merely reflective, emotionally disengaged souls. We are emotional beings, with real-life needs and interests. The things we should do in order to effectively satisfy those needs and interests in a deterministic world may be diametrically opposed to what we should do from a purely detached, philosophical point-of-view.

To wrap things up (and sorry if this reply is a bit rambling; I'm on a very early-morning flight), I think I can sum up my position best by addressing the point you make when you write: "I suppose my objection to your contextualism is that only one FW concept can actually play (for example - the role will have other facets) the role of guiding my behaviour towards other people."

I reply: I've just given reasons to think that there isn't just one FW concept that can play an important role in guiding my behavior. There are at least two such concepts. One the one hand, there's the "ultimate" concept I should act on in my more reflective, unemotional moments--let's say, at home, before I go to bed. Then, on the other hand, there's the "ordinary", emotionally engaged concept I should act on when I'm out there engaging in ordinary life.

I have no problem toggling back and forth between the two concepts, and indeed, I think it's pretty natural for reflective people to toggle between them as well. To use myself once again as an example, as I go about my life, I can't help getting angry at myself for my moral mistakes. I also presuppose in ordinary life that I have the capacity/free will to do better. But, for all that, as I lay my weary head down to rest at night, I can tell myself, "You know, maybe everything is determined. Maybe you shouldn't feel bad for what you've done. Maybe it's all just fate." Then I wake up the next morning, meet the day, and become angry at my mistakes yet again...and for what are good *instrumental* reasons (getting angry at oneself can improve one's behavior!).

This seems to me a compelling, nuanced picture of how our compatibilist and incompatibilist intuitions pull in different directions, and how they can both be simultaneously right and figure into our actions in a coherent (though "toggling") manner, waxing and waning (as Hume puts it) in proportion to our vacillating between (1) detached, metaphysical reflection, and (2) ordinary-everyday emotional engagement and human interests.

Hi Justin: Thanks for your comment!

I think you make a very good point, and that I probably should have made a much more qualified claim.

Consider the external-world skepticism case. Some people who were deeply convinced by skeptical arguments--Pyrrhoian skeptics--didn't just go on as usual. They profoundly changed their behavior to reflect their skepticism.

Further, I suspect that if I came to believe in ultimate FW incompatibilism (if I came to believe that we don't have libertarian free will), I expect that I might live a bit differently. I might be more forgiving, both towards myself and towards others. (Note: if I recall, some empirical studies show that people who don't believe in FW are apt to have less retributivist leanings with respect to punishment).

My real point is this: that believing we have no ultimate FW/MR does not *necessitate* giving up compatibilist FW--that it is possible (on instrumental grounds, emotional grounds, etc.) to preserve some robust compatibilist notions of free will and moral responsibility even if we have (and we believe we have) no ultimate FW/MR.

And I think this point stands. As seriously as one might take determinism--and as seriously as one might believe that ultimate MR requires libertarian FW--one can still retain a (legitimate) compatibilist notion, just as Hume can be convinced of epistemological skepticism in his more reflective moments and (legitimately) abandon it in an ordinary-everyday context. One can, as I note in my previous comment, "toggle" back and forth between different 'shoulds' (ultimate MR 'shoulds' and ordinary MR 'shoulds'). And insofar as this account simultaneously preserves incompatibilist and compatibilist intuitions--just like Hume's view preserves skeptical and non-skeptical intuitions--it seems to me these are compelling marks in favor of the view.

Anyway, thanks again for your comment!

Interesting dialogue. I think I understand the view a lot better now.

Let's go back to the analogy with epistemological skepticism. Do I know that I have a hand? Your answer is "yes" in the ordinary context but no in the philosophical context. But part of the skeptics argument is that only the philosophical context can properly ground any sense of "knowledge" worth the term. In other words, Sextus Empiricus or Barry Stroud might argue that if you can't properly ground knowledge, then there is no genuine knowledge. Likewise, libertarian-FW (Sextus or Stroud might argue) is the only thing that can properly ground free and morally responsible action. If you don't have it, you don't have free will in any legitimate sense of the term. In other words, once you condone the legitimacy of libertarian-FW, what's to keep from calling the other views a sham?

