It seems to me that incommensurability is really quite vague. — Ludwig V

.

Perhaps that's why Davidson was so keen to equate it with nontranslatability; at least this is something you can demonstrate. But he also maintained that there is difference between concept and language, so the question doesn't quite resolve. A very non-Davidsonian way of putting the question might be, "Is it the 'concept' part or the 'scheme' part of a 'conceptual scheme' that's allegedly incommensurable?"

moral certainty is a furphy — Banno

Thanks for introducing me to a new word!

Yes, there's a difference between having the courage of your convictions and being convicted beyond the shadow of a doubt that you're right and They are wrong.

Do you have a suggestion of how to justify a moral 'fact'? — AmadeusD

I was afraid someone would ask me this! The question has occupied me throughout my life, and I don’t know the answer. But since you’ve only asked for a suggestion . . . I suggest that facts about values are to be found in a different “world” than, say, scientific facts. I also suggest that we don’t arrive at moral facts using the standard philosophical questions, such as “What ought I to do?” or even “What is the good?” The world of values is, perhaps, one of spiritual recognition, more like being in love than achieving knowledge.

Gadamer is a philosopher who might also have good suggestions. He emphasizes the importance of tradition in talking about values, since it’s unrealistic to expect every single person to have a transcendental, mystical experience of the Godhead! No more would you expect everyone to prove general relativity for themselves. It’s often appropriate and necessary to take someone else’s word for it. Gadamer uses the metaphor of a well to describe the “world” of values, “which is at one and the same time the soil, source, and water of life, but which is not knowledge in the strict sense.” Invoking the importance of both tradition and non-strict-sense knowledge are of course like poking a hornet’s nest for some philosophers . . . So, a final word from Jean Grondin, who writes a lot about Gadamer: “To recognize that thought has limits is not to silence it, but to allow it to better apprehend itself and to open itself more easily to dialogue.”

my main interest here is that there seem to be true moral statements, and that for some of those it is odd to demand a justification. Talk of brute statements is a bit strong . . . — Banno

That works for me.

I think what you say about brute facts is correct, but it raises an interesting question in this context. Your other two examples -- about the words in the sentence, and the acceleration -- are more or less incontestable. No one, to my knowledge, disputes them (leaving aside the vexing question of interpretation). How, then, do we account for the fact that "It is wrong to harm people" -- supposedly also a brute fact -- has engendered endless debate over the centuries? The debates often focus on circumstantial codicils, such as "Is it wrong to harm people if it will help your family?" etc., but this is part of the reason why it seems odd to invoke "brute fact" here. If we're honest, the standard thoughtful response to "Is it wrong to harm people?" is "Usually, but it depends." I'm not saying that's right, only that it's the standard response, rather than "Yup, it's a brute fact."

I think many moral truths are facts, but I'm skeptical (and nervous) about justifying them with these kinds of analogies. Interested as always to hear your thoughts.

Hmm, I'm not sure I get your meaning. Isn't it perfectly true that, if only a few people decide to make exceptions for themselves and do whatever they please, this will not affect the human species in the way you described, and therefore this wouldn't serve as a motive for not so doing? A "society of murderers" probably can't exist, but one with just a few is quite feasible, as we know, sadly. I'm not seeing how respect for truth comes into it, at this point. But say more . . .

My "have to" is innocuous. I only meant that one would have to agree to some form of the categorical imperative in order to have the kind of motive you described.

But this is the Kantian problem of universalization. I have to first accept that my actions can serve as a "maxim" for others, before the general fate of humankind would matter to me. If I'm simply making an exception for myself, then I can murder as I please, since most other people will not, and the species will be fine.

Fair enough.

I express my intention to lie, — Joshs

Well, but that's just it -- you don't. We stipulate that your listener can't tell the difference. You may have the intention, but it's not expressed. BTW, I agree that we're going to need some appeal to intention as a way of explaining what's going on, but I'm not sure a hardcore OL proponent would.

