Blog Coordinator

« What Counts as Evidence for the Meaning of Free Will? | Main | Latest issue of Science features... experimental philosophy of free will »



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

What an excellent post Dan. You've made beautifully explicit large-scale themes about the debate that need to be addressed.

My take is that moral conservatism isn't essentially a naturalistic contribution per se. Yes, it might be informed by data--X-phi can tell us a lot about responsibility claims that people make. Of course it doesn't mean they make them *properly* in some final moral sense. But even widespread agreement about moral responsibility claims might just reflect the current milieu of intuition--and as we've seen in the past, shared intuition might just be a mask for prejudice. And while I think that moral realism might have evidentialist grounding in some instances (e.g., the near universal assent that selfishly-motivated first-degree murder is wrong), I am much more suspicious about evidentialism about claims of responsibility for such objective evil. Emotions tend to rachet responsibility intuitions up as cases get more personal, and that alone should make us more skeptical about a proper evidentialism going on there.

On the other hand I think that instances of indisputable freedom--physical freedom, social freedom, political freedom--form an evidential basis for analyzing freedom at least in some general sense. So the general picture of freedom metaphysically can be grounded in an evidentialist way. Or so I'd claim.

So in the tug-of-war here, I'd side with evidentialism about the metaphysics of freedom. No so much with responsibility. Conservatism about morality in general at least has the virtue of garnering the attention of a rational audience to some extent (witness the excoriation Peter Singer has endured for challenging that conservatism). But conservatism about responsibility claims is something I find highly suspect because it usually entails hidden assumptions about FW blended with emotional factors in such complicated ways that may well undermine any usefulness of the assumption.

So for me, evidentialism for freedom, logical consequences of that commitment for responsibility.

Wow, Dan, what a thoughtful and elegant post. Who trained you? He must be great.

Anyway, I think I might not accept both parts of the Pairing. Or something like that. So, I take it that it counts (significantly, although of course==and perhaps lamentably--not decisively) in favor of a view of free will/moral responsibility that, on it, our freedom/responsibility "does not hang on a thread", and further, that, on it, we wouldn't have to give up our basic metaphysical views if we discovered an empirical thesis, such as that causal determinism is true (or, for that matter, false).

I've argued in different contexts that Supercompatibiliistic Semicompatibilism has precisely this virtue, and that it counts in its favor that--unlike libertarianism and certain sorts of compatibilisms--SS does not render our status as free and responsible unacceptably fragile, nor does it issue in funky metaphysical flip-flopping.

So, given this approach of mine, it seems I am not a Pairer. Is this right? If this is correct, then we are methodologically fellow-travellers, even if we end up in (slightly) different places.

"That is, it might seem that only an ungrounded and unnatural hope could inspire us to think that the natural world is going to conspire with commonsense moral thinking to work out an unproblematic fit between them." DW

"On the other hand I think that instances of indisputable freedom--physical freedom, social freedom, political freedom--form an evidential basis for analyzing freedom at least in some general sense." VAW

Let me join Prof. White in praising the clarity of your meta-philosophical thinking, Dan.

However, I just don't see why as a philosopher I should kow tow to scientists. I think that Profs. Kane and Vargas take the wrong tack here. Suppose my understanding of my mental life doesn't square with Determinism/Materialism? Isn't that simply a reason to argue against the latter, a la Augustine and Descartes Shouldn't my intuitive belief that I am in control of my choice-making be added to the list of "instances of indisputable freedom?" Why are only cases of Compatibilist freedom philosophically relevant?

Call me unbalanced, but I'm siding with the philosophical conservatives every time. I take that to be in keeping with Socrates' practice of shunning nature in favor of discourse with his fellow men. (Funny though, I've never met a political conservative I didn't want to rebut.) Singer deserves every bit of the excoriation he has received. For moral conservatism is deeply ingrained in all of us and not only lends order to society but is an essential part of our self-image: we are the beings who realize that killing babies is wrong. Similarly, we are the beings who hold ourselves and each other responsible for our choices and the actions they cause. And there could be no "hidden assumptions" here regarding FW: responsibility conservatives must be up front about their commitment to the notion that we are in control of our choice-making. Because if we are not, there is simply no such thing as praiseworthiness/blameworthiness.

Thanks for the nice comments, gentlemen. A few initial responses.

Alan, if I’m reading you rightly you are not totally ill-disposed toward the conservativism in certain clear cases but you think that we ought to put the most weight on our naturalistic evidence in this domain. Is that right? But I think you may also be adding something (that I’m really interested in) about the status of the “good” invocations of conservation. When there is nearly universally agreement that selfishly motivated first-degree murder is wrong, it looks like you want to say that preserving that judgment in our theorizing will be making a naturalistically evidential contribution. Do you have any thoughts about how this goes? How does commonsense moral commitment get its bona fides as part of the evidence?