How would you respond to this skeptical line of argument?

Hi Marcus: thanks for the reply.

I find it interesting that you think you would be *more* forgiving if FW skepticism were true. I say FW skepticism rather than ultimate FW incompatibilism as I see Libertarians as ultimate FW incompatibilists as well, but I digress.

I find it interesting because it seems that at least many instances when we forgive someone we do so because of a sincere apology that a wrong-doer has given us. But, as i eluded to earlier, sincere apologies would seem to go out the window if FW skepticism (at the metaphysical level) were true and we believed that it was true. The concept of forgiveness itself seems to rely on an assumption of wrongdoing that would also seem to go if we lose Libertarian free will and believe that we do not have it, even for those who do not think that an apology needs to be a part of one's theory of forgiveness wrongdoing surely is ( but the wrongdoing would have to go as a result of obligation going via the ought implies can principle going, but again, I digress).

Another way of putting the point is that forgiveness in the world you describe doesn't really seem like forgiveness anymore. Feeling bad for someone as you suggested in the Hitler scenario is not forgiving them. It’s seeing them as a victim rather than the cause of wrong doing. This seems a lot like *not* holding the person accountable for what they did which seems to be something different than genuine forgiveness. Forgiveness proper seems to rest on an acceptance of having been done wrong and deciding to continue on in a relationship because of an apology or because you feel the person truly regrets what they have done. But, if we believe that the originator is the cause of the action that we are being accused of is it possible for us to truly regret what we have done? I don't see why. I can't regret something *you* have done because I'm not connected to your actions in the right sort of way. Similarly, if the originator is the ultimate cause of my actions it seems that I'm not connected to my actions in the sort of way that would allow me to regret something in the way I would now. So, it's interesting that you think you would forgive more, I have the opposite intuition. I might not take as much to heart because I wouldn't take people as seriously as I do now but that's different than genuinely forgiving someone for what they have done. I think we would be looking at them objectively, again these are all things that Strawson has mentioned in his famous essay. So if we choose your world we lose genuine forgiveness.

This is all relevant, I think, because you seem to be arguing that incompatibilism can be true on the meta level (and we can believe that it is) but that holding such a justified belief should not affect us on the ground, or, the effects would not be so profound; we could still forgive (was your example) and we may even do so more often. I just don’t see how that would be, hopefully I’ve given you reason to think it could be contentious. Many of the best accounts of forgiveness on offer imply that a libertarian reality be true, or at least not necessarily false (on the table so to speak). But it if determinism is true then it would not be on the table on the ground level.

Anyway, this was a bit of a ramble, my apologies. I hope you get the gist of my resistance to the contextualism on offer. The meta level seems to inform our ground level beliefs, or, at least many of us think that it should. And, many of these beliefs are about concepts that affect the way we treat one another.

Sorry for the spelling and grammatical errors, I'm responding from my phone at the gym.

I can refer you to a few papers that make similar claims, and much better than I have hear. Zac Cogley and Dana Nelkin have recently made some similar claims.

Hi Joe: Thanks for your reply!

Well, until about a week ago--when I was reading Hume and thinking about stuff to blog about this month--I would have accepted the skeptical argument!

Indeed, for most of my career I have been no friend of compatibilism (see e.g. my 2013 post at The Philosophers' Cocoon, "Trouble with Compatibilism", http://philosopherscocoon.typepad.com/blog/2013/06/trouble-with-compatibilism.html ). In fact, historically, I was so hostile to compatibilism that I stayed away from the free will debate like the plague. I believed that libertarianism is the only kind of free will worthy of the name, and I didn't believe we had it. Finally, it was only because I came up with Libertarian Compatibilism--what I thought (and think!) to be a physically and metaphysically serious case for libertarianism--that I entered into the free will debate.

Recently, however--since reading Hume--I've changed my tune. I'm no longer convinced that libertarianism is the only kind of free will worthy of the name. I'm still pretty sure its the *coolest* kind of free will worthy of the name, but for all that, three things got me thinking along the new (contextualist) lines that I explore in this post.

