Is this is a meta-ethical analysis of consequentialists theories in particular or do you think that deontology and virtue ethics are prone to the same meta-ethical problems? In particular, you seem to be imaging a moral system that allows the moral agent to say of another's actions, "you ought not do x because it is immoral", whereas I want my morality to tell me and only me, "I ought not do x because it is immoral". If I want to judge another person's actions, morality is the wrong tool for that. This is not to say I advocate relativism, realism or nihilism, just that morality is not the sort of thing that fits neatly into those distinctions. Perhaps this is covered in the non-committal word "usual", where you would consider my opinion to be an outlier and not subject to your analysis.

Moral language is deceptive because it has pragmatic use in other disciplines that make judgements of conduct (i.e., law and etiquette). If you're going to do an analysis of moral language, I would think you might want to at least acknowledge the dual-duty the language does or show how these uses are one and the same. I wouldn't rest too much on the naïve use of the words because such an analysis is susceptible to error.

I don't see how the truth conditions of moral statements are any more problematic than those of any other kind of claim. If that's a problem for the realist, then it's a problem for anyone who believes in truth conditions about anything.

Is this is a meta-ethical analysis of consequentialists theories in particular or do you think that deontology and virtue ethics are prone to the same meta-ethical problems? — Soylent

Meta-ethics is prior to normative ethics. Before we can discuss whether or not the the moral value of X is determined by duty or by consequences (or something else) we have to determine what it means for X to be moral (or immoral). You might say that X is immoral because it causes harm, but then what do you mean by saying that X is immoral? Until you can answer that the claim is vacuous. That's where meta-ethics comes in. And the position I've argued above is that the claim that X is immoral doesn't usually have just a single meaning. It can mean a lot of things depending on how it's used and what the claimant is trying to do with the sentence.

I don't see how the truth conditions of moral statements are any more problematic than those of any other kind of claim. — The Great Whatever

They're not. But the issue with truth conditions wasn't to distinguish moral statements from non-moral statements but to distinguish realist moral statements from non-realist moral statements.

If that's a problem for the realist, then it's a problem for anyone who believes in truth conditions about anything.

Sure, but the difference is that the realist can't make sense of his truth-conditions whereas the non-realist can. The subjectivist can say that the statement "X is immoral" is true if the speaker doesn't approve of X, and so we know what "X is immoral" means; it means that the speaker doesn't approve of X. But what can the realist say?

It's a fundamental problem with verification-transcendent truths (i.e. realism) as argued by Dummett.

Sure, but the difference is that the realist can't make sense of his truth-conditions whereas the non-realist can. The subjectivist can say that the statement "X is immoral" is true if the speaker doesn't approve of X, and so we know what "X is immoral" means; it means that the speaker doesn't approve of X. But what can the realist say? — Michael

But then you just have the problem of, when do you know someone approves of something? Certainly the non-realist makes no more sense here than the realist. (That is to say, moral claims are not made any more intelligible by translating them into another medium whose truth conditions are not any more transparent than the original). There's also the fact that semantically, things like 'X is wrong' just don't mean the same thing as 'I don't approve of X;' and if an anti-realist must claim this to satisfy his metaphysical prejudices (that is, commit himself to an empirical semantic error), then that is so much the worse for his position.

Meta-ethics is prior to normative ethics. Before we can discuss whether or not the the moral value of X is determined by duty or by consequences we have to determine what it means for X to be moral (or immoral). — Michael

I asked because your analysis seems biased towards a meta-ethical analysis that presupposes morality as consequentialist. The bolded section seems particularly troublesome absent a normative framework. A meta-ethical consideration might be something like, are actions the sorts of things that have moral worth or do we measure character? From there we can build a particular normative theory, but you jumped right to actions as the subject of moral judgement. Exemplified by this comment:

If someone tells me that some X is immoral then they are usually telling me not to do X — Michael

*edited*

But then you just have the problem of, when do you know someone approves of something? — The Great Whatever

This isn't what I meant. It's not about whether or not one can verify whether or not the truth conditions are satisfied; it's about whether or not one can make sense of what those truth conditions are. What are the truth conditions for the claim "X is immoral" under moral realism? If you don't know what those truth conditions are then you don't know what it means for the claim to be true. It's a vacuous claim.

