Ways of making the compatibility debate mostly irrelevant

A recent correspondence with the always thoughtful Justin Capes made me reflect on the options for those who would like to sidestep the compatibility debate. One way to sidestep the compatibility debate is to just declare that one’s allegiance and move on. But that’s not so much a side-stepping as a taking a side. So what other options are there? 

Here are the two options that came up in our conversation:

1. Agnostic Autonomism: This is the well known view of Al Mele’s on which one is agnostic about the compatibility debate, but on which one judges that it is more plausible that we have responsibility than not, so one goes on to offer both a libertarian theory and a compatibilist (success) theory.
2. Independent Justification Thesis (+ semantic agnosticism): Semantic agnosticism (which maybe should have been called “referential agnosticism”) is the considerably less well known view on which one is agnostic about whether terms like ‘free will’ and ‘moral responsibility’ have incompatibilist reference-fixing elements. By itself, semantic agnosticism doesn’t allow one to sidestep the compatibility debate. The core of the sidestep move is an appeal to the independent justification thesis, i.e., the view that whatever the reference-fixing content is for these terms, the practices, attitudes, and inferences characteristic of responsibility can be justified on grounds independent of those typically regarded as at stake in incompatibility debates (e.g., whether we have libertarian agency). [Analogs here include things like talk or race, marriage, and folk psychology—lots of folks think there is good reason to continue to think it is worthwhile to have organizations like the NAACP, spousal health insurance coverage, and talk about depression, despite the dubious status of things like biological race, marriage as property exchange, and the predictive status of folk psychology.] On this view, whether we call our theory “a theory of responsibility” or “a theory of responsibility*” doesn’t matter so much as the practical consequences are unaffected. So long as we can more or less keep blaming as we have been, and attributing forms of agency sufficient to support such practices, the incompatibility debate is mostly inert, except for bookkeeping purposes. 

Agnostic Autonomism (AA) has the advantage of preserving our commitment to our being responsible but requires multiple ontologies of agency. If one doesn’t find both accounts plausible, it looks like it is still going to matter how the compatibility debate turns out. 

The Independent Justification Thesis (IJT) has the advantage of a single picture about our practical commitments, with theoretical commitments that vary (or, perhaps more accurately, are “translated” or “framed” differently) depending on how the reference question sorts out. But that also means more weight on the question of just how much our practices and attitudes can be captured by the independent justification. So, whatever one’s story is about the independent justification (whether consequentialist, contractualist, virtue theoretic, or what have you), it better be good. 

Another strategy would be to deny that the compatibility debate as ordinarily understood is coherent or well-rendered. I take it that something like this is the idea behind Ted Honderich’s view, and Saul Smilansky’s “fundamental dualism,” about which there is more discussion here. 

Are there other strategies out there? What do you think about these strategies? Are there reasons to favor one over another? Is the compatibility debate sufficiently mined that it is (dialectically, philosophically, otherwise) useful to try to sidestep compatibility issues?

Comments

[Kip]

Manuel,

Here are some comments to try to get the conversation started.

Re: "Independent Justification Thesis," I understand this view to be something like: "the free will debate (or part of it) is a merely verbal dispute, as Hume alluded to, and in that case we don't care about the verbal dispute, we just care about our actual practices of praise, blame, punishment, reward, etc."

A number of comments:

1. I take it that you're attracted to, or defend, this view. To the extent that it says "mere labels don't really matter," I see an asymmetry. If words don't matter then it shouldn't matter whether we say "free will exists" or "free will doesn't exist." Yet the revisionist seems very interested to say that free will *does* exist, even if it is different than we thought it was (or we're allowed to perform some other kind of linguistic gymnastics by the language police).

2. The view says something like, "labels like 'free will' don't matter, as long as practices like praise and blame remain the same." But it seems to me that practices like praise and blame have an essential verbal characteristic. Much, or all, of praising/blaming is simply affixing certain labels to certain activities. So the IJT theorist seems inconsistent in this way.

3. The IJT theory reminds me of an oral argument that I heard in a patent appeal (I'm a patent attorney by day). In the oral argument, a lawyer in this case said to the appellate judge, "oh, your Honor, the interpretation of that little word is a merely semantics issue." (Keep in mind that virtually all appeals at this level involve multi-million dollar litigations.) The judge replied, "counselor, you do realize that, quite often in patent law, the most important cases turn upon mere semantics and little differences in word definitions, and for good policy reasons."