First, Hume's response to the external-world skeptical problem really spoke to me. I've long been persuaded by the external world skeptic...well, at least in principle. I don't think I can *know* (in the most ultimate sense of the world) that there's anything besides my own mind. Still, for all that, just like Hume, when I leave the philosophy room I claim to know things--and I think I do know things...as far as ordinary-everyday contexts are concerned.

In other words, I noticed that I seem to be a skeptic and anti-skeptic all at once, and that there's no real contradiction between them because the concepts of knowledge I seem to be deploying (ultimate/infallible knowledge vs. ordinary-everyday knowledge) seem to me to be about different things.

Second, I began to survey how I respond to the spectre of determinism (setting my libertarian leanings aside). Seeing as I was a hard-compatibilist for a very long time, it occurred to me that I always responded to my (firm) belief in incompatibilism analogously to how I respond to the external-world skepticism: namely, the way I describe in the post (when I take incompatibilism seriously, I find myself vacillating between not holding anyone responsible, and then entering back into real life...where I can't help holding people responsible, and where there's a real moral point to holding people responsible).

Third, and on a related note, it occurred to me that the way I vascillate maps onto the way Hume draws the distinction in the epistemic context: by contrasting (1) cold, dispassionate reasoning from (2) emotionally-engaged, practical reasoning vis-a-vis ordinary-everyday issues. And so it occurred to me that both contexts have a philosophical point to them. From a purely detached perspective, just thinking about the metaphysics, it seems clear to me that if we lack libertarian free will, there is no sense in which we are *ultimately* morally responsible for our actions. So far, so good. But now it occurred to me--again, following Hume--that there's a second way of thinking about moral responsibility that is *legitimate* and important. After all, as I point out in several comments above, even if we are not ultimately responsible for our actions, it can make a great deal of practical (and moral!) sense to *hold* people morally responsible. Holding people responsible can make them behave better. In contrast, when people don't believe they have free will and consider themselves not morally responsible, it turns out (determinism or no) that they behave morally worse.

In a nutshell, then, I am suggesting that there are two types of free will and moral responsibility.

The first, ultimate kind is purely principled, much as external world skepticism is purely principled (viz. [ultimate] knowledge is infallible, we don't know the external world exists infallibly; thus, we don't [ultimately] know the external world exists.

The second, non-ultimate kind of free will and moral responsibility is *instrumental*, much as fallibilism seems to me to do in epistemology. In epistemology, even if the skeptic is right that in some ultimate sense we lack knowledge, the fallibilist is right that we have knowledge...at least insofar as ordinary-everyday knowledge claims have instrumental *purposes* (e.g. coordinating human activity, serving as a norm of assertion, etc.).

By a similar token, I think that even if the incompatibilist is right that we lack free will and moral responsibility in an ultimate metaphysical sense, there are still instrumental purposes for talking about free will and moral responsibility...relative to our interests and aims in ordinary life.

That's how I'm thinking about it. The "ultimate" concepts of free will and moral responsibility have one purpose. They orient how we think about an ultimate metaphysical/moral question (i.e whether, at the end of the day, at a ground-floor moral-metaphysical level, we are free and responsible for our actions). The second, "compatibilist" concepts of free will and moral responsibility have very different purposes--orienting how we speak and think about these things...vis-a-vis ordinary-everyday life, with real people, important instrumental moral aims, emotions, etc.

It still seems to me that these are two different senses of "free will" and "moral responsibility" and that they both have a legitimate purpose, just as fallibilist and infallibilist concepts of knowledge have a purpose (the former orienting practical life, the latter concerning what true things, if any, we are infallibly justified in believing).

Does this help?

Marcus, you've certainly stirred things ups here! Good for you and all of us.

I'll keep it simple: would you describe your overall position as pragmatist, and if not, why not?

Marcus,

I agree with you: there are two different senses to the terms “free will” and “moral responsibility”. In a similar manner, I’m thinking there are also two different senses to the terms “causation” and “control”.

