Some Of Our Books


« In the News | Main | Ethics Discussions at PEA Soup: Gunnar Björnsson and Stephen Finlay, "Metaethical Contextualism Defended" »

November 19, 2010


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

Very interesting post, Jussi. I admit, though, that I am not quite sure I get what positions qualify as metaethical dualists, and so, who would be able to benefit from your point here. So rather than try to formulate a challenge, let me just ask for a clarification. Can we deal with the example of some fact being a reason?

We might think that the fact that some act would cause unnecessary pain is a reason not to do it. Now, for the non-reductionist about reasons, this is a reason regardless of what we care about, how we judge, etc. I take it that, in the spirit of Chalmers, you want to posit supervenience bridge-laws in a way that makes it not implausible to call a non-reductionist a naturalist here. Is that right?

If so, then you think we can mitigate the queerness objection. But isn't that objection just that the particular fact's being a reason gives it the property of 'to be avoided', or 'not to be done' or something? Whether we call it natural or not, saying that this fact about the world has such a property strikes some of us as weird, and so I guess I wonder why we wouldn't just then say that dualism is weird, whether it is naturalistic or not.

Did I totally miss the point? That may well be, but hopefully this is at least an opportunity to clarify!

Interesting post, Jussi. What about epistemological worries? I understand how I have epistemic access to my conscious experiences and how I could possibly discover laws that connect them up with physical, conscious-making properties. But I'm not clear on how things work for the "dualist in ethics." Thoughts?

Interesting post! I'm a little resistant to the analogy, though it's difficult to pin down the precise reason. Morality just seems more clearly "transcendental" than consciousness. There's a sense in which qualia have a kind of concrete being, as part of the fabric of the cosmos, which abstract moral properties seem to lack. The non-contingent nature of moral principles (in contrast to the psycho-physical bridging laws) is perhaps one sign of this.

I haven't read the Chalmers yet, so there's one important issue that's not clear to me.
Are the mental/physical laws causal? Does Dave say that my pain is caused by my neural firings?

That's not, I think, how moral dualists think of the 'because' (the one that Mackie was worried about).

I just lost one set of answers for some odd reason, but here's another go.


the issue is not quite whether reasons depend on what we care about, how we judge, and the like.

Rather, let's take an act of torture and the fact that this would cause pain. Intuitively, that fact is a reason not to do that act. The dualist (i.e., non-naturalist) believes that this is because there is an additional, normative 'counting against' relation between the fact that pain would be caused and the act of torture.

In contrast, the monist would say that the claim that that it would be painful is a reason not to torture is made true by some physical relation between the fact and the act - such as that the obtaining of this fact would in part explain why any of the desires of the agent would be frustrated. Now this reduction might make the reason-relation dependent on the desires of the agent but presumably there could be reductions where this wasn't the case.

Now, by Chalmers's lights, the dualist position would be a naturalist one if there were some principles that captures which facts give rise to reason-relations to which acts. As such, this doesn't make these relations any less queer (or the further normative properties like oughts that we also get with reasons). But, I've never quite been convinced by queerness worries anyway. But, yes, here there would be room for queerness objections to naturalist views. Hope this helps a bit, at least this seems to make the case that you didn't miss the point.


I share your epistemic worries - they have driven me often to some versions of anti-realism. But, it's not like there are no dualist attempts to deal with those issues by McDowell, Wedgwood, Shafer-Landau, Scanlon, Huemer, and many many more. So, yes, I would start from those. Personally, I think that the processes we use to think about moral properties are fairly standard deliberative procedures. The question of justification then is an external question about the reliability of these processes. So, we might never know whether we know but I'm fine with this.


I was thinking about the transcendental in the quote too. But if you think of transcedental in the Kantian sense (something that is a necessary precondition for something that actually obtains), then it's not clear that morality is transcedental in that sense. So, then the question is in what sense is it transcendental. For me, morality seems pretty concrete anyway...