John, I was trained by a philosopher whose mustache is one than which a greater cannot be conceived. And, yes, I’ve suspected that you might be willing to reject the pairing in virtue of some of your methodological commitments. But you don’t HAVE to. You might, after all, try to argue that your “stability” intuition about our moral status as persons is part of our naturalistic evidence. I would like to see how that argument goes, of course. There could be such an argument, however.

Robert, I appreciate your honesty and straightforwardness. But I will bet that your own approach involves affecting some sort of balance between conservation and naturalistic evidence. It just seems that you would be on the other end of the spectrum from Alan, giving considerably more weight to the preservation of commonsense morality than to the evidential fit with naturalistic considerations.

For everyone, let me emphasize that I am neither endorsing nor rejecting the methodology I have claimed to find in the contemporary debate. I only want to note what seems to me to be a consequence of the strategy: namely, that there appears to be room in our theorizing about freedom and responsibility for considerations that outstrip or circumnavigate traditional epistemology. That would be something to notice, if true. If we are willing to reject the pairing, then it seems that we can have no general principled reason to discount pragmatic factors in the free will debate.

To continue.

I think it counts in favor of a theory that it is elegant. But elegance does not pertain to truth; a more elegant theory is not more likely to be truth, unless, as Van Frassen argued some time ago, (say) God made the world simple. But still, it counts in favor of a theory that it is elegant.

So what counts in favor of a theory--what commends it--is broader than what can be understood purely epistemtically, where "epistemic" is understood in terms solely of truth.

So if "credibility" is construed broadly, to include anything that counts in favor of, or commends, a theory, then I would deny the Pair.

Nice post, Dan. It is not obvious that moral conservatism provides "a naturalistically evidential contribution to theorizing" (to use your language).

Suppose I have some idea of what the referent of "moral conservatism" is. If I am right, then moral conservatism is in the same boat as theorizing about free agency. Theorizing about either free will or morality requires determining what the ontological commitments of our theories and discourse are if we are at all concerned about taking ontological naturalism as a proper constraint on theorizing. If this is right, then the only reason I can think of for why we may want to put them (moral concerns and ontological concerns) on par is because we think that, for practical reasons, they are of equal importance. This sounds like John Bishop's reconciliatory naturalism. Bishop seems to regard both the naturalistic perspective provided by a scientific worldview and the ethical perspective provided by common sense morality (if there is such a thing) to constrain our theorizing about action and agency (including free agency). The goal is to reconcile these two perspectives (which correspond roughly to Sellars's "manifest" and "scientific" images of humans). This sort of approach is attractive and seems similar to what you describe. But if the perspectives are seen as being on an equal footing, that strikes me as a mistake.

Assuming realism about theories of free agency and morality, it seems that in the case of both free will and morality we have to first work out what the ontological commitments of our discourse are. We can go the familiar Quinean route or go the anti-Quinean route recently endorsed by Ross Cameron and others (e.g., Heather Dyke and John Heil--even Bishop and, earlier, Smart seem to favor similar strategies). The Quinean route is familiar ("to be is to be the value of a variable"). The anti-Quinean strategy uses truthmaking to help us sort out our ontological commitments. Specifically, we must ask how must things be in the world in order for us to be realists about some domain of discourse--for the sentences of that domain to be true? If we begin here, then it is not obvious that moral conservatism provides an additional "naturalistically evidential" constraint to theorizing. Rather, it faces the same constraints as our theorizing about free will. They are on par. Once we sort out how things must go ontologically in order for us to be realists about either domain (assuming we can be realists about both), we can then see if we can put them together in one big picture. But if we cannot, then it seems methodologically suspect to somehow privilege moral intuitions over ontological seriousness. (So I reckon I'm suggesting that Alan is right.)

My guess is that if we take ontological naturalism as a starting point in constraining our theorizing, the end result is that our project will look more and more like revisionism if we proceed in the manner I am recommending. In the worst case scenario, we will have to abandon realism about our free agency and common-sense moral discourse. But I presently think there is no good reason to think that result is inevitable.


So you were trained by F. Nietzsche?

John, as you know, I wasn't trained by Nietzsche. And that should tell you how strongly I feel about your 'stache.

I like the idea of linking my point here to something like the role played by the theoretical virtues. In fact, this is a point I have made in other contexts. I was surprised to find, however, that some folks do want to treat elegance and simplicity (for example) as making an evidential contribution. Part of what I am wondering in this post is if the free will community believes it maintains the conservativism BECAUSE it is evidentially relevant. So, your response now leads me to wonder if Flickerers think that appeals to elegance and the like are evidential.

