Some Of Our Books


« Ethics Discussions at PEA Soup: Erich Hatala Matthes' "History, Value, and Irreplaceability," with précis by Carolyn Korsmeyer | Main | What Mary Can Expect When She's Expecting »

November 17, 2013


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

Hi Jamie,

I have a question about "Metaethics and the problem of creeping minimalism": why doesn't minimalism about explanation make trouble for your proposal about what makes quasi-realism quasi? Ideally I would now explain what I mean by minimalism about explanation, but I'm not sure how to do that. Still, all the minimalist debunking of attempts at explaining the difference between expressivism and realism earlier in the paper makes me nervous about the prospects for your own explanation.

Hi Jamie,

It seems to me that the Cosmic Coincidence problem is actually worse for expressivists than for non-naturalist realists. Non-naturalist realists need some account of why we count as having beliefs about goodness rather than, say, some Kripkensteined cousin thereof. It's not clear exactly what they should say, and that's one of the most pressing challenges for their view. That said, the most obvious strategy -- i.e., some sort of principle of charity, perhaps supplemented by something like Lewisian reference magnetism -- has as a consequence that our moral beliefs have some systematic correlation with the moral facts. But this sort of explanation is obviously unavailable to expressivists (as you characterize them), since it amounts to an explanation of our moral beliefs partially in terms of the moral facts. So the problem looks particularly difficult for them. Does that seem right to you? (I haven't read your paper yet, so sorry if you address this there.)


I guess I just don’t believe there is a minimalist deflation of explanation. Put it this way: it seems to me that if an expressivist is really willing to accept non-naturalist explanations of all things metaethical, then it’s perfectly right to say this expressivist has become a non-naturalist. We should no longer try to find the secret hidden difference, at that point. Explanation is what it’s all about.


I think the question of why our beliefs are about the normative properties rather than being about the quormative ones is just a different question. Isn’t it? The Cosmic Coincidence question is about why we’re so accurate. Your question is an interesting one, which I haven’t thought about in this connection (that is, I’ve only thought about it as a problem for real realists).

Hi Jamie,

Of course I find all of your work very interesting and there's so much to ask. I just wanted to ask about the very last problem you mention for quasi-realism. I've always thought that quasi-realists have given a nice answer to problems like this since early days of Blackburn.

I take it that quasi-realists want to say that our talk about ethical truths and knowledge too us expressive of our attitudes. One thing that quasi-realists have argued this kind of discourse expresses are positive attitudes towards certain kind of improvements of attitudes and the attitudes that would result from those improvements. I take it that then the question of accuracy becomes a question of how well our actual attitudes track the attitudes we have positive attitudes towards - the ones we would have after the improvements we have positive attitudes towards.

Now, it doesn't seem accidental in this situation that our attitudes fairly well track the true judgments. After all, all this requires is that we tend to make the kind of improvements to our attitudes we want to make.

Maybe there's something wrong with this story, but at least there is a story and at least it doesn't seem obviously flawed. I've always been suspicious about the quasi-realists' story about supervenience too but again I take it that this requires systematic investigation. At least it sounds like a runner, that the purpose of our attitudinizing is co-operation which has an evolutionary advantage. Imagine we had a system of moralizing that didn't respect the supervenience constraint. In this case, this practice would be far less fit for purpose. So, it seems to me that there is a practical reason for the constraint on our attitudes and judgments. Again this seems prima facie appealing. But, I would be interested in hearing what you think is wrong with this story too.


Your expressivist story seems to me to be missing the target in two dimensions. First, it seems to be trying to explain improvement, when the task is to explain accuracy. And second, it doesn't really manage to explain improvement anyway -- if successful it explains why beings like us are bound to be pleased with what we take to be improvements, which isn't the same thing as explaining why we have improved.

As to supervenience, yes, that is the standard sort of story. It explains the wrong thing. It explains why we have attitudes that respect the supervenience constraint. But the challenge (and this is the challenge to the non-naturalist, after all) is to explain why these sorts of properties are necessarily connected with those sorts of properties.

Hi Jamie

thanks. I think there's a disagreement here between expressivists and others about what counts as an explanation of accuracy and supervenience. I take it that the expressivists will believe that all claims about accuracy and the explanations of why our beliefs are accurate are all normative claims very much made from the internal moralising perspective (I assume that the latter would be about the epistemic reasons to hold our moral views). Blackburn often emphasises the holism of moral epistemology and the Neurath's boat picture. The expressivists then give an expressivist account of what attitudes these claims about accuracy and explanations express, and it is at this point that the previous story about desire-like attitudes towards certain kind of attitudes and changes to them comes into picture. But, once this story is told, there isn't much left to be explained. Now, you might not be happy with this, but I don't immediately see what further explanatory burden there is for the expressivist.

