Some questions I had been thinking about came up at the Wisconsin Metaethics Workshop, and I wondered if people had any thoughts about them. As a moral irrealist, and indeed someone tempted by the error theory, I am inclined to say that I don’t believe any moral statements to be true. Thus, I am inclined to believe things like:
1) “‘Torturing children for fun is wrong’ is false.”
The LEM says that if 1 is false, then its denial must be true. Thus, the irrealist (or at least the error theorist) seems committed to:
2) “It is not the case that torturing children for fun is wrong”
And I am inclined to believe it too.
That raises two questions:
A) Are 1 and 2 moral statements?
B) Does 2 imply that torturing children for fun is permitted or ok?
I want to say no to both. The answer to B is easier. 2 better not have that implication. For the error theorist does not believe that anything is (morally) ok. 2 is just the sort of moral statement she denies to be true.
One way to see this is, a la Harman (the Elder)’s comments at the WMI, is to note that certain statements may seem to make presuppositions, and yet the presuppositions are inappropriate all things considered. Suppose I say, “John has not stopped beating his wife.” The presupposition seemingly revealed in ordinary discourse is that John is still beating his wife. (“John has stopped beating his wife,” after all, appears to be true if the statement above is true.)
But the presupposition can be defeated explicitly, for example by my saying “John has not stopped beating his wife because he never started!” And it can also be defeated implicitly. (Or perhaps, in some contexts it doesn’t even arise.) In realms we do not think of as realms of fact the presupposition seems inappropriate. “No flavor of ice cream is really best” does not presuppose that all flavors are really equally good. (Of course there is a sense in which they are, namely they are all not at all “really good”. But there is a sense in which the very point of the statement is to deny that such predicates as ‘good’–if taken literally or in their ordinary usage–are correctly applied to things like ice cream flavors.
The same can be said (by the error theorist) of statements like 2, I think.
What about A? Are 1 and 2 moral statements? Well, what is a moral statement? 1 and 2 do express propositions and do employ the moral vocabulary. As such it is natural to think of them as statements about morality, and thus, perhaps, as moral statements. “It is illegal to park there,” by analogy, does seem to be a statement about the law. Still, ‘legal statement’ may seem unclear or awkward. Or take, “Those flowers are red.” Perhaps it does not seem odd to think of that as a color statement. But how about a ‘flower statement’? That’s just not something we’d recognize. ‘Statement about flowers’ is better. By analogy, it seems like the error theorist might be able to live with 1 and 2 being statements about morality, if not moral statements.
Furthermore, if we treat 1 and 2 as moral statements then it looks as though the error theorist believes many many moral statements to be true—has many moral beliefs—but nevertheless doesn’t believe in morality! It seems unfair to saddle her with inconsistency or paradox. (And we wouldn’t, I think, for “coolness irrealists,” e.g.) So I’m inclined to say that the error theorist does not believe any moral statements to be true.
Perhaps at this point it looks like going further would be to engage in a fruitless terminological quibble. Does this all seem right so far? (I have many more thoughts on this, including some big complications related to that bogey of bloggery, moral incoherentism. But I’m sure I have already violated several norms of bloggery by posting such a long entry. Luckily, I don’t think it’s wrong to do so. Still, I do not want to be kicked off.)
Hey Don,
When did you join up? Was there an announcement? This is an outrage!
Anyway, good questions. I’d say that given your account of moral statements, the error theorist should be OK with saying that some moral statements are true. ‘Killing isn’t wrong’ should come out true for the error theorist, since there’s no such thing as wrongness. The statements that turn out false are the ones that presuppose that there are moral properties, and not all moral statements (in your sense) do that.
I wonder what happens to deontic logic if error theory (as you describe it) is true. We can’t define wrongness in terms of permissibility in the normal way; we can’t define ‘permissible’ as ‘not wrong’, since if error theory is true, every act is not wrong and no act is permissible. Will moral terms be interdefinable in some other way?
Ben, which is the outrage? That I joined, or that there was no announcement? If the latter, it is coming. If the former, c’est la vie!
I agree that ‘Killing isn’t wrong’ should come out true for the error theorist, since there’s no such thing as wrongness on her view. But my question is whether to treat it as a moral stt. if it does not make the relevant presupposition. I’d be inclined not to, for the reasons I gestured at, but I’m wondering if there are reasons I didn’t mention cutting the other way, and whether it matters.