There's also the fact that semantically, things like 'X is wrong' just don't mean the same thing as I don't approve of X

They don't just mean that. But that's exactly the point I've been making. Moral claims mean a lot of things – do a lot of things. One of which is, usually, to report on some putative objective fact. The problem I find is that it fails in this task, hence error theory succeeds where realism doesn't.

What are the truth conditions for the claim "X is immoral" under moral realism? If you don't know what those truth conditions are then you don't know what it means for the claim to be true. — Michael

What are the truth conditions of "X approves of Y?" How are those any less mysterious than the truth conditions of the sentence this is supposedly translated from?

They don't just mean that. But that's exactly the point I've been making.

They don't mean that at all. A claim about murder is a claim about murder, not about oneself. Of course one may in saying murder is wrong imply something about oneself, even purposefully; but we can do that with any language, without saying that the language itself literally says or means that. I would say that someone who doesn't understand that claims about murder are claims about murder, rather than the speaker, doesn't know how to speak English -- and so the anti-realist of the stripe you are describing is forced to pretend not to know their own languages because of metaphysical prejudices about purported problems with the reality of moral properties (which his supposed solution does not actual solve anyway).

Moral claims mean a lot of things – do a lot of things.

But meaning and doing are different. If I say, 'you didn't take out the garbage,' one thing I can do with such a statement is imply that I don't like someone. But that is not what the sentence means. It means that the addressee didn't take out the garbage.

What are the truth conditions of "X approves of Y?" How are those any less mysterious than the truth conditions of the sentence this is supposedly translated from? — The Great Whatever

They might be "as mysterious". But we're not discussing what it means for X to approve of Y. We're discussing what it means for X to be immoral. Obviously to engage in any (successful) discussion on what a claim means there must be some common understood language else we'd never get anywhere. So for the sake of argument I'm assuming – reasonably, I would say – that we all understand and agree on the meaning of "X approves of Y". But we can't say this about the meaning of "X is immoral" because the meaning of "X is immoral" is the very thing being questioned. The subjectivist has made an attempt to translate this claim into one that is presumably understood and agreed upon. Can the realist make such an attempt?

But meaning and doing are different. If I say, 'you didn't take out the garbage,' one thing I can do with such a statement is imply that I don't like someone. But that is not what the sentence means. It means that the addressee didn't take out the garbage.

It can mean either. Which one it means in this context depends on what you're trying to do with the claim.

We're discussing what it means for X to be immoral. Obviously to engage in any (successful) discussion on what a claim means there must be some common understood language else we'd never get anywhere. So for the sake of argument I'm assuming – reasonably, I would say – that we all understand and agree on the meaning of "X approves of Y". But we can't say this about the meaning of "X is immoral" because the meaning of "X is immoral" is the very thing being questioned. The subjectivist has made an attempt to translate this claim into one that is presumably understood and agreed upon. Can the realist make such an attempt? — Michael

This makes no sense. We have to have a common understood language, yes. We call this language English. That same language that lets us understand what "X approves of Y" means lets us understand what "X is immoral" means. What do you mean, that meaning is the very thing being questioned? Why is it more questionable than "X approves of Y?" The subjectivist has taken one sentence and given another, which doesn't even mean the same thing, and which is just as mysterious as the first. So how is that an account? How does it help? What light does it shed on anything? If knowing English is what it takes to understand English sentences, where does the subjectivist get off saying no one can understand the English sentences he thinks are problematic because of his metaphysical prejudices?

It can mean either. Which one it means in this context depends on what you're trying to do with the claim.

No, it can't mean either. That is not what those English words mean. I claim that 'You didn't take out the trash' means that the addressee didn't take out the trash, not that it means something else. One can of course use the meaning of the words to imply something that is not strictly said by the sentence.

I asked because your analysis seems biased towards a meta-ethical analysis that presupposes morality as consequentialist. — Soylent

I don't see how it does. I'm not saying that one ought not do X because it has bad consequence and I'm not saying that one ought not do X because it contradicts a duty. I'm saying that when people say "X is immoral" they are often telling others how to behave. Their justification for making this claim might be in deontological or consequentialist terms, but that's a separate issue to the one I'm raising.