You’ve been referring to the two different senses as “compatibilist” and “ultimate”, and I think those are very descriptive terms. I’ve also heard those senses referred to as “modest” and “ambitious” or “weak” and “strong”.

I believe we’re all on the same page – we’re just using different terms to refer to the same ideas.

Perhaps what it all boils down to, is simply the source of the forces which are exerting control. If everything in reality is controlled solely by the four fundamental forces of physics (4FFOP), then only the “compatibilist/modest/weak” sense truly exists for those four terms. If other forces are exerted from other realms (e.g., “brute forces” as you might call them, from the realm of life), and those forces transcend into the realm containing the 4FFOP, add together with the 4FFOP, and thereby affect the path forward, then the “ultimate/ambitious/strong” sense of those four terms may exist.

As you’ve said, both senses have a legitimate purpose. If it turns out that the “ultimate/ambitious/strong” sense truly exists, however, I’m thinking there will no longer be a need for the “compatibilist/modest/weak” sense. The primary reason we presently need the distinction between the two senses, is to enable us to be on the same page when we’re communicating with one another. In other words, we can’t simply debate whether or not “free will” exists – we need to debate whether or not a certain sense of FW exists.

I think it’s fair to say that almost everyone agrees that the “compatibilist/modest/weak” sense exists for those four terms, so what we’re really debating, is whether or not the “ultimate/ambitious/strong” sense exists.

V. Alan White: Thanks for the very good question! I guess my answer is that it depends on how you're understanding pragmatism.

I certainly don't endorse any kind pragmatism about truth--the idea that truth is nothing more than usefulness (or some such). I vacillate between a correspondence theory and deflationary theory of truth.

I suppose I'm a pragmatist in the sense that I think is that certain concepts are more useful than others. I don't think, for instance, that the concept of a mereological sum is all that useful. For instance, we can talk about things like the mereological sum of the Sun and my bicycle--but this is not a very useful concept (except in some very circumscribed areas of philosophy). The concept of "chair", on the other hand, is very useful: it enables us to figure out which things to sit on (chairs!) and not to sit on (e.g. fire).

That being said, I still think the object of many concepts--both useful and non-useful ones--is to represent reality accurately (or truthfully). There are *truths* are the mereological sum of the Sun and my bicycle. Many of them are just not truths worth caring about (which is why we don't walk around talking about mereological sums in ordinary life)! Similarly, there are truths about chairs--and, I claim, there are two fundamentally different *types* of truths about free will and moral responsibility.

If I'm a pragmatist in any sense, I'm a pragmatist in the sense that I think there is more than one "sound" way to represent the world. Indeed, I think we can conceptualize the world in many ways--some useful, some not. My claims in this post, and in my comments, are (1) there are two useful ways to conceptualize free will and moral responsibility, and (2) insofar as concepts have truth-conditions, there are *truths* about both types of phenomena (ultimate FW/MR, ordinary-everyday FW/MR) the two types of concepts pick out.

Does this make me a "pragmatist"? In one sense (viz. there are multiple ways to conceptualize phenomena, some useful some not), I guess it does. But, in a deeper sense--at the level of truth--I don't think I am. I don't think truth is reducible to usefulness.

Hi James: Thanks very much for your comment. I think we are very much on the same page.

The only thing I'm not sure of is this. You write: "As you’ve said, both senses have a legitimate purpose. If it turns out that the “ultimate/ambitious/strong” sense truly exists, however, I’m thinking there will no longer be a need for the “compatibilist/modest/weak” sense."

Here's why I'm not sure of this. Suppose we have libertarian free will, and so are (in my way of conceptualizing things) are ultimately morally responsible for all of our actions at ground-floor metaphysical level. This still leaves open very many questions at the ordinary-everyday level--for instance, what makes ordinary adults worthy of moral praise/blame for our actions but not, say, children, the mentally infirm, or people who are manipulated/coerced by others. Serious questions of freedom and moral responsibility at *these* ordinary-everyday levels still exist even after we have settled questions at the ultimate metaphysical level.

