Bentham's Blunder (Part 4)

A response to "Moral realism is true"

This is the fourth post in a series of posts responding to Bentham’s Bulldog’s post, Moral realism is true. Here are the previous posts in this entry:

Part 1

Part 2

Part 3

---

9.0 A response to “Classifying anti-realists”

In the next section, BB says:

Given that, as previously discussed, moral realism is the view that there are true moral statements, that are true independently of people’s beliefs about them, there are three ways to deny it.

BB then lists noncognitivism, error theory, and subjectivism as the three ways to deny realism.

BB is, once again, incorrect. These are not the only ways to deny moral realism. I’m a moral antirealist, and I don’t endorse any of these positions. BB is echoing a claim made by Huemer in Ethical Intuitionism that there are only three possible antirealist positions. Huemer, like BB, is simply incorrect.

All three of these positions rely on a semantic thesis:

1. Error theory: Moral claims express propositions that purport to describe stance-independent moral facts

2. Noncognitivism: Moral claims do not express propositions and therefore cannot be true or false

3. Subjectivism: Moral claims are true or false relative to some standard

All three positions turn on a position about the meaning of moral claims. 

They all appear to presume that there is a category, “moral claim,” and that all members of this category share the same semantic content, in that they, as a category, refer to stance-independent moral facts, express nonpropositional content only, or express claims that are true or false relative to some standard.

Here’s the problem: I don’t think there is any category like this. I don’t think there is a category of “claim,” a “moral claim,” that shares some specific semantic content. As an aside, I don’t even think there is a moral domain at all, so I’d have considerable objections to there even being a well-defined category of “morality” in the first place. But setting that aside, the characterization common among these positions of language and meaning is one that seems to rely on a conception of language that I don’t share. I endorse a language-as-use view, according to which words, phrases, and claims themselves don’t mean anything: it’s the people using language that mean things. This is captured by my slogan that “words don’t mean things; people mean things.” I cannot emphasize enough that I mean this literally: I do not think words mean anything. I think the people using words mean things, and the words are their means of conveying what they mean. 

When we talk about the “meaning of moral statements,” I take this as an awkward and non-literal characterization of what people mean when they make moral statements, where “moral statements,” would be operationalized as some rough attempt at capturing a subset of ordinary discourse. I do think that when people make moral claims, they mean things, but:

1. I take facts about what people mean to be empirical questions

2. I endorse folk metaethical indeterminacy, i.e., I think that with respect to metaethical considerations, ordinary people rarely mean to express claims that determinately fit a realist or antirealist analysis at all

Mainstream analytic philosophers addressing questions about the meaning of moral claims tend to either (1) appear not to treat such questions as empirical at all, and instead see such claims as a priori. I reject this view, or (2) may nominally regard such questions as empirical but believe armchair philosophy is sufficient to settle such questions reasonably well. I disagree with (2) as well. As such, I either fundamentally reject the philosophical presuppositions behind the approach most philosophers take to addressing metaethical questions, or, at best, believe they’re using embarrassingly bad methods for addressing those questions: I am adamantly against generalizing from the armchair. If you want to know what nonphilosophers mean when they make moral claims, you need to engage in empirical research.

…And that’s what I did. I came away with the impression that, whatever they mean, they are not making claims that determinately fit error theory, noncognitivism, or subjectivism. These are not the only logical possibilities, pace Huemer. Huemer is mistaken about this. Alternative possibilities are not novel, nor did I invent them. Loeb describes one possibility: moral incoherentism: people may hold conflicting presuppositions at the same time, and, as a result, their moral claims could be incoherent. This is not strictly speaking error theory, subjectivism, or noncognitivism, and therefore reflects a genuinely unique type of antirealism. Michael Gill proposes folk metaethical variability and indeterminacy: it could be that there is considerable variation in the metaethical presuppositions implicit in ordinary moral discourse, in which case no single antirealist view would be correct; one would need to be a pluralist: insofar as people speak like realists, one might be an error theorist, and insofar as they use moral language to express nonpropositional attitudes, one might be a noncognitivist. Arguably, this would simply yield a hybrid view: a combination of the three possibilities BB lists. However, Gill’s other proposal, indeterminacy, reflects a genuine departure from the three: in those instances in which a nonphilosopher’s moral claims neither determinately comport with a realist or antirealist analysis, there may be no fact of the matter about whether the person’s claim fits the error theorist, noncognitivist, or subjectivist analysis.

Even when this occurs, you can still believe there are no stance-independent moral facts. Denying that there are such facts does not require a commitment to a determinate semantic analysis of the meaning of moral claims. Antirealists are not required to share BB or Huemer’s assumptions about language and meaning. Once again, this is a point I’ve already addressed on my blog, so if you want to see my develop on this point more, I do so here.

In short: these three views are not the only antirealist views that one can endorse. Antirealism only requires you to reject moral realism. It does not require you to make a claim about the meaning of moral claims. You could be agnostic about the meaning of moral claims, or endorse indeterminacy, variability, or incoherentism. Fundamentally, the problem with the claim that these are the only three options is that all three positions rely on substantive assumptions about language and meaning that are not an inherent feature of moral antirealism as a position, and one is therefore not obliged to endorse one of these positions to qualify as an antirealist.

