Stance-independence and the Frame Game

Twitter Tuesday #31

1.0 Introduction

Other people have noticed that some of the terms and concepts non-naturalist moral realists use may not be meaningful, e.g., terms like “external reason” or the concept of an irreducibly normative fact.

I try to avoid the mistake of saying that the notion of a stance-independent moral fact is unintelligible. The main reason for this is that there is a division in the academic literature between naturalist and non-naturalist accounts of moral realism. While I find non-naturalist conceptions of stance-independent moral facts to have questionable intelligibility, it is trivially easy for naturalists to produce intelligible accounts. So long as moral facts are identified with certain “natural” facts, and those facts are descriptive, then the naturalist can easily secure a set of intelligible claims about the “stance-independent moral facts.” 

The naturalist can easily explain what a stance-independent moral fact is in a way that is coherent. A naturalist could hold that moral facts are facts about what increases wellbeing, then point to an action or policy that would increase wellbeing, and say that such an action is “morally permissible” or “morally good.” This would be true, since the action in question increases wellbeing. If that seems trivial to you, you’re in good company.

My problem with naturalist accounts is nothing new or unconventional. I have roughly the same reaction that many non-naturalists and antirealists have: that such accounts typically sacrifice much of what I took morality to be about, much of what made it nontrivial, in the process of naturalization. The SEP describes this sort of objection here.

Over on twitter, Ex-Muslim asks:

Is Atheistic Moral Realism Incoherent? 

Can people to [sic] explain to me what “stance-independent reasons” are and how they could exist in the natural world? 

So far I have yet to hear a coherent explanation of that concept 

A few comments. First, I’m not sure why the emphasis is on atheistic moral realism. I don’t think theism helps realists make more sense of the position. Second, secular moral realists will often be non-naturalists. As you can see from the 2020 PhilPapers survey, non-naturalism is nearly as popular as naturalism:

2.1 Metcalf’s response

There were several notable responses to this comment. Thomas Metcalf responds:

"There is lots of evidence for the Theory of Evolution." Don't we know that that's true, without having to specify individual audience members' desires?

This is not a good question for demonstrating that the notion of stance-independent moral (or normative) facts make sense. There’s a lot of ambiguity that would need to be unpacked, mostly centering on the notion of “evidence.” Depending on how this ambiguity is resolved, a “yes” or a “no” could both be viable responses available to an antirealist that thinks stance-independent normative concepts aren’t meaningful. First, one can understand “evidence” in descriptive terms. If so, one can affirm a “yes” to this question but hold that there being lots of evidence for the theory of evolution in no way suggests that there are stance-independent normative (in this case, presumably epistemic) facts. Another approach would be a pragmatist or similar approach, where our values are inextricably bound up in our epistemic evaluations. While I’m perfectly happy to say there’s an external world and that there are descriptive facts about that world, truth cannot be separated from value, and in a certain sense, one couldn’t on such a view talk of “lots of evidence” in some quasi-normative sense without reference the real or hypothetical perspective of some agent or epistemic evaluative standard. 

Note a recurring theme: one cannot categorically state whether stance-independent normative facts are “unintelligible” or not with respect to the sorts of views I (and some others) have about what is or isn’t a meaningful notion, because the term “stance-independent moral fact” is used to refer to very distinct notions (naturalist and non-naturalist notions), and is thus, without specification, ambiguous, and ambiguous in a way that matters a great deal to one’s assessment of the intelligibility of the concept. A similar ambiguity lurks behind Metcalf’s question. Ambiguity is everywhere, and language (coupled with the poor methods of analytic philosophy, insofar as it fails to adequately grapple with language) is the root of the problem.

For a somewhat related take on how to address these sorts of concerns, see Cowie who proposes construing epistemic reasons in terms of evidential support relations.

2.2 Ex-Muslim’s response to Metcalf

Ex-Muslim responds to Metcalf by stating:

Evolution is a true scientific theory. 

However, if a person wants to make the extra claim that you “ought to believe in evolution” then that is making a jump to the normative domain. 

In that case, if you “ought” to believe that depends on your desire for having true beliefs.

I agree, and this is a good response. Metcalf replies:

But if I don't desire to have true beliefs, then I can truly say "there is no evidence for evolution." Right?