Indeed, consider the kinds of phenomena that compatibilists theorize about: whether (and if so how) our actions are reasons-responsive, or reflective of deep-selves, etc. Even if we have ultimate free will/moral responsibility, these compatibilist issues still matter! The mentally infirm individual, for instance--the person suffering from massive delusions--may be free and responsible in a metaphysically ultimate sense (viz. libertarian freedom), but still not free or responsible in a manner specifically relevant to ordinary-everyday moral praise and blame.

This is why I don't think either way of conceptualizing things "trumps" the other. We can have compatibilist FW/MR without ultimate FW/MR. We have it *with* it. Similarly, we can have ultimate FW without compatibilist FW/MR (viz. the delusional or manipulated person). Etc. The distinctions are at least partly orthogonal. They are about different stuff (i.e. ground-level moral-metaphysics vs. moral concerns of ordinary life). Or so it seems to me.

Does this make sense?

Hi Justin: Thanks for your comment, and sorry for not replying to it sooner! (For some reason your comments keep getting caught in Typepad's spam filter).

Anyway, I think you make a really nice point when you write: "I find it interesting because it seems that at least many instances when we forgive someone we do so because of a sincere apology that a wrong-doer has given us. But, as i eluded to earlier, sincere apologies would seem to go out the window if FW skepticism (at the metaphysical level) were true and we believed that it was true. The concept of forgiveness itself seems to rely on an assumption of wrongdoing that would also seem to go if we lose Libertarian free will and believe that we do not have it, even for those who do not think that an apology needs to be a part of one's theory of forgiveness wrongdoing surely is ( but the wrongdoing would have to go as a result of obligation going via the ought implies can principle going, but again, I digress)."

I think you're absolutely right. One of the reasons why I think it is important to distinguish these different types of FW/MR (viz. ultimate vs. everyday) is that each type raises different issues for how we should respond to our moral-metaphysical predicament. And indeed, as I will now suggest (following you!), I think it is important to distinguish between the two types of FW/MR precisely so that we can appreciate interaction effects each might have on the other.

As you note, if libertarianism (ultimate FW) does not exist, there are probably ways in which we should be less forgiving (sincere apologies aren't ultimately coming from the person, only their physical programming). On the other hand, if libertarianism does not exist, there are--as I note--probably grounds to more forgiving in other respects (e.g. the criminal/liar cannot ultimately "help" themselves).

I think this is exactly right. It is only by distinguishing carefully between ultimate and everyday FW/MR that we can think carefully, and soundly, about how different combinations thereof should affect our behavior (viz. forgiveness, punishment, blame, etc.).

This also speaks to your next point. You write: "Another way of putting the point is that forgiveness in the world you describe doesn't really seem like forgiveness anymore. Feeling bad for someone as you suggested in the Hitler scenario is not forgiving them. It’s seeing them as a victim rather than the cause of wrong doing."

You can probably imagine what I want to say here. I want to say there are different *kinds* of forgiveness: ultimate forgiveness and ordinary-everyday forgiveness. On the one hand, I think you're right that if Hitler's actions are the result of him lacking libertarian free will, we're not *ultimately* forgiving him (since forgiveness requires fault!). Be that as it may, it is (I think) entirely natural to say to someone, "I forgive you for your sins. They're not your fault. It was fate that you committed them."

The important thing here once again, though, is that I think the kind of forgiveness here is very different than the ultimate kind. Rather than a principled/metaphysical kind of forgiveness, it's a kind of (Frankfurtian?) reactive-attitudes forgiveness, viz. "I'm not going to be angry at you, because it was fate that you lied to me").

Now again, you might say, "That's not real forgiveness!", and in a sense I might agree with you. As Hume notes, the kind of reasoning we're engaging in when we set deep metaphysical/epistemological issues aside is "vulgar." But, although it is vulgar--in the sense that it sets ultimate issues aside--it is no less real or legitimate. Again, I think there is a natural (if vulgar!) sense in which it is correct to say, "I forgive you...because none of it was truly your fault."