Coincidentally, there’s an article in the new Phil. of Science by Lorenzo Lorenzetti called “Functionalism, Reductionism, and Levels of Reality” that’s apropos. Lorenzetti talks about theories rather than conceptual schemes, but I think his questions apply.

Suppose A and B are rival theories, but B can be functionally reduced to A. (This would apply to all the examples Wang gives, I think.) So why not invoke the extension/intension distinction and “maintain that the reduced and reducing entities are coextensive and identical but have a different intension”? As he says, this would be a basic Kripkean response to the question of radically different theories. But the problem is, this doesn't explain the asymmetry of the relationship. A and B don’t just offer different descriptions; the claim is that B can be reduced to A (crudely, that A provides a better explanation) but not the reverse. Lorenzetti calls this “the puzzle of identity,” because we seem to want to say two contradictory things: that A and B wind up talking about identical entities, and that they aren’t the same because the reduction relation is asymmetrical.

Lorenzetti has some good ideas about how to resolve this, but my question concerns the astrology example. I’m realizing that, unlike a traditional pair of “incommensurable” scientific theories, astronomy doesn’t actually claim to offer a reductive explanation of everything contained in astrology. Since astrology is talking about human behavior, among other things, a truly alternative explanation would have to go far afield from astronomy and invoke some psychological/biological laws. What, then, do we want to say is the relationship between astrology and astronomy? “Asymmetrical” doesn’t seem to cover it. Any ideas?

Okay, forget the OL response, which isn’t really important. The question remains, ”If I promise you something but have no intention of delivering – that is, I’m lying – this use is indiscernible from a sincere promise. So what makes the difference in meaning?” Unless you’re wanting to say that there is no difference in meaning? So what would be the relevant difference between the truthteller and the liar, on this view? I’m not sure myself; I think appealing to “meaning” makes more sense; but I’m willing to be convinced.

The meaning of a word is its linguistic use. That's what Wittgenstein tries to show in his Philosophical Investigations. How can meaning be anything else? — Michael

Easily. If I promise you something but don’t mean it – that is, I’m lying – this use is indiscernible (in that moment, and assuming a talented liar) from a sincere promise. So what makes the difference in meaning? Indeed, our aggrieved “ordinary language” response to such a situation, if it's revealed, is, “You didn’t mean it!” So what’s going on here?

Thanks for wading through this with me. It does seem obvious that some kind of one-to-one correspondence, a la Tarski, would help bridge the gap between states of affairs and facts. Certainly it would make my question about quantifying over statements pretty much moot. I'll keep thinking.

We all agree to the fact that coffee is delicious, and a great way to start the day. — Banno

Coming in a little late on this, but help me out: I know you can't mean, literally, that everyone agrees that coffee is delicious. So what is the use of "fact" here? Are you saying that, if absolutely everyone really did agree about this, it would be a fact?

I feel, in reading all this, that I'm even more uncertain than when I started in spite of spilling so many words. — Moliere

My sentiments exactly! But I think this is one of the great benefits of doing this kind of intensive and text-based discussion: You understand from the inside out, so to speak, why these questions are so difficult to resolve.

If not an entire garment! Yes, enough of meta-ethics for now.

Just a general comment on where this argument has gone. Clearly, there’s a strong sense on the part of many of us that a statement like “You shouldn’t kick puppies for fun” or “Torturing children is wrong” qualify as moral facts. Moreover, many of us also seem to believe that what makes these statements facts, as opposed to beliefs or preferences or yada yada, is some built-in obviousness, whether resulting from a personal moral intuition or a pragmatic/OL view that we clearly know how to use these sentences and there hasn’t been any problem until the philosophers came strutting in, so . . . no problem.

Either way, on this view, trying to demonstrate a moral fact, or prove its existence, would be bootless. We don’t need arguments, we need . . . what to call it? A certain kind of hermeneutics, I guess, that includes a built-in interpretation. You either see it or you don’t. So if this is right, it’s very confusing why radical skepticism about morality could continue, among brilliant thinkers. What part of this did Nietzsche not understand? Was J. L. Mackie unfamiliar with the linguistic practices of his community?

