Brian Cutter has written a blog post presenting the case for psychophysical harmony.
I’ve already raised several objections to psychophysical harmony, which you can see here:
Part 1 I argue that to the extent that psychophysical harmony poses a threat to nonreductive physicalist accounts which grant that there are phenomenal states, this elevates the appeal of accounts which deny that there are phenomenal states (such as illusionism).
Part 2 I argue that the primary form of harmony that the argument presented in Cutter and Crummett’s (forthcoming) paper, normative harmony, relies on questionable assumptions about normativity. A normative antirealist can readily reject normative harmony.
Part 3 I point out that all of the premises in an argument must be propositions. Any argument or psychophysical harmony requires at least one premise that draws on unintelligible concepts like phenomenal states, which would render the premise in question nonpropositional. As such, there can be no argument for psychophysical harmony.
Cutter’s blog post offers a concise account of what the supposed puzzle of psychophysical harmony is:
Now for the puzzle of psychophysical harmony. Psychophysical harmony consists in the fact that experiences are correlated with physical states and with one another in strikingly fortunate ways.
This strikingly fortunate correlation is only striking or fortunate if there is a genuine correlation between two sets of seemingly disconnected phenomena. Experiences have been reified on Cutter’s view, into something metaphysically distinct, autonomous from, and irreducible to a description of the functional states, physical mechanisms, and so on that produce or characterize them. One solution to the problem of psychophysical harmony is to simply deny there is anything to harmonize: if there are no phenomenal states, then whatever experiences actually are, they don’t call for any distinct phenomena to harmonize. In their paper, Cutter and Crummett dismiss this possibility in a footnote.
Here I want to focus on Cutter’s push for normative harmony:
[...] which occurs when the descriptive role of experience is harmoniously aligned with its normative role.
Cutter offers the valence associated with pain and pleasures as a paradigmatic example. There are the descriptive features associated with pain states: avoidance, withdrawal, and so on, but, critically:
Pain also has a normative role: it is non-instrumentally bad; it is an experience we have reason to avoid and to eliminate/reduce when we have it.
Its descriptive and normative roles are, according to Cutter, “nicely aligned.”
Yet just as there are no phenomenal states, the normative role Cutter believes pain and pleasure play is also a phantom. We are told that:
Generally speaking, pain is associated with the behaviors it gives us reason to perform, the very behaviors it justifies or rationalizes. The same goes for pleasure.
2.0 Normative reasons
Here we face the dreaded hydra that will not die: the notion that facts can “give us reasons.” In this section, I’ll address this concern independent of Cutter’s remarks. So this is going to be one big digression that isn’t directly addressed to Cutter or Cutter’s arguments. As such, the imagined critics I respond to here are not Cutter, but are instead reflective of a general template of reaction I’ve received when pushing back against some of the notions Cutter appeals to in the post. Cutter’s remark that pain “gives us reasons” is simply the impetus for a more general objection to this and similar turns of phrase.
I find the use of “gives us reasons” and related language to be an argot characteristic of contemporary analytic metaethics. How does a fact give a reason? What does “give” mean here? Its use strikes me as metaphorical, or metaphysically obscure. Yet when I’ve asked people to explain what this language means, this isn’t taken as a curious puzzle to itself be further explored, but as an impertinent or stupid question to be dismissed as quickly as possible. I’m typically given one of three styles of response:
(a) it’s meaning is obvious, which may be accompanied with the suggestion or outright insistence that I already know what it means and I am being disingenuous
(b) It’s a primitive concept the meaning of which cannot be explained (in this case, primitive is a misnomer; all that typically means is that it isn’t composed of further concepts, but I am not asking what concepts it’s composed of, I’m asking what it means) (c) The Mutually interdefining merry-go-round. Philosophers conceal the meaningless of their terms and concepts in networks of interconnected terms and phrases that create the illusion of conceptual depth. When I ask what these sorts of reasons are, I’m often told that they are considerations that “count in favor.” But it’s not clear what that means. Requests for further clarification may run you in a loop of similarly obscure and turns of phrase that superficially appear meaningful but never seem to bottom out in any grounded appeal to terms, concepts, or phenomena I can make sense of.
This set of response strategies represent a bulwark against efforts to unmask the meaningless of many of the terms and concepts employed in contemporary analytic philosophy. When challenged, one is routinely confronted with an unusually incurious obstinance or even hostility towards this line of inquiry. Here’s one common response I’ve received:
Are you saying you don’t know what it means for something to count in favor?
Such questions are often presented as a kind of rhetorical gotcha, since it’s supposed to be obvious what the term means. The person presenting this kind of objection, along with their imagined audience, are expected to take their side. And they’re often right: this can win over an audience, and satisfy the critic that you’re being unusually stubborn and possibly disingenuous. However, I think this sort of question is driven by a misunderstanding about what my objection is, and a more general misunderstanding, or at least poor framing, of meaning.
To respond to this imaginary critic: no, I am not saying that I don’t know what “counts in favor” means. I’m saying that I don’t know what you mean when you use the phrase.