I'm not sure I know what Chalmers thinks of causation. He seems to prefer the term 'arises' and sometimes is ok with 'emergence', and never quite puts it in terms of causation. Whether these arising relations are or need to be causation is another question. There is a section on whether phenomenal properties themselves have causal powers or whether they are epiphenomenal, and there Chalmers seems to be saying that no one really quite knows what causation is in the first place. If we accept a Humean view or a nomic view of causation, then it comes up trivially that the physical causes the phenomenal/moral but this won't give rise to any extra worries. In contrast, if we think that causation is a sui generis relation then maybe (Chalmers is fairly sympathetic to causation being an extra relation on page 86).

Searle of course thinks that the relation between the physical and the phenomenal is a causal one.

"Firstly, it is true that Chalmers’ argument for property dualism do not seem to apply in metaethics. After all, he argues from the conceivability of zombies to the failure of logical supervenience of the phenomenal on the physical. This means that the physical does not entail the phenomenal which thus has to be an additional feature of the world. Yet, in metaethics, it would be awkward to propose that there are ‘zombie worlds’ or ‘inverted spectrum worlds’ which would be physical duplicates but would lack moral properties or have different ones."

Hey Jussi,
Quick question about that awkwardness. If there are zombie worlds, aren't there worlds that are physically indiscernible from ours with no phenomenal states? It's not obvious to me that our zombie twins have moral status if they have no phenomenal states.

(Fwiw, I've always thought that the moral case is an interesting test case for zombiephiles. I'm about as good at deriving an "ought" from an "is" as I am at deriving a "phenomenal" from a "non-phenomenal-is", but (maybe like you), since I'm not tempted to draw from this any conclusions that would support the error-theory or the possibility or morally inverted worlds, dancing duties, morally zombified worlds, etc..., I don't see why we'd rush to draw any conclusions about the mind from the (alleged) conceivability of zombie worlds.)

I'm not sure this adds much to Doug's point, but I'll say anyway that it looks like the issue will come down to what we are to understand by 'natural' and 'naturalism'. At least Russ S-L wants to cash these out in terms of epistemic methodology. Since we have first-person access to phenomenal states, it might be plausible to think that we can use the methods of the best natural sciences to discover laws that connect the physical with the phenomenal. But since we don't have first-person access to moral properties, and since moral properties are not scientific properties, it's hard to see how the methods of the natural sciences alone will help us discover a lawlike connection.


I once posted about zombies here:

I agree with you but I think we can deal with this worry in the following way. We should add that moral properties do logically supervene on physical and phenomenal properties. This means that, if Jackson's right, then moral could be reducible to physical plus phenomenal. And, the dualist in metaethics cannot argue for the distinctness of moral properties by showing that they vary independently of the physical and phenomenal. Some other argument has to be given to avoid the Jackson argument.


Chalmers would definitely oppose this statement you make:

"Since we have first-person access to phenomenal states, it might be plausible to think that we can use the methods of the best natural sciences to discover laws that connect the physical with the phenomenal."

he explicitly argues that because of the epistemic asymmetry (first personal epistemology of the consciousness), the best natural sciences can never discover the bridge-laws that connect the physical with the phenomenal. Rather, because of this asymmetry, all sciences can do is to assume such laws.

Also, it's not clear whether moral epistemology is first or third personal. My immediate reaction is that neither really.

Okay, well, I didn't really mean it's critical whether the 'because' is causal (even though that's what I said). I'll trade in my original question for one of Richard Chappell's: Is the 'because' contingent?

Yes. Experiences naturally supervene on the physical but not logically. So, in all the worlds that share our psychophysical laws you get the same phenomenal facts. Yet, in the words that lack these laws (which are logically possible) you either get different phenomenal facts or no phenomenal facts at all even if the same physical facts obtain.

One might think that this is a relevant difference to the moral case. And, often Chalmers writes as if logical supervenience, and thus the necessity of the bridge-principles, would be sufficient for reductive explanations and materialism. But, as I argued in the OSME paper, you might think that logical supervenience is only necessary but not sufficient for reduction.

This would mean that the contingency/necessity of the because only addresses the extension of the because but not its nature.

Hi Jussi. Yes, you're right that 'discover' is not the right word. But I'm pretty sure 'assume' isn't either. (Chalmers doesn't just assume that there are laws relating the physical with the phenomenal, does he? Surely this is an inference one draws from repeated observation of relations among phenomenal states/properties and physical states/properties?)