I’ll stop there, and only add: I don’t believe there are definitive answers to be found on this question within philosophy, but we have to keep asking. May I quote T. Nagel, one of my faves? “[Problems like this] are probably not soluble, but they are irresistible, and the attempt to solve them has yielded over the history of the subject, and continues to yield, brilliant and fascinating philosophical responses and theories, all of which have something wrong with them.”

To say of some normative statement, that it is true, is itself to make a normative statement, isn't it? — Banno

Glad my post was helpful. Yes, I think you’re on the right track, and let me say again that I am not Prof. Logic and can easily get muddled up myself.

The two ranges of possible variables aren’t mutually exclusive in any deep sense. We just have to specify them. I think there used to be an idea of a “universal class” which was supposed to mean “everything,” but that’s no help. We need to know, in any given formula, what the existential quantifier is quantifying over. (I’m going to use capital E rather than the reversed E cos it’s easier.) ‛Ex’ is paraphrased as ‛There exists an x such that . . .” or words to that effect. But absent an agreement on what “existence” is going to cover, we’re left with all sorts of ambiguities and puzzles. Quine said, famously, “To be is to be the value of a bound variable.” This way of stating it shows, I think, that what counts as “being” or “existing,” for logical purposes, is up to us. There’s no prior, obvious, “right” definition of what “exist” means – and that’s certainly been historically true in philosophy.

And yes, valid propositions can be constructed using either or both of the suggested quantifier ranges – you can say true things about both states of affairs and statements.

Now here is where you’re off track a bit: You say, “ ‛T is a normative fact’ is not itself referencing a statement – it is, rather, referencing a fact.” But the viewpoint I’m advocating makes a sharp distinction between “facts” and “states of affairs.” The rain outside isn’t a fact, it’s a state of affairs. The proposition “It is raining”, if true, is a fact. (See Wittgenstein: “The world is the sum of facts, not of things.” I think this is wrong – and so did he, eventually – but he’s using the distinction in the same way.) So, when you reference a statement, you’re also referencing a (potential) fact, but not a state of affairs. And this is exactly where we’ve all been debating. Let’s substitute “true statement” for “fact.” We get ‛T is a normative true statement’, and what we want to know is whether asserting this is equivalent to asserting the truth of T.

If we’re allowed to use statements as bound variables – that is, if Ex can quantify over “facts” and “statements”, not just “objects” or “states of affairs” -- then it looks like we can quote the statement without committing ourselves to its truth, or even to whether it’s true-or-false. But if only states of affairs can count as existing, then we have to make what is (to me) an awkward translation – and the exact way to do this is, again, what we’re kicking around. Should we say ‛(Ex) n(x)’, with ‛n’ meaning normative? But if x can’t be a statement or a fact, how do we translate this? What would it mean for a state of affairs to be normative? What are we predicating normativity of?

Lastly, yes, the “refurbishment” you suggest makes the question even sharper – but of course it firmly commits us to quantifying over statements.

Where I would like some help is in understanding whether Banno's objection, quoted above, must be correct. Are we making a normative statement in the sense that we're talking about truth, or in the sense that we're talking about whatever the normative behavior is? Does it require both truth and normativity to create a normative statement? Have I even made a meaningful distinction? I think so, but . . . see above re my logical competence.

Moliere, I echo Banno's appreciation for your careful reading.

As Banno says, there’s a lot in your post to think about, so just a couple of initial reactions.

The question for me would be whether this still counts as a conceptual relativism, or not? — Moliere

I agree, you’ve managed to (in a good way) blur the distinctions sufficiently that this question is now key. My response is: I’m not sure, and I’m beginning to worry that this could lead to a merely verbal/historical dispute about how we ought to divvy up our terms.

The consideration of poetry translation is very good. I’d love to know what Davidson’s reply would be.

About “phlogiston” and meaning change: Really? This is a rather eccentric use of “meaning,” isn’t it? I’ll grant you that phlogiston now has vastly different connotations and employments than it originally did, but has the meaning actually changed? Or perhaps I’m not understanding you deeply enough.