There is a difference, and it’s an important one. When people employ terms and phrases, they often imagine that they have “a meaning,” and that anyone involved in the discussion ought to know “the meaning” of the term or phrase. But this isn’t how language works. Words and phrases are flexible, and their meaning is highly sensitive to the contexts in which they’re used. It makes no sense to ask me whether I know what the phase “counts in favor” means. The phrase itself has no meaning at all; rather, it has possible usages. Incidentally, there are an infinite number of possible uses. What matters is the communicative intent of whoever happens to be using the phrase in a given exchange. Now, what someone could insist upon is that the term “counts in favor,” has some kind of typical usage. That is, they could say:
Are you saying you don’t know what ordinary people typically mean to communicate when they use the phrase count in favor?
Once this clarification is made, the question loses all of its rhetorical force. Here’s why. First, suppose I do know how the phrase is typically used by ordinary people. Well, so what? Is that how they’re using it? If so, why not tell me what it means, then, and embarrass me by exposing my ignorance of terms with clear meanings that figure into everyday discourse? Nobody ever does this, though. And, in any case, suppose the answer is “No.” I don’t happen to know how the term is typically used. Why should I be embarrassed by this? Analytic philosophy’s whole focus throughout the 20th century to now has been on the conceptual analysis of ordinary terms and phrases. If the way “knowledge,” has been handled is any indication, it’s not at all obvious what philosophical commitments are implicit in ordinary terms and phrases. One might even say the central task of analytic philosophy is to find out! So no, I don’t know what “counts in favor” means when it's used in everyday discourse. Critically, I also take such questions to be empirical. If we wanted to know how people used the phrase, we’d have to go do empirical research. I don’t know of any experimental philosophy investigating folk notions of “counts in favor” or “reasons,” and, even if there were any, I doubt most analytic philosophers are paying careful attention to such literature. I’d also suspect such literature would quickly confront at least three problems:
(1) Participants frequently didn’t interpret questions about these terms and phrases in the way researchers intended
(2) Participants frequently used these terms and phrases in a variety of ways, indicating that the terms and phrases are polysemous and highly context-dependent, as one would expect if they appreciated the highly context-sensitive nature of language and the importance of pragmatics
(3) These terms and phrases didn’t seem to be used in a way that readily comported with the categories and distinctions consistent with their use in contemporary analytic philosophy
This is all moot, though. I don’t think philosophers are using terms and turns of phrase in their ordinary sense, not least of which because (a) I don’t think there is an ordinary sense and (b) if there is, I doubt it has any philosophical depth to it; such terms and phrases are conceptually shallow, and don’t commit ordinary language users to substantive philosophical positions; rather, the terms and phrases serve primarily to facilitate proximal goals consistent with their current context of use. Terms and phrases don’t themselves mean anything; rather, they reflect intersubjectively shared markers with reliably overlapping patterns of usage in local, largely improvised contexts for facilitating the practical goals of the person using them. The whole notion that people use a term like “counts in favor” in some rigid, fixed way is predicated on a notion of how language works that I reject and that I think has little going for it, philosophically or empirically.
The trick, then, is that philosophers equivocate between the motte of putative ordinary usage and the bailey of rarefied philosophical obscurantism. That is, they will use turns of phrase that look like ordinary language, and are presented rhetorically as if they the person using those terms and phrases were merely invoking everyday “commonsensical” terms and phrases. However, what philosophers are actually doing is extracting words and phrases from everyday discourse (or phrases that closely resemble the sorts of things we might find in ordinary discourse), remove them from their everyday contexts of usage, redeploying them in an altogether new context, and then treating them as if their meaning in their original contexts of usage carries over into the new context.
Yet that isn’t how language and meaning work. Extracting words and phrases from a particular discourse—in this case, ordinary discourse—and redeploying them in a philosophical context strips them of whatever meaning they had in that discourse, for the simple reason that words and phrases only derive their meaning in virtue of their contexts of usage. One can no more take a term like “reason” or a phrase like “counts in favor,” as it appears in ordinary language, redeploy it in a philosophical context, and claim to have preserved its meaning than one can extract rules from chess or soccer from the games themselves and reply to them in new contexts.
One can readily imagine the absurdity of such misbegotten redeployments of meaning outside a context of usage by envisioning the police taking a bishop away in handcuffs because the bishop was caught walking in a straight line rather than walking diagonally. Biologists carefully affixing crowns to the heads of frogs whenever one leaps over another, because they reached the other side of the pond.
These are profoundly silly examples, but it’s also silly for philosophers to take terms and turns of phrase that appear (or allegedly appear) in ordinary language, and, without knowing how they’re being used and the degree to which their usage turns on the circumstances of everyday interactions, use those terms outside that context in a wholly new context, then insist that they retain the same meaning (whatever that is). Then, when asked what these terms mean, insist their meaning “is obvious,” but falter when asked to explain what that obvious meaning is. Extracted from their everyday contexts of usage, these terms and phrases lose their meaning, and are instead inserted into the broader network of philosophical terms and contexts this person is using in a kind of pantomime of the grounded and practical nature of everyday discourse. Philosophers are, in effect, creating a whole new shadow lexicon: a set of terms and phrases that superficially appear meaningful because philosophers have sculpted them into a set of interlocking webs of phrases, connected by networks of arguments and distinctions, but which never touch ground in the real world.