I think the judge made an excellent point: disputes framed in terms of words are *governed* by the definitions of those words. This is true for a 500 million dollar breach of contract case, as much as it is true for the free will debate about whether "free will" (or any other term) exists. It's not fair to either party, at the end of centuries of debate, to say "well, the definition doesn't really matter, or the old definition said that person A won, but we're going to use a new definition and now person B wins." Etc. In other words, it's not fair to switch the rules of the game in the middle of the contest.

4. One last point, I think the IJT theorist wants to accommodate compatibilists and no-free-will theorists. On the one hand, the IJT theorist wants to acknowledge that free will has incompatibilist aspects, and that libertarianism doesn't rescue free will. On the other hand, the IJT theorist wants to reassure everyone, in a conservative manner, that much/most of what we hold valuable about our traditional socio-political practices of praise/blame/punishment/reward remain untouched. If the IJT theorist can disguise this compromise by slightly twisting the meanings of words like free will and moral responsibility, all the better.

That said, I think the IJT theorist is going to run up against the long run consequences of denying free will. Suppose free will doesn't exist. Suppose, however, that society still blames/praises/punishes/rewards in largely the same way, for pragmatic reasons - even if this involves some subtle dishonesty or distortion of our language practices. That view is unsustainable in the long run. In the long run, computers and technology will enable society to largely predict and eliminate crime before it happens, while also crafting and designing people to live predetermined lives. In other words, we'll slowly move to a more "medicalized" society. Sam Harris uses the example of a "moral pill" that a person can take to make the person act perfectly moral. Greene and Cohen, as well as Richard Dawkins, make similar points about treating criminals more like buggy computer programs in the future.

If Harris, Greene, Cohen, and Dawkins are right, then our social blame/praise practices will not remain conserved. That's an *independent* line of attack on the IJT theorist, apart from attacks on its liberties with language and definitions.

[Al Mele]

Manuel,

Another strategy is to argue that there’s no such thing as the concept of free will. (I don’t know how widely circulated Peter van Inwagen’s draft, “The Problem of Fr** W*ll,” is, but you’ll find the idea there.) I’ve been thinking off and on about writing something on this idea. One practical problem with doing that is that there is a lot of disagreement about the nature of concepts. (Maybe there’s no such thing as the concept of concept.) It might turn out that according to a range of understandings of concepts, there’s no such thing as the concept of free will and that, given some other way of understanding concepts, there is such a thing. It seems to me that the x-phi literature on free will gives us good reason to think that there are different but overlapping “understandings” (let’s say) of free will out there in the real world and that some of them are compatibilist whereas others are incompatibilist. Should we say that some of these understandings are more in line with *the concept* of free will while the others are further off target? And should we say the same sort of thing about the “understandings” of professional philosophers who disagree about what free will “is”? Is it more credible that there is no such thing as the concept of free will and that something else is going on?

Anyway, if you believe both that there’s no such thing as the concept of free will and that the “compatibility debate” is a debate about the concept of free will, you can explain why you’re side-stepping that debate.

Here’s a related strategy. Maybe you saw it coming. One can be *agnostic* about whether there is such a thing as the concept of free will.

[John Fischer]

Manuel,

As Al Mele wrote above, you will have seen this coming!

I think we need to take a more refined approach to "the compatibility question", distinguishing between compatibilism about (say) causal determinism and regulative control and causal determinism and guidance control. I *do* think we can productively sidestep questions about the relationship between determinism and regulative control by rejecting PAP. Of course, there remain interesting and difficult questions about the compatibility of causal determinism and guidance control, although I have always thought that here the compatibilist is on at least slightly friendlier terrain.

It is perhaps interesting to pause to reflect on why it might be desirable to leave these debates behind. You mention that we have "mined" the issues, and so forth. I agree that some debates just get tired and not much progress gets made. On the other hand, I am often amazed and "in awe" of the fact that some debates keep their vitality--new things can be said about some ancient problems. I think there is something to be said for engaging the ancient debates in new ways, or bringing to bear new methodologies and/or insights. But then again it is not obvious that the compatibility questions are the only interesting questions in our area--by no means is this obvious! Why have they dominated discussion so much, and should they continue to?

For example, I have always been interested in the role of poverty and other kinds of deprivation in moral responsibility. One way of getting at these issues is to try to get a really clear understanding of the relationship between responsibility and (say) manipulation and also mere causal determination (ha--I stuck the "mere" in), and then to seek to compare the situations under poverty and (say) psychological abuse to those of manipulation and "mere determination". This gives some role to traditional compatibility issues, while also expanding our horizons. I am inclined to think that we can still learn from the compatibility debates, but that we should not necessarily stop there.