I find this specific way of phrasing a question unfortunate. Why make a statement then ask “right?”? Why not just ask the question in a more open-ended way, rather than a leading way (e.g. “Suppose I have no desire to have true beliefs. If so, do you think it would be true or false that I had no evidence for evolution?”)?

Setting aside my concerns with the phrasing of the question, I think the best response at this point would be to try to get clear on what’s meant by “evidence.” One can agree that there is evidence in a descriptive sense, but not that such evidence “gives” us “epistemic reasons” to believe in the theory of evolution independent of our desires. 

Ex-Muslim does eventually get to this point, asking: “How are you defining evidence?” I think that’s where I would have started. The discussion becomes much more involved back-and-forth at this point, so I won’t post it all here. Much of the dispute centers on their assessment of modus ponens:

P1: If P, then Q.

P2. P.

C: Therefore, Q.

Metcalf asks whether, even if one believes P1 and P2, they have “no reason to believe” the conclusion, “unless they desire true beliefs.”

This is going to turn on what we mean by a reason. You could think that we can have “reasons” to believe the truth of a valid argument with true premises independent of our desires. You might also think that if we desire to  believe what’s true, and we judge a valid argument to have true premises, that our desire “gives us” a reason to believe the conclusion. But I don’t think we should even agree to this. What does it mean for a desire to “give a reason”? What is a reason? Reasons, even on a stance-dependent account, strike me as inscrutable in the mouths of philosophers, and, I suspect, they strike me this way because they continue to retain vestiges of what makes the non-natural normative realist’s invocation of “reasons” unintelligible.

 

2.3 A digression to discuss the Frame Game

I don’t think there is any such thing as a “normative reason.” Neither a stance-independent nor a stance-dependent “normative reason.” I think we can make sense of reason talk in terms of descriptive facts about the relation between means and ends. This looks a bit like a stance-dependent account, where if you desire something this “gives you a reason” to pursue it, but I don’t think desires literally “give” reasons; this strikes me as an excessive and confused reification of the notion of a reason, as though reasons were something separate, above, and apart from desires and the relational facts associated with those desires. What I don’t think is this:

1. People have desires

2. There are facts about what actions would achieve a person’s desires.

3. There is a relational fact about desires and those actions that would achieve those desires.

4. Something about (1)-(3), whether it be the desire, or the relational fact, or whatever, “gives one a reason” to perform the action in question

On this view, “reasons” emerge from, or exist in addition to (a) our desires (b) facts about what actions would achieve those desires. Talk of “giving” and “having” reasons strikes me as, at best, metaphorical, and at worst, simply confused nonsense. I believe philosophers have walked themselves into a way of speaking about reasons that unnecessarily reifies “reasons,” treating reasons as things-in-themselves, as separate and distinct entities or properties or phenomena or facts, that exist in an eliminable way above and in addition to mundane descriptive facts about goals and means of achieving those goals. My view is that:

1. We have desires

2. There are facts about which actions would be conducive to those desires

3. There are no further facts

Note that (1) and (2) are descriptive facts. My view is that there are only descriptive facts. If we’re to speak of “normative facts,” normative facts are either a type of descriptive fact (about the relation between means and ends) or they are nonsense. There are no irreducibly normative facts.

Many critics of stance-dependent accounts of reasons criticize such accounts on various verbalistic grounds, twisting and contorting ordinary sentences to make the stance-dependent normative reasons realist look stupid or confused. Ironically, this works only by continuing to project either the realist’s own mistaken metaphysical presumptions, or their confusions about language, or both onto the antirealist. Unfortunately, antirealists seem to go along with this, tacitly granting the realist’s presumptions.

This is why I’ve repeatedly said that moral antirealists concede too much to moral realists. Antirealists fall into the same bad habit that many people do when engaging with analytic philosophers: accepting their framing of the dialectic.

I call this the Frame Game. As a friend and colleague told me recently, one of the questions they like to pose to their students is to ask them, if anyone had set up the labels for “moral realism” and “moral antirealism,” who do you think it was: realists, or antirealists? In other words, whose interests does it serve to frame the debate in such terms?

I feel drawn towards a particular answer: realists. I don’t think morality “isn’t real.” Yet even moral antirealists will tacitly accept this framing, such as in Case and Lutz’s recent book-spanning debate titled “Is morality real?” 