To bring this back to the initial problem, recall that the critic objects to my questions about what they mean: don’t I know what counts in favor means?
This isn’t a meaningful question. No, I don’t know what it means, because words don’t mean anything. As I’ve said, and will say many times again: words don’t mean things, people mean things. It’s not a question of whether I know what counts in favor means. The problem is that I don’t know what they are trying to communicate when they use that phrase.
I can tell them what I would likely mean if I used the phrase. I can describe some plausible stories about what I think nonphilosophers might mean if they used the phrase. I can even conduct empirical studies on how nonphilosophers use the phrase. What I can’t do is read their minds.
If you know what you mean, then why not just tell me? Why am I the bad guy for asking?
This irritating attempt at denigrating me by suggesting I don’t know “what words mean,” displays a misunderstanding of my concern, and reveals the motte and bailey in action. The motte is the supposed ordinary usage of the term. Audiences are supposed to have an impression like this:
The phrase “counts in favor” seems perfectly benign to me. All three of the words are frequently used by nonphilosophers, and the phrase as a whole is familiar enough to me. Of course nonphilosophers use terms like “counts in favor” in everyday language, if not all the time, then often enough that it clearly means something. And, even if I can’t tell you exactly what they mean, it’s absurd and excessively demanding to ask someone to define completely familiar and ordinary terms and phrases. So this Lance guy, in asking people to explain what they mean, is, for whatever reason, either stalling because he doesn’t have any real objections, or pretending he doesn’t know what words mean, or something. I’m going to side with the person who is using this perfectly ordinary language…
Yet I am not questioning whether the use of the words “counts,” “in,” and “favor” are used in meaningful ways in ordinary discourse, nor am I questioning whether “counts in favor” is or could be used meaningfully in such contexts. I’m asking how the philosopher using the phrase is using it in the specific context of their philosophical argument, where I take it the term is endowed with additional philosophical content that I don’t take for granted is a feature of the ordinary use of the phrase or the terms that compose it.
What philosophers do, in other words, is coopt language that does mean something in one context (everyday usage), use it in another context where it’s meaning is no longer clear (philosophical usage), and then insist that if you know what it means in everyday usage then you must also know what it means in philosophical usage, even though it’s never made clear what it means in ordinary usage, and, whatever it does mean, it’s an open question whether that meaning transfers to the new context.
What philosophers are relying on is the impression or feeling that the terms or phrases are meaningful in their everyday usages to forestall objections to their specific usages of them in philosophical contexts without doing any work to establish that that their isomorphic usages actually share the same or similar meaning. This is effectively like saying that because a particular symbol or string of letters means something in one language, that therefore it means the same thing in an entirely different language.
This would be absurd, yet the comparison isn’t that far from what’s actually going on: philosophers have, in effect, crafted their own paralanguage the meanings of which are obscured by the illusion that the way they’re using terms or phrases maps onto some mythical “ordinary usage,” the contents of which is an empirical question, but for which they have virtually no empirical data to support claims about any given usage.
Instead, what they rely on is the philosophical method itself, which purportedly captures that usage, but is, in reality, simply a projection of their own theory-laden conceptions of the meanings of terms and concepts onto ordinary people without doing any of the work to show people actually use the terms or phrases in question in a way consistent with what philosophers are attributing to them. Again, such questions are empirical, yet philosophers routinely either
(a) insist they aren’t empirical, which is absurd. If we’re talking about facts about what people intend to communicate, such questions are manifestly empirical. If we’re not talking about that, then it’s unclear what we are talking about but it’s likely to be questionable philosophical assumptions about how language and meaning work or
(b) claim they are empirical but that philosophers are justified in appealing to experience, intuition, and reflection to make generalizations about what ordinary people mean. They aren’t justified in making such claims (as I argue here) and, even if they were, such justification would itself require empirical data which they don’t have and which, when it has been obtained, tends to illustrate that they’re mistaken.
Similar objections apply to terms like justifies and rationalizes. However ordinary people use these words, philosophers use them in specific and obscure ways. Imagine if philosophers didn’t use any ordinary terms or concepts for the technical elements of their theories, but instead invented a whole new set of terms and phrases that involved stipulating what they did mean. What would the content of these stipulations consist in? I have no idea, but I suspect a lot of “intuitions” derive their force from subtle equivocation between the impression of meaningfulness of some term or phrase that has some analog in ordinary contexts that is then reused in a philosophical context. This impression allows philosophers to avoid having to explain what they mean in those philosophical contexts.