[James Gibson]

Following up on Al Mele's reference to van Inwagen: In case you do not have the paper he mentioned, there's a relevant video online that you can access and which touches on these issues. Here's the link: (http://www.youtube.com/watch?v=ImvEqnbfdIw).

[Tamler Sommers]

Al please write on this idea. I think I've never agree with anything more in my life.

[Manuel "Taxonomizer Bunny" Vargas]

Hi Kip: Thanks for getting the conversation thread going! And yes, I embrace a version of IJT/Semantic Agnosticism as part of the Total Package of Vargas Views (to vary a phrase that has come up in talk about John F’s work!), but quite apart from the particular form of it I adopt, my interest here is to just sketch the essential features of the approach, separate from the larger commitments of my account.
As even a partial reply to your questions is fairly involved, and too long to drop into a comments thread, I’ve posted a longer reply here: http://www.usfca.edu/fac-staff/mrvargas/Files/IJT-cy.html

In brief: I take that there are cases where we have practices where the meanings of words are super-crucial (including, and maybe especially in the law). And, one can certainly be interested in providing an account of words or concepts. However, the IJT theorist is explicitly /not/ in the business of giving an account of the meanings of words (or concepts)—that’s precluded by the semantic agnosticism, after all. Rather, the IJT theorist wants a (normative) theory of what it would take to be licensed or warranted in, say, resenting, withholding interpersonal warmth, speaking scornfully about someone, and so on. With an account of those things in hand, the IJT theorist is happy to integrate that account into whatever account about meaning or concepts triumphs.

Al: That’s great. I wasn’t aware that van Inwagen had such a view, but like Tamler, I'm very sympathetic to worries about concept talk in general. It is one of the reasons why I’ve taken to talking more about reference and connotational content, rather than concepts, when characterizing revisionism. (Not that I don't give into concept talk, even so.) For what it is worth, semantic agnosticism, conceptual agnosticism and meta-conceptual agnosticism (that’s the view you suggest at the end, right?!) strike me as sibling views, perhaps with some mutual entailments lurking. [This makes me think that someone should organize a conference around people who are agnostic about all the stuff that was traditionally viewed as central to the free will debate. I’m agnostic about who should do that, though.]

John: Thanks for re-joining the party! You are surely right that the familiar compatibility debates have continued to provide a fruitful resource for thinking about all sorts of issues. Despite exhaustion in some quarters, it is surely a mistake to think we can’t continue to learn from carefully reflecting on aspects of the traditional compatibility debate. A quick question for you, though: as you think of it, is your “refined approach” rightly characterized as a sidestepping of compatibility debates? Maybe we are splitting hairs here, but I would have thought your approach is to precisify compatibility questions for the purposes of taking a stand on them. That is, I would have thought it was more apt to say that you aren’t trying to avoid compatibility debates (about fw and moral responsibility) but are instead trying to improve them, whereas Al and I (wearing our respective agnostic hats) are offering ways to avoid these debates whether or not they are improved. This isn’t meant as an objection to you (or Al!) but just as a way of thinking about what the varied strategies are here.

I’m inclined to think that we’ve now got four or five options on the table: (A) accept some version of compatibility debate whether (i) conventional (lots of folks) or (ii) suitably refined (John F); (B) reject the coherence of the debate (Saul S & Ted H, perhaps; maybe the PVI view on one gloss offered by Al); (C) go agnostic autonomist (Al); or (E) go for IJT+semantic (& maybe conceptual?) agnosticism.

James- Thanks for the link.

[John Fischer]

Manuel,

Good--I am fine with your characterization. But I am indeed trying to avoid the *traditional* debates about the relationship between casual determinism (and also God's foreknowledge) and regulative control. I'm trying to "side-step" those debates, which have been going on for a long time. But, yes, I do want to take a more nuanced approach with an eye to taking a stand on behalf of compatibilism about the relevant doctrines and guidance control (and moral responsibility).

[V. Alan "I Left My Heart in SF" White]

This debate has always been about values of concepts at some basic level. Say we inhabit a deterministic world. Then compared against another where we have positively defined indeterministic powers we might have less control (L intuitions) or thankfully we avoid having negative values of no control (due to say luck intuitions against the mere opportunities of simple indeterminism). We align with worlds when worlds present primarily the metaphysics of individual agents representing those values.