Note how unclear such a question is. Is what real? If by “morality,” one means human judgments, practices, institutions, and other features of human psychology that are studied by psychologists, historians, anthropologists, sociologists, legal scholars, and so on, then obviously it exists, and this is a silly question to ask. If one means “stance-independent moral facts,” then one may think these don’t exist, but why would that prompt one to conclude that morality doesn’t exist? This question is needlessly ambiguous and it fosters confusion. It can (and is) leveraged to imply that moral “antirealists” don’t have normative moral standards or don’t care as much about human flourishing or honesty or truth or goodness or whatever else people may invoke to try to make antirealists look stupid and evil. It is a bad question. It is a bad question because it is unclear and its lack of clarity is routinely leveraged to wage a rhetoric war on moral antirealists.

If antirealists think antirealism is correct, and that it doesn’t carry the harmful practical implications attributed to it, they have a responsibility not simply to defend antirealism on philosophical grounds, but to draw attention to and reject the rhetoric directed against the position. This could be partially achieved by renaming the position or drawing attention to biased aspects of this framing and encouraging moral realists to publicly and very explicitly make it clear that the unfortunate labeling of the positions is an entrenched legacy of the history of contemporary metaethics, but that they don’t think antirealists are brooding and sinister abominations lurking in sewers. The problem is that many realists do seem to think or at least imply that we actually are brooding and sinister abominations lurking in sewers. And they may even maintain that, to at least some extent, the labeling and framing of these positions isn’t biased, but an accurate characterization. Rather than combat this, antirealists have simply argued that antirealism is true, rather than targeting the dark and foreboding clouds that casts the position as perfidious and unappealing.

This is something proponents of any philosophical position need to understand. This is a distinctly American way to put it, perhaps, but they’re marketing a product. You can list all the wonderful ingredients in your all-natural philosophical position, or tout its superiority over the the competitor, but if your product is called “Sinister Sid’s Oubliette of Eternal Torment,” people are probably not going to want to check it out, aside from a handful of people with a penchant for whips and latex.

The current “Frame Game” in the realist versus antirealism dispute is one in which realism is depicted as the “commonsense” default position, and antirealism is depicted as some kind of “radical” skeptical view. This is how Bentham’s Bulldog frames moral antirealism:

Well, in this article, I’ll explain why moral anti-realism is so implausible — while one always can accept the anti-realist conclusion, it’s always possible to bite the bullet on crazy conclusions. Yet moral anti-realism, much like anti-realism about the external world, is wildly implausible in what it says about the world.

BB goes on:

We do not live in a bleak world, devoid of meaning and value. Our world is packed with value, positively buzzing with it, at least, if you know where to look, and don’t fall pray to crazy skepticism. 

As if the antirealist thinks we live in a “bleak world, devoid of meaning and value.” This is rhetoric par excellence. While this is a deeply embarrassing way for BB to begin a critique of moral antirealism, it’s almost as embarrassing that antirealists tend not to push back against framings like this. When an antirealist is told they’re “biting the bullet,” they tend to just accept this. Some even enjoy the idea, as though they’re setting dangerous frontiers by staking out a position far outside “commonsense.” Only moral realism is not a commonsense position, we have no good reason to think there’s much (if any) presumption in its favor, all things considered, and there’s absolutely no reason to grant that we’re “biting the bullet” when we endorse moral antirealism. Again, as I’ve argued before, philosophers misuse “biting the bullet” as a rhetorical move. A moral antirealist can both reject moral realism and reject that doing so involves biting any bullets. The notion that one is “biting a bullet” involves tacitly accepting that one is giving ground, compromising, or accepting an unpalatable implication. But I don’t think rejecting moral realism involves giving any ground, compromising on anything, or accepting any unpalatable implications. If anything, I think it would be quite a pill to swallow moral realism, and I see myself in just as good if not better a position to say that moral realism involves “biting the bullet” with his wonky concepts, dubious epistemology, and its often extravagant metaphysics. I think non-naturalist moral realism is less plausible than telepathy. And I don’t think many of us would be inclined to agree that if you deny paranormal powers that you’re “biting the bullet.”