When philosophers attempt to criticize the quietist for allegedly not knowing what some term means, the quietist should respond that they are not claiming to not know what the term means; rather, they are claiming not to know what that philosopher means when they use the term. And it is entirely reasonable to ask someone what they mean. If they take themselves to be using the term in some conventional or typical way, that’s fine: then they are welcome to say what that conventional or typical usage is. If they aren’t able to tell us, it’s hardly our fault for asking, and it suggests that the reason they can’t tell us because they themselves don’t know the answer, in which case they are hardly in a good position to criticize us for not knowing what they mean: apparently they don’t know what they mean!
Returning to Cutter’s remarks, recall that, according to Cutter, Pain has a “normative role,” namely, that it is “non-instrumentally bad.” What does Cutter mean by “bad”? This may seem pedantic, but it isn’t. If by “bad” we mean something consistent with antirealism, then it’s unclear whether there is any problem of normative harmony. Likewise for Cutter’s use of “reason,” “justifies,” and “rationalizes.” I reject all realist usages of these terms, yet I suspect Cutter’s account relies on those assumptions. This is especially apparent given that Cutter regards normative error theory as one viable response (I’ll skip over the response that precedes this, the contingent normative roles response, though check the blog post if you’re interested).
3.0 Cutter’s response to the “Normative error theory response”
Cutter claims that one response to normative harmony is to endorse normative error theory. This entails not so much an account of how normative facts correlate with physical facts, but a denial that there are normative facts of the relevant kind.
It’s a bit strange to limit the objection in question to error theory in particular, rather than antirealism. Error theory is typically involves a semantic account of the meaning of moral claims, and consists of something like the following claims:
(1) Normative claims purport to describe stance-independent facts
(2) There are no stance-independent facts of the relevant kind
(3) So all normative claims are false
Yet someone could reject normative harmony because they don’t think there are normative facts without endorsing (1) and (3). That is, moral antirealism only commits you to (2), and not (1) or (3). Why does Cutter present error theory as a possible response, rather than antirealism without the commitments of (1) or (3)? I suspect it’s because the objections Cutter raises turn on (1) and (3) and not on (2), even though (2) is the most important of the claims, and would, if true, constitute a rejection of realism without having to appeal to contestable claims about the meaning of folk terms.
Normative antirealism would therefore probably be a more appropriate critical response to address, rather than error theory. Error theory represents only one of many possible antirealist responses to Cutter’s claims. Why restrict one’s rejection of antirealist positions to one particular account?
I reject error theory because I don’t think it’s the case that ordinary moral claims uniformly and determinately refer to stance-independent moral facts. So I would reject (1). This in no way prohibits me from endorsing (2). Facts about what reality is like don’t turn on contingent facts about how people use words, and, if one were to insist that this isn’t what (1) entails, but instead reflects some analytic fact about the meaning of moral claims, fine: I deny that, too. Antirealism does not require endorsing (1). So it’s weird that critics of antirealism often focus on specific accounts of antirealism that hinge on specific semantic theses; such criticisms can give the impression of refuting antirealism by refuting the semantic theses, when all they’re doing is refuting the semantic analyses; what they are not doing (at least much of the time) is refuting the outright denial that there are stance-independent moral facts.
This may seem like hair splitting, but it isn’t. I’m a normative antirealist, but I am not a normative error theorist. I see a substantial difference between my position and error theorists, who I think needlessly mire themselves in various philosophical (and in some cases, rhetorical) problems by insisting on (1) when this is unnecessary and probably false and in practice tend to concede far too much to moral realists, including e.g., the intelligibility of non-naturalist realism, normative reasons, and other notions, something I don’t grant.
In any case, the normative error theorist (and the antirealist) claims that:
[...] there is no correspondence between the normative role of an experience and its descriptive role because there are no normative facts at all.
This is where I think Cutter makes a serious misstep. Cutter proceeds to claim that error theory is committed to the view that:
Pain is not bad, and does not give one reason to avoid and eliminate it.
This is why error theorists get themselves in trouble. The error theorist may, in fact, be committed to using the term “bad” in the sentence above in the same way as the realist. And since the error theorist thinks:
(1) Nothing is bad in this respect
(2) This is the correct account of what “bad” means
(3) That therefore nothing is bad at all
This makes error theory look silly, even though it isn’t. It makes it look like the error theorist is saying they don’t oppose pain, or want to avoid it, or think there should be less of it, or not want to inflict it needlessly on others, and so on.
That is, to say that pain isn’t “bad” may, for the error theorist, only entail that it doesn’t have some kind of stance-independent badness. Yet it does not follow from error theory that the error theorist doesn’t dislike pain, want there to be less of it, and so on. The error theorist could, in fact, have exactly the same normative standards and psychological attitudes, emotional states, and so on towards pain as a moral realist. And yet because they endorse the goofy notion that there is some kind of analytic fact about the word “bad” that fixes its usage, and obliges them to use the word in that way, they are forever barred from calling pain “bad.”
The error theorist has effectively preemptively rendered themselves vulnerable to normative entanglement, and normative entanglement is exactly the angle Cutter takes in objecting to error theory. Once again, objections to antirealist accounts hinge on conflating metaethical considerations with normative and psychological implications (in the latter case about the attitude and character of the antirealist).