One of the strategies you can employ if you insist there’s only three antirealist positions is an argument by process of elimination: if you can demonstrate some deficiency in each of these forms of antirealism, you can declare victory for realism by default. I think it’s dialectically important, as a results, for BB, Huemer, and others, to sideline positions like mine and show they’re not legitimate possibilities. This allows the process of elimination arguments to go through. I have yet to see any good argument for why these are the only three options, though. This appears to be, once again, a matter of assertion.

Although I don’t endorse any of these antirealist views, critiques of all three are usually illuminating, in that they’re not very good. Even when moral realists can tee up the target position to knock it down, they still fail. So let’s go through BB’s critiques of each of these positions.

10.1 Noncognitivism

Noncognitivism is the view that moral claims do not express propositions (statements that can be true or false) and that therefore moral claims are neither true nor false. Compare to utterances like “Eww!” (an emotive expression) or “Shut the door!” (an imperative). These utterances cannot be true or false. Noncognitivists hold that, appearances notwithstanding, moral claims likewise express only nonpropositional content like this. BB quotes a previous article that explains why BB thinks noncognitivism is implausible. The argument is a reiteration of the Frege-Geach problem.

The problem is supposed to be that, if moral claims didn’t express propositions, then we shouldn’t be able to place moral claims into syllogisms, arguments, lines of reasoning, and other chains of inference that only make sense if the phrases they contain are propositions. BB gives an example:

“It’s wrong to torture infants for fun, most of the time,”

is neither true nor false

The statement

“If it’s wrong to torture infants, then I shouldn’t torture infants

It’s wrong to torture infants

Therefore, I shouldn’t torture infants”

Is incoherent. It’s like saying if shut the door then open the window, shut the door, therefore, open the window.

I’ve always found this to be an atrocious and silly objection to noncognitivism. Perhaps a very crude and flat-footed form of noncognitivism is vulnerable to this criticism. According to such a view, the only appropriate use of “X is wrong” amounts to something like “X? Boo!” or “Don’t X.” So this would make certain utterances like the ones BB provides appear incoherent. And it probably does successfully indicate that such a view is incoherent. Only, is a noncognitivist obliged to think that ordinary moral language is as rigid and inflexible as this?

No. People may use moral claims primarily to express nonpropositional claims even if one can shift into using them in a propositional way in various contexts. How reflective of ordinary thought and language is a syllogism like the one BB offers, anyway?

Not very. It is the artificial construction of the philosopher. Most people don’t speak in modus ponens most of the time. And if terms and phrases are flexible and change based on context, it could still turn out that in most instances, the primary use of a moral claim is to express nonpropositional content, even if there are conceivable contexts where there is nothing especially strange about treating them as propositional. In other words, the phrase “X is wrong,” in ordinary language need not always and in every case mean exactly the same thing, for it to still be the case that moral claims primarily serve to express emotions or issue commands. There may be little reason to prioritize the assertoric functions ordinary moral claims can play if these are secondary to and parasitic on their primary uses. At best, Frege-Geach only reveals a problem for rigid semanticists, but similar problems emerge for rigidity in every direction, in that one can always find apparent pockets of English moral discourse that are hard to reconcile with an opposing view, provided one wishes to adhere to the uniformity and determinacy assumptions, as outlined by Gill. The real problem here is a problem for all rigid semantic accounts: there was never any good reason to think all moral claims share the same uniform and determinate set of semantic characteristics in the first place.

It’s also worth noting that there are a variety of hybrid accounts and various approaches to language that circumvent the rigidity and crudeness of classical noncognitivist views. The objection BB offers is, at best, only serviceable as a critique of crude, flat-footed, early forms of noncognitivism from decades ago.

BB goes on to cite Huemer, who provides other reasons to regard moral claims as propositions:

(a) Evaluative statements take the form of declarative sentences, rather than, say, imperatives, questions, or interjections. 'Pleasure is good' has the same grammatical form as 'Weasels are mammals'. Sentences of this form are normally used to make factual assertions. )] In contrast, the paradigms of non-cognitive utterances, such as 'Hurray for x' and 'Pursue x', are not declarative sentences.

Everyone recognizes this and it was never in dispute. The mere fact that something is expressed as a declarative sentence doesn’t mean that it functions to express a proposition. People use declarative sentences to express nonpropositional content in English: e.g., “You will not go to that party,” “This cake is delicious!”. I don’t think that what a person means isn’t determined by grammar; I think it’s determined by what they’re trying to do with their words.

Note, too, that these remarks only concern English. Such claims may or may not extend to the 7000+ other languages in the world.

(b) Moral predicates can be transformed into abstract nouns, suggesting that they are intended to refer to properties; we talk about 'goodness', 'rightness', and so on, as in 'I am not questioning the act's prudence, but its rightness'.