But when we go outside those values as metaphysically pinned-down to individual agency to questions of social responsibility, we leave the easy arena of what-means-what as inhabiting a certain possible world description of agents. We ask broader questions of what wider possible worlds we wish to inhabit that exhibit values of agency that might crisscross deterministic and indeterministic ones. I take it that it is that attraction that is basic to compatibilism. That is, the embrace of values that transcends more entrenched ones that attach to particular metaphysical descriptions is the central issue. I put it to my 101 students this way: there are two possible spacetime diagrams of human nature as deterministic or indeterministic as metaphysically open or closed to the future. All compatibilists agree that while those diagrams are of metaphysical interest (and across all possible worlds), they do not capture the value of what we mean by "free will" and "responsibility." Insofar as the motivation of their common claim is to value something else than those pictures of individual agency, then all compatibilists agree that significant moral values lie elsewhere.

As you know I argue this in my zombies paper (presented in your simply wonderful SF conference) that directly poses L intuitions against compatibilist ones in one world where metaphysics takes a pragmatic back seat to how to treat equivalent agents that can be empirically distinguished as indeterministic or not. My point in that paper is that values win and any particular metaphysic loses. Which may be consistent with your thesis here.

[Neil]

One way of making the compatibilism debate irrelevant is to argue that we lack free will for some *other* reason. I believe that free will is compatible with determinism, but nothing turns on my being right if am right it is incompatible with luck. Others might argue that absence of mental causation does the same job, or epiphenomenalism or situationist or more broadly naturalism (like Bruce Waller). Some of these claims make compatibilism less irrelevant than others: some are claims about the actual world only and therefore leave it open that if compatibilism is true, free will is possible. Others close off even that unactualised possibility.

[Robert Allen]

Manuel,

The Revisionist's suggestion that we abandon our belief in the significance of freedom of the will is a controversial claim if ever there was one. Could we simply sever our conception of free agency from (what seem to be) the more basic notion of free will? Is it even possible to act freely sans a free will? Isn’t free agency dependent upon freedom of choice, given that intentions are taken to cause as well as guide behavior? To the extent that Determinism and Divine Foreknowledge are inconsistent with PAP and self-control, then, given that they belong to the concept of FW, we have no philosophical choice but to tackle the compatibility issue, nettlesome though it is.

John Perry, in his recent defense of Revisionism, maintains that the belief in free will is an artifact of discredited religious dogmas. But does it not belong to common sense as well? Were the definitions of St. Augustine, St. Anselm, Duns Scotus, et al. mere stipulations, formulated whole cloth for religious purposes? Or were they not based upon a widely shared intuition, viz., that control of our minds in choice formation is the sine qua non of our being free and responsible agents?

Suppose I came to believe that my choices were being caused by someone else, content to otherwise leave me alone (risking the occasional case of akrasia). Should I then conclude that, since he has left my skills intact and deprives me of no opportunities I would otherwise have, that my agency is yet free? Or should I believe that my bodily movements, well-timed and skillful though they may be, share in my will’s loss of freedom? Competence and opportunity are surely goods. But would they alone make for true freedom, if our choices were beyond our control? In positing compatibilist free agency in lieu of a defense of free will, Revisionism only appears to point up, rather than rebut, Kant’s charge of “wretched subterfuge.”

Even in Perry’s PAP dispensing alternative, we can discern a belief in precisely the sort of control the Revisionist is supposed to abjure. To wit:

'It seems that I would think of things as follows. If I choose the skim rather than the lite, it will annoy my wife. I don’t want to annoy my wife. If I wanted to annoy my wife, that’s what I would do. Since I don’t want to annoy my wife that’s not what I’m going to do. It would be crazy to pick the skim rather than the lite, unless I wanted to annoy my wife.'

On whose authority was this resolution made- just who is the ‘I’ engaging in this act of self-examination? Whence the confidence that precisely those desires and preferences will become directives? (As a Humean, it cannot be the belief that the two are necessarily connected.) Why would anyone think that they are conatively related unless he has reason to believe that it’s wholly up to him what becomes salient, decisive? Further, whence the preference for comity and reasonableness? Here too we must employ the language of free choice, lest we invite Kant’s charge. For if he had not freely chosen these values over competing inclinations, if he’s just ‘always been that way’, then he hardly deserves credit for his cooperation.