Cutter notes that since error theorists deny that “Pain is bad,” that:
The main problem with this response is that it is extremely implausible. It seems about as self-evident as anything in philosophy that excruciating pain is bad, or that at least some pleasures are good.
Only, it’s not. The problem with error theory isn’t that it denies pain is bad: this is ambiguous, and on many construals of what this entails or implies, it’s not even true. What the error theorist denies is that pain is stance-independently bad (and to simultaneously affirm that moral claims are committed to stance-independence). They are not obligated to deny anything else other than this. Of course, this generalizes to more than just pain. The error theorist denies there are things like badness, goodness, rightness, and wrongness, where these are understood to possess distinctive metaphysical properties that (for whatever reason) the error theorist finds objectionable. It has absolutely nothing to do with their attitude towards pain, how much they care about their own pain or the pain of others, and so on.
I suspect the notion that error theory is implausible insofar as it denies “pain is bad” derives almost all of its force from the implication that this is what an error theorist who denies “pain is bad” is committed to.
If the error theorist wants to concede linguistic territory to the realist, then perhaps they can’t or won’t say that pain is bad in some way consistent with antirealism. Even if they did, they could be accused of revisionism, or changing the subject, or using the word “bad” in some meaningless way. And rightly so: they would be doing so if they’re correct that the “correct” account of “bad” just is the realist account.
The problem with error theory is thus not that it denies that pain is stance-independently bad, but that it agrees with realists that this is the only meaningful sense in which moral claims could be bad in a moral sense.
Once we deny that, too, Cutter’s argument no longer has any force. And I do deny that. An antirealist like me can maintain that:
(1) Pain is bad
(2) It’s just not stance-independently bad
The problem with error theory isn’t its denial of stance-independent badness, but its semantic commitments. Drop these, and I am curious to hear what Cutter’s objection to my form of antirealism would be. It isn’t the case according to my view that pain isn’t bad (though I do think that pain doesn’t “give one reason to avoid and eliminate it” because I suspect this terminology appeals to notions of normative reasons that I don’t accept; again, I do have my own conception of “reasons” where they are understood in terms of consistency relations between goals and ways of acting consistently with those goals, but this approach involves a rejection of irreducibly normative facts; unfortunately, without further clarification I don’t know if Cutter’s account appeals to irreducible normativity).
4.0 It’s always language, isn’t it?
The problem here is, as always, one of language. Error theory mires itself in a rhetorical trap. The error theorist looks like they’re saying pain isn’t “bad,” but what practical implications follow from this?
Absolutely none whatsoever.
The error theorist can think, act, and do pretty much anything the realist can do except use certain terms or phrases. Imagine, for instance, we agree that the word “carrot” requires a commitment to the view that carrots are magical sentient beings from an alternate dimension.
Suppose you don’t think there are any such things. This would make you a “carrot error theorist.” You don’t think there are any small orange root vegetables that are also magical sentient beings from an alternate dimension. Since you’ve agreed that “carrot” refers to such beings, what does this cost you? It would commit you to the following view:
There are no carrots.
That seems absurd, right? Because there clearly are carrots. I just ate some carrots for dinner (I didn’t make that up for illustrative purposes. I really did have carrots for dinner). Also, here is a picture of some carrots:
I don’t know about you, but I’m very confident the sorts of things in this picture exist, whatever we call them.
Now suppose due to some ancient superstition or whatever, the vast majority of people are convinced carrots are magical sentient beings from an alternate dimension and, when confronted with evidence that they aren’t, they refuse to budge on the matter: carrots just are magical sentient beings from an alternate dimension. In this case, this property is a non-negotiable feature of carrots. If carrots aren’t magical sentient beings from an alternate dimension then there are no carrots.
If you’re like me, you probably don’t think those things people call carrots are magical sentient beings from an alternate dimension. But if we discovered this is how most people used the term, what would the implication be? If we’re committed to using “carrot” in its ordinary sense, we are committed to the view that:
There are no carrots.
Again, absurd, right?
No. Not at all. All this amounts to is denying something like:
There are no orange root vegetables that are also magical sentient beings from an alternate dimension
What it does not require us to do is think that these things:
…don’t exist. We can still insist that the orange root vegetables depicted in these photos are real. We’d simply have to call them something other than ‘carrot’.
That’s it. That’s all we have to do. All “carrot error theory” amounts to is recognizing that people tend to use a particular word to refer to something that doesn’t exist. It doesn’t require us to deny the existence of orange root vegetables like this:
The same thing is going on with moral error theory, only in a way that is less obvious, and allows moral realists to give people the false impression that the error theorist is committed to more than what they’re committed to.
The moral error theorist believes that normative moral claims attempt to describe a stance-independent moral reality. There is no such reality, so all such claims are false. This technically entails that sentences like:
It is morally bad to cause unnecessary pain.
…are false. Yet think about what this amounts to, from the error theorist’s point of view. It amounts to something like:
It is not the case that there is a stance-independent normative fact according to which pain it is morally bad to cause pain.