On the Wang paper, you’ve succeeded in highlighting passages that do seem to further the conversation, for which I’m grateful. The point about the difference between truth-values and truth-value status still eludes me, though – I’d asked into it earlier in the thread, I think. Why would different assignments of “either-true-or-false”, rather than different assignments of “true” and “false”, make any difference to the question of scheme-content dualism? I don’t see why the one is more alien or difficult to translate than the other.

This is pretty similar to my objection to the WMT vs. CMT section. You write that the “difference of meaning isn't one of distributing ‛...is true’ across sentences, but rather is a different kind of difference.” And then you say, “Whether we ought to call this a radical or incommensurable difference I'm still on the fence about.” Exactly – Wang hasn’t convinced me. I’m off the fence, on the side of “no.”

I think Kuhn -- perhaps sensing the possibility of Davidson-style objections about translatability -- preferred to compare the concepts rather than the language. Davidson's reply is that "the mind's ordinary categories" contain a structure which would then have to be paired somehow with the organizing structure of the language, hence "doubling the work." But if conceptual scheme and language do co-vary, then there's no need to say everything twice, so to speak.

BTW, thanks for all your interesting thoughts on this topic.

. . . and I'm probably missing some.

To be sure, a ritual for American vegans on Thanksgiving! Free the turkeys . . .

While I slept my tofu-heavy sleep, this discussion has done really well for itself. I am no longer sure of my position here. “Legitimate ambiguity,” as Leontiskos puts it, now seems about right to me on the entailment question. So just one more comment: Could the ambiguity lie in the fact that we haven’t specified the universe of discourse – the objects over which we want our predicates to range?

Let me explain (coughs nervously and consults his old logic notes). Statements are perhaps best understood, for logical purposes, as not existing in the world of space and time – the old logical distinction was “subsistent” vs. “existent”. So we could, following this idea, quantify either over the universe of space/time objects, or over a different set, in this case the set of statements, or facts. My original reading of Bob Ross’s syllogism was that it quantified over both things and statements, whereas Banno’s objection is that we have to take it as referring only to things or states of affairs -- the subjects of facts, rather than the facts themselves, which are of course statements. There is a vast literature on quantifier variance which I’m only casually familiar with, but here is the difference, as I understand it, for our purposes:

If we exclude statements as bound variables in themselves, then “X is a normative fact” and “It is true that X is a normative fact” are equivalent. This is Banno’s position, if I’m understanding him correctly. But if we allow statements into our universe of discourse, we get a different interpretation. “X is a normative fact” and “The statement ‛X is a normative fact’ claims to state a truth” now say two different things, because they quantify over different ranges, in the first case a state of affairs, and in the second case a statement. I don’t think we’d need to know, or claim, anything about the truth of the statement in order to talk about it, provided we allowed ourselves to talk about statements at all as a separate class.

I think the example often given of this (I’m taking it from Copi & Gould’s Readings on Logic) is: “Sentences having ‛ghosts’ as a subject-term are not really about ghosts . . . but about some people’s statements about ghosts, or perhaps certain ideas about ghosts.” Substitute “normative fact” for “ghost” and this makes the case pretty well, though C&G say (or said, in 1972) that this interpretation can lead to several “odd consequences.” And, thinking it through, I'm unsure whether it requires the presumption that the subject-term doesn't exist, which moral realists would deny.

That said, I am an indifferent logician at best, and I’m open to correction here by my betters.

As for moral facts, I also think there are such things, but emphatically disagree that demonstrating their existence is as easy as Banno says:

That one ought not kick puppies for fun is a moral statement.
It is a true statement that one ought not kick puppies for fun.
Facts are true statements.

Therefore there are moral facts. 

The second premise merely imports the conclusion, thus begging the question. If there were no moral facts, then premise 2 couldn’t be true. But if it is “a true statement” that we shouldn’t kick the puppies, then this true statement is a moral fact. Circular, no? How have we established that premise 2 is a true statement? Is it meant to be obvious? But if it were, then we’d already know there are moral facts, and thus no proof would be required.