What antirealists should do is reject the framing analytic philosophy has foisted on the field, and they should stop abiding by the rhetorical moves realists make. This is why I emphasize normative entanglement, and appeals to what most philosophers think, and all these other features of what one might think of as the sociology of philosophy. Simply put, marketing matters. No matter how bad other people’s arguments are, and how good yours are, if one’s opponents can successfully depict one side as righteous and sensible, and the other as an edgy teenager’s rebellious phase, people who don’t spend years delving into the dispute are going to look more favorably on the former. 

I worry many philosophers would like very much to convince everyone that antirealism is some kind of weird “fringe” position, and that one should keep this in mind when considering it: whereas the realists dress and act normal, and are upstanding citizens who pay their taxes and grace us with chiclet teeth, we antirealists shamble around in cellars, scribbling treatises by the light of a single flickering incandescent bulb. Realists may deny they do this, while incredulously asking us again and again, after we’ve made it crystal clear that we don’t think there are “objective moral facts,” whether what Hitler did was “objectively wrong,” as if this weren’t, intentional or otherwise, casting us in a negative light via guilt-by-association and the very easy and real tendency for people to misconstrue the antirealist’s response to such questions as moral hesitance or complacency or lack of commitment, as if we’re any less horrified or opposed to genocide than the realist. It has become more than a little tedious to constantly have to disentangle questions or comments that imply antirealism somehow impairs one’s ability to oppose genocide. 

It doesn't, but the constant line of questioning can give people that impression. Realists may believe such questions are a reasonable and benign way to attempt to bring our own realist predilections to the surface, and in that respect such questions may seem warranted. I’m certainly willing to grant that such questions are typically well-intentioned. The problem is that, well-intentioned or not, realists still have a responsibility not to be reckless, and not to present lines of questions that could mislead audiences.

Just imagine the reverse. What if 95% of what antirealists did is go around asking “So if it was an objective moral fact that Hitler was good, would you agree Hitler was good?” or “If you were convinced it was objectively morally good to torture babies, would you torture babies?” Such questions can and do serve important purposes, either to illustrate problems with the relation between stance-independent moral facts and motivation, but they can also be misused and overused or used negligently or followed by the antirealist with a careless and sloppy effort to portray the realist as one syllogism away from slaughter. I’ve employed questions like this on occasion, but I try to remain self-conscious to not overuse them or not drag out their rhetorical elements too hard. 

I confess to a bit of fighting fire with fire, and employ such questions largely to illustrate that one can (if they wish) make a concerted effort to depict realists as evil or potentially evil. It’s less effective, but if antirealists were more popular, and there were enough unscrupulous antirealists, such efforts probably could be successful. There’s probably enough ambient worry that realists are pushy and rigid absolutists that one could probably leverage that, too, to cast realists in an even worse light.

Just imagine enough high profile cases of awkward and stilted responses from unprepared realists to questions like “Suppose you came to believe that it would be morally good to commit genocide. Would you think it was morally good to commit genocide?” This is a bullshit question, because the answer is a trivial “yes”: this is just asking “If you thought X, would think X?” A realist blindsided by this sort of question may stumble in their response and give a careless reply. They might even just say “Yes” only to have the realist say “Aha! Then as we can all see, realists are one belief away from endorsing genocide!” I’ve called this the “Nuremberg” objection to realism: that insofar as one “outsources” their moral agency to stance-independent moral facts, then it remains an open epistemic possibility that anything could be justified, including genocide. Insisting this isn’t possible because the moral facts are “necessary” is no help: the argument can always be framed in terms of epistemic possibility or belief, i.e., in the latter case, if you believed it was an objective moral fact that… Such questions may, on reflection, not have much teeth to them, but then again neither do objections to antirealists that exploit normative entanglement. 

Ironically, realists have taken advantage of this kind of rhetoric for many years, but antirealists typically haven’t countered this with similar rhetoric. The irony I see in this is that realists strike me as having acted worse than antirealists, at least in this respect. Of course, this is all colored by my personal experience, which is undoubtedly biased towards instances of realists employing rhetoric against antirealists, so take it with a grain of salt. I’m not that confident about this, but I’m confident enough that I’d bet my framing if the situation is roughly accurate.

