Some Of Our Books


« Ethics Alert | Main | Parfit's CtM, Chapter 11: Contractualism »

August 03, 2006


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

I'm a big fan of that paper. If I don't remember too wrong much of the latter part of the paper discusses how we are able to expressivism/cognitivism distinction when all this minimalism has creeped in. I take it that the realism/anti-realism distinction is a separate one, and for that reason I'm slightly sceptical about the way you phrase the Euthyphro-contrast. For one, it assumes that realism must go hand in hand with some form of Putnamian externalism about mental states.

When it comes to the realism/anti-realism distinction, I'm quite happy to start from Crispin Wright's criteria from Truth and Objectivity on what kinds of truths, facts, properties, and so on we have within the subject-matter of a particular discourse. You bring up the Euthyphro contrast (in a different way) which Wright also discusses but there are others like whether the truths in the discourse are evidentially constrained, whether the discourse exerts Cognitive command, and how wide of a cosmological role the facts of the discourse have.

I'd been wondering about the Euthyphro as a more generalized instrument for exploring irrealism, but due to my woeful education on this topic I hadn't known that Wright had also treated it. Is this in Truth and Objectivity, or is there another cite(s) for it? Has any other metaethicist gone into this?

Dear Andrew,

yes - there's an excellent appendix on one of the chapters of Truth and Objectivity on the Euthyphro and its role in the realism debates (chapter 3 I think). That also created some discussion as a result. There's been more discussion about the contrast on the basis of the so-called 'best-opinion' or responce-dependence accounts of metaethics. There's a good section about this in A. Miller's Introduction to Metaethics.

Uriah, thanks for your very clear overview of Dreier's paper.

I think, however, that Dreier is too fast in addressing the issue of mind-independence. I’m not quite sure that we can accept the following conditional as a *platitude* about the notion of a fact:

(a) It is a fact that P if and only if it is mind-independent that P.

(And I would have the same kind of doubt about any strictly analogous schema relative to the notion of a property or of truth that were proposed as a platitude).

The reason of my scepticism is that it does not seem to me that, to understand the everyday (pre-theoretical) notion of a fact, a speaker has to accept the validity of (a). I do not believe that (a) is conceptually true.

There are indeed many instances of (a) which are presumably false – e.g. ‘it is a fact that I feel pain if and only if it is mind-independent that I feel pain’. (Suppose that, now, it is a fact that I feel pain).

The reply could be that I’m just using a *tick* notion of mind-independence (MIND-INDEPENDENCE), while I should use a *thin* one (mind-independence). But what kind of thin notion of mind-independence could ever make the above biconditional true?

Someone might suggest that we have different platitudes for different notions of facts (properties and truth). For instance, the following schema might appear to express a platitude:

(b) If it is a physical fact that P, then it is mind-independent that P.

Perhaps (b) just expresses a platitude. But a more relevant schema, for our concern here, is:

(c) If it is a moral fact that P, then it is mind-independent that P.

Can we accept all instances of (c) as conceptually true? To understand the everyday notion of a moral fact, should a speaker accept the validity of (c)? Well, I’m not sure.

While we are naturally (pre-theoretically) realist about *physical facts* (this might mean that (b) is a platitude), it seems to me it is at least debatable whether we are naturally (pre-theoretically) realist about *moral facts*.

But what does it mean the expression 'mind-independent' in (c)? The tick/thin objection still apply. Am I referring to MIND-INDEPENDENCE or to mind-independence?

Well, a very weak or very thin notion of mind-independence that applies to facts is probably given in terms of the satisfaction of this condition:

(mi) It can be a fact that P and someone can judge that not-P.

(Briefly, facts are mind-independent in the minimal sense that we can make mistake in evaluating whether they obtain or do not obtain).

Plausibly, when 'mind-independence' is cashed out in terms of (mi), one should accept (c) as a platitude about moral facts.

But is this sense of 'mind-independence' sufficently strong to qualify all those who accept (c) as moral realists? I do not think so. For instance, a moral *relativist* and a moral *constructivist*, can accept (c) without being ipso facto a realist.

In conclusion, it seems to me that, if Blackburn, Gibbard, Timmons and Horgan have recently became moral realists, this does not depend on the sole fact that they have accepted minimal, or thin, notions of truth, of a property and of a fact. For these notions in themselves - I believe - do not yeld any substantive form of realism.

