In the original post, he takes perspectives to be sets of true propositions. This makes it easy for him to say when two perspectives disagree: they disagree just in case one contains a proposition and the other contains its negation. However, he often talks as if propositions need to be evaluated for truth relative to a perspective, so it’s not clear how to make sense of the claim that perspectives are sets of *true* propositions. If perspectives are sets of true propositions, and if a proposition is true relative to a perspective just in case it is a member of the perspective, then Raleigh’s appeal to truth isn’t doing an work: perspectives are just sets of propositions.

However, this would make hash of his attempt to secure propositions that are true from all perspectives. No proposition is a member of every set of propositions. There are far, far too many sets! Raleigh runs into trouble, because he tries to make perspectives set-theoretic constructions, on the one hand, while still trying to make them answer to objective reality, on the other hand. This is the first indication that he’s running together two sense of “fact” that it’s important to keep distinct. In the original post, Raleigh says that reality is “made up” of perspectives. This makes the problem immediate. He tries to back off from this in the comments, by taking true non-perspectival propositions to be propositions true from all “relevant” perspectives. I suspect that it will be hard to say which sets are relevant in a way that’s not circular.

After being pushed by Joe, Raleigh explicitly brings up the second understanding of “fact”: a structured bit of reality, or something that can be the case. However, he immediately tries to square this with his earlier understanding of “fact”: a true proposition. This is a mistake. Call facts in the first sense “states of affairs” and facts in the second sense “true propositions.” One nice, but controversial, way to go here is to take a state of affairs to be an object’s having a property or two or more objects’ standing in a relation. Instead of taking reality to be constructed out of propositions, or sets of propositions, take it to be constructed out of states of affairs.

The worries about whether or not perspectives are closed under logical consequence follow immediately from the confusion between the two conceptions of facts. As Joe and Raleigh both point out, logical consequence holds between propositions, not between bits of reality.

Talking about indexicality is confusing on either understanding “fact.” There are indexical expressions, but no indexical contents or bits of reality. It’s a bad piece of terminology, but it’s one Raleigh doesn’t need. The distinction he wants is the distinction between facts (in whatever sense he decides upon) that have perspectives as constituents and those that do not.

Suppose Raleigh irons out the confusion between the two different sense of “fact” that seem to be in play. I echo the question that has been raised several times. What is the concept of perspectival disagreement supposed to get us?

“The ordering among knowledge states that involve the same content does not need to be a linear ordering. For example, some multi‐grade adjectives may not impose a linear ordering on their domains.”

It will be tempting, nonetheless, to think of them as plausibly linear (i.e. a theory that takes them all to be linear could, for all that, be plausible) or often linear (a theory that allows multi-grade epistemic strength could nonetheless allow for sets of contexts in which the varying, epistemically relevant independent variable was linearly ordered, and that these sets are common among our dataset). contrastingly, it seems to be only very strange context in which contrasting standards of taste are linearly ordered. so we might have something like a spectrum: standards of tall (always linear), epistemic standards (plausibly/commonly/usually linear) and standards of taste (rarely and degenerately linear).

Your “impossible task” is my knock-down objection. I think relativism is a non-starter for just the reasons we (redundantly) laid out so clearly. (And just to spell out the rest of my position,) I suggest that the relativistic account does not provide a satisfactory account of faultless disagreement, because it makes the only disagreement present in matters of taste non-conversational disagreement. And… blah blah blah you know how the line goes.

“True propositions” are linguistic items, sorry. I take it we can agree that propositions are linguistic items. Adding “true”, if you’re a correspondence-theorist, is a property of propositions that holds only if they bear a certain relation to the world. But it sounds like you want something closer to states of affairs?

The more I think about it, the more I worry that there’s a semi-serious muddle here. “Closed under implication” will only make sense if we’re talking about propositions. There is no sense in which states of affairs are true: they either obtain or fail to obtain. And I don’t know what to make of this sentence:

“however, the metaphysical counterpart of the linguistic debate over relative truth is to claim that ‘it is not the case that licorice is tasty’ is a perspectival, but *not indexical*, fact. It is a perspectival fact about the world.”

Assuming you’re serious about this being a metaphysical matter, it’s nothing more than a category mistake to say that ‘licorice is tasty’ is indexical. On your view, licorice’s being tasty is a perspectival state of affairs (if I may). But then so is my growing up in Schenectady, and there is no difference in the amount of ‘disagreement’ between the metaphysical counterpart of first-person perspectival states and matter-of-taste perspectival states; you might as well be a metaphysical version of a contextualist. Relativism only gets going once we’ve got the context/circumstance distinction, which you claim (I think) to lose by going metaphysical. If that sort of distinction is one you want to capture, then maybe you’ll want to allow for two sorts of perspectiviality, one corresponding to context and the other to circumstance. But I’m quickly losing track of what the advantage of going metaphysical was in the first place.