The same seems to go for the supervenience challenge. By explaining the constraint on the attitudes and practice, the expressivist seems to have explained the constraints on properties - after all the properties are projections of our attitudes on the expressivist view (all property talk is after all done from the internal moralising perspective). Again, this might not be satisfactory but it would be good to say more about why this isn't enough.

I guess the idea is that there are no explanatory mysteries from the internal moralizing perspective, and the external metaethical perspective can give an account of why the internal moralizing functions in the way it does. The further external metaethical explanatory questions seem to drop out here - formulating these questions requires metaethical claims which the expressivists are going to accept only as internal to the practice.


Let's see if I have better luck with the spam filter if I sign in to my Typepad account to comment.

You may be right but I have not seen any such explanation of our normative accuracy. As I understand you, the explanation is supposed to be normative. So, how would it go? I’m having trouble imagining it. “We are so awesome, we just have to have the best possible normative beliefs.” Hm. I don’t think explaining our normative accuracy is a part of normative theory. Also: Neurath’s boat is a metaphor used in stories about justification. My question isn’t about justification. It’s about the explanation of our accuracy.

“The properties are just projections of our attitudes.” Well, if this is meant in a Humean constructivist way, or a Response Dependent Properties kind of way, then great, this will answer the question nicely. But that’s not expressivism. The advantage that Humean contructivists have over expressivists on this question is that the former, but not the latter, can happily say that our attitudes and our constraints on them *make the properties what they are, so they can explain why the normative properties are linked to the natural ones. But expressivists just don’t believe that!

My thought was that 'real' non-naturalist realists can (and probably should) appeal to considerations of charity to answer the Cosimic Coincidence challenge, but expressivists can't do the same.

Hi Jamie

it's better put in Gibbard, Thinking How to Live pages 233-235 even if I admit that this too is very brief. But there's a short story of how accuracy judgments are normative judgment and an expressivist way to understand these judgments in terms of planning to rely on people with certain qualities.

Okay, first, Shoe has set this thread to Moderated, but the effect is not what I expected. Well, anyway, I can now clear my own comments from the Typepad spam filter.

Jeremy: I think I get it. Charity in interpretation is one way to be confident that a whole community of speakers cannot all be radically mistaken; expressivists won't have that resource. Hm. Why wouldn't expressivists have that resource, exactly? This is pretty murky to me.

Jussi: I'll have to look at those pages (I don't have a copy here today). But they are about knowledge attribution, I believe. The problem I have in mind is not about knowledge at all, but about accuracy, and not about attribution either. Do those pages really offer an explanation of why human beings are very accurate in our normative judgments?

I thought your view was, roughly, that quasi-realists aren't allowed to appeal to the distribution of moral properties in their metasemantic account of moral belief. I would have thought that any charity-based metasemantics violates this prohibition.

Oh, so, charity says, what determines what 'right' refers to is (in part) what is right. But that means rightness is playing a ground-level explanatory role, which expressivists don't believe ever happens.

Hmmm. Yes, I guess so. If that's how Charity works. Not sure about that part...

Yeah, I don't know how 'charity' works. But I sort of know how Lewisian reference magnetism works, and it seems like the sort of thing expressivists definitely can't appeal to

Hi Jamie

it comes down to what you expect from such an explanation. In some ways you are right. Gibbard does address attributions of reliability. For him, it seems like such claims are normative claims (as I assume they have to be for the expressivists). As such claims, they express planning states. In effect, we plan to rely on other people with certain qualities and their advice and we express these claims by saying who are reliably right and why. Furthermore, from the external perspective there is an explanation of why we have the practice of making reliability claims on certain grounds.

As far as I see, from the expressivist's perspective this is all there is to explain why human beings are very accurate in our normative judgments. I also assume that you will think that there is more than this for the expressivist to explain. But, it's not clear to me what that explanatory challenge is and why the expressivist must be able to say something more.


I don't see how any of that is supposed to be an explanation of why humans are so accurate in our normative judgments. Maybe you didn't mean it to be -- you say "this is all there is to explain", as if the fact that humans are very accurate had been expressivistically explained in some way.

Suppose a zoologist wonders why baboons are good at telling when things are very hot. You tell her, "Well, when one judges that a baboon is reliable in its temperature judgments, one plans to trust the baboon's temperature judgments when the occasion arises." The zoologist, oddly enough, does not think this answers her question. You go on, "Now I will tell you why zoologists have the practice of making temperature-reliability claims on the grounds that you do make them." She is getting annoyed. "Well," you say in your defense, "it's not clear what your explanatory request is and why I must be able to say something more."