"X is immoral" they are often telling others how to behave. — Michael

That's where the presupposition comes in from my perspective. As a meta-ethical analysis, where did this idea that morality is about telling others how to behave come from? Deontology and virtue ethics would disagree with that foundation. Deontology and virtue ethics would stand as a principled (universal, non-relative) theory, that prescribes personal action specific to the character of the agent (relative). I cannot tell you what is immoral because the prescriptions of conduct are not universal and truth-apt to an observer, but I can give you truth-apt claims as a foundation for the normative theory that supply truth-apt prescriptions through introspection.

What do you mean, that meaning is the very thing being questioned? — The Great Whatever

I mean that we're actively asking the question "what does 'X is immoral?' mean?" and trying to answer it. That's what meta-ethics is.

Why is it more questionable than "X approves of Y?"

Because we're questioning it. If we were questioning the meaning of "X approves of Y" then the meaning of "X approves of Y" would be questionable.

If knowing English is what it takes to understand English sentences, where does the subjectivist get off saying no one can understand the English sentences he thinks are problematic because of his metaphysical prejudices?

I don't understand this.

No, it can't mean either. That is not what those English words mean. I claim that 'You didn't take out the trash' means that the addressee didn't take out the trash, not that it means something else. One can of course use the meaning of the words to imply something that is not strictly said by the sentence.

The meaning of a sentence isn't simply to be understood only as what follows from combining the (most prominent) dictionary-definition of its component words. Language doesn't work like that.

This makes no sense. We have to have a common understood language, yes. We call this language English. That same language that lets us understand what "X approves of Y" means lets us understand what "X is immoral" means.

Yes. We understand what words and sentences mean by understanding how they're used. And the statement "X is immoral" is used to express one's disapproval, to report on one's disapproval, to prohibit behaviour, and/or to report on a rule. The problem is when one then says that this rule is non-constructed or that moral properties are something other than the traditional empirical properties that we're familiar with. Once you adopt realism you attempt to subtract the empirical use of language from the meaning of the phrase, but as the meaning of the phrase is its empirical use there's nothing meaningful left.

I mean that we're actively asking the question "what does 'X is immoral?' mean?" and trying to answer it. That's what meta-ethics is.

And in answer you give another sentence: it means, 'I disapprove of X.' But what does that mean? Recall that you chastised the moral realist for trying to push the question back rather than resolve it. What makes the anti-realist's answer any different? In addition, it has the vice of being empirically false. Even if unilluminating the moral realist at least doesn't express gross semantic ignorance in saying that 'murder is wrong' means that murder is wrong, rather than something else.

Because we're questioning it. If we were questioning the meaning of "X approves of Y" then the meaning of "X approves of Y" would be questionable.

But there is no reason to question moral claims over claims of approval. The reason one is questioned and not the other is to satisfy metaphysical prejudices that cannot allow for moral facts to be real. Thus a problem is created where there is none.

I don't understand this.

Competent speakers of English understand what 'X is immoral' means. It means that X is immoral. The subjectivist must deny this, and say that it instead means something else.

The meaning of a sentence isn't simply to be understood only as what follows from combining the (most prominent) dictionary-definition of its component words. Language doesn't work like that.

I didn't say it was. However, contrary to your claim, 'you didn't take out the trash' doesn't sometimes mean, something else. One can imply something else by exploiting the meaning of the sentence; but that doesn't mean the sentence means whatever you want it to imply in any context.

And the statement "X is immoral" is used to express one's disapproval, to report on one's disapproval, to prohibit behaviour, and/or to report on a rule.

No, it is not. For those purposes, we have sentences of the form, 'I disapprove of X.' A statement 'X is immoral' is used to say that X is immoral.

The problem is when one then says that this rule is non-constructed or that moral properties are something other than the traditional empirical properties that we're familiar with. — Michael

What is 'ordinary?' Everyone is perfectly familiar with moral properties, at least as much as the supposedly more real empirical properties the subjectivist tries to replace them with. So it is impossible not to be prescriptive here, because if the subjectivist looks to how the language is actually used, he will notice that people use sentences like 'X is immoral' and mean that X is immoral, not something else. So moral properties are mysterious? We understand 'traditional, empirical' properties better? Says who? Ah, says the subjectivist. But why should we listen to him? Again, because he has a certain metaphysical agenda.