Rather, if Blackburn, Gibbard, Timmons and Horgan are now realists it is because they have also accepted schemata like, possibly, (c) in which 'mind-independent' is interpreted in a sense *stronger* than (mi). If these philosophers are realists it should be because that have *inflated* the minimal notions of truth, a property and a fact.

If I'm right, this might help to clarifly the puzzle of creeping minimalism.

I’m delighted to see that PEA Soup has room for meta-meta-ethics.


The reason I’m unhappy with this formulation is that an unbearably sophisticated irrealist might say: “Oh, I think an account of what normative belief consists in will appeal to normative facts – it’s that that an ACCOUNT of what normative belief consists in will not APPEAL to NATURALISTIC FACTS” - or something annoying like that.

So far I’m not very worried. Have the unbearably sophisticated irrealist tell us what the account is. If he does think there is a (realist) explanation of moral belief, etc., then I no longer believe him when he says he’s an irrealist.

My distinction was designed to get expressivism on the irrealist side, but having done that I am willing to call anything on that side ‘irrealist’ and everything on the other side ‘realist’. The terms are fuzzy, in my opinion. (For example, Mackie is usually called ‘anti-realist’, but Dummett’s taxonomy counts him as realist.)

I completely agree that your (a) is not a platitude. I’m unclear on what you think the significance is of this fact. My view is that minimalist-friendly expressivists like Gibbard and Blackburn (and me, I guess, and Timmons) believe that there are moral facts that are mind-independent. But they aren’t moral realists. So we need something other than the usual philosophical conception of mind-independence to distinguish realists from irrealists.


Let’s go creepy on explanation for a second.

Consider Kim’s (1994) distinction between “explanatory internalism” and “explanatory realism.” The former holds that what make explanatory statements true are some relations internal to one’s theory, or one’s thought process, or somesuch. The latter holds that what make explanatory statements true are real-world explanatory relations between the event picked out by the explanandum term and the event picked out by the explanans term. So for explanatory realism, explanatory statements have real-world truthmakers, whereas for explanatory internalism, explanatory statements are true, or at least correctly assertible, roughly in virtue of their inferential role. (That’s “short-armed” inferential role – I’m suddenly remembering I’m talking to a Harman student!)

OK, so in a first stage, the moral realist says she explains moral beliefs in the realist sense of explanation, whereas the irrealist explains them only in the internalist sense of explanation. That is, the moral realist offers an EXPLANATION of moral beliefs, whereas the irrealist offers only an explanation. But then unbearably sophisticated irrealist no. 1 says that she also has an EXPLANATION. Because what’s the difference between explanation and EXPLANANTION? One suggestion is that the difference is that EXPLANATION is explanation with truthmakers. But our irrealist went minimalist about truthmakers, so she does have an EXPLANATION in that sense. You might say EXPLANATION involves more than just any truthmakers – it involves real-world truthmakers. But unbearably sophisticated irrealist no. 2 went minimalist about real-worldliness, so they have that too. Suggestion 3 is that the real-world truthmakers be mind-independent. But unbearably sophisticated irrealist no. 3 went minimalist about that too. And so on and so forth.

What am I missing? Why can’t the whole minimalist gambit be run in the domain of explanation and the truth conditions of explanatory statements?


Thanks for that Wright reference – I’ll have to check it out. Also interesting point about the relevance of content externalism. I’ll have to think about that. On the face of it, though, it’s not obvious to me that individuating contents independently of vehicles and disallowing the converse entails content externalism. Not obvious, but does sound right somehow. Like I said, I’d have to think about it.


Sorry, too long for this Tasmanian internet café. I’ll catch up with you later…

Uriah's positive proposal doesn't seem to have been addressed here yet. I don't see how it could work if the goal is to get contemporary expressivists on the irrealist side, as Jamie reasonably tries to do. Blackburn and Gibbard are adamant that the following is false: "According to the irrealist, part of what makes it the case that the fact of slavery’s wrongness is the fact it is is the fact that Jane judges that slavery is wrong." No, what makes slavery wrong is, among other things, the reduction of autonomy, the theft of the fruits of one's labour, and the suffering it involves.
The whole point is that slavery's wrongness can be, and can be thought of as, judgment-independent in spite of the fact that to judge slavery to be wrong is to express some complex attitudes. The attitude must involve something like disapproval of slavery and the disapproval of those who do not disapprove of slavery, my potentially different past and future selves included.