All I wanted to point out was that a moral "naturalist/nonnaturalist" debate between you and, say, R S-L would likely come down just to how 'natural' is cashed out. For example, if I'm understanding him right, R S-L would cash out 'natural' in epistemic terms, namely, as that which is discoverable wholly through a posteriori inquiry. Since neither phenomenal properties/states, nor moral properties/states, nor (thereby) any bridge laws/principles connecting these to the physical are discoverable wholly through a posteriori inquiry, both would be considered non-natural on Russ's understanding of 'natural'/'nonnatural'. So, you can deny that 'natural' should be cashed out in terms of that which is discoverable wholly through a posteriori inquiry. In that case, I'm thinking Russ would likely say (a) that the burden is now on you to specify what else it is that makes a property "natural," but (b) that in any case, such a debate wouldn't be very substantive.

Well, assume seems to be the right word for natural sciences if you look at the discussion of the neurobiological sciences for instance on page 115. Of course Chalmers himself takes part in the discovery of the psychophysical laws but this is not an investigation in any of the natural sciences. He doesn't need to assume the laws because he has a first-personal access to his experiences. But this wouldn't work from the third-person as we don't have third-personal access to phenomenal states.

I just think that the a priori/a posteriori distinction might not be able to do the work to carve the natural/non-natural distinction. First, as Williamson for instance has argued the whole a priori/a posteriori distinction is deeply suspect. Second, I think phenomenal properties would go to the a posteriori side (think of Mary - could she gain information of them through a priori reflection?). If the properties would be a posteriori, the bridge-laws would be too. Of course with morality things are differently.

As to regards to what Russ would say, for (a) I just gave an alternative proposal from Chalmers which seems to be much closer to an intuitive understanding of natural. With (b), I think I would be with Russ. We got two interesting questions, one ontological question about monism vs dualism - are moral properties over and above physical ones?, and another epistemological question - do we know of them a priori or a posteriori? The question of naturalness is far less interesting than these.

But, in this situation, I don't see what the point of dualists is to insist that their view is a non-naturalist one. I think Chalmers is right that there's going to be room for naturalist dualism.

Hi, Jussi,

Could you say a bit about what you mean by a law and why we might think there are laws connecting physical properties with moral properties? In the phenomenal case, it's pretty easy to see why you might think there are laws connecting physical and phenomenal properties, even if you think latter can't be reduced to the former, e.g. by noticing that when you take acetaminophen, your headache goes away. If the case for naturalistic dualism depends upon the claim that there are physical-moral laws, it seems that we should have some reason to think they exist.

Hi Jussi,

Thanks again for the interesting post.

Yes, I wasn't implying that you wouldn't have any response to a non-naturalist like R S-L, but rather just trying to locate the likely 'naturalist'/'nonnaturalist' fork at which you and a non-naturalist would diverge.

About this:

He doesn't need to assume the laws because he has a first-personal access to his experiences. But this wouldn't work from the third-person as we don't have third-personal access to phenomenal states.

That brings us back, I think, closer to what I think was Doug's point earlier. From first-person access to phenomenal states/properties and third-person access to physical states/properties, it looks one would be justified in drawing inferences about the relations among these and state these relations in terms of laws/principles. But as you suggested earlier, one might think we have neither first nor third person access to moral properties/states, and so it's not clear how one is to know or be justified in believing any moral-physical laws/principles, or even that there are any. (Again, if I'm understanding him right, this is where R S-L would invoke self-evidence.)

Sorry about the late responses,


most dualists in ethics seem to be some kind of Rossians so it's easy to start from their view.

So, Ross, for instance, gave a list of prima facie duties of fidelity, reparation, gratitude, non-maleficence, justice, beneficence, and self-improvement. He also thought that 'the moral order... is part of the fundamental nature of the universe'. So, it seems like he thought that there are laws that tell us that, in so far as and act is an act of fidely, reparation, and so on, it is an obligation. That is, these laws capture how these features of acts make them morally required which is a new, distinct property of these acts.

Of course, you don't have to be a pluralist in this sense. You can also be a monist - for instance a utilitarian would say that there is a law that acts are right in virtue of maximizing well-being. You could even be a particularist and think that the law only connects disjunctions of whole world descriptions to moral properties.