If something as basic to practical reasoning as value formation is taken to be beyond one’s control, then philosophical rigor enjoins Hard Determinism, instead of an attempt to persuade us to refashion our own self-image. At the end of the day, should I become convinced that I am not in charge of my decision-making, that my choices can be attributed to something besides exercises of self-control, I can no longer regard myself as a free agent, no matter my skills and opportunities.

[Kip]

Hi Manuel,

Thanks for the reply and thoughtful feedback.

I see now, in hindsight, that some of my criticism was really directed to revisionism, and not IJT - I tended to conflate the two in my mind.

With a cleaner separation in mind, I can try to summarize my points as:

1. You propose that the compatibility debate is irrelevant. To which I would reply: why? What makes it irrelevant.

One answer (A) is: well, I'm an IJT theorist, and I'm just not interested in it, I'm interested in defending our traditional moral/blame practices. And so I declare, almost by fiat, that compatibilism isn't relevant, because it's not relevant *to me*.

Another answer B: the compatibility debate is really irrelevant because of specific reasons X, Y, Z. For example, it doesn't really matter what words we use, as long as we use the right social/political practices, because of reasons X, Y, and Z.

It's not clear to me whether your IJT view is based on type A or type B answers. If type B, it's not clear to me what the reasons are that would make compatibilism irrelevant. If you defend a type A answer, then I think IJT theory will not be very attractive to people who don't share your particular interest (or lack of interest) in the compatibilism debate.

One more thing: my replies here necessary focus on disagreement, which can be more enlightening than simply nodding my head in agreement. That said, my disagreement shouldn't disguise the extent of my agreement: I agree that, in many ways, words and labels are less important than our physical practices and socio-political order. I also agree that much or most of our current practices of praise/blame etc. *can* be justified, independent of free will, for pragmatic or consequentialist reasons (putting aside Pereboomian efforts at quarantining/rehabilitation) - we can't let serial killers run free bc they lack free will.

But I think this justification is time sensitive: with advances in moral pills, prediction, genetic engineering, computer prediction, education, etc., I think that our traditional practices will start to seem more barbaric, uncivilized, and unnecessary. This is a move that Harris, Dawkins, Darrow, and Greene/Cohen, and others have made throughout the years, and I think it's pretty fatal to both compatibilism and to IJT theorists - in the long run.

[Fritz Warfield]

Sensible ways of sidestepping "the" compatibility debate may well vary depending on which of many compatibility debates that sometimes get called "the" compatibility debate one is wanting to sidestep.

[Manuel "Man(uel)splainer" Vargas]

Robert- Thanks for the questions. There is a lot more there than I can reasonably hope to answer in a blog thread. However, I will note that it sounds like you are presuming the truth of libertarianism. That's fine, but not really an engagement with the question on the table here, which has to do with the choices available to those who, unlike you, don't take compatibility questions to be settled.

Kip- Thanks for the thoughtful reply. I think of the IJT-approach as somewhere between your A & B. It is like A, in that it specifies what it is interested in (practices, roughly), and invokes semantic agnosticism to signal indifference about terminological issues. However, the moral status of these practices is surely one of the drivers of interest in the compatibility debate (although not the only one). So, it isn't as though the IJT-theorist isn't plausibly on about stuff of interest to many theorists in this domain. The IJT approach is a bit like B in that there are particular reasons why the debate is rendered irrelevant, i.e., the substantive commitments (terminology aside) about whether and which practices are justified can be rendered on any outcome for how reference/meaning issues are sorted. That's a kind of argument about why irrelevance is licensed, but it is indirect. Final point: I agree that the defense I offer for responsibility practices is a "here and now" rather "for all time" justification. See p. 238 of BBB.

Fritz: I agree, and sorry for appearing to suggest otherwise! I was presuming everyone on a specialist blog like this would understand that "the compatibility debate" is just shorthand for "the set of arguments concerning the compatibility of, variously, on the one hand, free will and/or moral responsibility with, on the other hand, determinism and/or indeterminism." But I should have made that clearer and, you are surely right that the details will matter depending on which of these debates one is focused on. (I'll note here that how one operationalizes the meaning of free will, whether as a control condition on responsibility or as deliberative choosing or something else, will affect the relationship of how one understands the significance of the various compatibility debates to one another.) That said, it sounds like you have views about which approaches are more and less sensible ways of sidestepping some of these debates. I'm sure I'm not the only one who'd love to hear your thoughts about these things.