In short, any Euthyphro-style distinction seems bound to fail to capture the irrealist/realist distinction. What it does get at is the distinction between constructivists on the one hand and realists and expressivists on the other - after all, it is definitive of constructivism that our judgments are constitutive of the facts in question. Ironically, this type of arguments have been used by Blackburn to claim that response-dependent views lead to objectionable relativism if they avoid vacuousness.

About ordinary "realist" judgments there are two kinds of story to be told. One is the causal provenance of the judgment. The other has to do with the correctness of the judgment. In principle these are independent, but there is an epistemological merit to having the causal story bring in, or trade on, the correctness story.

It seems to me that a distinguishing mark--like Jamie, I think these categories are fuzzy--of the "irrealist" about moral or other normative judgment is the conviction that, about this class of judgment, there is a causal story to be told but no correctness story to be told. The unbearable sophistication comes in when the irrealist insists that, sure, there is a correctness story to be told, and tells one: the judgment that slavery is wrong is correct because slavery reduces autonomy, abuses human beings, etc. But the irrealism just goes meta: there is a causal story to be told about _this_ explanation, but no correctness story. And if he is called on this, he can go meta again, if need be.

The root conviction of the irrealist, as I am describing him, is a form of naturalism: that ultimately, all the explanations have to be in natural terms, and there are no normative terms among the natural terms. So at some point, he will not be willing to countenance explanations of the correctness of moral judgments, or at least explanations of the correctness of those explanations, or at least ... etc.

"My view is that minimalist-friendly expressivists like Gibbard and Blackburn (and me, I guess, and Timmons) believe that there are moral facts that are mind-independent."

Not to sound completely clueless here, but I would have thought Blackburn and Horgan & Timmons mean to deny that there are intrinsically valuable states of affairs in the universe. Is the claim here that they would say that there are "minimal facts" that are mind-independent, but that their irrealism rests (or at least, ought to rest) on the contrast with robust facts about cars and stars?

(This is my first post so apologies if I've broken any conventions).

A couple of points on the minimalism issue:

1. Antti seems correct to say that the mind-indpendence claim is not going generate the required distinction, since modern expressivists adopt the so-called 'internal' reading of such claims. So no expressivist would except that what makes something wrong is that we (or some tutoried version of ourselves) judge it wrong. To accept this is to endorse a sensibility that forms moral attitudes in response to beliefs about those very attitudes - but this is no sensibility that we want to endorse. (This is Blackburn's line - see 'Spreading the Word' ch.6). So I cannot agree with Uriah's original suggestion.

2. It might be worth asking what the ambitions of one's minimalism are, in other words whether there is a principled reason to 'go minimalist' about some notions but resist the levelling-down urge for others. It seems to me that where ordinary discourse employs a term (as it does for 'true' or 'fact') it is usually in a theoretically non-committal sense - so, for example adding 'is true' to one's moral judgements doesn't make one a moral realist (What Blackburn calls 'Ramsey's ladder'). This is the core minimalist insight. But where a notion doesn't feature in the oridnary discourse (as, arguably, a notion like 'representation' does not) it seems less apt for a minimalist treatment. This is because part of the minimalist treatment involves showing what everyday discursive role a term might have (so 'is true' serves to emphasise or generalise, for example) and it is because this everyday discursive role imports no substantial meta-theory that the notion is 'minimal'. But if a notion has no everyday discursive role then it cannot have a minimal, meta-theoretically neutral role. So perhaps a notion such as 'representation' cannot be 'minimalised' in the way 'truth' and the others can be. (This is not to say that it must have a non-minimal sense' only that it can't have a minimal one.)

3. As the above suggests, there may be scope for casting the realist/anti-realist divide in terms of representation. The distinction would be between discourses which involve states of mind that represent the way the world is, i.e. that aim to be responsive to the way the world is; and those which involve other (perhaps affective) states of mind. This would transmute the issue into question of different psychological roles that each of the various states play. I think this is Blackburn and Gibbard's line - (I think) both suggest that ultimately the distinction may be psychological rather than metaphysical. (Note that this involves a slightly re-alignment; for on this view the error-theorists are grouped with the realists, since they hold that the same state of mind is involved).