Don’t know much about deontic logic (he sang) but I know what I like. I mean, I’d suggest that if we were to accept the ET, we’d probably want to introduce new meanings for ethical statements, since the discourse would still be useful, as would the making of inferences within it.
d
FWIW, I wouldn’t bother trying to clarify the meaning of “moral statement.” RAther, just say that it is your opinion that there are certain predicates (right, wrong, permitted, good, bad, etc.) which (necessarily) apply to nothing.
Atomic uses of these predicates are then false, and the laws of logic can work out the rest of it. If you really want to, you can define “moral statement” as all those statements including these predicates that come out false.
Hi Don,
This might not be much different from what you’ve said, but why not say, that 1) is ambiguous between something such as,
3) There is a property of wrongness and killing children lacks it.
and
4) There is no property of wrongness and so killing children cannot have it.
Let ‘moral statements’ be statements of the form of 3), and statements of the form of 4) be ‘non-moral’ statements. The grounds for the distinction is the idea that moral statements all include the assertion that there are moral properties. Then you can say your position is not a moral position, being of the form of 4), but also that statements of the form 3) are all false, given the truth of 4).
Thanks,
Heath: I’m not so much trying to clarify the meaning as hoping to stipulate in a non-misleading way, a meaning for ‘moral statement’. I should have made that clear. I’m not sure about “apply to nothing”. That, in a way, may relate to Robert’s suggestion, so see below, please. On atomic uses, here’s what I have found troubling about that way of setting things up (though I am not sure I fully understand what you mean by “work out the rest of it”). I am trying to define a view which has as it’s central tenet the denial of certain sorts of claims (about goodness, virtue, rightness, etc.) I can’t do that by JUST denying all atomic uses (to borrow your shorthand) and all of their logical implications. For I deny other statements (some of which may APPEAR to be the denials of those) as well. So I deny that pleasure is a good thing. But I also deny that pleasure is a non-good thing (on the reading which presupposes that there are any good things.) And certainly some people want to say things like, “Pleasure is not good,” and mean to imply that it is not among the things–of which there are some–that are good. Likewise I want to deny that one ought to pay one’s taxes. But some who deny that mean to convey that it is morally permissible not to pay them. So I think I need the story about presupposition–or something like it. That’s why I am inclined to (at least semi) stipulate that by ‘moral statement’ I mean a statement that makes the relevant presupposition (and not just one that employs the moral vocabulary.
Robert: The worry I have about defining my view as the denial that there are moral properties is that I aim to defend a particularly radical form of moral irrealism, one which opposes not just moral property realism, but also the sort of moral truth, correctness, right answer, correct way to live, etc. that constructivists and many contemporary non-descriptivists accept. Korsgaard says that she is not a (“substantive”) moral realist because she denies that there are moral facts and (if I am remembering right) properties. But she believes in moral duties, correctness, etc.
Wow, ran out of room. I’m windy. Sorry.
Kors. may be confused or she may not be. But that’s HER issue, not mine. My view (call it anti-objectivism if you like) denies that there is moral correctness (of the sort she defends) and even minimal moral truth of the sort defended by the usual non-descriptivist suspects like Blackburn and Timmons. I think that almost all the going views have in common the acceptance of moral facts or properties or truths or correctness or right answers, etc. (And I have a guess as to why that’s so, actually, which I take on in a recent paper.) My form of irrealism opposes all of them. (Ayer’s emotivism, the error theory, and incoherentism are all on my side of the divide, realism, constructivism, etc.–even relativism–are on the other.) I think this distinction is important, and maybe clearer–or can be made clearer–and perhaps even more useful than the vexed realism/irrealism divide. But then I need to get away from property talk to make the distinction I’m aiming for (without raising questions for Korsgaard and others that aren’t my concern!)
Hi Don. It’s great to have you aboard. (Don’t worry, Ben…the official “welcome” is coming!)
A few things, though I’m not confident any of it will be helpful.
Don, can you treat moral statements just as, well, statements that assign a certain kind of evaluation (viz, moral evaluation)? Whatever else one may want to say about evaluative predicates, they are evaluative predicates because they are assigning some kind of positive or negative evaluation. For example, we can evaluate certain kinds of things (e.g. actions, feelings, etc.) as:
Right wrong
Correct incorrect
Apt not apt
Wise unwise
Obliged not obliged
etc. That these are all predicates that assign a certain kind of evaluation is, I take it, something that everyone agrees on. (Or is this, in the end, what you will want to deny?)