People reify and speak in metaphorical terms all the time. This is a common feature of English and probably most other languages (if not all of them). It demonstrates very little. 

(c) We ascribe to evaluations the same sort of properties as other propositions. You can say, 'It is true that I have done some wrong things in the past', 'It is false that contraception is murder', and 'It is possible that abortion is wrong'. 'True', 'false', and 'possible' are predicates that we apply only to propositions. No one would say, 'It is true that ouch', 'It is false that shut the door', or 'It is possible that hurray'.

This recapitulates the same crude criticism of noncognitivism outlined above. If a noncognitivist has a rigid and inflexible conception of language, this criticism may land. But they don’t have to think language is that rigid.

I also question how common or representative such remarks are of ordinary moral language. These sound more like the sorts of things philosophers say. Note, again, that such considerations only emphasize English, as well. They may or may not reflect how speakers of other languages tend to use moral language (if they use moral language at all).

Finally: 

(d) All the propositional attitude verbs can be prefixed to evaluative statements. We can say, 'Jon believes that the war was just', 'I hope I did the right thing', 'I wish we had a better President', and 'I wonder whether I did the right thing'. In contrast, no one would say, 'Jon believes that ouch', 'I hope that hurray for the Broncos', 'I wish that shut the door', or 'I wonder whether please pass the salt'. The obvious explanation is that such I11ental [sic] [mental?] states as believing, hoping, wishing, and wondering are by their nature propositional: To hope is to hope that something is the case, to wonder is to wonder whether something is the case, and so on. That is why one cannot hope that one did the right thing unless there is a proposition-something that might be the case-corresponding to the expression 'one did the right thing'.

There are a few more examples but they’re more or less of the same flavor. This kind of reasoning tells you how people can use moral claims. That’s important. It at least means that there are instances in which people saying things like “murder is wrong,” aren’t merely expressing an emotion or issuing a command. And insofar as noncognitivism supposes that that’s all and only what people do with moral language, well, it’s probably not a very sensible position. Note, however, that such “tests” are purely armchair tests: by thinking about how it seems reasonable for us to use certain terms and phrases, it appears some of those terms and phrases seem to comfortably play propositional roles in at least some contexts.

What does this tell us, though? If we suppose that questions about what people mean when they make moral claims are empirical questions, then Huemer, BB, and others are not employing the best methods for determining how people use moral language. As the famous phrase goes, no plan survives contact with the enemy. Just so, no simple armchair hypothesis survives contact with people. People are complicated and messy, and so is language. All of the armchair considerations in the world could crumble the moment we conduct actual empirical research on what people are trying to do with moral language. And that’s just it: why suppose they’re only trying to do one thing? Either assert or not assert propositions? While my own twist on noncognitivism may not be orthodox, we could discover as a matter of empirical fact that in most ordinary contexts in which people make moral claims, the primary function of those claims is to express nonpropositional content. And suppose that the use of declarative sentences involved for rhetorical and persuasive reasons, not because people really thought there are “moral facts” (stance-independent or otherwise). If so, then the standard grammatical rules of our language, and many ways in which people talk, could both give the impression that moral language “is propositional” in some contexts, even if in most actual instances of use, they don’t serve to express propositions.

What then? Do moral claims express propositions or not? The answer would, in these cases, turn out to be, “it depends.” Even if people use moral language in a propositional way some of the time, we may still ask why people use language in that way: for what purpose? To what end? Such questions are not easy to answer from the armchair. These are, ex hypothesi, empirical questions, and are best addressed using empirical data. Why, then, do BB and Huemer rely on armchair considerations? This strikes me as a strange way to address empirical questions, if they are, in fact, empirical questions.

Suppose instead that questions about what people mean aren’t empirical questions. What are they questions about, then, if not what people mean? Are these supposed to be a priori questions? If so, about what? English sentences? There’d be something profoundly strange about the question of the meaning of moral claims not being an empirical question, but philosophers may buy into such a notion. Even if they do, we still face what I will call the empiricist’s dilemma: either 

(a) these questions are empirical, in which case empirical methods are the best means of addressing them, not, as BB and Huemer do, by engaging in armchair theorizing

(b) these questions are not empirical, in which case they face a somewhat different challenge.

With respect to (b), suppose we declare that the cognitivism/noncognitivism question is not an empirical question. Now suppose we discover that, in 100% of actual, real-world instances in which people say things like “murder is wrong,” and “honesty is morally good,” they themselves intend only to express emotions or issue commands. That is, their communicative goal is to convey nonpropositional content alone. What are we to conclude?

One possibility is to suppose that these are not actual moral claims. Since these people are only expressing nonpropositional content, and moral claims are, a priori, propositions, then it would appear that nobody ever actually makes moral claims in everyday situations. This strikes me as a reductio ad absurdum: if you’re so committed to “moral claims” being propositional that this would lead you to conclude that nobody ever actually makes moral claims in ordinary contexts, there’s something wrong with your theory.