Finally, though I'm sort of working this stuff out as we go, I don't think you have me quite right when you say, "This is all relevant, I think, because you seem to be arguing that incompatibilism can be true on the meta level (and we can believe that it is) but that holding such a justified belief should not affect us on the ground, or, the effects would not be so profound..."

Although I've been arguing that it's important to distinguish ultimate from ordinary FW/MR, I think your general point--that they might have "interaction effects"--is spot on. It is important to distinguish both types precisely so that we can think carefully about how the existence or non-existence of ultimate FW/MR might affect (perhaps in revisionary ways) ordinary FW/MR. Perhaps, as you note, if we have ultimate FW/MR, we should change *some* of our practices for ordinary FW/MR -- and if we don't have ultimate FW/MR, we should change our views on FW/MR in others.

These are precisely the reasons why I think distinguishing between the two is important. If we adopt a "monist" picture--either compatibilism or incompatibilism tout court--we may fail to properly appreciate these types of important (and potentially very revisionary) relationships!

Hi Marcus, If you don’t mind I’d like to go back to your thought experiment to suggest some answers and ask some questions (I’ve wanted to use the Matrix to talk about free will but haven’t gotten around to it). There are two very different ways to imagine a Matrix. In one the Creators create a very complex program with rules that allow the AI beings within it (and humans plugged into it) to experience and act within a consistent world—indeed, a world with rules (or laws) designed to match the laws of our real world, except that they can be ‘bent’ by some. This type of Matrix might allow some Creators to foresee everything that will happen (other than rule ‘bending’?) or not; its rules might be deterministic or not; it might have ‘grown’ from an initial setup into later stages in various ways, some controlled by Creators, some not. This is the Matrix of the movie, and it’s pretty clear that no one is permanently controlling everything that most agents decide to do. This Matrix is similar to a universe created by God with laws of nature, perhaps allowing God foreknowledge, but this God is not controlling or causing everything that happens (no predestination).

The other way to imagine a Matrix is the way you seemed to be suggesing, one in which the Creators control everything that happens, like a God who foreordains all.

Now, I think free will is an illusion in the latter Matrix, because I think manipulation or complete control by other intentional agents does undermine free will (and MR), but it is also metaphysically different from determinism (or action in accord with laws and earlier states ultimately beyond one’s control). So, I don’t see any incompatibility between free will (or MR) and being in a Matrix, deterministic or not. Nor do I see any reason to be skeptical of one’s senses (or knowledge gained through one’s senses) in such a Matrix (or because such a Matrix is possible, for all we know). One’s knowledge is based on concepts and categories learned within the consistent rules of the Matrix (or whatever one’s world is) and corresponding to the objects established by those rules. (I don’t know Kant well enough to say with confidence, but it seems like he thinks we have knowledge of the phenomenal world and our categories are not just convenient fictions, and we don’t know much about the noumenal world, except perhaps the categories it would have to have in order to ‘stand behind’ the phenomenal world.) We cannot—or at least do not yet—know what the ultimate constitution of our world is. It could be bits or strings or quarks, or who knows what. But whatever it is, we have every reason to believe that it provides a consistent set of rules or laws that allows us sensory, and other kinds, of knowledge. And it allows us a robust kind of control over our actions that we call free will. (By the way, Neo doesn’t know that ultimate reality is what he thinks it is when he leaves the Matrix any more than he knew it inside the Matrix.)

So, are you suggesting that we’d lack free will in my kind of Matrix (or just your kind), and would we lack it only if it had deterministic rules? And do you really think compatibilism can only get us a pragmatic sort of free will that is no better than what we’d have in your complete-control-by-Creators version of the Matrix? From the armchair (well, with a little x-phi data), I’d predict that people would react very differently to these two kinds of Matrix.

Thanks for that, Marcus!

I had a few further questions.

You say: "Now again, you might say, "That's not real forgiveness!", and in a sense I might agree with you. As Hume notes, the kind of reasoning we're engaging in when we set deep metaphysical/epistemological issues aside is "vulgar." But, although it is vulgar--in the sense that it sets ultimate issues aside--it is no less real or legitimate. Again, I think there is a natural (if vulgar!) sense in which it is correct to say, "I forgive you...because none of it was truly your fault.""