Suppose you think (as I do) that the reference of moral terms is causally regulated by natural facts about human welfare. Then the gap you see between evidentialism and conservatism won't open: moral considerations will be considerations about natural facts, and in discovering further such facts, we may refine our moral terms.

What Neil has expressed so succinctly gets to the heart of what I was saying (in my typically long-winded fashion). I would just add that there may be no competition at the end of the day once we sort out the ontological commitments of our theories.

Neil and Andrei:

Okay, good. This is what I want to think more about. Even if you do think that the moral terms are regulated by the natural facts, does this give you a reason for thinking that commonsense moral judgments ought to get to play the role in theorizing that I claim that they do in fact play in the contemporary debate?

Neil, it strikes me that your language is entirely compatible with Aristotelian virtue theory. (I'm serious--that is not a snarky remark.) Would your view embrace that?

I ask because I think virtue concerns about character should connect with classical action-theory views to be morally adequate. (I have a form of virtue utilitarianism in mind here--maximization of human flourishing, basically.) Is that at all relevant to what you're thinking?


"Moral terms are regulated by the natural facts" is suggestive, but, as you know, still vague. I'm inclined to accept some sort of supervenience thesis: no difference in moral terms (including moral responsiblity language) without some difference in the natural facts. But this still leaves room for considerations such as "not hanging on a thread" playing a role in the fixing of the referents and the construction of adequate theories, no?

Hi Dan, Alan and John,

The view I have in mind is lifted from Richard Boyd. He uses it to defend a consequentialism, but as he concedes there is no direct linkage from the causal theory to the moral theory. I actually think it is entirely compatible with some kind of virtue theory, where virtues are dispositions to act in ways that promote human flourishing. Indeed, I suspect consequentialism needs the virtues, since a calculating machine would be less reliable at promoting welfare than someone with phronesis. The story is a historical story rather than a synchronic (supervenience) story. That entails - I *think* - that reference will be fixed by all the synchronic stuff plus some historical stuff, which entails, in turn, that supervenience underdetermines reference. But I am getting outside my comfort zone.

The important question is does this get an adequate reply to Dan. It ensures that moral intuitions are unlikely to be radically astray, but the causal regulation is ongoing, and new discoveries, including surprising ones, are possible. I think we can without embarassment invoke our moral judgments as truth-apt but also be open to the revision of moral judgments from other directions, including considerations about free will.


I think that, at best, this creates a parity between theorizing about morality and free will. If that is correct, then it is not obvious that there is any reason for us to take our thinking about morality as constraining our thinking about free will apart from an antecedent commitment to the notion that morality trumps metaphysics. If anything, I think there is good reason to take our thinking about morality to be constrained by our thinking about free will (especially if we agree with Elizabeth Anscombe's claim to the effect that it is not profitable to do moral philosophy if we have not already done some spadework in moral psychology).

This is not to say that reconciliation is not unimportant. Reconciliation is one of the desiderata of our theorizing about free will and a motivation for thinking about the problem. But I don't see any obvious reason why considerations of morality should actually constrain theorizing in the same way ontological considerations do.

Nice post, Dan. Just caught up with it. We've talked about this a little bit: I'm completely on board with your methodology. I just don't think it will get you your beloved libertarianism. For one thing, it relies on the claim the libertarianism reflects our ordinary moral commitments. I'm not sure that's the case even in our neck of the woods, and I'm virtually certain it's not in other less individualistic cultures. (I'm distinguishing between our phenomenological commitments--which might be libertarian--and our moral commitments, which (in your view) would be that libertarian free will is required for moral responsibility.)

But even if you were correct, you'd have the additional burden of showing that the depth of these commitments can outweigh our commitments to evidentialist values which are quite deep. This is where most people would hop off board, stopping either at the compatibilist or hard incompatibilist stations. So I agree that it's one big balancing act, but it's going to be tough to show that the balance leads to wackjob libertarianism.

As luck would have it, by the way, I'm going over the copy-edits right now of this discussion in my book. But it's about your 2004 axiological paper. Anything I should add?

Thanks again, people. This is good stuff that is definitely helping me.

First things first, though: Tamler, it is great to hear that you are copy-editing the book! That means it should be coming out some time soon, right? Congrats!

Now, Tamler, you are jumping the gun, I think. I did mention that I want all of this to serve my nefarious libertarianism but I haven't said how just yet... one step at a time! ("step one: cut a hole in the box...")

What I was hoping to get out of the reflection on the going methodology is simply an openness to the rejection of the pairing. I'm just introducing these considerations as a way of tenderizing the meat, as it were, with respect to epistemicism and/or alethism. I can't tell if I've yet succeeded on this very minimal score.

There are three questions in particular that I'm not sure how everyone is answering.

The first is: is it true that the contemporary contributors to the free will debate deploy something like the methodology I've described?

The second is: should we be deploying this methodology?