2.4 Returning to Ex-Muslim

Ex-Muslim eventually expresses a view much like mine:

My view is that reason basically reduces down to your desires. 

To say that I have a reason to accept P1 basically means that I have the desire to have true beliefs and hence I accept P1. 

I don’t even know what the word “reason” is supposed to mean outside that context.

Metcalf’s reply is unhelpful:

Okay, I guess it seems irrational to me to accept P1 and P2 without accepting P3. And "irrational" means I have a reason not to do it.

But what does Metcalf mean by a “reason”? Ex-Muslim has provided a reductive account where reasons are understood in terms of desires. It’s fine to say that reasons are something other than desires. Okay, then what are they?

3.0 Ben Watkins’s reply:

Watkins responds by saying:

I don’t know what it means to say it’s incoherent.🤷‍♂️

(1) We have more reason to help those in need than we do to harm them further.

(2) We have more reason to have consistent beliefs than to have inconsistent ones.

Neither (1) nor (2) are problematic for naturalism to my mind.

I’m puzzled by the first remark. I’m happy to take Watkins at his word, but I’d be curious to know whether Watkins thinks that terms or concepts can be incoherent, or meaningless, or unintelligible in other contexts. I don’t know exactly what Ex-Muslim means, but one could mean something like this, in principle:

The concept of a normative reason is, as a matter of analytic fact, a concept about the relation between desires and means of acting in accord with those desires. To say that one has a reason to do something is to say that it would accord with their desires. This means that if someone has a reason to do something, they must, by definition, have a desire to do it.

Is this a good account of “reason”? No. But one could hold such a view. And if they did, they would regard it as nonsense to say that one has a reason to do something, but has no desire to do that thing. By stipulation, to say that they have a reason just is to say that they have a desire to do so. On this view, “reason without desire” would be incoherent in much the same way “he’s a bachelor but he’s married” would be incoherent. I don’t know if Watkins would accept this way of characterizing incoherence.

What’s interesting about the second part of Watkins’s remark is that he states that neither of the remarks he provides about reasons would be “problematic for naturalism” on his view. Since he doesn’t explain what he means by a reason, there’s little way to evaluate this claim.

4.0 Secular Outpost and incoherence

Secular Outpost asks: 

What do you mean by incoherent? Why do you think it is incoherent?

Ex-Muslim responds:

Thank you for asking that. 

I’m asking them to explain it to me so that I can get the concept of what they are referring to. 

Whenever I conceive of a “reason” it is connected to and dependent on a person’s will and desire.

I have a similar sort of reaction. To me, the notion of a “reason” in the sorts of normative contexts realists invoke it just wouldn’t make sense unless it were indexed to one or another real or possible evaluative standpoint. I don’t know if I’d insist on cashing that out in terms of “desires”, but that’s a decent enough approximation that with a bit of retrofitting it I probably endorse something more or less the same. It’s reasonable to ask, if this isn’t what someone thinks about reasons, what they do think. And “it’s not reducible to desires” isn’t very illuminating. That’s fine, as far as it goes: you’ve told me what you don’t think. Now what do you think about reasons?

Note that what we don’t tend to get in these exchanges in any comparable non-desire based reduction, and perhaps that’s because proponents of external or normative reasons of whatever kind don’t think reasons can be reduced in this way. If that’s the case, why not just come right out and say so?

5.0 Finnegans Take

Finnegans Take asks:

Do you also reject the existence of stance-independent epistemic reasons too?

Ex-Muslim affirms that they do, and adds that:

Yes. 

They strike me as very incomprehensible. 

I literally don’t understand how they could exist.

I agree, but I worry about saying anything like “I don’t understand.” This is routinely misinterpreted by people to be the claim that because you, personally, don’t understand something, that therefore it isn’t capable of being understood:

I don’t understand what they mean, therefore it’s unintelligible.

This is an annoying misrepresentation of the claim, so it’s best to avoid language that could be leveraged to foist it on the proponent of the unintelligibility thesis. Finnegans Take doesn’t make this mistake, I’m just flagging this as a general issue with this kind of framing of unintelligibility. Don’t make it personal. If you think something sounds like nonsense, say it sounds like nonsense. I don’t think I’d I don’t personally understand someone if they said “twas brillig and the slithy toves,” rather, I think nobody understands this, because there isn’t anything to be understood. If that’s the position you take towards certain philosophical conceptions of normativity, then it’s best to directly say that.