4. In response to Andrew: I guess Blackburn and Gibbard would say that there are moral truths that are mind-independent, just as there are car-truths that are mind-independent. The contrast will be in the nature of the state of mind issued by these judgements; the first an affective attitude, the second a putative representation.

I can’t get a grip on what you mean. What is an example of an explanation that an irrealist is happy with, which would be realist if it were an EXPLANATION in the same words? As I recall, the explanations given by expressivists are very different.

I agree with all that, though it’s not clear that no Euthyphro distinction will work. (The ones mentioned here won’t work.)

Andrew Lee;

… I would have thought Blackburn and Horgan & Timmons mean to deny that there are intrinsically valuable states of affairs in the universe.

Hm. ‘Intrinsically valuable’ as opposed to instrumentally? But Blackburn, Timmons, Gibbard, all think there are lots of things that are valuable and not instrumentally valuable. (It’s a normative question which things those are, of course.)

I think that's probably a good stab at what I had in mind. But do those expressivists in the quasi-realist orbit say that there are lots of things that are valuable and not instrumentally valuable or do they say there "ARE" lots of things that are valuable and not instrumentally valuable? My readings of them have always been closer to the latter, although I have to admit it seems Gibbard's comments about "the need for practical faith" on this issue in Thinking How to Live do tend toward amphiboly.

Hi everybody – my profuse apologies for the delay in responding. I’ll work my way backward.


I don’t think an irrealist explanation would become realist once upper-cased throughout. What I had in mind was the converse: you can take any realist explanation and make it acceptable to the irrealist (compatible with irrealism) by distinguishing it from its upper-cased version. So the irrealist can accept any realist explanatory statement, though the realist cannot accept any irrealist explanatory statement.

Perhaps that should be enough to distinguish realists from irrealists. Perhaps instead drawing the distinction by saying that the irrealist embraces a practice-to-truth direction of explanation whereas the realist embraces a truth-to-practice direction of explanation, we could drawn it by saying that the irrealist embraces a practice-to-truth direction of explanation whereas the realist denies that there is any such explanation.

Antti (and Neil):

You’re totally right that my Euthyphro-style question is too naïve. What I had in mind is ultimately something fancier, where the irrealist explains moral facts by appeal to moral judgments not just of oneself, but also of one’s counterparts, temporal parts, other community members, as well as general social practices etc. etc. Contemporary expressivists reject the straightforward explanation of facts in terms of judgments, but they accept the more complicated explanation in terms of a whole bunch of facts about the representation side in the representation/reality divide.


The way I read Jamie’s paper, the explanation he has in mind are not causal – at least not primarily. They are some kind of “in virtue of” explanations (which is something I tried to be more explicit on in my cumbersome formulation of the Euthyphro-style question). One thing I worry about regarding your characterization of the spirit of irrealism is that a hard-core naturalist realist of the Cornell brand could readily accept it.


I too didn’t think Jamie was trying to draw a platitudinous distinction, or a distinction in terms of platitudes, between realism and irrealism. The project is to find an English statement that the realist accepts and the irrealist rejects. Whether that statement is a platitude shouldn’t matter for this particular project.

I'm a little late to the discussion, but I hope someone is still around, because I am still confused about how Dreier's "explanation" explanation is supposed to work. As Uriah glossed it:

Dreier’s solution is that the difference between realists and irrealists is in whether their explanations of normative judgments (or concepts) appeal to normative facts (or properties).

And, I take it, an adequate "explanation" of normative judgments is supposed cite that in virtue of which it is true of someone that they make these judgments.

However, Dreier's test seems to me to be ambiguous between (a) explaining that in virtue of which someone makes a token moral judgments, and (b) explaining that in virtue of which someone makes the type of judgments: moral judgments.

But, if I am understanding the operative notion of explanation, (a) cannot be right, since neither the realist nor irrealist will appeal to the fact that, e.g., what Hitler did was good in explaining the neo-Nazi's belief that what Hitler did was good. They don't think this is a fact, so, by their lights, it cannot figure in an adequate explanation of that in virtue of which it is true that the neo-Nazi makes the judgment.