I would say that! I agree that it is natural to say something like “I forgive you...because none of it was truly your fault” but I don't think it is correct (if determinism, etc.), and if it’s not correct, then how can it be legitimate in most circumstances? It seems like you're arguing taht we should reject the monist picture on pragmatic grounds. But why don't I think it's correct? Well, if to forgive someone requires that I must see the act of the actor as an act of wrongdoing (if it’s not wrong what are you forgiving them for?), then I don't see how we are forgiving in the world you describe. It sounds more like we are understanding. Understanding wouldn't go if the context changed, in fact it might get better and I think you might agree with that. Understanding is much different than forgiving, wouldn't you say? That's where the problem lies for me. Once agents understand that things are not “up to them” in the ultimate way this affects the contexts you are appealing to. I'm claiming that once that understanding *does* get better, via determinism being true and my justifiably believing it to be so (your example), that it forces us into Strawson's objective stance in other contexts. Now, I grant that this might sound lovely to some when discussing anger and blame as it might not be appropriate to direct these negative reactive attitudes due to lack of desert (desert requires ultimate FW), but it doesn't seem so nice when we consider the positive attitudes, our "achievements", and other moral concepts like moral obligation, right, and wrong. And once we are in the "on the ground" context I don't see how we take others seriously, why see them as morally responsible agents, or take them as seriously as we do now? For instance, I think it would be real hard to take the compliments of another in any meaningful way if I knew they were programmed to do so. It would be similar to a complex robot saying I looked nice today. Or, maybe the suggestion is that I treat robots and persons the same? (I'm being tongue-in-cheek) ;)

There is some experimental work suggesting folks are more willing to cheat or steal once primed with anti-FW claims. This suggests that the FW appealed to in every day contexts is ultimate FW which suggests the monist picture, that there is one sense of FW being appealed to and not having *it* affects our lives. I can grant that other notions are real but I reject that using them would be correct if it's the ultimate FW that's grounding our practices.

One last thought: if the contexts are not exclusive, and I gave at least some reason why they might not be, then why think that the conditions that ground ultimate FW are not the same that ground our MR practices and concepts like apology and forgiveness and moral obligation (among others)?
If one believes that ultimate FW grounds MR (libertarians, many skeptics), then there would not be any morally responsible agents in your world, or so it seems. We will need a way to ground our moral responsibility practices; we need to have a way of saying that some emotional responses are appropriate and others are not. But whatever story you give would likely be rejected by skeptics and libertarians which suggests that your view is a compatibilist view and only compatibilist.

Are you suggesting that the ultimate incompatibilist would have to give up the view that such concepts (obligation, forgiveness, apology) rely on libertarian ability. The staunch incompatibilist thinks that these concepts rely upon the world being the right sort of way. To claim that some acts are wrong, others obligatory, and that some emotional responses are appropriate the staunch incompatibilist would seem to ground these practices in their notion of ability. So I guess I am wondering why the libertarian, as the term is commonly used in the literature, *must* accept your compatibilist story of all of the concepts being grounded on the compatibilist notion of ability? Further, why would your story compel them given their views on ability already? For a libertarian to adopt the view that you’re suggesting they would seem forced to hold the view that the concepts our practices rely upon don’t themselves rely on a incompatibilist understanding of ability. But many reject that *that* compatibilist ability is enough to ground our concepts. Or, maybe I am missing something?

Sorry to be so long-winded. I'm just trying to get a handle on the features of your position. I'm actually developing a nuanced libertarian position myself so it's helpful for me to get clear on where our intuitions and arguments diverge.

One last question: how is your view different from semi-compatibilism? Let's define semi-compatibilism as "the idea that moral responsibility is compatible with determinism, whether free will is or is not compatible with determinism".

I have a bunch of questions and comments about your response but given that I am being very long-winded I should probably stop.

Thanks again for the back and forth, much appreciated!

The comments to this entry are closed.

Books about Agency


3QD Prize 2014: Marcus Arvan