Finnegans Take responds:

I'm honestly baffled when I hear people day [sic] this - there are very few truths more obvious to me than something like "You should seek out evidence for your beliefs." What is controversial or problematic about that?

What’s controversial or problem about this is, from my perspective, that it’s not clear what “should” means. If you mean that it would be conducive to the goals of someone who is trying to obtain true beliefs, great: we’ve reduced normative language (“should”) to a descriptive fact about the relation between ends (desire to know what’s true) and means (seeking evidence). But if this isn’t what’s meant, then what is meant? I should seek evidence for my beliefs regardless of whether this would serve my interests or be consistent with my goals? Okay, why? Unfortunately, the answer I typically get from non-naturalist realists is something like “because it’s good” or “you just should” or “why makes no sense here. It just is a fact that it’s what you should do, so asking why you should do what you should do is confused.” I don’t think I’m the confused one here. Let’s suppose someone did insist I just “should” comply with this or that epistemic norm. And let’s suppose I don’t.

What are the consequences? I’ve had people tell me something like “Then you’d be doing something irrational.” Okay. And? So what? One of my favorite remarks on this matter comes from David Lewis, who asked:

Why care about objective value or ethical reality? The sanction is that if you do not, your inner states will fail to deserve folk-theoretical names. Not a threat that will strike terror into the hearts of the wicked! But whoever thought that philosophy could replace the hangman? (Lewis, 1996, p. 307)

If the best that the realist can say is that if I reject epistemic realism I’m being “irrational” or if I reject moral realism I’m being “immoral,” well, so what? I don’t want to be irrational-in-the-realist-sense or moral-in-the-realist-sense. I want to achieve my goals. That’s the only thing that motivates me. I suspect it may be the only thing that motivates anyone, realists included. It’s fine and well if they want to comply with these stance-independent facts. Go ahead. I don’t want to. So what are the direct consequences for me if I don’t? As far as I can tell, absolutely none whatsoever. Sure, realists can condemn me and call me names. Sure, someone could impose artificial consequences, e.g., “If you don’t comply with them, I will punish you.” But these are all contingent and extraneous and don’t in any way show that the position itself has any direct, meaningful implications at all. This is why, at bottom, realist accounts don’t simply strike me as meaningless or false, but as practically irrelevant. I simply wouldn’t care whether they were true or not. 

In any case, if the notion is that it strikes someone as obvious that one “should” employ various epistemic practices, independent of their goals, that not only doesn’t seem obviously true to me, it seems obviously false to me. And I’ve seen nothing from realists by way of arguments that can circumvent the simple fact that I don’t share their intuitions on the matter.

The exchange that follows between Ex-Muslim and Finnegans Take illustrates precisely the sort of issue I have. Ex-Muslim says that they interpret the claim that “you should seek out evidence of your beliefs” as a claim that presumes a person has a desire to seek the truth. We may not explicitly add this in everyday situations because we assume people have such goals. Finnegans Take asks why they should think that the statement does presume something about desires. It’s a good question. I don’t know if it presumes it one way or another. We may not need to take stand on what people mean or are trying to mean when they make these sorts of claims. There will be some fact about a person’s communicative goals, but those facts may be exhausted by various social and practical objectives that aren’t determinately cashed out in accord with one or another philosophical theory.

In any case, Ex-Muslim states that while they’ve offered an explanation, Finnegans Take has not. Finnegans Take responds:

I would say I mean exactly what I'm saying - that someone should do X, even if they don't want to. The "should" being used here is exactly the same "should" you would use in relation to someone with a stance-dependent reason. It's not a different kind of reason.

Unfortunately, this does not discharge the mystery around the use of “should” in this context. On the desire-based view, someone could drop the should and say:

It would be consistent with your goals to seek evidence for your beliefs.

Realists don’t mean this. So what do they mean? Is there a way to reconstrue any normative terms like “should” in non-normative language? If so, how? If not, why not? This response doesn’t explain what a stance-independent reason is, which leaves the explanation (if there is one) incomplete.

References

Lewis, D. (1996). Desire as belief II. Mind, 105(418), 303-313.