This can probably all be sorted out somehow, but it needs doing. Boils down to two choices for you, I think: 1) is the difference betweeen relativism* and contextualism* important to my project? 2) Is there really an advantage to going metaphysical?

No inconsistent set of facts is closed under implication.
If you and I disagree about whether licorice is tasty, then it seems like (a) neither of us has a false belief, and (b) it’s either the case that (b1) we disagree on whether a proposition is true, or (b2) we both believe a proposition indexed to ourselves to be true, and would accept that the same proposition indexed to the other is true. (b1) is relativism. If you take b1 to be correct, and you concatenate the true beliefs that both of us have into a single set of beliefs, you’ve got an inconsistent set.

by asking the relativist to retain a sense of disagreement between two conversants, but to insist that the set of true beliefs on the table during such a disagreement be part of a set that is closed under implication, you’ve given the relativist an impossible task. Perhaps closure under implication can be held up as a theoretical virtue of rejecting relativism, but it can’t be held as one of the desiderata that relativism is supposed to accomodate. CuI is incompatible with relativism.

Several things.

First, It’s a shame that you didn’t read the exchange between Brian and I. This was was discussed at length. Your first and third questions is answered there, where we discuss the metaphysical sense of disagreement. I’ll refer you to those comments.

Second, Facts are things that are the case. Not languagy things. languagy things can’t be the case. If facts are true propositions, (as I’m happy to say they are) this is a metaphysical concept of truth, not a linguistic concept of truth. Additionally, sets of facts aren’t facts. Perspectives are sets of facts. This, by the way, seems like the theoretically cleanest way to construe perspectives, and to say the things I should be able to say about them, but i suspect that some people who work in aspect may be unhappy with this construal of things. Construing times (temporal perspectives) for instance, as sets of facts, seems to fail to capture something essential about relationship between one time and another. But any adjustment they’d make is, I think, going to preserve the sorts of things I want to say about perspectives, and make them even less mistakable as languagey entities. No matter, though. Facts aren’t languagey things.

Finally, Closure under implication.
All of the facts that make up a single perspective are compatible with one another. The union of two perspectives (which are of the same sort) will always be an inconsistent set. So a single perspective is closed under implication, a pair of perspectives is not. This is why the question of ontology of perspectives arises; by allowing multiple perspectives into reality, without qualification, we risk allowing contradictions into our ontology. This can be solved by giving ontological privelege to one of the perspectives, or denying the existence of irreducibly perspectival facts (where all purportedly perspectival facts become ‘p at t’ or ‘p to s’ or ‘p in w’ in order to remove the contradiction). Both solutions have their problems. So…suppose from my perspective beets are tasty and licorice is tasty. From this I can derive within the scope of my perspective that beets are tasty, but I can’t assume that that conclusion is true from your perspective. All of that holds so long as ‘licorice is tasty’ is a genuine first-personal-perspectival fact; if there are no such facts, then all of the facts that hold from my perspective should hold from your perspective, and the union of the two should be closed under implication.

The reason that this is supposed to be interest/worth building a seminar around, is that my concept of perspectival disagreement provides a clean explanation of what’s going on in faultless disagreement. If there are irreducibly first-perspectival facts, then first- personal perspectives will disagree with respect to the facts. If there are not (and everything that looks like a fpf of the form ‘p’ is actually an objective fact, or a contextualized fact, ‘p for s’, or is some sortof nonpropositional entity, as the expressivist would have it)then all perspectives will agree on all of the facts (and there won’t be an interesting metaphysical concept of perspective left). FPF’s can have faulty disagreement, when one believes p, the other beleives ~p, and p (if true) is an objective fact (and so is true from both perspectives or neither). But their disagreement is faultless if it reflects one of the factual disagreements between their perspectives.

But if you’re talking about facts, maybe you want something linguistic anyway.

(Caveat: I skipped the Brian-Raleigh exchange in its entirety.)

There’s a lot here, but here are a couple things to get going on.

1) I’m not sure there’s any useful sense in which metaphysical things can disagree. It strikes me that only language-y things can disagree. Perspectives, on your view, are facts, and facts are language-y things.

3) Do you want these perspetives to be closed under implication?

2) The last paragraph ostensibly takes up my central claim, that the disagreement Macfarlane (and you, and all these relativists) produce is unworthy of the name. But I don’t think it’s deallt with satisfactorily. A way to put this is to ask: insofar as the utterances disagree, is (a) more like (b) or (c).

(A) Alphonse: Capers are tasty.
Beatrice: Capers are nasty.

(B) Clarissa at t1: Bush is President.
Clarissa at t2: Bush is not President.

(C) Derrick: Capers are members of the anchovie family.
Ephraim: Capers are a type of plant.

If Macfarlane/you say that (a) is more like (b) than (c), then we’re talking about something pretty dull and mundane–hardly worth building a seminar/book around. But if (a) is more like (c) than (b), your/Macfarlane’s strategy won’t do the trick.