If so, then can you say that a statement is a moral statement only if it contains a predicate (in an extensional context?) that assigns a postive or negative moral evaluation? If so, then I take it that your view (and Ayer’s, and error theorists) would hold that no sentence that assigns a positive moral evaluation can be true.
On this view, neither (1) nor (2) are moral statements (which seems plausible to me), since (1) assigns a negative *semantic* evaluation to a sentence and (2) assigns a negative *semantic* evaluation to a proposition. (However, a sentence like ‘Torturing children for fun is wrong’ would be a moral sentence, because the predicate ‘is wrong’ assigns a negative moral evaluation to an act type (or maybe its instances).)
Also, a caution about Heath’s suggestion:
For something like this to work, it would have to be beefed up to say something about the kinds of things that can be in a predicate’s extension. For example, the sentence ‘Rocks are wrong’ is akin to the sentence ‘Rocks are true’. Neither is false, since both are meaningless, since rocks are not the kind of thing that can fall within the extensions of ‘right’/’wrong’ or ‘true’/’false’ respectively.
Sorry, I wrote too quickly. I meant:
B) Does 2 imply that torturing children for fun is permitted or ok?
I want to say no to both.
If you don’t think that the “literal” or “ordinary usage” of “wrong” can be correctly applied to the act of torturing children for fun, then on what grounds can you deny that that act is permitted? Or, put differently, what does “ok” mean to you, and how is that meaning different from the ordinary usage of “morally permissible”?
Dan, thanks for you thoughts. I’ll get back to you shortly.
ASG: Thank you, too. ‘Ok’ means ‘morally permissible’ (putting aside required, for simplicity here) or ‘permitted by morality’, and since (putting it a little crudely) I don’t believe in morality, I don’t believe that anything is ok.
In other words, I don’t think the moral predicates are correctly applied to anything–actions, states of affairs, character traits, etc. Nothing is morally permissible, on my view, nor is anything morally impermissible or required or good or virtuous, vicious, etc.
Suppose you had started with my denial that eating meat is morally permissible (because nothing is). That’s also a claim I am inclined to make. You wouldn’t want to attribute to me–an irrealist–the view that eating meat is IMpermissible. It’s not permissible, and it’s not impermissible.
By ‘ok’ you might have in mind ‘not forbidden by morality.’ That’s fine. Not forbidden, I accept. I don’t deny the law of the excluded middle! But not permitted either. If I don’t believe in the morally impermissible, I can hardly be required, it seems to me anyway, to accept the morally permissible.
In fact, that’s the point about presupposition. The denial of “Don has not stopped beating his wife” is not “Don HAS stopped”! It’s, “It’s not the case that Don has stopped beating his wife.” The latter needn’t presuppose (and better not!) that I ever started. That, I guess, is a matter of pragmatics, and even if such a presupposition would ordinarily be appropriate, I can eschew it as long as I am clear, I think. The same goes for the denial of a moral claim. ‘Not impermissible’ may look like a double negative, reducing to ‘permissible,’ but that involves the presupposition that the moral predicates are correctly applied to anything, which I deny.
(Again, I’m putting aside incoherentism for the sake of this discussion.)
d
Sorry, it’s late and I’m new to blogging. I should read what I wrote after midnight. I didn’t mean to talk about the denial of “Don has not stopped beating his wife”. I meant to say (just) that another way to SAY that is to say, “It is not the case that Don has stopped beating his wife.” Both are compatible with my never having started, though the latter formulation may bring out the presupposition I think it fair to eschew in this case.
Here’s how the ill-edited sentences should have read:
In fact, that’s the point about presupposition. The denial of “Don has not stopped beating his wife” is not “Don HAS stopped”! “It’s not the case that Don has stopped beating his wife,” needn’t presuppose (and better not!) that I ever started.
I’m going to bed. I’m sure it’s 09:21 somewhere, but not here.
d
OK, that does clear it up; I guess I was reading the original post as suggesting that a negative answer to (B) was different from, but implied by, (2), when they are really more or less saying the same thing.
Does a moral irrealist have any grounds on which to object to someone else’s conduct at all, then? If you personally encountered someone torturing a child for fun, and stopping him entailed, say, a 5% chance of serious bodily harm to yourself, what would you do?