Once you adopt realism you attempt to subtract the empirical use of language from the meaning of the phrase, but as the meaning of the phrase is its empirical use there's nothing meaningful left. — Michael

Except the objectivist is honest about how people use the phrase, whereas the subjectivist makes up other meanings to suit his philosophical agenda.

Recall that you chastised the moral realist for trying to push the question back rather than resolve it. What makes the anti-realist's answer any different? — The Great Whatever

I chastised the moral realist for doing this when obligation is involved. If we say that iff X is immoral then one ought not X, and if we define "X is immoral" as "X is harmful", then we're saying that iff X is harmful then we ought not X. But this is refuted by the is-ought problem. Therefore either "X is immoral" means "X is harmful" but "one ought not X" does not follow from "X is immoral" or "one ought not X" does follow from "X is immoral" but "X is immoral" doesn't mean "X is harmful".

I specifically said "if one wants to abandon the notion of an objective obligation and argue for a more limited account of moral realism then one can make a case for a stipulative, or even lexical, definition of 'immoral' in objective terms."

Even if unilluminating the moral realist at least doesn't express gross semantic ignorance in saying that 'murder is wrong' means that murder is wrong, rather than something else.

But nobody is saying that "murder is wrong" means something other than that murder is wrong. That one can explain the meaning of a phrase without just employing disquotation is not that it is wrong to employ disquotation. You can say both that when you say "I am a bachelor" you are saying that you are a bachelor and that when you say "I am a bachelor" you are saying that you are an unmarried man. You can say both that when you say "murder is wrong" you are saying that murder is wrong and that when you say "murder is wrong" you are saying that you don't approve of murder. Just as one can say both that when you say "chocolate is tasty" you are saying that chocolate is tasty and that when you say "chocolate is tasty" you are saying that you enjoy the taste of chocolate.

Competent speakers of English understand what 'X is immoral' means. It means that X is immoral. The subjectivist must deny this, and say that it instead means something else.

As above, he doesn't deny it. We can agree that "X is immoral" means that X is immoral, just as we can agree that "I am bachelor" means that I am a bachelor. But that doesn't mean we can't then clarify that "I am a bachelor" means that I am an unmarried man or that "X is immoral" means that I don't approve of X.

You've presupposed that "X is immoral" doesn't mean "I don't approve of X" to then claim that the subjectivist's claim that "X is immoral" means "I don't approve of X" entails that "X is immoral" doesn't mean that X is immoral. But of course this fails, because if "X is immoral" and "I don't approve of X" mean the same then if "X is immoral" means that I don't approve of X then "X is immoral" means that X is immoral.

I didn't say it was. However, contrary to your claim, 'you didn't take out the trash' doesn't sometimes mean, something else. One can imply something else by exploiting the meaning of the sentence; but that doesn't mean the sentence means whatever you want it to imply in any context.

I'm not saying that a moral sentence means anything in any context. I'm saying that in the context of a moral discussion the claim "X is immoral" can mean more than one thing – including an expression of or report on one's attitude.

What is 'ordinary?' Everyone is perfectly familiar with moral properties

But we're not. We're familiar with empirical properties because we can see and hear and feel them, or make use of them in a mathematical description to predict phenomena. There's nothing like that with these supposed moral properties. If you say that X is immoral, and intend to report on an objective fact, à la realism, then I literally have no idea what you're saying. You might as well talk about magical or spiritual properties.

Except the objectivist is honest about how people use the phrase, whereas the subjectivist makes up other meanings to suit his philosophical agenda.

And I have agreed that people often use the phrase to (try to) report on objective moral facts. But because they can't make sense of what it would take for these claims to be true then their claims are vacuous and so fail to report on anything. Hence error theory.

As a meta-ethical analysis, where did this idea that morality is about telling others how to behave come from? — Soylent

I didn't say that morality is about telling others how to behave. I said that when someone says "X is immoral" they are often telling me how to behave (as well as, potentially, telling me or expressing other things).