So it must be (b). But doesn't this have a similar problem? What if we are realists yet think all of the neo-Nazi's moral judgments are false, will we appeal to any moral facts in explaining that in virtue of which it is true that the neo-Nazi makes moral judgments? It seems that by Dreier's test we must in order to remain moral realists, but what moral facts could be at all relevant to this explanatory project?

Thanks to Uriah and Matthew for thinking about details; I agree that they need more work, and maybe my suggestion doesn’t pan out.

One major problem is that I just don’t know what the Realist explanation of moral judgment is supposed to be. (I do more or less understand how Cornell-type explanations will work, and it seems to me that they will count as Realist on my criteria; it’s the non-naturalist versions I don’t understand.
So, Uriah, I don’t know whether “realist explanation and make it acceptable to the irrealist (compatible with irrealism) by distinguishing it from its upper-cased version.” I originally thought not, but on reflection I shouldn’t be confident without having a clear idea of what that explanation might be like.

Matthew, let’s think about a mundane case. Suppose Max judges, “Pluto is made entirely of gold.” What is the explanation of what it is in virtue of which Max counts as making this judgment? Won’t it refer to Pluto, and to gold? I think it will. Then won’t an erroneous moral judgment still have an explanation that refers to moral properties? I expect it to. But maybe not?

Hi Jamie,

Just in case it would help: Graham Oddie offers a model for thinking about non-naturalist explanations of value experiences and, by extension, value judgments in his "Value, Desire, and Reality".

Here is a rough summary from my review:

"In chapter 6, Oddie uses recent work on properties to elucidate and defend the argument that evaluative properties are irreducible to natural ones because the natural "properties" that "underlie" evaluative ones are gerrymandered conglomerations and that such conglomerations are too incoherent to count as self-standing properties...In chapter 7, Oddie argues that evaluative properties can cause value experiences. He builds on Steven Yablo's work on causation and the discussion of properties in the previous chapter to respond to Harman-inspired arguments that evaluative facts play no substantive role in explanations of value experience."

Jamie asks:
Suppose Max judges, “Pluto is made entirely of gold.” What is the explanation of what it is in virtue of which Max counts as making this judgment? Won’t it refer to Pluto, and to gold?
Two initial reactions:

First, in the original suggestion due to Gibbard, I thought it was reference to normative facts rather than reference to normative objects and properties that was supposed to make one a normative realist. He writes, “To explain belief in natural fact adequately, we must assume a natural world of which we are a part. We must start with a realm of naturalistic facts. To explain belief in normative facts, in contrast, we need not start with a realm of normative facts…”(2003: 183). So what motivates the change?

Second, even in the mundane case, I would have thought that explaning Max's judgment won't refer to Pluto and gold any more than explaining that in virtue of which Clark counts as judging, "Krypton is made entirely of kryptonite" will refer to Krypton and kryptonite. Will they? I'm unsure because I am unsure what Jamie means by "refer to" in this context.

At one point Jamie writes, “By Fine’s criterion, a realist about the moon should say that belief that the moon is a quarter of a million miles away consists in some fact that includes the moon, presumably the fact of standing in some representation relation to the moon”(2004: 44). However, it seems wrong to me to say that these judgments consist (even in part) in the fact that the person stands in a representation relation to the thing of which he judges. For, in the Krypton case, it seems that there is not the requisite object for Clark to stand in such a relation to.

Or, am I missing something?

Thanks Brad, that does make sense to me. I don’t know why Oddie thinks normative properties aren’t natural, given that they cause experiences, but aside from that it fits with what I expect a realist to say.

“So what motivates the change?”
I don’t know that there’s any ‘change’, but we do have to choose between a property criterion formulation and a fact criterion formulation.
Well, it’s important what is actually used in realist explanations. If properties are actually used in realist explanations, then our criterion should be a matter of properties. If facts are used, then our criterion should be a matter of facts. And if both are used, then our criterion should permit both. Does this seem wrong or problematic?
It would be really helpful if you’d give just a hint of what you think the explanation is. According to causal theories of reference, the explanation for what’s happening when Max thinks that Pluto is made entirely of gold is going to refer to gold and to Pluto. Obviously, you think this is wrong; what’s your alternative?