Hi Don,
I have a brief clarificatory question about your position. I’m wondering what the relation between the discussion of error theory is, on the one hand, and your position, on the other. I can see why there is a puzzle about the relation between Mackie-style error theory and your two statements (1) & (2), but I’m also tempted to take Ben’s approach to tackling it. But if your view denies not just moral properties (realist-style) but also everything from Horgan & Timmons to Korsgaard, I’m puzzled as to why you want to say that the statements embedded in (1) and (2) are false, rather than simply devoid of truth-value. I suppose my question is: do you have a view about what would make the statements embedded in (1) and (2) come out true?
Ok, that’s it. I’m quitting my day job.
Dan: First, I have no problem with labeling (or thinking of) moral terms as evaluative and saying that no moral evaluations are correct, whether positive or negative. Perhaps some would say that their being evaluative is tantamount to their being (primarily) expressive (of attitudes, not propositions—sheesh!) and not descriptive, but I don’t mean to be opining on that. The error theorist holds that (at least) a certain class of evaluation—moral ones—are all false. Then statements like, “It is not wrong to cheat on your taxes,” and “It is not permissible to cheat on your taxes,” would, if understood in their non-evaluative sense, be correct according to the ET.
I don’t have a view on the claim that no such sentence “can be true” or that the moral predicates, “(necessarily) apply to nothing”. Insofar as I am tempted by the ET, I’d at least say that the sentences are all false and that the predicates apply to nothing.
One quibble, (and it’s my own fault for not being clear, I now see): On my view, 1 and 2 could be moral statements, if uttered in a context in which the presupposition discussed above is appropriate. When an ordinary person says, “‘Torturing children for fun is ok,’ is false.” (to make the example more plausible), the error theorist should think the person is making a moral statement. In THAT context (or with that presupposition, or on that use . . . ) the ‘is false’ really does just serve as a truth value flipper. The ET should say that the person’s statement is equivalent to “Torturing children for fun is not ok,” I think.
(Someone who knows more about linguistics than I may be able to say [and probably has said] what the relationship is between the pragmatics, semantics, truth conditions, and permissible [required] inferences, when all this stuff about presupposing makes the scene. I’d welcome clarification on this, or a reference that can help me to sort it out. But I’m fairly certain that the moves I am making can be done, however it all comes out!)
Finally, your suggestion about “Rocks are true,” and the like being neither true nor false but meaningless points to the issues about incoherentism I am trying to avoid at this point. (But that’s why I keep saying, “the error theorist should say” or “I am inclined to think”. I’m hedging, because I take incoherentism seriously. For those who aren’t familiar with the term—and why should you be?—think about what the traditional emotivist [in contrast with the error theorist] would say about moral truth. Moral incoherentism may have similar implications.)
ASG: I don’t agree that a negative answer to B is, “really more or less saying the same thing,” as 2, but I do want to claim that there is a reading of 2 (appropriate when the ET utters it) according to which a negative answer to B is also appropriate.
You also ask:
Does a moral irrealist have any grounds
on which to object to someone else’s
conduct at all, then? If you personally
encountered someone torturing a child for
fun, and stopping him entailed, say, a 5%
chance of serious bodily harm to
yourself, what would you do?
I’d stop him, of course, if I could. I am against that! Strongly! And there’s nothing wrong with stopping him (or anything). I don’t value my own interests above all else, and nothing about MI makes selfishness a sensible or necessary default position.
Do I have “grounds” to object? That’s complicated, and afield from the current topic, though I’d be happy to discuss it. The short answer is: There are not grounds that anyone would have to accept on pain of irrationality or anything like that. But it’s possible to object, and even to reason with one another in much the same way realists recommend (and I don’t think doing so is ill motivated or silly). We irrealists have values; we just don’t think they—or any others—are correct. If we can find shared values and shared commitments regarding how to reason, we can do so. And that’s not at all unlikely (or at least not much less likely on my view than if realism is true).
Thanks to you both and I’ll get back to Ezra soon.
d
Hi Don,
Ahh, I knew that bogey of bloggery would reappear!
You write,
Yes, I was wondering if you could be helped by a principled distinction in terms of the *direct* speech acts these sentences are conventionally used to perform. But, certainly, these can be used to perform indirect speech acts.
There are certainly various relations between sentences and states of affairs. But I was thinking that the former is a sentence that is directly about another sentence, and the latter is a sentence that is directly about a certain act type. I really don’t think much rests on such a distinction, though, as long as you have available to you the distinction between direct and indirect speech acts.
