Explaining Away Incompatibilist Intuitions

Dylan Murray and I have our article with this title published now in early view at PPR here (if your institution does not have a subscription, you can email me for a copy of the paper). 

Some of you may have seen some of the results if you've read our chapter in New Waves in Philosophy of Action, but we discuss results of a second study in section 4, and we'd be especially interested in what people think of our argument that our results put pressure on the Consequence Argument (in section 4.2).

And some of you participated in this discussion where I offer some reasons we should care about people's intuitions about free will and responsibility,  but we'd be especially interested in what people think of our discussion of verbal disputes and reflective equilibrium in section 1. 

For a glimpse of what we say, I'll cut and paste the first 3 paragraphs (minus notes) of the paper and two paragraphs from section 4.2 below the fold (but the full discussion is, well, fuller!). 

Debates about free will remain mired in “dialectical stalemates” (Fischer 1994).  Compatibilists typically agree that if free will were what incompatibilists say it is – a type of freedom that requires having an unconditional ability to do otherwise or being the “ultimate source” of one’s actions – then it would be incompatible with determinism.  Incompatibilists typically agree that if free will were what compatibilists say it is – a type of freedom that requires a less metaphysically demanding set of capacities, such as reflective, rational self-regulation of one’s actions – then it would be compatible with determinism.  Each side believes that the other is wrong about what free will is, and about what conditions are required for having it, but they agree on which conditions are compatible with determinism and on which are not.  To avoid this morass, we might try banning the term ‘free will’ from discussion and proceeding instead by using the terms ‘freedom_I’ and ‘freedom_C’ for incompatibilists’ and compatibilists’ respective conceptions of it (Chalmers 2011).  Doing so would likely avoid some confusion.  But the debate would surely persist, because much of it is verbal – not in any pejorative sense, but in that much of the fundamental impasse between compatibilists and incompatibilists just is over what we mean by ‘free will’; about which conception of free will our inquiry concerns.

We take it that philosophical investigations of concepts used in everyday, non-philosophical life are typically concerned with, and are at least importantly constrained by, the ordinary or “folk” understanding of those concepts. This is certainly the case when the concepts are normative.  The default method for theorizing about many normative concepts, wide reflective equilibrium (WRE), takes as inputs our normative principles, background scientific theories, and pre-theoretical (but reflective) judgments, or intuitions, about relevant cases, and then attempts to develop a philosophical theory that is maximally consistent (and, ideally, mutually justifying) among those inputs.  WRE builds upon the pre-theoretical understanding of the concept under investigation, as revealed by intuitions about specific cases, so that our final theory – even if it deviates from that understanding to reach equilibrium with other inputs – is recognizable and relevant to the normative roles that the concept plays in everyday human life.

‘Free will’ plays a central role in the conceptual scheme that we use to navigate the normative world via its connections to ‘moral responsibility’, ‘blame’, ‘autonomy’ and related concepts.  Theorizing about ‘free will’ in isolation from the ordinary conception thus risks being an academic exercise about some other, technical concept divorced from people’s actual practices of assessing praise, blame, reward, and punishment, and from their understanding of themselves and their place in the world.  Imagine, for instance, that freedom_C (and not freedom_I) is what ordinary people care about having, and worry about not having – that freedom_C is what they refer to when thinking about, speaking of, and otherwise using the concept ‘free will’.  If so, would the fact that freedom_I is incompatible with determinism support any interesting or important version of incompatibilism?  We cannot see how.  Philosophers are of course free to discuss some other, stipulated notion of ‘free will’, but for philosophers interested in escaping the dialectical stalemates, the starting point that compatibilists and incompatibilists should both be able to agree on is figuring out what people mean by ‘free will’ – ‘freedom_I’ or ‘freedom_C’....

... Thus, the debate again appears to bottom out in a verbal dispute, shifted from the meaning of ‘free will’ to the meanings of ‘choice’ and ‘ability to do otherwise’.  It is relatively uncontroversial that unconditional conceptions of choice and the ability to do otherwise are incompatible with determinism and that conditional or dispositional conceptions are compatible with determinism.  The question is which conceptions are relevant, and again, we believe that the philosophical debate should concern those that are used in, and actually matter to, ordinary human practices, especially regarding ascriptions of responsibility....

Hence, the evidence suggests that most non-philosophers do not share van Inwagen’s intuition that the sense of having a choice and the ability to do otherwise relevant to free will and responsibility is unconditional.  This is not to claim, of course, that most people explicitly have in mind any specific conditional analysis of the ability to do or choose otherwise.  Many people may simply have an implicit understanding of contingent events in general, including human decisions, such that they could have happened otherwise only if something leading up to them had happened otherwise.

Comments

[Paul Torek]

Eddy,

I like your points about the normative connections of free will. I've always felt that the normative was the dog of free will, and we should stop letting the metaphysical tail wag it. I suggest you add to your list of normative dimensions, the idea that free will is a presupposition of rational deliberation about what to do. This might already come under "autonomy", but I think it deserves explicit mention.

Thanks for sharing your paper.

[Josh May]

Hi Eddy,

This is a really interesting paper. I've gone back over it quite a few times since first encountering it at the summer seminar at Florida State.