But, it is quite clear what the explanatory request is. It is a request for an explanation of why baboons are so accurate in their judgments of when things are very hot. You haven't offered one. As to why you must, well, perhaps you needn't, but then you should just say, "Yes, I see what you are looking for, but I don't have any explanation."

There is, of course, a good Darwinian explanation of why baboons are so reliable in their judgments of which things are very hot. Some metaethicists think there is a good Darwinian explanation of why human beings are so accurate in forming normative judgments. Those metaethicists are not expressivists, nor are they non-naturalist realists.

Hi Jamie,

Enoch offers the following sort of reply to the corresponding challenge to non-naturalism:

(1) There is some respectable naturalistic explanation of why we tend to make the sort of normative judgments that we do.
(2) Together with some plausible assumptions about what the normative truths are, this explanation gives us an explanation of why we tend to make accurate judgments.

I tend to think that this is a good reply to the (or at least a) cosmological explanatory question, and it seems that quasi-realists could make equally good use of it. But I take it that you disagree. If so, what is missing?

(I also think that this sort of reply leaves semantic and epistemological worries untouched. But you set the epistemological worry to the side, and quasi-realism might perhaps be able to avoid the former by not requiring any (substantial, external) account of reference.)


Well, all I really want to say is that Nonnaturalists and Quasi-realists are in the same boat with respect to this issue.

But in fact I don't think that explanation is sufficient. I'm pretty sure Enoch doesn't either -- as I recall he says it manages to reduce his cost and answer some epistemological questions, but does not claim to have provided the full explanation.

Anyway, the reason is that in general you can't explain why a process is accurate by explaining why it tends to say what it says and also making is plausible that what it says is true. That will sometimes leave the accuracy looking like an amazing coincidence. Are we just lucky that the forces that shaped our normative judgments led us to accuracy rather than systematic error? Well, maybe we are -- that's just to say there is no explanation of our accuracy.

My colleague (and David's old co-author) Josh Schechter discusses this in "The Reliability Challenge and the Epistemology of Logic".

Jamie Dreier has a nice piece related to this: "QUASI-REALISM AND THE PROBLEM OF UNEXPLAINED COINCIDENCE." He starts off with a language like English save its lack of normative predicates and lack of ability to evaluate at all. No unexplained coincidence (or accuracy) there. Then he considers successive stages where non-predicative ways of evaluating are introduced, then predicative, etc., until we get to the quasi-realism we know and love. At no point does it seem like the additional complexity in the expressive power of the language introduces a coincidence (or accuracy) problem. After all, at no stage do we introduce something normative with extensional purport that has the potential of being accurate or not.

This seems like good reason to think that any appearance of an explanatory problem for the quasi-realist is illusory. No explanation needed. Or at least, realism has more to explain that quasi-realism, for in building realist language we would introduce normative terms with extensional purport, and thereby introduce an extra explanatory burden for any accuracy. So at least we quasis are not in the same boat.

Don't you like this argument? You probably aren't satisfied. Jamie wasn't either, if memory serves. But what's missing?

Great suggestion, Matt; what a brilliant and convincing writer that guy is. (I did link to the PhilPapers page for that paper, actually. The paper itself is behind a paywall, but I'm happy to send it to anyone who requests a copy.)

The final sentence of that paper is:

So my conclusion is that, pace Street (and Schechter), we can see that the Problem of Unexplained Coincidence is not a problem for quasi-realism, but we, or at least I, cannot yet see quite why it is not.

I would be happier if it were clear from the account of the difference between lightweight, deflated properties and heavy-duty robust ones why match-ups of the latter need no explanations.

Hi Jamie, I have a question on your first worry about non-naturalism: that it divorces normative judgements too sharply from motivation.

Do you think this worry might be appeased by a desire-as-belief theory, or some nearby variant? If we think of desires as both having motivational import and as being representations of the normative, they might provide a link between the normative and motivation that the non-naturalist can exploit.


I don't think the explanation would go like that. I'm going to drop the baboon example, but think of ordinary situation in which you know that one of your friends has very strange and inaccurate moral views and another person you know has very accurate moral views. You are asked to explain why the second person is more accurate. The explanation might say things like this person doesn't let their selfish interests or biases get in the way of their judgments, they are more coherent and reflective, they are more sensitive to the suffering and situation of others, they care more about animals and biodiversity, they are better informed about many things, they've taken part in more open discussions, they consider ethical questions from many perspectives, they grew up in loving family, and so on and so on.