This might still sound “implausible” or “obviously false” or whatever, but this is, I suspect, because of mistaken assumptions about the philosophicalimplications or consequences of holding this view. Again, take the statement:
Pain is bad.
In claiming that this statement is false, one may have the impression that the error theorist is:
(1) Less concerned about suffering than a moral realist who would affirm the statement
(2) Less reluctant to cause people pain than a moral realist who would affirm the statement
(3) Suffering a disturbing lack of concern about their own or other people’s welfare
(4) Claiming that they or others “have no reason” to avoid causing pain to others
(5) Claiming that they or others “have no reason” to avoid pain themselves
And so on. This person is interpreted as saying something so manifestly ludicrous it isn’t even worth taking seriously. Here’s the problem:
None of these claims, nor similar claims, follow from error theory.
Error theory has few if any practical implications at all, and certainly not the kind that are so obvious and so overwhelming as to make the position obviously absurd or self-evidently false. Without exception, these associations tend to be pragmatically implied by denying that “pain is bad,” or rely on distinctive, narrow, technical meanings of terms or phrases that the error theorist may reject, but which give the false impression to audiences that the error theorist is rejecting far more than what they’re rejecting. For instance, in saying that the error theorist denies that we “have a reason” to avoid pain, all this amounts to is denying that we “have a reason” given a distinctively realist account of what it means to “have a reason.” What it does not mean is that the error theorist thinks it isn’t in our interests to avoid pain, or that they’d prefer to avoid pain, or that people benefit with respect to their own goals and interests (most of the time) to avoid pain, and so on. Again, the implication in claiming the error theorist thinks:
There is no reason to avoid pain
…is that the error theorist is committed to something wildly ludicrous and implausible. But this is a false implication that is not entailed by error theory.
This “meaning overflow” wouldn’t be a problem if these additional (mistaken) inferences about what error theory entailed or implied played little or no role in why people thought error theory seemed implausible. However, I suspect these mistaken inferences are doing most or even all of the work in causing people to regard error theory as counterintuitive or implausible.
In other words, I suspect that the common reaction to error theory: namely, that it is “counterintuitive” or has “radically implausible” implications, and so on, are based largely (and in some cases entirely) on misunderstandings about its implications.
For some reason, this entanglement, or conflation by way of pragmatic implication, seems to have escaped the notice of realists and antirealists alike, so I’ll try another way to clarify just what I take the problem to be.
When people are told that error theorists deny that “pain is bad,” this is interpreted not merely to be a sophisticated philosophical claim about the semantics of ordinary moral discourse and the metaphysical status of moral claims (which is all error theory amounts to), but is mistakenly interpreted to entail or imply a variety of other claims or commitments the denial of which would be implausible or ridiculous or reflect poorly on the character of whoever endorsed them.
To put this even more concisely:
Objections to error theory tend to result from confusions and misunderstandings about language.
For comparison, imagine the following exchange:
Alex: I can’t believe that murderer killed four people.
Sam: I don’t think what they did was so bad.
Most of us would recoil at what Sam said. Why? Because we don’t take Sam to mean something like:
Sam: I don’t think that there are stance-independent moral facts according to which that particular set of actions was wrong independent of the goals, standards, or values of individuals or groups.
This would be a weird response, and would be completely unlike the sorts of things an ordinary person is likely to say or to mean in ordinary contexts. It also violates norms of relevance. Typically, when someone makes a claim, one expects a cooperative interlocutor to respond in a way consistent with the conversational aims of the speaker. In the exchange above, Alex appears to be expressing incredulity at the thought of a murderer killing four people.
It’s plausible, even without context, that Alex is probably horrified and opposed to these actions, and is (in part) expressing this, alongside expressing propositional claims about what took place. In other words, in most contexts in which an exchange like this would occur, there is a strong expressive element to Alex’s remarks. A cooperative conversation partner would be expected to pick up on this not-so-subtle facet of Alex’s remarks, and respond accordingly. It’s downright weird to respond to that sort of remark by pivoting to an obscure statement about the metaphysics of moral claims.
Instead, most of us would take Sam to be stating or implying:
(a) A normative claim about the moral status of the murders. Namely, that Sam isn’t personally opposed to them on normative moral grounds
(b) Certain attitudes towards the murders. Namely, Sam isn’t personally as upset, incredulous, horrified, outraged, and so on as others plausibly would be.
Even if Sam also happened to be an error theorist, and thought:
(c) There are no stance-independent moral facts. So technically what this person did was not stance-independently wrong
This is not the sort of thing most people would assume about what they meant, nor is it the thing that people would care about. People would likely react to Sam’s remark with shock and outrage, not because Sam thinks (c), but because infer Sam is (or at least may be) expressing (a) and (b). In other words, the normative and attitudinal implications are doing most (I suspect virtually all) of the work in making Sam’s remarks objectionable.