As far as the discussion of verbal disputes, I'm unsure about how it connects with the paper's main focus on folk intuitions. Early on, you suggest that the philosophical debate rests on a verbal dispute but then back off, saying that it's not a "mere" verbal dispute in any pejorative sense. Rather, it's just that we need to do some xphi to see what we mean by "free will" and so forth. But later on you say it does rest on a verbal dispute. Is this still just the idea that we need to figure out what we mean by the term or have we truly gotten into the more problematic (pejorative) kind of verbal dispute?

In general, it's unclear to me how your explanation of folk intuitions (the bypassing error theory) connects, if at all, with the dialectical stalemates we see among theorists. I think I recall you saying at the seminar that you thought there isn't much of a connection. That is, you don't want to say that incompatibilist philosophers make the error that the folk are making about bypassing.

That seems like the way to go, but then we don't seem to have a theory which captures both the mixed thinking among philosophers and ordinary folks. It's not clear that we should capture both with the same theory or that the one should somehow connected to the other. However, I should say that John Maier and I are working on a paper that does just that. But it requires taking both compatibilist and incompatibilist intuitions in ordinary thinking as part of our conceptual competence (not a performance error). It also involves reporting some new data and a critique of the studies in your paper... :)

[Joe Campbell]

Eddy,

I'm sympathetic with the verbal dispute line but I wonder about the theory of meaning underlying much of what you say.

For instance, you write: "We take it that philosophical investigations of concepts used in everyday, non-philosophical life are typically concerned with, and are at least importantly constrained by, the ordinary or 'folk' understanding of those concepts."

I'm fine with the first part (philosophers are concerned with understanding folk concepts) but I wonder about how to cash out the second part. Here's an example.

If you did some empirical research you might find that the folk differ about their responses when it comes to the concept of "women." Experimental data might suggest that some in the community think that women are inferior to men whereas others do not. Does this mean that they have different concepts of "women"? If someone thinks that being a woman entails being inferior to men does that mean that "inferiority" is part of his concept of "women"? If someone says "My wife is inferior because she is a woman" is the proper response: "You and I disagree about the concept of 'women'"? Or is it: "No, they are not; you are wrong"?

Here's a case in which some of the folk, perhaps (unfortunately) a large percentage (especially within some linguistic communities), assume that there is an entailment that many of us more progressive members deny. In this case, I'd say the progressive members are correct about the lack of entailment but that's not my main point.

The main point is that it seems wrong to suggest that the two subgroups have different concepts. They have the same concept of "women" but one group wrongly thinks that an entailment follows and the other group doesn't.

It can't be the case that every time that one subgroup in a linguistic community thinks that an entailment holds that others deny, that this is the result of a difference in conceptual understanding. Otherwise people would constantly be talking passed each other and communication would be impossible. In order for communication to be possible there must be some shared part of the understanding (which I would classify as the concept). Otherwise ALL disagreements would turn out to be verbal disputes and there wouldn't be any real disagreements at all.

[Dylan]

A few thoughts on Josh and Joe's comments:

We don't think that philosophers are making the bypassing mistake (though maybe it’s part of the story for why some felt an initial pull toward incompatibilism in their philosophical infancy). Rather, that non-philosophers make the mistake has inflated the numbers of how many non-philosophers have seemed to have incompatibilist intuitions, and that's relevant to the philosophical debate just insofar as non-philosophers' intuitions are relevant to that debate. I don't think we've ever had any more direct connection in mind between the bypassing mistake and the existence of the philosophical stalemate than that: just that figuring out what the non-philosophical intuitions are might help to break it (perhaps, e.g., because incompatibilists need those intuitions to side with them).

We do kind of back off of the verbal dispute line in the paper (but, as Josh notes, try to remind people of it a few times later on). Partly, framing things in those terms just helps to bring out the potential relevance of the empirical work: what sorts of disputes could x-phi have more sway over? For my part, I do suspect that much of the free will debate is a verbal dispute, but hesitate to commit to strong forms of such a claim for a few reasons: (i) it’s not entirely clear where we should draw the line between real and verbal disputes, especially (ii) in cases where the concepts play important social and normative roles, and (iii) I agree that it’s not entirely clear what our theory of meaning for these concepts should be. I’m inclined to think that ‘free will’ and ‘moral responsibility’ are considerably more deferential to ordinary understanding and usage than, say, scientific concepts, and in a way that has something to do with (ii). Joe’s ‘women’ example provides a nice comparison here: it seems similar in all these respects. On the other hand, Sally Haslanger has made some very strong arguments for why we should be fairly externalist about social concepts like ‘race’ and ‘gender’ precisely because of their normative component…

Perhaps another way of putting the point about verbal disputes is this: most parties to the debate would agree that the compatibilist’s necessary and sufficient conditions are compatible with determinism, and that the incompatibilist’s are not. To the extent that hang-up over the term ‘free will’ prevents that point from being fully appreciated (and I think it often does, in practice), the dispute is verbal. What we’d continue to disagree about even after appreciating that point, though, is which set of necessary and sufficient conditions are relevant to the role that figures importantly in our social and normative practices. That’s less obviously a verbal dispute than an empirical debate, but it is still a disagreement about which set of necessary and sufficient conditions characterizes the concept that actually figures in those practices. (Of course, we can also go on to ask purely normative questions about which set of conditions are in fact fair to employ, and the like, instead of questions about the practices themselves, but these questions might have to defer to ordinary understanding and usage considerably, too (especially if they’re supposed to be answered through reflective equilibrium).)