Deontology and virtue ethics would disagree with that foundation. Deontology and virtue ethics would stand as a principled (universal, non-relative) theory, that prescribes personal action specific to the character of the agent (relative).

It's still about telling people how to behave so I don't see how what I've said conflicts with this.

The usual approach to meta-ethics is to argue for a single understanding: realism or nihilism or relativism, and so on. This is far too simplistic and fails to consider the complexity of moral language. — Michael

I don't understand the criticism. For one, realism, nihilism, and relativism aren't theses about moral language. Realism is constituted in part by a claim about moral language, and relativism may be depending on which of the several ways that term is used you are intending to pick out here, but these views are primarily about the metaphysics of ethics. For another, people argue realism or nihilism or relativism because the way these are defined (again, depending on what you mean by relativism), they are all mutually exclusive. You can't be any two at once, so of course the decision is treated as exclusive. Third, adopting one of these views does not rule out having other metaethical views as well. So it is odd to think that adopting one of these views is simplistic on the grounds that they do not cover the full complexity of moral language. They aren't even supposed to do that. It's like complaining about the electrician because he didn't fix your sink.

I don't understand the criticism. For one, realism, nihilism, and relativism aren't theses about moral language. — Postmodern Beatnik

They make claims about what it means for X to be immoral, which is to make a claim about what "X is immoral" means.

For another, people argue realism or nihilism or relativism because the way these are defined (again, depending on what you mean by relativism), they are all mutually exclusive. You can't be any two at once, so of course the decision is treated as exclusive.

I wouldn't say that this is exactly correct. Realism and nihilism are mutually exclusive, as realism argues that some X is inherently immoral whereas nihilism argues that no X is inherently immoral, but nihilism and relativism are not mutually exclusive as the claims "X is not inherently immoral" and "X is non-inherently immoral" are compatible.

Third, adopting one of these views does not rule out having other metaethical views as well.

Leaving aside the fact that this seems to contradict your immediately prior claim that "you can't be any two at once", I'm not saying that these views rule each other out. I'm actually saying the opposite; that they're compatible (the dichotomy between realism and nihilism notwithstanding).

So it is odd to think that adopting one of these views is simplistic on the grounds that they do not cover the full complexity of moral language. They aren't even supposed to do that. It's like complaining about the electrician because he didn't fix your sink.

To the extent that what they're supposed to do is determined by what the people who promote them try to do with them, I think that this is evidently false. People do indeed try to argue for just a single meta-ethical view and claim that this meta-ethical view successfully makes sense of (all) moral claims. Indeed, this seems to be the most prominent approach to meta-ethics. Many people identify as realists or prescriptivists or subjectivists or relativists, and so on.

So I'm a bit confused here. You start by questioning my claim that multiple meta-ethical views can work in tandem and then claim that multiple meta-ethical views can work in tandem. Forgive me if I missing something here.

They make claims about what it means for X to be immoral, which is to make a claim about what "X is immoral" means. — Michael

I think it is a mistake to tie these questions so closely together. It is completely possible to think both that there are real moral facts, but that our moral language does not concern itself with them. That is, one could be a non-cognitivist about actual moral language while nevertheless thinking that there are moral facts that we ought to be thinking and talking about.

I wouldn't say that this is exactly correct. Realism and nihilism are mutually exclusive, as realism argues that some X is inherently immoral whereas nihilism argues that no X is inherently immoral, but nihilism and relativism are not mutually exclusive as the claims "X is not inherently immoral" and "X is non-inherently immoral" are compatible. — Michael

But the views are defined more precisely than this. Robust moral realism is the conjunction of three claims:

• The semantic thesis: Moral assertions are truth-apt.
• The alethic thesis: Some such assertions are true.
• The metaphysical thesis: Those such assertions that are true are made true by objective features of the world, independent of subjective opinion.