If, instead, we suppose that these are moral claims, then what? We might say that meaning isn’t determined by the goals of speakers, and that these people’s moral claims actually are propositions. This, too, strikes me as very strange: it would mean that while in practice, in everyday contexts, nobody ever used moral claims to express propositions, they nevertheless do, in fact, express propositions. This seems to me to rely on an insane view of language, where the language itself means things, rather than the people using it. If a philosopher wants to defend such a view, they’re welcome to it, and if noncognitivists and cognitivists alike want to go for something like this, I’d object to both. I don’t believe words and phrase have any sort of meaning independent of their use in actual contexts. Even so, views like this may be popular among philosophers, so they may opt for something like this. And that’s just it: what if I don’t accept this view of language? This brings us back to the claim that only the three views outlined by BB are available to the antirealist. If it turns out that these views rely on specific commitments to particular views in philosophy of language, and you don’t endorse those views, then what? Do you just not get to have a metaethical view? That’d be ridiculous. You certainly can have a metaethical position without endorsing a specific view about language and meaning.

None of this should be taken as some sort of defense of noncognitivism. I don’t endorse noncognitivism. However, I think criticisms of noncognitivism stem from and appear plausible only insofar as one endorses views of language that I don’t share: namely, the presumption that there is a category, “moral claim,” that shares a uniform semantics, such that all moral claims are propositional or nonpropositional. If we examine the actual patterns of use of moral claims in ordinary discourse, where “moral claim” is operationalized to refer to the sorts of things people actually say, it’s an open empirical question what people mean when they make those claims. Critics of noncognitivism rarely, if ever, engage with empirical literature on how people use language. If they think that the meaning of moral claims doesn’t turn on empirical considerations, critics like me will reject those presumptions. If they think they are empirical questions, then why aren’t they engaging in or consulting empirical research? Perhaps armchair considerations are sufficient to establish that it ordinary English does appear to allow, without it seeming strange or improper, to speak of moral claims in ways that treat them as propositions. Fair enough: perhaps so. This does not, by itself, show that moral claims are propositions. It only shows that they can be. And perhaps in other contexts they’re not. If that’s enough to refute noncognitivism, perhaps it’s refuted. But it doesn’t establish cognitivism, either, if cognitivism is the view that all moral claims are propositions. Perhaps the cognitivism/noncognitivism dispute is a false dichotomy based on incorrect conceptions about the rigidity of language.

In sum: insofar as noncognitivists buy into the same conception of language as realists do, they may very well be vulnerable to these charges. But there are ways of thinking about moral language that don’t require such views, and that still allow for a view of ordinary moral thought and discourse that is at least similar to noncognitivism. I suspect, then, that insofar as Huemer’s critiques of noncognitivism “work,” they work largely on the basis of sharing in the same mistaken presumptions about language and meaning as the target noncognitivist.

10.2 Error theory

BB begins with this description of error theory:

Error theory says that all positive moral statements are false. 

This is a bad description of error theory, though BB later clarifies what the error theorist thinks to some extent. Error theorists specifically hold that all first-order moral claims involve an implicit commitment to one or more false presuppositions and are therefore systematically false. The most common form of error theory holds that moral claims purport to describe stance-independent moral facts, and since there are none, then such statements are false. This is a bit like if one were to go around making claims about the various features of unicorns, even though unicorns didn’t exist:

1. Unicorns like cupcakes.

2. Unicorns can magically heal you.

3. Unicorns dislike violence.

Since these claims implicitly presuppose that unicorns exist, then if unicorns don’t exist, these statements are false. Just so, if moral claims presuppose that there are stance-independent moral facts, but there aren’t any, then such claims are false.

BB’s description doesn’t clearly convey this initially, but later adds:

The error theorist has to say that the meaning of those terms is exactly the same as what the realist thinks. 

What about BB’s objections to error theory? 

Error theory is best described as in error theory, because of how sharply it diverges from the truth. It runs into a problem — there are obviously some true moral statements. Consider the following six examples.

1. What the icebox killers did was wrong.

2. The holocaust was immoral.

3. Torturing infants for fun is typically wrong.

4. Burning people at the stake is wrong.

5. It is immoral to cause innocent people to experience infinite torture.

6. Pleasure is better than pain.

The only way in which this objection would work is if Bb meant that there are obviously some stance-independently true moral statements. So error theory holds that there are no stance-independent moral facts. BB’s objection to this? That it’s obvious that there are such facts. This is not an objection. It is simply an assertion to the contrary. I could simply retort: “No, it’s obvious there are no such facts.” After all, it seems obvious to me. That’s literally true. 

BB’s response here is very feeble. At best, it’s simply a reiteration of some kind of Moorean foot-stomping insistence that it’s just obvious that moral realism is true. A realist could offer this response to any argument at all for an antirealist position, so what’s the point in doing it here? Why not just say: “It’s obvious moral realism is true. QED.” and end his post there? 

If, in the absence of any specific criticism of a metaethical theory, one always just falls back on reiterating that realism is obviously true, what’s the point in describing these positions and framing one’s response as though one is offering a distinctive response to various alternatives? 