Very helpful analysis. A lot of people do seem to overlook or forget that Davidson doesn’t dispute conceptual relativism on the grounds that conceptual schemes can’t be relative, but rather on the grounds that the “very idea” can be shown to be either incoherent or contradictory. He’s not trying to fall back on some traditional/foundationalist One True scheme/content distinction.

I’m not sure I’m with you on the T-sentence interpretation. Why does the translation have to be “IFF”? Wouldn’t a more modal understanding be closer to what Davidson means?: “s can be true if p” We arrive at the same conclusion – that it’s incoherent – but without claiming that only being p renders s true.

The section in Wang on WMT vs CMT is by far the weakest in the paper. We could exercise charity ourselves here and agree to ignore it!

So incommensurable schema would be incommensurable sorting of the same stuff. They must be about the very same stuff. — Banno

Yes, that’s the key issue. Davidson says it very simply and effectively: “Strawson’s many imagined worlds [the first, relatively harmless metaphor] are seen or heard or described from the same point of view; Kuhn’s one world is seen from different points of view.” And the argument is that you literally can’t conceptualize “one world,” aka the very same stuff, in this way.

I think Kuhn means more like establishing a one-to-one correspondence between concepts. — Apustimelogist

This would indeed be the best way to salvage Kuhn’s argument. I’m sure Davidson’s insistence on the co-variance of concepts and language was, in part, meant to moot the point. But do you agree with this statement from Davidson?: “If conceptual schemes aren’t associated with languages in this way [that is, strict co-variance], the original problem is needlessly doubled, for then we would have to imagine the mind, with its ordinary categories, operating with a language with its ordinary structure.” What do you think Kuhn might reply to this?

if someone says, "X faith-proposition is false, and here is a proof for why," Aquinas thinks the proof can be addressed and refuted, even though no contrary proof for the faith-proposition is possible. — Leontiskos

Ahah, I see. Thanks for the clarification. Argumentation is possible within the "world" made possible by articles of faith, but the articles themselves, and that world, can't be demonstrated, only defended from (necessarily false) refutations. Is this closer?

I agree, it's tricky. I think the question is whether adding "It is true that . . . " to a statement adds or changes anything of substance. In a non-normative context, would we say that "It is raining" and "It is true that it is raining" express two different propositions? I don't think so. But including normativity seems to genuinely add something. "You shouldn't pick your nose" and "It is true that you shouldn't pick your nose" arguably do say two different things. The one is normative, in that it's telling you what (not) to do. The other doesn't adjure or command or affirm an obligation or anything like that; it only reports on the truth of the first statement. Or so it appears . . . . and hopefully Bob Ross will tell us if this is what he meant. Maybe I'm too full of tofu turkey to think clearly about it tonight! I'll see how it looks in the morning.

No, it makes sense. The claim would be that the statement "T is a normative fact" states something non-normative, something factual, because it's a claim about a statement, not the reality the statement refers to. It's about normativity, not itself normative. I'm not sure I agree, but I think it can be said coherently.

Oh, OK, so you meant that "T is a normative fact" is a non-normative fact. Got it.

Yes, I figured you didn't mean the scientists themselves thought this way. But it still remains important, I think, not to separate self-understanding from an allegedly more accurate account, here or anywhere else. Wouldn't that be a too-firm dualism of the sort that a lot of postmodernism rejects?

As for Husserl, we'd need a new thread! What you say about his project isn't wrong, but the distinction between what is external and what is phenomenologically present preoccupied him throughout his writings. I don't think he doubted for a moment that an external world was there to be encountered. What was important was the bracketing process, the epoche, without which the questions can't be meaningfully posed. But no one will ever have the last word on Husserl! :wink:

If this appears as a solipsistic rejection of an authentically external world, then perhaps it was never such a world we were concerned with in constructing our sciences. — Joshs

Interesting thought. But it depends on who the “we” is here. I think the evidence is overwhelming that scientists have always – until very recently – understood their project as trying to understand the authentically external world. They may have been wrong to do so, but let’s be careful not to read back into their projects a (post)modern view of science. For me, a more convincing challenge to the traditional idea that science constructs a picture of an external world is the actual work of physicists today. Someone who understands quantum physics better than I do, please correct me, but it seems to be the case that, while an independent external world remain ontologically likely, it’s no longer believed possible, on epistemological grounds, to know anything about it that isn’t observer-dependent. I guess Kant would be happy!