I'm not sure I have anything sophisticated to say about that which makes it true that someone counts as judging that Pluto is made entirely of gold or that Krypton is made of kryptonite or that the moon is a quarter of a million miles away, but I'm inclined to think that it is the same kind of thing in each case. And this is what bothers me about saying that it will be a fact including Pluto and gold in the first case--for there are no facts including Krypton and kryptonite to serve a similar explanatory role in the second case. I'd be happy with the relevant constitutive facts being the fact that Max represents Pluto as made entirely of gold, the fact that Clark represents Krypton as being made of kryptonite, and the fact that we represent the moon as being a quarter of a million miles away. But I'm not sure that this is probative with respect to the distinction between realism and irrealism--for won't the quasi-realist also want to say that, e.g., Mary's belief that stealing is wrong is constituted by the fact that Mary represents stealing as wrong?


...but I'm inclined to think that it is the same kind of thing in each case.

I'm inclined to think that the two cases differ in a very significant way, although of course that's consistent with their both belonging to some interesting common kind. Maybe it's because I've been indoctrinated, but it seems pretty clear to me that when Max believes that Pluto is made of gold, the facts about Max that make this true of him include Pluto; in particular, there is some causal chain of reference with Max and his thought on one end and Pluto on the other. And likewise for gold.

... for won't the quasi-realist also want to say that, e.g., Mary's belief that stealing is wrong is constituted by the fact that Mary represents stealing as wrong?
I don't know, that seems like a very obscure claim to me. (How can her belief be constituted by a fact?) Off hand I don't see that quasi-realists are committed either way. (Maybe representing just is believing -- was that your point? I'm prepared to agree that there is a sense of 'represent' that just means 'believe'.)


Sorry, yeah, my point about constitution (read: that in virtue of which it is true that) was just that there is a sense of 'represent' which means 'believe' and the quasi-realist will want to have both of these on his side in capturing the aspects of ordinary ethical discourse which tempt people to realism.

So, what do you do with nonreferring terms like 'kryptonite'. What is it in virtue of which it is true that Lois believes that the moon is made of kryptonite? (I'm assuming here that there is no such thing as kryptonite). Presumably, on the causal view, it will be a fact that includes Lois and the moon, but is the third element something like words in a comic book?

What is it in virtue of which it is true that Lois believes that the moon is made of kryptonite?

That doesn't seem like a good example, since Lois doesn't exist either. So, there is nothing in virtue of which she believes that! (Or if there is, because it's true-in-a-fiction, it's something very different from what it is in virtue of which you or I could suspect that Pluto is made of gold.)

Suppose Uriah gets confused and comes to believe that Superman loves kryptonite. Now he has a belief which we express using failed referential expressions. I think it's very hard to see what it is in virtue of which he has that belief. I bet it has something to do with his causal connections with a canonical version of a story. It seems unlike ordinary mistaken belief, about real things.

Yes, the Lois example is bad. I meant to give an example of a real person believing that a real thing has some nonreal property. So, perhaps, Uriah's believing that the moon is made of kryptonite. (Assume that Uriah came to mistakenly believe that...sorry Uriah!). Alternatively, we could have a real person believing of a nonexistent thing that it has some real property. So, perhaps, Uriah's believing that Krypton is made of gold.

The point is only that it strikes me as strange to think that the account of that in virtue of which Uriah has those beliefs will be that much different from the account of that in virtue of which he has a true belief that, e.g., Jamie is a smart person. But maybe we are, at this point, just evincing theory-laden intuitions.

In my case, it's not intuitions at all, but theory! I freely admit it.

Matthew, maybe we can talk about this in Madison. (And I hope to see some other PEA Soup regulars and irregulars, too. Is Chris Heathwood the only Souper on the Mad Meta program?)

Heathwood's the only Souper on the program, but there will be several other PEA-ers in attendance, including Boisvert, Portmore, Sobel, and me (and I wouldn't be surprised if Kawall showed up as well).

Sadly can't make it to Madison this year - will be off to St. Andrews that Monday. So I'll have to lose to you guys in poker some other time...

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.