At least in ordinary contexts, these things go some way towards explaining why one person is more accurate ethically than the other. You might not agree with everyone of the explanations and you might have others to add. But, in normal life, explaining accuracy in these ways is natural. I take it that the expressivist can accept this and then give an expressivist account of what plans are being expressed when we give explanations like this.


I don't, because to me it looks like that story forges no connection at all between the normative properties and the motivation. Suppose it's true, there are these hybrid or dual-aspect states that motivate us and represent the normative world. Well, I figure there are also hybrid states (maybe you and I never have them) that both motivate us and represent the world of ceramics. That doesn't entail that reasoning about ceramic is practical reasoning. So why would reasoning about normative properties be practical reasoning?

Maybe the idea is that what is interesting about normative judgment is not its subject matter (the non-natural properties) but the functional role of its concepts and modes of presentation. Ah, now I get it! But the non-natural properties turn out to be a free wheel in the theory, in that case. So it seems to me, anyway.


My main answer is: all of that is also available to a non-naturalist to say. So, quasi-realism is in the same boat with non-naturalism.

I think your story explains something different from what I was asking about. Why is Shelly-Ann so fast? Well, she practices and trains and eats and devotes her life to it, and she has an ideal human natural endowment. But that's why she runs fast for a human. Why do humans run fast? That needs a completely different kind of explanation. (And this is why I like the question about detecting temperatures of things.) So, your story may give a good explanation for why one friend has more accurate views than another, but it won't explain why human beings have the ability to form accurate normative beliefs in the first place.

Hi Jamie,

Thanks for your reply. Let me try once more. Won't the quasi-realist want to accommodate claims such as "That lying is wrong explains why we correctly judge that lying is wrong, and not vice-versa"? Isn't that enough to have a (deflated) explanation of the judgment in terms of the (deflated) moral fact? What does the non-naturalist explanation amounts to, if it amounts to something more than that?

Juan, my advice to the quasi-realist is to deny that. Lying's being wrong does not explain why we correctly judge that lying is wrong, is what I recommend.

Maybe there's some pressure to agree with the explanatory claim that I'm not feeling?


My thought was that this follows from the other minimalisms that your quasi-realist accepts. In general, what explains that I correctly judge that P? Part of the explanation has to be the fact that P, just because S correctly judges that P only if (and perhaps if as well) it is a fact that P. Given that the quasi-realist accepts (deflated) moral facts, why would he deny that?

I get the main idea. What is it for someone correctly to judge that p? Well, it's for that person to judge that p, and p. So p will have to be part of the explanation.

Okay, sure. Because, after all, it's a normative fact! Normative facts always (?) have normative explanations.

But my criterion isn't to do with what explains correct normative belief.


How's this for a first stab at your final problem. As you note, a central pillar of expressivism is that its metaethical explanations do not make any reference to values, moral truths, etc. Now, this is not an accidental or peripheral feature of the view: Blackburn (for instance) claims that the definitive feature of any pragmatist approach ( According to him, the expressivist, qua pragmatist, begins by denying that the philosopher's job is to "vindicate" our moral reliability.

My thought is that the realist's burden might be more serious because she (standardly) operates under very different methodological presuppositions. That is, realism is premised on the anti-pragmatist idea that normative thought and language function to "map" mind-independent features of reality. When one begins there, one incurs the burden of providing a meta-ethical explanation of how the mapping succeeds. But when you are a Blackburn-style pragmatist, your inquiry begins by denying that such explanations are even required, let alone possible. This may appear evasive, but I think it's motivated by the observation that the 'representationalist' method has run its course in metaethics. According to the expressivists, we just have to stop thinking that way, or we'll keep spinning the same wheels in the same mud.

So, perhaps accusations like "you've failed to explain [supervenience/normative reliability/etc...]!" are implicitly relativized to the methodological assumptions under which the target is operating.

I think we agree that there is no explanation of accuracy for the quasi-realist (and so disagree with many other posters here). And we agree that accuracy need not be explained for the quasi-realist - there is no problem. But that thing that convinced me that there is no problem--your argument from modifying simple language--also explained to me why there is no problem. Namely, in the simple language there is no accuracy to be explained and each step on the way to quasi-realism introduces no accuracy-realted explanatory burden. What I'm puzzled about is how that argument convinced you that there is no problem without explaining to your satisfaction why there isn't one.