Constructivists—including the sort of relativist I take it you mean to be discussing—deny the metaphysical thesis. Nihilists deny both the metaphysical and alethic theses (with error theorists accepting the semantic thesis and non-cognitive nihilists denying it). This can then be further complicated by whether one is a fictionalist about (certain parts of) ethics and what sort of fictionalist one might be (a hermeneutical fictionalist or a revolutionary fictionalist). One can also be a non-cognitivist about the origins of morality (aka "sentimentalism") while nevertheless being a cognitivist about moral language. One can also be an error theorist about some discourses while not being an error theorist about other discourses. And of course, that's just the tip of the iceberg.

Leaving aside the fact that this seems to contradict your immediately prior claim that "you can't be any two at once" — Michael

There is no contradiction. Just because one cannot be both a realist and a nihilist doesn't mean that one cannot be both a nihilist and, say, a non-cognitivist (the cognitivism/non-cognitivism debate also being a metaethical one). Indeed, one can be both a nihilist and a non-cognitivist (as most non-cognitivists are). And as I already alluded to, one could have a view on which one was a hermeneutical non-cognitivist (our moral language does not involve making truth-apt assertions) and a revolutionary realist (there are moral facts, and so our moral language ought to be revised so as to be about them). Again, these are just two possible examples.

I'm not saying that these views rule each other out. I'm actually saying the opposite; that they're compatible — Michael

And my point is that no one has ever said that adopting any single metaethical view thereby rules out all other metaethical views (just the ones that are inconsistent with the one already adopted). That's why I said "I don't understand the criticism" rather than "I think your claim is mistaken."

People do indeed try to argue for just a single meta-ethical view and claim that this meta-ethical view successfully makes sense of (all) moral claims. — Michael

Please cite a professional moral philosopher who does this and where they do it. Even a robust realist—who must address multiple issues in defending his view—nevertheless does not cover all aspects of metaethics simply in virtue of being a realist.

I think it is a mistake to tie these questions so closely together. It is completely possible to think both that there are real moral facts, but that our moral language does not concern itself with them. That is, one could be a non-cognitivist about actual moral language while nevertheless thinking that there are moral facts that we ought to be thinking and talking about. — Postmodern Beatnik

I can't make sense of this. This seems comparable to arguing that there are real 'bachelor facts' but that our 'bachelor language' does not concern itself with them, i.e. to claim that "X is a bachelor" is not truth-apt but that being a bachelor is a fact that we ought to be talking about. It seems to me that if "X is a bachelor" is not (or, rather, is never) truth-apt then ipso facto being a bachelor isn't a fact, and if being a bachelor is a fact then ipso facto "X is a bachelor" is (sometimes) truth-apt.

Furthermore, to suggest that an explanation of the meaning of "X is immoral" can differ from an explanation of what it means to be immoral seems comparable to suggesting that an explanation of the meaning of "X is a bachelor" can differ from an explanation of what it means to be a bachelor. If "X is a bachelor" means that X is an unmarried man then ipso facto to be a bachelor is to be an unmarried man, and if to be a bachelor is to be an unmarried man then ipso facto "X is a bachelor" means that X is an unmarried man.

So I'd say that the two are inextricably tied.

But the views are defined more precisely than this. Robust moral realism is the conjunction of three claims:

The semantic thesis: Moral assertions are truth-apt.
The alethic thesis: Some such assertions are true.
The metaphysical thesis: Those such assertions that are true are made true by objective features of the world, independent of subjective opinion.

Constructivists—including the sort of relativist I take it you mean to be discussing—deny the metaphysical thesis. Nihilists deny both the metaphysical and alethic theses (with error theorists accepting the semantic thesis and non-cognitive nihilists denying it). This can then be further complicated by whether one is a fictionalist about (certain parts of) ethics and what sort of fictionalist one might be (a hermeneutical fictionalist or a revolutionary fictionalist). One can also be a non-cognitivist about the origins of morality (aka "sentimentalism") while nevertheless being a cognitivist about moral language. One can also be an error theorist about some discourses while not being an error theorist about other discourses. And of course, that's just the tip of the iceberg.

...

Just because one cannot be both a realist and a nihilist doesn't mean that one cannot be both a nihilist and, say, a non-cognitivist (the cognitivism/non-cognitivism debate also being a metaethical one). Indeed, one can be both a nihilist and a non-cognitivist (as most non-cognitivists are).