Shouldn't the first P2 be "normative fact" rather than "non-normative fact"?

We might be on the same page here, pretty much. Like you, I’m not sure we can maintain physical/causal closure, in our current understanding of what this means, and also give a convincing account of emergence. Yet that’s what many philosophers seem to require. Instead, we may need to rethink whether a commitment to the fundamentality of physics really precludes strong emergence. You know all the problems with trying to do this, I’m sure. I think so-called top-down causation is the biggest roadblock. But I’ll say it again – we need a lot of help from neuroscientists here, who are surely still decades away from even a good theory-based hypothesis about how all this works.

The allegory of the riverbank and the text thereabouts seem to say that there is an interchange between scheme and content, whereas to carry their case Wang must show a separation. — Banno

Right.

I don't see how there could be any scheme-neutral content, any more than there could be a thing-in-itself; for no sooner do we start to talk about it than we place it within a scheme. — Banno

You find many of the same problems as I do in Wang, concerning scheme-content dualism. I don’t think the arguments are there – on that front. Earlier in the paper, though, he discusses the question you mention concerning alternate truth-values and/or truth-value status in a given language. I didn’t address that section at all in my OP, but it’s very interesting. Maybe someone would like to take it on . . .

In setting out the argument Davidson sets out a position on the nature of language. — Banno

I think it’s right on target to see Davidson in the context of defending (or assuming) a particular view of language. Indeed, this is one point where Wang seems to misunderstand him, or make too-hasty equations of terminology. For instance, he attributes to Davidson “the identification of conceptual schemes with sentential languages.” But Davidson explicitly does not do this. He uses words like “association” and “relation” rather than identification, and says that language and scheme will “co-vary,” but the thrust of his argument relies on clear differences among language, concept, and scheme.

That said, Davidson upholds Tarski-truth as the model of how propositions work. He also believes language must refer. Is it all we can say about language? The cetaceans are a good counter-example. So is human music. Musicians generally don’t think that abstract music either states propositions or refers, but it seems impossible to get rid of the idea that music is nonetheless a language. Or, if that’s questionable, that it communicates. What, then, does it communicate? What are dolphins “talking” about?

I think you’re offering the cetaceans as an example of a genuinely incommensurable conceptual scheme – if they can be said to have a language. And your guess is that Davidson would simply deny them that designation. Or perhaps we could convince him that here is a case of partial translation. Might this not be closer to what’s going on with the cetaceans? You say we haven’t been able to make progress in the direction of translating dolphin sounds. I’m sure you’re right; I don’t know much about it. But don’t we treat all animals, even much less intelligent ones, as if they are “saying something” when they make their various noises? And it isn’t just fanciful. I certainly can tell the difference between when my cat is “saying” Please Feed Me and when she’s saying Eff Off, I’m Sleeping.

So here’s what we would need to ask Davidson: Do you require mental propositional content [or fill in whatever term you like for subjectivity] in order to constitute a language? Does the dolphin have to have the idea of Look, A Tasty Fish? Or are we willing to accept a functional/behavioral sense of what it means to communicate through language? I think we non-philosophical humans have already settled that for ourselves: We don’t require that our pets know what they’re talking about -- quite literally. What they know, if anything, is mysterious. But what they mean to tell us is often something even a child can quickly pick up. So: partial translatability? We get some, but not most, of what they say. And the scheme-content dualism remains in place, since whatever translatability is possible is down to the sharing of concepts between two languages – provided you give “concept” a free pass as a mental entity.