Just the same, when the error theorist is described as committed to the notion that:
It’s false that pain is bad
Why is this taken to be obviously silly and absurd? Is it because the error theorist denies that there are stance-independent moral facts? Or is it because such a remark implies absurd and repugnant analogs to (a) and (b), namely:
(a) A normative claim about the moral status of the pain whereby they do not personally find anything objectionable about causing or experiencing pain
(b) An attitude toward pain. Plausibly one of indifference or lack of concern
These are the sorts of things most of us (myself included) would find repugnant and objectionable. And yet neither of these have anything to do with error theory, and aren’t entailed by it.
The error theorist is, just like someone who denies carrots are magical sentient beings from an alternate dimension, merely denying that a certain word, as it is used by ordinary people, is used in a way that correctly describes some feature of what the world is like.
That’s it. It’s a thesis about how other people use a word, and whether when they’re using it that word accurately describes something in the world. It has nothing at all to do with their attitude towards pain. It doesn’t mean they’re indifferent to pain. It doesn’t even mean they are one iota less concerned with pain than moral realists are. In short, even if a moral error theorist confidently insists that “pain is bad” this has zero implications for how the error theorist treats people, what their personal convictions and attitudes are, how they feel about pain and suffering, and so on.
If we take an error theorist that denies “pain is bad” and a moral realist who insists pain is stance-independently bad, we have no reason to suppose that a sincere commitment to these positions would entail that the former would behave in a meaningfully different way if they observed other people suffering (short of empirical evidence of some causal or correlational relationship between endorsing error theory and behaving this way; the point is that error theory doesn’t logically entail such attitudes or behavior).
I am confident that almost all of the force behind insisting that error theory is wildly implausible, or has absurd implications, or seems obviously false, and the converse: the insistence that moral realism is “self-evident,” or “obvious,” or “intuitive,” derives from the conflation between the false implication of unpalatable normative and attitudinal implications of antirealist positions—and the false implication that realism offers some kind of affirmation of or endorsement of the normative and attitudinal implications antirealist positions ex hypothesi deny—not because of the error theorist’s stance on semantics of folk terms or the metaphysical status of moral claims.
Here’s one attempt at an intuition pump that lends support to this proposal. If you’re already a moral antirealist, then this may be irrelevant. But suppose you are a moral realist. Presumably, you think that pain is stance-independently bad.
Now suppose you were convinced moral realism was false, but that most people speak or think in ways that commit them to the notion that moral realism is true.
Would you stop caring about pain? Would you stop avoiding your own pain? Would you become indifferent to the suffering of others? Would you become a nihilistic monster with no regard for human life? A violent and manipulative person with no guilt or remorse?
I doubt it. I suspect you’d feel and behave pretty much the same way you already do, minus the extra metaphysical baggage. Moral realism lends nothing to the badness of pain, or the goodness of joy. It is utterly superfluous. Perhaps we’re mistaken about the metaphysical status of claims about stance-independent moral facts. Yet a firm commitment to such a belief has little, if any, genuinely unpalatable implications.
A second, related point is that if embracing antirealism did cause some kind of normative and attitudinal erosion, we should expect antirealists to behave much worse than moral realists. There is little compelling evidence that they do. One response to this may be that despite claiming to be moral antirealists, people who make such claims actually think and act “like moral realists.” Absent an argument, that moral realists would act better is precisely what I am denying, so this claim gets us nowhere, as any presumption that moral realists would behave better than antirealists in virtue of endorsing moral realism is the very thing I’m contesting and that would itself need to be demonstrated empirically. In other words, if what I’m contesting is that those who claim to be moral realists will behave better than those who claim to be moral antirealists, one cannot point to the fact that a person claiming to be an antirealist isn’t really committed to or acting consistently with antirealism because they’re behaving as well as a realist, one would simply be begging the question about whether realism leads to better behavior. This is an empirical matter, and it cannot be resolved by ex post facto.
In short, I think Cutter’s objections to error theory on the grounds that it’s “extremely implausible” would, insofar as it has the power to convince anyone, rely almost entirely on normative entanglement: the inappropriate conflation between metaethical considerations and normative (and psychological) implications. I’m not sure how people would evaluate error theory or other antirealists accounts if people could evaluate error theory in the absence of normative entanglement; perhaps they’d still object to it.
Instead, I suspect what we’d find is that people would see little reason to object to error theory, regarding it as little more than an academic dispute about abstruse metaphysical matters that have little or no practical implications, or even interesting implications. People would drift on to other discussions, and the moral realist/antirealist dispute would largely fade from public exchanges because neither side could gain much rhetorical leverage over the other.
Even so, I think antirealists should abandon error theory. Error theory unnecessarily ossifies the meaning of terms like “good” and “bad,” ceding their use (in moral contexts) exclusively to moral realists. This leaves the error theorist in a tricky position: so long as they want to avoid accusations of equivocation or revisionism, they have to avoid agreeing that anything can be morally good or bad, right or wrong. They could endorse fictionalism, but they could be accused of being disingenuous. They could come up with their own alternative vocabulary, but this would still leave them vulnerable to the direct accusation that they deny “pain is bad,” and might make their position look bizarre and unappealing.