I'm not sure how this runs contrary to what I've said. I'm not saying that any two meta-ethical approaches are compatible (indeed, I concluded nihilism by pointing out (what I believe to be) the flaws of realism); I'm saying that multiple meta-ethical approaches, e.g. expressivism, subjectivism, prescriptivism, relativism (as arguing for non-inherent moral facts), and nihilism (as arguing against inherent moral facts), can all correctly make sense of moral language (and where appropriate, moral facts).

Please cite a professional moral philosopher who does this and where they do it.

I don't think I can, so I'll withdraw the first two sentences of my opening post. I'm now just making a case for a multi-faceted meta-ethics and not casting possible aspersions.

But with that in mind, I'd be interested in which professional moral philosophers have argued in favour of multiple meta-ethical approaches to explain morality and moral language.

I can't make sense of this. — Michael

Let's take a simpler example. Suppose that agent subjectivism is true: statements of the form "x is permissible/obligatory/wrong" mean something like "I approve of/demand/disapprove of x." Just because language developed this way doesn't mean that there are not actual moral facts of which we are ignorant, however. And if we discovered them, we would need to either come up with new words to talk about them or reappropriate our moral language for them. As we have other ways of expressing approval, demands, and disapproval, and as one might think that the new discourse ought to replace the old discourse, it is not entirely bizarre to think the second option (the revisionary one) might be advocated by some.

This happens in other disciplines, as well. Sometimes we take the first option and abandon the old discourse (this is what happened when we abandoned phlogiston discourse in favor of oxygen discourse). Other times we take the second option and reappropriate the old discourse while revising what the terms mean (this is what happened when non-Euclidean geometers decided to redefine the necessary and sufficient conditions for being parallel within their discourse; Kant and Einstein can also be seen as having done this with Newton's notions of time and space, albeit in rather different ways).

I'm certainly not endorsing the combination. But it doesn't make sense to ignore it as a possibility.

Furthermore, to suggest that an explanation of the meaning of "X is immoral" can differ from an explanation of what it means to be immoral seems comparable to suggesting that an explanation of the meaning of "X is a bachelor" can differ from an explanation of what it means to be a bachelor — Michael

You are bewitching yourself with language here (and mistaking the project of metaethics at the same time). One task of metaethics is to explain the meaning of "X is immoral." It is not a task of metaethics to explain what it means to be immoral, however, unless we are using the latter expression to stand in for something more complicated that does not present the same surface difficulties (such as the alternative usage of "mean" in the second expression"). Metaethics is about the meaning of moral expressions, about whether or not a certain type of fact exists, and about the relationship between the facts and our expressions. But there is no presupposition in the subject itself that the expressions must line up with the facts in one of only two ways.

I'm not sure how this runs contrary to what I've said. — Michael

Again, it doesn't. What I am suggesting is that you are hunting snipe. For any two views that are not mutually exclusive, no one says you can't hold them both at the same time. And for any views that are mutually exclusive, it would be a mistake for you to argue that they can be held at the same time. And of course, it has long been recognized that different views might be appropriate for different discourses (or contexts of discourse). In short, I don't see you as having any real targets in your sights.

I'd be interested in which professional moral philosophers have argued in favour of multiple meta-ethical approaches to explain morality and moral language. — Michael

Sure. Let's start with David Hume. He is a non-cognitivist about the origins of morality (this is his sentimentalism, which says that morality is more properly felt than judged and that the origin of morality is in the passions). Nevertheless, he is a constructivist about moral practice (our moral sentiments influence us to adopt various personal and interpersonal moral practices, with varying degrees of compliance). He is also a moral naturalist of sorts (what we count as virtuous and vicious relates directly back to our nature, though our circumstances also play an important role and nature is not the external imposition that Aristotle would have it be). Furthermore, he is a cognitivist about moral language (though morality begins in the sentiments, our primary use of moral statements has come to be the expression of propositions). He is then an error theorist insofar as he thinks those statements have come to be reinterpreted as presupposing the existence of facts that go beyond what is actually the case (so the majority of moral practices are justifiable, but the common justification is mistaken—complete with a few mistaken practices as a result). While he is both a constructivist and an error theorist, he does not adopt them both in relation to any single discourse or praxis (thus he is not violating the "no two at once" rule). Similarly, he is both a cognitivist and a non-cognitivist, but about different things (the origins of morality and moral expressions). He may also be open to the possibility that not all moral expressions are cognitive (as are nearly all cognitivists; it is the non-cognitivists about moral language who are more likely to take a universal approach here—though they need not).