BB also says a few other things. Consider this remark:

The error theorist has to say that the meaning of those terms is exactly the same as what the realist thinks.

Has to? BB frames this like the error theorist is making some sort of concession. They’re not.

The error theorist has to think that when people say the holocaust is bad, they’re actually making a mistake. However, this is terribly implausible. It really, really doesn’t seem like the claim ‘the holocaust is bad’ is mistaken. 

If by “bad” you mean “stance-independently bad” then it does seem mistaken to me. I don’t think anything is “stance-independently bad.” In fact, it seems obvious to me that it isn’t bad. If that strikes BB or anyone else as horrifying or repugnant, then I think they’re making the same mistake as I keep pointing to of engaging in normative entanglement. Not thinking something is “stance-independently bad” doesn’t require you to not oppose it with just as much passion, or to feel just as much repugnance, as a moral realist. An error theorist can, without any inconsistency, be exactly as opposed to murder, torture, and so on as the realist. There are no substantive practical implications to thinking the things on BB’s list aren’t stance-independently immoral, or bad, or whatever. BB and others constantly entangle metaethical claims with normative claims, giving the false and misleading impression that if you deny the realist’s metaphysical views, that you’re somehow an awful, terrible, no good person with repugnant views. I’m not the first or only person to draw attention to this problem. Joyce alludes to normative entanglement in the SEP entry on moral antirealism:

The last example (“Stealing is not morally wrong”) calls for an extra comment. In ordinary conversation—where, presumably, the possibility of moral error theory is not considered a live option—someone who claims that X is not wrong would be taken to be implying that X is morally good or at least morally permissible. And if “X” denotes something awful, like torturing innocent people, then this can be used to make the error theorist look awful. But when we are doing metaethics, and the possibility of moral error theory is on the table, then this ordinary implication breaks down. The error theorist doesn’t think that torturing innocent people is morally wrong, but doesn’t think that it is morally good or morally permissible either. It is important that criticisms of the moral error theorist do not trade on equivocating between the implications that hold in ordinary contexts and the implications that hold in metaethical contexts.

I believe all or most of the rhetorical force BB and others get out of depicting the error theorist as thinking that torturing babies is “not morally wrong,” derives from just this conflation: the pragmatic implication that the error theorist’s views are truly ridiculous and awful and repugnant. Yet if the error theorist merely thinks that a certain metaphysical thesis is baked into everyday discourse, that that thesis is mistaken, but that this has absolutely no practical consequences at all and has no impact on their attitudes, behavior, or judgments, what exactly so objectionable about the error theorist’s view? Merely rejecting baroque metaphysics is no cause for pearl clutching horror, yet such pearl clutching seems to lurk beneath the incredulous declarations that the error theorist’s view is “implausible.” WHY is it implausible? I believe BB and others mistakenly find it implausible because they fail to disconnect the error theorist’s claim from the implications they and others project onto it, and I think that’s because they themselves think things only are or could be valuable if they’re valuable in a realist sense. Neither the error theorist nor anyone else is obliged to share in this view. The realist seems to partially impose their own preconceptions onto the error theorist or other antirealist’s response and worldview, then stand back in shock and horror at what the error theorist appears to be saying, thinking, and feeling. This is a partial failure of imagination, and an insistence on partially imposing one’s own philosophical views on other people. I call this the halfway fallacy, and discuss it here:

The halfway fallacy occurs when one argues that a particular position contrary to their own has one or more flaws or undesirable characteristics, but those flaws or undesirable characteristics are only applicable to the position if some set of claims you believe are true are in fact true, but you haven’t argued for and those the objections are directed at are free to (and in many cases probably would) reject. In other words, the problem occurs when one holds certain presuppositions that those who hold the view are free to reject (it might also be the case that they aren’t merely free to reject these presuppositions but do, and it may even be that rejecting the presuppositions is a natural and synergistic feature of the contrary view).

One might also call this the fallacy of unshared presumption. The central problem with this form of reasoning is that an argument for or against some position is based on considering the implications of someone holding one or more views contrary to one’s own, but, critically, not considering that they reject certain other presuppositions you hold.

I believe BB and other realists are routinely guilty of the halfway fallacy when criticizing error theory and other antirealist positions.

Then we get this:

Any argument for error theory will be way less intuitive than the notion that the Holocaust was, in fact, bad.

What is this? An argument? It sounds like a slogan. It also seems like it’s declaring, in advance, that no argument for error theory will be intuitive, without even having to consider what the content of those arguments actually is. How on earth can BB possibly know this? Even if it were true, what is the implication? If an argument’s premises need to be intuitive for BB to accept them, then is BB implying no argument for error theory could be convincing in principle? If not, then is BB only suggesting that the premises of arguments for error theory will always be “way less intuitive” than the claim that something is stance-independently bad (note BB still just says “bad” without specification, as if antirealists like me don’t think these things are “bad”), but that BB would still consider accepting such arguments anyway? If so, then what’s the point of saying this?