The third is: does such a methodology reveal a kind of latent openness to a rejection of the pairing.

I'm reading Neil as answering yes to the first question, offering a qualified yes to the second, but hesitating about answering yes to the third. I'm reading Andrei as at least leaning toward a no to the second question.

Remember (back to Tamler's preemptive strike) that my bigger claim is going to be that once we give up the pairing there are going to be no general grounds for resisting the invocation of non-true-directed considerations in our theorizing about freedom and responsibility.

So do any of y'all want to be explicit about how you are tempted to answer the three questions above?

Hi Dan,

Thanks for this post. I am very much in favor of striking a balance between ‘moral conservatism’ and ‘naturalistic evidentialism’ and so I do reject the ‘pairing’ (to use your handy jargon). But the desire for this sort of balance itself seems to me to be reminiscent of an already widely shared view (outside of the free will debate) that the meaning of any philosophically interesting term should be sensitive to both conservative conceptual factors and reformist naturalistic factors. For example, David Lewis famous said that philosophical terms should be sensitive to two vectors of meaning: 1) USE – the functional role occupied by the term in the various folk platitudes associated with the term; and 2) ELIGIBILITY – the kinds and relations out there in nature that are candidates to be meant or referred to. The question of how to balance use versus eligibility has raged for years in many philosophical debates. But the respective positions have perhaps been best developed and studied in the debate over eliminitivism about beliefs. Steve Stich was an early proponent of eliminitivism, saying that there is nothing out there in nature that answers to the folk concept of belief (in From Folk Psychology to Cognitive Science). But he recants that simple view, and provides wonderful insights into how to balance use versus eligibility, in Deconstructing the Mind and in recent papers, such as The Flight to Reference. So in short, I think you are absolutely right to note the need to balance conservative conceptual/moral factors and reformist naturalistic factors. I agree this does seem to run afoul of the pairing.

To address your three questions above, I say 'yes' to all three. But I would add that the debate about free will is far from unique here. It is common practice among many, if not most, philosophers across a range of debates to strike this sort of balance between use and eligibility (and thus to also reject the pairing).

I'm late too and perhaps not following this as closely as I could be.

Isn't it the case that even in science the pairing doesn't strictly hold? For one thing, simplicity (or elegance, as John calls it) plays a role in theory choice but it is difficult to tie this down to evidence and argument. Yet once you throw in simplicity, libertarianism is in a bind, for the simple reason that every libertarian theory has exactly one more necessary condition than its compatibilist counterpart.

I agree with Chandra. Seems like a lot of philosophers reject the pairing, although few are explicit about it. Of all the philosophers who write on free will, the only one I can think of who seems clearly to reject your rejection of the pairing is Smilansky. Others might as well, but the whole idea of illusionism seems to be premised on a clear distinction between the 'truth' about free will and any pragmatic benefits that come from believing that free will exists.

Hi Dan,

I guess I'm answering yes to one and two of your questions, but not to three. Considerations that might be advanced to show that we ought to reject the pairing are probably going to be considerations that ought instead lead us to think that one of the relata - the moral pole - is rather different to what we thought. The pairing will remain, but one of the pairs will change (naturalistic fallacy? What naturalistic fallacy?)

I’m still not seeing why philosophers attempting to understand FW must be sensitive to “the kinds and relations out there in nature that are candidates to be meant or referred to.” Isn’t this methodological constraint tantamount to question begging in favor of Materialism? Why must FW turn out to be a natural phenomenon? Neil has a book coming out with a priori evidence against Agent Causalism- that concerns me. Why should I be worried about what our friends in laboratories are up to? Let’s suppose that they do one day manage to correlate brain states and choices. It is still open to me to argue that we are dealing with coincidences and that the true cause in each case is some person, Ockham’s Razor notwithstanding.

Yes, I wasn't imagining that the methodology is unique to the free will debate-- especially since it appears to bear at least a family resemblance to Rawlsian reflective equilibrium. But I hadn't made the connection to Lewis' USE/ELIGIBILITY point. That's really helpful, Chandra.

I wonder if epistemologists will be as cavalier about giving up the pairing as many (though not all-- Neil, Andrei, etc.) of us seem to be?

Robert, speaking as someone sometimes in the lab, things are worse than you think! We are not correlating choices with brain states, we are causing choices by manipulating brain states. We are demonstrating in detail the role of different parts of the brain in different aspects of decision-making and action initiation. However, the debate about compatibilism is more or less untouched by all this. Brain processes may be indeterministic, for all anyone knows right now, but that they cause or constitute choice cannot be doubted.

Can Dan or someone remind me what the pairing thesis says, so I can catch back up with this interesting discussion.

The comments to this entry are closed.

Books about Agency

3QD Prize 2014: Marcus Arvan