I think some of the other comments suggest that accuracy can be both heavy-duty and light-weight (creepy). And they push lightweight accuracy explanations within the first-order discourse. Your comments with Jussi suggest to me that the sort of accuracy you ask after--general ability to be accurate in some domain--might only have a heavy-duty reading. If so, then non-nattys and quasis can't explain that accuracy. But when we start with simplified English (stripped of evaluation) and contrast what it would take to build a quasi-realist language with what it would take to build a realist (purport) language, that explains to my satisfaction that realists have an explanatory burden vis-a-vis accuracy that quasi-realists lack. For their language engineering introduces terms with extensional purport and ours does not.

Does any domain have an explanation for light-weight general accuracy (of the sort you are interested in) that quasi-realism and realism lack? That would be odd, for they should enjoy the same light-weight explanations as anyone else. Maybe your temperature example? But that seems to tap into notions of heavy-duty accuracy.


That's what I suspected you would say. But how does the moral fact enter into an explanation of the judgment for the non-naturalist? In the paper you give the following example:

(J) Julia believes that knowledge is intrinsically good.

And you say that the difference between Moore and Gibbard is that Moore (but not Gibbard) would say that "(J) must consist in Julia’s standing in a certain doxastic relation to knowledge and intrinsic goodness." But Gibbard need not deny that if he is the right kind of minimalist. For instance, Gibbard could say that of course (J) consists in Julia’s standing in a certain doxastic relation to knowledge and intrinsic goodness, because Julia believes that knowledge is intrinsically good, and that is what it is to stand in a doxastic relation to knowledge and intrinsic goodness. Now, Gibbard will go on to say that, in addition, (J) consists in Julia being in a certain sort of planning state, so maybe you want to say that this is where the difference lies, because the Moorean would not want to add that. But why would the Moorean not want to add that? The Moorean cannot be forced to deny the functional "hotness" of judgments like (J), right?

Blackburn does deny that there is any need to vindicate our reliability. But I’m not asking for vindication. (I’m not exactly sure what it is, but I think it’s epistemic.) I’m asking for an explanation.
I know Blackburn thinks we don’t need one. But that just leaves the pattern of accuracy as a massive coincidence. It could be a massive coincidence, but explanations are better.
Mapping to mind-independent features of reality is a nice metaphor. When we spell it out, I think, it turns out to be something no quasi-realist wants to deny. Do our beliefs map to the moral truths? Sure – we are accurate, and that’s a nice 1:1 mapping! (We’re not perfect, so it’s a makeshift map.) Are the moral truths mind-independent? Of course they are. That’s what our independence counterfactuals say, and we endorse them.

As to methodological assumptions: here is mine. When there are big coincidences, like supervenience (which is a coincidence that echoes all through modal space), or very high accuracy, a theory is better when it can explain them. I think that’s a pretty good assumption, and it’s the only one I need.

Well, hm. Maybe I am wishing for something that doesn’t make sense, when I wish we could see why there is no explanation needed. But, I feel like the difference between the lightweight and the heavyweight leaves it still kind of obscure why lightweight coincidences don’t need explanations. I think this often happens: we have a demonstration – and induction is often the type of explanation, in fact -- that something is true, but we’re left feeling unsatisfied because we can’t see why it’s true. However, it’s possible that I’m searching for the patron saint of modesty, who blesses all and only those who do not bless themselves.

The difference, I think, is that the expressivist denies that the doxastic relation thing explains what it is for (J) to be true. He doesn’t have to deny it, he just says it is no part of the explanation. He thinks the explanation is the thing about planning, emotional dispositions, or whatever.


I see, so the idea is that, although both the Moorean and the expressivist accept that (J) consists in the doxastic relation thing and that it also consist in the planning thing, the Moorean thinks the doxastic thing is more basic whereas the expressivists thinks the planning thing is more basic. But just to be clear, the expressivist does not think that the planning thing is more fundamental in the sense that it explains the doxastic thing, but rather that it explains (or fundamentally explains) (J), right? But if so, to what camp would someone who accepts both the doxastic and the planning things and think that they are both on a par explanatorily belong? Are not some possible (or maybe even actual) expressivists or realists like this?

Sounds right. (Now I’ve lost my ear for ‘consists in’, so I’m not 100% comfortable with it, but it seems innocuous.)
Someone who thinks both explanations are on a par? Hm. I find that a little hard to understand. What if there were a planning state, just like the ones you and I have, but that had no normative properties as any part of its fundamental explanation? I guess really what I’m wondering is, does the person who has this dual explanation view think the normative properties themselves are carrying any actual explanatory weight?
If so, that seems like a realist view to me. But once we understand what it’s saying, it doesn’t matter very much what we call it (or whether we classify it as Realist).

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.