Note, again, the use of “intuitive” as if intuitions aren’t features of people. Claims can’t be “intuitive” or “counterintuitive” in and of themselves; they can only be intuitive to someone. BB then says, “Let’s test these intuitions,” but for some reason repeats a handful of seemingly unrelated principles from earlier. There may be a typo or duplication here or something. I’m not sure, so I’ll move on and maybe BB will clarify.

10.3 Subjectivism

BB next turns to subjectivism, which he characterizes as follows:

Subjectivism holds that moral facts depend on some people’s beliefs or desires. This could be the desires of a culture — if so, it’s called cultural relativism.

This is ambiguous between agent and appraiser relativism. Agent relativism holds that moral claims are made true by the standards of the agent performing the action or the culture of that agent. Appraiser relativism holds that moral claims are true or false relative to the standards of whoever is evaluating the moral action (or principle, character trait, etc.) in question. Critics of relativism routinely focus only on agent relativism, critiquing it to the exclusion of appraiser relativism, as though the latter didn’t exist. BB does that here. 

10.3.1 Cultural relativism

But BB also gets a little silly with the rhetoric. Look at how BB opens this discussion:

This is an embarrassing subheading. In any case, BB goes on to describe cultural relativism as follows:

Cultural relativism is — as the sub-header suggested — something that I find rather implausible. There are no serious philosophers that I know of who defend cultural relativism. One is a cultural relativist if they think that something is right if a society thinks that it is right.

BB makes no clear distinction between agent and appraiser cultural relativism. The last sentence indicates that this is agent cultural relativism, not appraiser cultural relativism. An agent cultural relativist thinks that if a particular culture regards an action is good or bad, then it is good or bad. That culture “fixes” or sets the moral standards for itself. An appraiser cultural relativist instead holds that when we say things like:

“That’s morally wrong.”

This means something like:

“That violates my culture’s standards.”

The appraiser cultural relativist can and does judge people according to their own culture’s moral standards, not according to the standards of the culture whose practices they’re evaluating. This distinction is mentioned in the SEP entry on moral relativism:

[...] that to which truth or justification is relative may be the persons making the moral judgments or the persons about whom the judgments are made. These are sometimes called appraiser and agent relativism respectively. Appraiser relativism suggests that we do or should make moral judgments on the basis of our own standards, while agent relativism implies that the relevant standards are those of the persons we are judging (of course, in some cases these may coincide). Appraiser relativism is the more common position, and it will usually be assumed in the discussion that follows. Finally, MMR may be offered as the best explanation of what people already believe, or it may be put forward as a position people ought to accept regardless of what they now believe. There will be occasion to discuss both claims below, though the latter is probably the more common one.

In any case, appraiser relativism does not carry the implication that if some culture or individual thinks torture is okay, that you must think it’s okay for them to torture people. That’s only a feature of agent relativism. I elaborate on the agent/appraiser distinction here.

Does BB’s critique “work” against agent cultural relativism? 

Sort of. If you really think that if some culture thinks it’s okay to torture babies then it is okay for them to torture babies, well, I’m not going to get on board with that. In that respect, I share the revulsion and opposition to such a view that BB likely shares. But I don’t think this makes the view false as a metaethical theory. I reject such views because:

1. They don’t reflect how I speak or think

2. I don’t endorse the normative and practical implications of these views

3. Insofar as they purport to capture ordinary thought and language, I doubt they capture any more than a small subset of some populations

I bet BB would agree with (a), (b), and (c). Note, however, that relativist accounts traditionally purport to capture ordinary moral semantics. In this respect, insofar as they succeed at doing so, they could turn out to be descriptively true, and, insofar as they are true in this respect, most critiques that appeal to one’s horror, shock, and outrage, would simply fail.

Suppose the relativist holds this view:

When ordinary people make moral claims, those claims serve to express a propositional statement about what is or isn’t consistent with the standards of that culture and, if it is consistent with the standards of that culture, this is what it means for the action to be good or permissible.

Now suppose that this is, in fact, what people mean. When they say things like “murder is wrong,” they mean “murder is inconsistent with my culture’s moral standards,” and further we establish that what it means for an action to be right or wrong just is that it is right or wrong according to that culture’s moral standards. If one makes the further normative leap to suppose that everyone is bound by this normative-value-fixing feature of ordinary moral language, then presto: agent cultural relativism is true, whether you like it or not.

In other words, whether agent cultural relativism is true or not could just turn out to be true in virtue of the semantics of ordinary moral language and the associated metaphysics of moral-fact-fixing features of language. If this is just how language and metaphysics plays out, then, and that it really is the case that each culture “decides for itself” what is “good or bad” relative to that culture, well, tough shit. Facts about how people speak aren’t refuted by horror or repugnance.