Error theory exposes itself to a kind of insurmountable rhetorical disadvantage. At the same time, it’s unclear why error theorists would be so committed to the view that moral claims refer to stance-independent moral facts. Is this an empirical claim? If so, there’s little evidence it’s true (se e.g., Davis, 2021; Pölzler & Wright, 2020; Pölzler, Zijlstra, & Dijkstra, 2022). Is it an analytic claim? I haven’t heard any convincing case that it’s true, and don’t think there is one. And, at any rate, the error theorist could (and probably would) still deny that there are stance-independent moral facts regardless of how ordinary people used moral language. If we discovered most people used moral language like a relativist or an expressivist, the error theorist is probably not going to start thinking there are stance-independent moral facts after all. If anything, this would lower their confidence further, since they already rejected moral realism. If they also thought that moral realism didn’t even have a presumption in its favor as a feature of commonsense moral thinking, this would be yet another mark against it. So the error theorist would probably continue to deny that there are stance-independent moral facts regardless of how ordinary people used moral claims.
I would be curious to hear Cutter’s critique of a straightforward assertion that there are no stance-independent facts. Such a position would not be vulnerable to the charge that it denies “pain is bad,” since it doesn’t entail any position on this claim.
5.0 A partial solution?
Cutter also says:
A second problem is that, while it sidesteps the puzzle of normative harmony, it does nothing to address the other main form of psychophysical harmony we discuss in the paper (“semantic harmony”— see below). It is therefore a partial solution at best.
This remark is ambiguous. There is a perfectly innocuous reading: someone who wanted to address the puzzle allegedly posed by psychophysical harmony wouldn’t be able to fully address the puzzle by appealing to error theory, because error theory could, at best, only address one form of harmony (normative harmony) but not another (semantic harmony). This is true, but this isn’t a problem with error theory.
The second reading would be to interpret this remark as an indication that there is a deficiency or problem with error theory itself. That it’s a fault of error theory that offers only a partial solution.
We should reject the second interpretation. Note that Cutter only states there is “A second problem” but not that it is a problem with error theory. It would be, at best, a problem for anyone who wanted to resist psychophysical harmony. The problem would therefore be with the consistency and adequacy of a hypothetical interlocutor’s overall position, not with error theory itself. I think Cutter could have made this more explicit.
It’s also worth noting that error theory need not be devised or appealed to specifically for the purposes of rejecting psychophysical harmony. I already denied that there are stance-independent normative facts (moral or otherwise) before encountering the argument. So while I do think antirealism offers a solution, that solution isn’t ad hoc. Again, Cutter doesn’t suggest otherwise, but it’s important to emphasize that we need not endorse positions because they allow us to resist views we don’t want to endorse. We can endorse those views on independent grounds. In this case, that my prior endorsement of antirealism circumvents normative harmony is incidental and to its credit.
6.0 Conclusion
Cutter comes back around to the fact that dualism isn’t necessary for their argument to go through at the end of the post, in the section: “Dropping the Dualist Assumption Doesn’t Solve the Problem.” According to Cutter:
The argument goes through even if we assign a substantial prior probability to physicalist views of consciousness, as long as we grant that there is at least an epistemic gap between the physical truths and the phenomenal truths, such that alternative psychophysical correlation patterns are a priori coherent epistemic possibilities (whether or not they’re metaphysical possibilities). Most physicalists nowadays accept such an epistemic gap.
I’m willing to grant this for the sake of argument. However, while we could take this to be a point in favor of the argument, we could also take this to be a point against the explanatory gap. That is, in fact, just how I’ve reacted to their argument: it provides yet another reason to think that popular physicalist accounts shouldn’t break towards the rejection of physicalism, but should dig their heels in further.
An argument that relies on certain background assumptions and positions is only as strong as those background assumptions and positions. Cutter and Crummett’s case for psychophysical harmony rests on the notion that there are phenomenal states, and elements of the argument appeal to stance-independent normative facts or at least normative reason-giving facts. If there isn’t a good case for these positions, then there isn’t going to be a case for much or all of psychophysical harmony. While many philosophers would want to grant as many assumptions as possible, and most probably are sympathetic to both realism and the existence of phenomenal states, I’m not, and don’t think anyone else should be. As such, my objections aren’t so much downstream of the foundational assumptions behind psychophysical harmony, but to those assumptions. As far as I’m concerned, attempts to make a case for psychophysical harmony never get off the ground.
References
Cutter, B. & Crummett, D. (forthcoming). Psychophysical harmony: A new argument for theism. Oxford Studies in Philosophy of Religion.
Davis, T. (2021). Beyond objectivism: New methods for studying metaethical intuitions. Philosophical Psychology, 34(1), 125-153.
Pölzler, T., & Wright, J. C. (2020). Anti-realist pluralism: A new approach to folk metaethics. Review of Philosophy and Psychology, 11, 53-82.
Pölzler, T., Zijlstra, L., & Dijkstra, J. (2022). Moral progress, knowledge and error: Do people believe in moral objectivity?. Philosophical Psychology, 1-37.