I don’t have any specific references off-hand, but Steven Davis’s anthology *Pragmatics: A Reader* has a number of good articles concerning the relations among pragmatics, presupposition, speech acts, truth, semantics, and more.
Since some of my comments at the WMW may have been part of what Don was alluding to in his original post, I thought I’d confess that I wasn’t 100% upfront about my attitude towards the error theory in those comments. I actually believe something even more uncompromising: in my view, the error theory is inconsistent with certain truths of (deontic) logic; so the error theory is itself logically self-contradictory.
The reason is simple. In standard deontic logic (SDL), whenever p is a logical truth, ‘It ought to be that p‘ is also a logical truth. Many metaethicists think that this feature of SDL is indefensible, but I’ve defended in print (e.g. in Section 5.1 of Chapter 5 of my book), and I stand by what I said. So, in my view, there are lots of logical truths in which ‘It ought to be that …’ has largest scope. Moreover, in my view, this fact is itself a truth of logic — even though it clashes with almost every version of the error theory.
Now, I think that there can be disagreements in which both sides are (at least in some sense) reasonable, even though in fact one of these sides is saying things that are logically self-contradictory. Some logical truths just aren’t obviously true (let alone obviously truths of logic). It seems to me that the dispute between Don and an adherent of SDL is going to have to like this.
Don, I’m not sure how much explaining of your overall view you want to get into, but I’ll ask a couple of questions that this thread (and in particular Ralph’s comment) provoked. I’m interested in just where, to the extent you accept ET, you see the E occurring. Is it moral claims or statements themselves that you take to be erroneous? If so, what is the moral that gives rise to the error?
By way of contrast, I wonder if you also would hold to an error theory about, say, prudential claims:
Or how about instrumental practical rationality?
General practical reasons claims?
Or theoretical reasons statements?
I suppose what I am thinking is that (as e.g. Hampton argued) much of what ET’s (Mackie, notably) object to in moral claims isn’t distinctive of moral claims per se, but more generally normative claims or reasons-statements. In your case, is it normative discourse generally that you find erroneous, or more generally the normative? If the former, aren’t you in general going to want to hang on to SDL? And if the latter, is that not going to handicap you in making claims about the warrant for your account vis-a-vis realist accounts?
Sorry, that last bit I mangled. It should be: is it normative discourse generally that you find erroneous, or more specifically the moral? If the latter, aren’t you in general going to want to hang on to SDL? And if the former, is that not going to handicap you in making claims about the warrant for your account vis-a-vis realist accounts?
Don and Ralph,
I think it is a very strong presumption against any metaethical view that the error theory cannot be stated within it. It just can’t be incoherent to think that moral talk is like talk of Santa Claus or moxibustion. It may be wrong, but it isn’t incoherent. Is it?
Regarding Korsgaard and your radicalism (which I like), I think my proposal can be modified pretty easily to include views such as hers. The implication of moral statements is that there’s a right answer to such questions. And that implies that there is something that makes them right. If you deny that last bit, you’ve got a metaethical view. Or so I was trying to say.
Ralph,
I’m not sure that irrealism is inconsistent with SDL. One may accept that all tautologies ought to be true while denying that any action ought to be done. It’s a general feature of tautologies that their being true is independent of our actions. Act in any way you please and it will still be the case that, say, if the sky is blue then the sky is blue; there’s nothing you can do to change that. That a tautology ought to be true therefore imposes no obligations on anyone to act in one way or another. So if we define irrealism as the view that there are no actions that ought to be done, then it’s consistent with the axiom of SDL to which you refer.
Thanks again, to all.
Yet another reason to read Ralph’s book! Ralph, I’ve ordered it. Give me a few days!
My initial reaction (which is consistent with Robert and Campbell’s suggestions) is to suspect that either:
a) The feature of SDL to which you point does not involve a MORAL ought. (It is hard to believe that it does!) Yet these are the only oughts I have denied so far.
OR
b) That it is inconsistent with the error theory is reason to doubt that this feature is correct. As Robert suggests, the ET should at least be on the table. (One person’s modus ponens is another’s modus tollens!)
More, including replies to others’ questions/comments, coming soon. I am truly grateful for the input from all concerned. (And yes, Ralph, your comments at the WMW did help to focus my thinking on this stuff!)
Oh, and people should check out Campbell’s weblink. he . . er . she . . is hot!
d
Ooh, he IS hot! But can he play the blues?