Here’s where I got on the soap box.

I say to Joe: If the relativist wins the semantic dispute, that suggests that there are first-personal perspectives in which facts reside; facts that are false from other perspectives.
Joe says to me: Fine, but that’s not disagreement. That’s just different perspectives being made up of different propositions.
Raleigh: That *is* disagreement. Two perspectives disagree insofar as a proposition true from one is false from the other. perspectives are sets of facts, and non-identical sets disagree on those propositions that make them nonidentical.
Joe: Fine, but that’s not the kind of disagreement relativists want.
Raleigh: Sure it is.
Joe: No it isn’t.
Raleigh: We must settle this with Jello wrestling!
Joe: You’re on!

Well…that’s mostly what happened anyway.

So, in writing this post I definitely came to see what was right about Joe’s insistence. We have to distinguish between conversational disagreement and perspectival disagreement, and the phenomenon I was describing was certainly *not* conversational disagreement. Nonetheless, I’m inclined to say that there are three sorts of conversational disagreement.

(a)
Joe: I’m from Schenectady.
Raleigh: No I’m not!

(b)
Joe: Obama is president of the United States.
Raleigh: No he’s not!

(c)
Joe: Licorice is tasty.
Raleigh: No it’s not!

Ok, so here are three sorts of conversational confrontations. Which of them count as real disagreements?
Well, B certainly does, because the proposition in question is not a first-personal perspectival fact. If it’s true, it’s true for all first-personal perspectives (because its truth conditions are not wrapped up in a perspective), so one of us is wrong about the propositions that are true from our own perspective. In this case, Raleigh is wrong, because he thinks ~(Po) is true from his perspective when Po is true from his perspective.

(a) certainly isn’t a disagreement. Now, I DID say that indexical propositions were among those over which perspectives disagree, and that is true. Different times, on my reading, disagree over which moment is the present (and over whether/by how much a moment is past/future). and different first-personal perspectives DO disagree over who is ‘I’ and who is ‘thou’. But for most speakers, it’s evident that the ‘I’ indexes to the speaker, so the content of Joe’s assertion is ‘Joe is from Schenectady’. So, as I said, different perspectives disagree over propositions over perspectives; but this isn’t plausibly expressing such a disagreement. Possible weird cases in which it would be? Maybe Raleigh is a solipsist that thinks that all I-facts are his I-facts. In which case, Raleigh would be disagreeing with the proposition he thinks Joe is asserting (though he’d be wrong. facts about what proposition Joe is asserting are, by the way, not first-personal perspectival facts).

Now, what about (c). Here, the proposition in question is a proposition about the world, but its truth conditions are tied up with a perspective. ‘licorice is tasty’ is a perspectival fact. My perspective is made up of lots of propositions, including ‘Raleigh is from Kansas City’, ‘Joe is from Schenectady’ ‘I am Raleigh Miller’ and ‘it is not the case that licorice is tasty’. Joe and I are not disagreeing over whether I am Raleigh Miller, and we are not disagreeing over whether Joe is from Schenectady, so there’s nothing left for us to be disagreeing about. So long as I know that, in (a), Joe is expressing that Joe is from Schenectady, I will not take him to be contradicting any of the propositions that make up my perspective. however, the metaphysical counterpart of the linguistic debate over relative truth is to claim that ‘it is not the case that licorice is tasty’ is a perspectival, but *not indexical*, fact. It is a perspectival fact about the world. And Joe’s perspective and my perspective disagree over the fact. So when we assert our beliefs, we are giving voice to the discrepancies between our perspectives. I think the best way to describe disagreement over relative truth is to say that this is what’s going on. It has all the markings of disagreement because (a) it has the superficial linguistic structure of disagreement over aperspectival facts, and (b) we take the propositions that are (i) true from our perspective and (ii) not *about* our perspective to be true from other people’s perspectives as well, until we are shown otherwise.

Is that *genuine* disagreement? Ok, fine, now I fall back to Stewart’s answer. Locate ‘genuineness’ where ever you want. But this is what I think is really going on, and it bears enough similarity to other sorts of disagreement to reward analysis that takes it to be disagreement.

P.S. I think the resistance to faultless disagreement comes from the assumption that if there is faultless disagreement, the world is made up of contradictory propositions.
What may be unclear is: This is not a dispute about linguistics anymore. This is a dispute about whether reality is fundamentally first-perspectival, in the ways that people want to say it is fundamentally temporal or fundamentally modal. There are reasons to think it is; there are probably much better to think that it isn’t. Also, if we take seriously that *this* is the real disagreement (and if we take the contradiction of allowing such perspectival facts simpliciter into reality to be inadmissible) then we will be extremely impressed by the parallel for this platitudinous anti-realism about first-personal perspectives and JME Mctaggart’s argument for the unreality of time.