John Mackie is famously an error theorist and a utilitarian. He believes that our actual moral language presupposes moral realism and that objective moral facts do not exist. He is a cognitivist about moral language, however, and so he adopts error theory: ordinary moral language expresses truth-apt sentences, and they all turn out to be false (because they either presuppose or attempt to make reference to facts that do not exist). But he also takes morality to be an indispensable feature of everyday life, so we need to invent and agree to some sort of moral edifice to fill the role that objective moral facts were supposed to fill. With no objective moral facts, the most agreeable candidate to build this edifice on is utility. Thus it makes sense to adopt a form of utilitarianism in both our private and public lives (which is a sort of revolutionary constructivism). Obviously, I am cutting the argument dreadfully short here. I just want to show how he combines various views.

Richard Joyce, meanwhile, is an error theorist and a revolutionary fictionalist. He is a cognitivist about moral expressions, and he takes those expressions to presuppose facts that do not exist (thus they are not true—note that Joyce says they are "not true" rather than "false" because he is a Strawsonian about presupposition, whereas Mackie is a Russellian). Accepting the semantic thesis while rejecting the metaphysical and alethic theses makes him an error theorist about moral discourse. Like Mackie, however, he thinks that moral discourse is too important to simply discard. His recommended solution, then, is that we treat it as a fiction: moral facts don't really exist, but let us act as if they do in order to retain the benefits of having a moral edifice (one that is easier to modify so long as we keep in the back of our minds the fact that it is ultimately a fiction).

So that's three examples. Obviously, I picked from a closely related set of views as they are, in my opinion, some of the better and easier to understand examples. Raimond Gaita has a complete mess of an attempt to bring various threads together (both metaethical and normative) in his Good and Evil: An Absolute Conception, but the book does not lend itself to summary. Sharon Street is another great example, in part because she is a more sophisticated Humean (though in Hume's defense, she has the advantage of being able to draw on nearly two and a half more centuries worth of reflection and observation). Her work is excellent, however. I would highly recommend it.

Suppose that agent subjectivism is true: statements of the form "x is permissible/obligatory/wrong" mean something like "I approve of/demand/disapprove of x." Just because language developed this way doesn't mean that there are not actual moral facts of which we are ignorant, however. — Postmodern Beatnik

But if "X is moral" means "I approve of X" then what does "actual moral fact" mean in the claim "there may be actual moral facts"? The very approach above presupposes a meaning of "actual moral fact" that differs from the meaning of "moral" as explained by agent subjectivism, but as we've accepted, for the sake of argument, that agent subjectivism is true then this presupposition makes no sense.

As we have other ways of expressing approval, demands, and disapproval, and as one might think that the new discourse ought to replace the old discourse, it is not entirely bizarre to think the second option (the revisionary one) might be advocated by some.

But whether or not it would be useful to revise the way we use moral language doesn't change the fact then when people, pre-revision, used the sentence "X is moral" they were reporting on their approval of X (assuming, for the sake of argument again, that agent subjectivism is true). So this revisionist account seems to me to miss the point.

You are bewitching yourself with language here (and mistaking the project of metaethics at the same time). One task of metaethics is to explain the meaning of "X is immoral." It is not a task of metaethics to explain what it means to be immoral, however, unless we are using the latter expression to stand in for something more complicated that does not present the same surface difficulties (such as the alternative usage of "mean" in the second expression"). Metaethics is about the meaning of moral expressions, about whether or not a certain type of fact exists, and about the relationship between the facts and our expressions. But there is no presupposition in the subject itself that the expressions must line up with the facts in one of only two ways.

I still don't understand the distinction. If I explain that "I am a bachelor" means "I am an unmarried man" then I have explained that me being a bachelor is me being an unmarried man. If I explain that "X is moral" means "I approve of X" then I have explained that X being moral is me approving of X.

And thanks for the suggestions. I'll check them out at some time.