I should say that in Chalmers' case the basic psychophysical laws connect phenomenal properties to functional properties which he takes to be multiply realisable physical properties. I believe that in the same way the moral dualist could first connect the moral properties to multiply realisable (on the physical level) functional properties.

Why do we think that such laws exist in the moral case? Well, it seems like acts do not just happen to be right and wrong - rather we can and need to always specify some considerations of why they are wrong. This is what has led ethical theorists to attempt to capture these right-making, good-making and so on considerations in ethical principles. The dualists just give these principles their own metaphysical reading.


when I said that we don't have neither first nor third person access to moral properties, I didn't mean to suggest that we don't have an access to moral properties at all. I'm fairly confident that we know that torture, rape, and so on are morally wrong and that helping others is good and right. Rather, it is just hard to tell whether this access is first or third personal.

Hi Jussi,

I haven't thought through all of these issues, but just for effect, I'll press one more time on the epistemology. Thus far in the thread, you've said that you suspect we have neither third person access to moral properties/states, nor first person access to moral properties/states, nor do we know moral claims/principles a priori and, so likely would also say that neither are moral claims/principles self-evident. Can you see why one might think there would be a potential problem justifying or knowing moral claims or moral-physical principles/laws on the "naturalist" view you're suggesting? (I'm not denying that you might have a response. So far, though, it seems like we disagree about whether there even is a potential problem.)


Dan - I can see that there is a problem - that the dualist needs to provide an epistemology. I just don't think that the claims you have before the 'can' claim are useful ways of approaching epistemology in this context.

For what it's worth:

a) I don't think moral claims/principles are self-evident. This would mean that believing them itself would justify the belief. I've never felt attracted about this and I say more about it here:

b) a priori/a posteriori distinction. The way Williamson discusses this is that, first, we are used to distinguishing between two roles experience can play in cognition: evidential and enabling. So, experience plays an evidential role if I know by seeing that this shirt is green, but the same kind of experiences play merely an evidential role in my knowledge that all green things are coloured - here I need the experience to acquire the concepts of green and coloured but there after I can go on by deliberation.

Now, traditionally, the idea has been that if the experience plays an evidential role then we have a posteriori knowledge, and if it plays merely an enabling role, then we can still have a priori knowledge in question.

Yet, it seems to me that, just like in the case of Williamson's counterfactual knowledge, in ethical knowledge experience can play a role which is strictly speaking neither evidential nor merely enabling. This is in part because our past experience and imagination of possible cases play a similar role here as in case of counterfactual knowledge. For instance, consider when I think about whether torture is wrong. Part of this judgment is considering the distress and suffering of the tortured person. Here I rely on my past experiences of seeing people in pain and being in pain myself. Are these experiences evidence for the wrongness of torture, or do they merely enable me to make the judgment? Neither seems to be the case strictly speaking. And the consequences of classifying this kind of knowledge as either a priori or a posteriori are going to be implausible more generally (see Philosophy of philosophy sec 5.5 here).

So, here seems to be another example of knowledge that isn't quite either a priori or a posteriori. If so, then the S-L way of looking at naturalism is going to be in trouble.

c) same kinds of things can be said about the first person/third person knowledge. When I think of the wrongness of torture, I do feel it within me in the Humean way and need to do so, but still I need to be able to give reasons and justification that is also acceptable for everyone else. So, in this knowledge, we seem to have elements of both first-personal and third-personal knowledge just like in say mathamatical knowledge. So, this distinction doesn't seem to workable here either.

But, now the question is, are there any respectable positions in epistemology available for the dualist if she rejects self-evidence and the distinctions of moral knowledge being a priori/a posteriori and first/third personal? It seems to me that there are. For instance, the kind of reliabilist social externalism I outline here:

seems to fit nicely the kind of dualist metaphysics I talked about in this post.

Thanks, Jussi. I can certainly see the direction you'd go.

The comments to this entry are closed.

Ethics at PEA Soup

PPE at PEA Soup

Like PEA Soup

Search PEA Soup


  • Unless otherwise indicated, the views expressed in any given post reflect the opinion of only that individual who posted the particular entry or comment.