Of course, it isn’t repugnance or horror at the implications of agent cultural relativism that serve as direct or decisive evidence that it’s false. Rather, it’s that such an account is counterintuitive. That is, it may not seem to us that this is what people mean when they make moral claims. And that may very well be evidence that this isn’t what people mean. Fair enough: if this isn’t what people mean, and agent cultural relativism is in part supposed to be a descriptive account of what people mean, then it is to that extent false. But again, I would’ve thought that facts about what people mean are empirical questions. While our intuitions may provide some evidence of what people mean, and may be a fairly reliable guide, the final arbiter will be empirical data, not armchair hypotheses. Suppose it turned out that, as a matter of fact, this is what people mean: they really do speak and think like agent cultural relativists. If one holds a view of language where ordinary usage fixes the subject matter, then it simply follows that, as a matter of descriptive fact, agent cultural relativism correctly captures the meaning of ordinary moral claims. 

This is where things get weird. Agent cultural relativism isn’t, strictly speaking, characterized as an account of ordinary language. It seems to carry normative moral implications. If one thinks that when people make moral claims, those claims refer to what is morally right or wrong relative to their cultures, this isn’t enough: one must also think that those claims in fact determine what is morally right or wrong for members of that culture. But how could that be a feature of the meaning of the claims? Simple: it can’t. This is a further, normative thesis that isn’t and couldn’t be a feature of ordinary moral language because it’s a normative thesis, not a thesis about the meaning of moral claims; it’d be one thing to say that people who make moral claims intend for those claims to be interpreted as having these normative implications; it’s quite another to say that their claims in fact carry those normative implications: the latter could not in principle be a feature of the meaning of the claims in question, for the same reason that if, when I say “that is morally obligatory,” I mean “that maximizes utility,” this actually makes it the case that maximizing utility is morally obligatory. Simply put: agent cultural relativism is not a purely metaethical theory at all; it is both a metaethical theory and a normative moral theory. 

Ironically, it is the normative implications of “agent” relativism that are so repugnant, not the metaethical implications. Even more ironically, what makes these normative implications so objectionable is the fact that they function the way moral facts function for moral realists. What do I mean by this? I mean this:

Moral realists hold that there are stance-independent moral facts, and typically maintain (unless they’re some kind of naturalist) that these moral facts have some kind of “authority” over us: they compel us to act, they “bind” us, they “require” us to comply, and so on.

Agent relativists hold that there are stance-dependent moral facts that likewise bind us in more or less the same way. Whereas for the realist there are facts that bind us that are true independent of anyone’s standards, for the agent relativist there are facts that bind us independent of our own (or our culture’s) standards: namely, the standards of other individuals and cultures. If the members of another culture thinks it’s okay to torture babies, then agent relativists grant that it is okay for them torture babies in such a way that you must be okay with them torturing babies.

This may very well be a ridiculous view, but it’s ridiculous because it compels us to honor the moral values of other individuals or cultures, even if those values conflict with our own values. 

Let’s grant that agent relativism is a highly objectionable view. What about appraiser relativism? Unfortunately, BB doesn’t offer a critique of appraiser relativism. As such, BB has not adequately refuted subjectivism. Appraiser subjectivism and cultural relativism don’t carry the normative implications outlined above. They can simply construe moral claims to be reports of the speaker’s standards or the standards of their culture. Such statements carry no necessary implications about how anyone else ought to act. For instance, suppose a psychopath says:

It’s morally good for me to torture babies for fun.

The appraiser relativist interprets this as:

It’s consistent with my moral standards to torture babies for fun.

If it is consistent with the psychopath’s standards, this would be trivially true. But it also carries no normative or practical implications for anyone else: all such a truth amounts to is the truth that the psychopath has accurately reported their personal values. You don’t have to honor or care about those values at all, because, for the appraiser relativist, your own moral claims reflect your values (or your culture’s). It’s much harder to show that such a position carries repugnant implications. I haven’t seen anyone convincingly do so.

10.3.2 Individual subjectivism

BB next turns to individual subjectivism:

Individual subjectivism says that morality is determined by the attitude of the speaker. The statement murder is wrong means “I disapprove of murder.” 

This appears to be appraiser individual subjectivism. Technically, the culture/individual and agent/appraiser distinctions are orthogonal to one another: one can be an agent cultural relativist or an appraiser cultural relativist, and one can likewise be an agent individual subjectivist or an appraiser individual subjectivist. BB only addresses agent cultural relativism and appraiser individual relativism. BB says he’s given objections in a previous article, and quotes himself as follows:

If it’s determined by the moral system of the speaker the following claims are true.

“When the Nazi whose ethical system held that the primary ethical obligation was killing jews said “It is moral to kill jews,”” they were right.

“When slave owners said ‘the interests of slaves don’t matter,’ they were right.”

“When Caligula says "It is good to torture people,” and does so, he’s right”

“The person who thinks that it’s good to maximize suffering is right when he says “it’s moral to set little kids on fire””

Additionally, when I say “we should be utilitarians,” and Kant says “we shouldn’t be utilitarians,” we’re not actually disagreeing.

BB seems to mix up agent and appraiser relativism here. BB’s initial characterization appears to be appraiser individual subjectivism. This is because, if someone says “X is wrong,” then they mean “I disapprove of X,” then their claims only amount to reports about their personal standards. This does not, in itself, entail that those claims fix or determine what is morally right or wrong for that person. Yet these examples seem to indicate that BB is discussing agent individual subjectivism. First, BB says:

If it’s determined by the moral system of the speaker the following claims are true.