I think there’s a third possibility why we tend to “default” against strong emergence. This ties to your point that a good model for strong emergence is usually taken to be “one that works with supervenience.” What we’re really wanting is a model that works with scientific understanding in general, that respects the successes of similar inquiries in chemistry and biology. Supervenience may be the ticket, or it may not, but we need something that doesn’t invoke new processes or entities that violate physical and causal closure. I’m not sure this is possible. Consciousness, if and when it reveals its secrets, may turn out to involve concepts that make “physical” and “mental” ludicrous, and “physical closure” a profound misunderstanding. But “popular metaphysical assumptions” seems a bit brusque to describe the bind we’re in. They’re popular largely because up till now they’ve done such a good job.

Thanks for the reference to Searle on this. The distinction he makes between internal and external understandings of promising is valid and useful, but I don’t think it gets him out of trouble here. Let’s grant that, from the internal, institutional viewpoint, “Ought I to keep my promise?” is tautologous, more or less equivalent to “Are triangles three-sided?”. The person I’m imagining – the deceptive promise-giver – is presumably going to say something like, “Fine, but my promise was not a ‛promise’ in your sense. I said some words to deceive, well aware that my listeners would assume I was speaking from within the institution of promising. But I was not.”

So – was he? We’d have to agree with Searle that the “full force” of the expression “I hereby promise” excludes any reference to a description of mental states. But then how shall we describe the difference between the sincere and the deceptive promise-maker? The performative expression is identical, the intention (if I’m allowed that word) is not. Searle says that this makes “the relation between promising and obligation . . . very mysterious,” but I don’t see why. Insincere, deceptive people are common. What sets them apart from genuine folk are their intentions or mental states. Are they “really” promising? Yes, no, and maybe all seem like possible answers.

In a way, though, none of this is central to the ought-is problem, which, if we follow Hume, is strictly a logical one. It is also much deeper than a simple question about entailment. The more closely we look, the more we realize that we’re interrogating the very meaning of “ought.” Does a true “ought” have to be categorical, in Kant’s sense – that is, without any “if” premise? To my mind, Kant’s thoughts about this are still the gold standard, but that’s enough for now.

I love the Searle example, and how much good philosophical conversation it has provoked. But . . . . it is sleight of hand. The missing "if" premise is: "If you believe you ought to keep your promises (or fulfill your obligations)." Sadly, it's perfectly possible to engage in the speech-act of making a promise without having or feeling the slightest sense of obligation to follow through. The sense in which making a promise "puts you under an obligation" may not apply to you at all, in your opinion. Others will disagree, of course, and call you names, but oh well. There's still no logical entailment.

The world speaks back to us through the apparatuses, language and practices we use to make sense of it. — Joshs

Interesting. I like the metaphor. Can you expand on this a little bit? It seems really important to get a precise sense of what "the world" would have to consist of, in order for us to understand how it's separable from apparatuses, language, etc., and how it can have the kind of agency that could "speak back."

Are you sure Rouse is arguing here for the impossibility of pre-linguistic or pre-theoretical experience? I read him as saying two related but different things. First, he says that once we arrive at Wang’s “common-sense experience,” it’s an experience of something that is no longer “innocent” but inculcated with theory, the web of belief, and all the other familiar metaphors. That's true, given how Wang defines common-sense experience. But does Rouse claim there is no possibility of an experience prior to this? I couldn't find a passage that argues for this.

The second thing I see Rouse doing is questioning what he calls the “near side” of the scheme-content duality. That is, our schemes and theories are no more innocent of empirical input than our experiences are of conceptual input. But again, this wouldn’t necessarily show that the empirical input has no theory-independent existence.

I'm fine with considering this an "ought" statement in the same way that a moral one is. I also agree that the colloquial, illocutionary nature of the statement is confusing, though I think we could sharpen it up if we needed to.

But the problem is the same old one: There's an implied hypothetical between "I command you to" and "you ought to," namely "You ought to do this IF you want to keep your kneecaps intact" or some such. It's perfectly possible, though unlikely, that the hitman could reply, "I'm OK with broken kneecaps," in which case we haven't managed to derive a pure "ought" from an "is." This example certainly clarifies that the ought-is problem is logical, not psychological. Since just about no one wants to be injured in this way, the command has a lot of psychological force -- but no logical entailment without the "if" premise,

Thanks!