This is ambiguous and consistent with both agent and appraiser relativism. What matters is what it is that’s being determined as true. If when a person says “X is wrong” they just mean “I disapprove of X,” the only thing that’s being determined to be true when the person makes moral claims are facts about whether they approve or disapprove of the claims in question and whether that disapproval is consistent with the claim being made, e.g., if they say “X is wrong,” and they disapprove of X, the fact that they disapprove of X is what makes “X is wrong” true.

If this is all subjectivism amounts to, BB’s examples are all extremely misleading: they only appear counterintuitive, horrifying, repugnant, and awful if the reader fails to interpret them in line with the appraiser individual subjectivist’s analysis of the meaning of these claims. Let’s translate them in accord with that analysis:

“When the Nazi whose ethical system held that the primary ethical obligation was killing jews said “It is moral to kill jews,”” this statement meant “We approve of killing Jews.” Since they did approve of killing Jews, their statement was true.

“When slave owners said ‘the interests of slaves don’t matter,’ this meant “We don’t care about the interests of slaves.” Since they didn’t care about the interests of slaves, this statement was true.”

“When Caligula says "It is good to torture people,” and does so, he means “I like torturing people.” Since Caligula does like torturing people, this statement was true.

“The person who thinks that it’s good to maximize suffering is right when he says “it’s moral to set little kids on fire,” because he means “Setting little kids on fire will maximize suffering and that’s what I want to do,” and since it will maximize suffering is what he wants to do, what he says is true.

As you can see, once you translate all of these statements into the appraiser individual subjectivist’s analyses, all of these statements are true: they’re true reports about the attitude of approval or disapproval of the speaker. By obscuring this fact, BB’s remarks give the impression that if you endorse those statements because you agree with the subjectivist, that you think killing Jews, slavery, and baby torture are good. But this is not an implication of the appraiser subjectivist’s analysis of the meaning of these statements. 

The force of BB’s objections rely on actively misleading readers. I rarely say this, but BB’s “arguments” here are not just terrible, they are unethical: even if BB isn’t aware of how misleading his remarks are, he has a minimal moral obligation not to be so negligent and sloppy in his presentation so as to mislead audiences. I believe BB’s objections here rely so extensively on rhetoric and misleading framing that, if doing so isn’t intentional (which would definitely make it unethical), it’s still culpably negligent. In short: I’m not just saying BB is wrong here (he is), I’m saying BB’s remarks are so misleading as to be unethical. It is morally wrong to misleadingly imply that philosophers who hold philosophical positions contrary to your own are okay with genocide and torture when this impression is entirely the result of your own negligence. BB and other realists should stop doing this: stop implying that antirealists are okay with genocide and torture, when you ought to know better. 

BB’s final remark is also an interesting one:

Additionally, when I say “we should be utilitarians,” and Kant says “we shouldn’t be utilitarians,” we’re not actually disagreeing.

I suppose this is supposed to indicate some kind of counterintuitive implication of individual subjectivism. It isn’t. Once again, let’s translate it:

Additionally, when I say “I approve of us being utilitarians”” and Kant says “I disapprove of us being utilitarians” we’re not actually disagreeing.

This is a disagreement. It’s not a disagreement about what’s true. It’s a disagreement about what to do. Suppose you and a friend want to order food. You’re in the mood for pizza. They’re in the mood for sushi. If you say “We should get pizza,” and your friend says, “We should get sushi,” does this require you to be a gastronomic realist in order to disagree? No: you need not think there is some stance-independent fact of the matter about the “correct,” food to get, where the “correct” food to get has nothing to do with your preferences. Nothing strikes me (and, I hope, not you) about construing these remarks in subjective terms:

You: “I would prefer we get pizza.”

Them: “I would prefer we get sushi.”

Neither of you disagrees about what’s true. But you do disagree about what to do. When realists claim that relativism and other views entail that nobody “disagrees,” they are implicitly employing a narrow conception of “disagreement” that involves only disputes about what’s true. But people also disagree on matters of coordination: peoples’ goals conflict, and when this happens, they run into conflicts. When you want to pay $5000 for a car, but the car salesperson wants you to pay $6000, you disagree about the price of the car. Such “disagreements” don’t require anyone to think there is a “true” car price.

Subjectivists can and do disagree in this respect: If I want to live in a world without torture, and a psychopath wants to live in a world with torture, we don’t agree about what we should do. We don’t have to think there are stance-independent moral facts for us to have conflicting attitudes about what to do, nor would it be senseless or insane or a waste of time or self-contradictory to argue. Arguing likewise doesn’t require that one argue about what’s true. Arguing can involve negotiation about what to do.

Realists, and philosophers more generally, have an arbitrarily narrow conception of “disagreement,” the rest of us are not obliged to endorse. Once again, I have an article addressing this, called “What’s true vs. what to do.” Check it out.