The photograph thing is clever. I had matter-of-factly observed that we can't deduce p from someone's asserting p -- never occurred to me to imagine deducing p from my believing that p.

But this is another point of divergence between the first- and not-first-person cases: we regularly infer p from the fact, which we (let's say) infer from what they say, that someone trustworthy, either in general or just in the matter at hand, believes that p. It is plausibly our principal way of gathering knowledge. This sort of inference is clearly reasonable even if we grant the point at the top, that there's no logical but only probable implication to be had here.

But in the first-person case? 'It must be true because I believe that it is' is a non-starter. Not only not deductive but not even probable, and in fact just not conformant with our standards of rationality. This looks wrong, on second thought. Of course someone could say, 'I'm probably right about this, because I'm usually right about this sort of thing.' No, the issue is whether they could take their own belief that p as evidence that p, and if it's evidence then it could count as a reason for believing that p!

That's pretty interesting.

Propositions can't be true prior to the existence of a criterion (hence the need for a criterion) — TheMadFool

This looks like a theory of truth, not a theory of how we know what is true.

Absolutely everyone agrees to all of this.

They say nothing about themselves. The say something about the weather. — Ciceronianus the White

Then you agree the Moore sentence is not a contradiction. So what's wrong with it? Why is it something no one would ever say?

In the interests of comity, I'm going to speak here with a looseness I'm immediately disavowing.

I believe the sticking point is this: meaning, referring and believing are part of our frame not yours, part of folk psychology, not neuroscience. That's why you'll find no takers for your claim that when someone says it's raining they're talking about something that's going on in their head. No one anywhere will assent to that. Ask a non-philosopher and they'll tell you that people who believe things like that get locked up and put on heavy meds.

If you restrict yourself to telling the neurophysiological story of how rain is detected by our senses, how the brain generates predictions about the near future based on this new data, how we reflexively respond to seeing someone heading for the door without an umbrella and certain parts of the brain spring into action to prepare and then cause yet another complex subsystem to emit the sound "It's pouring outside!" -- tell that story and mostly people will be fascinated, marvel at what science has learned, and have no problem.

It's not entirely your fault, of course, because the early modern philosophy canon we're still obsessed with is full of groping attempts at psychology (as we can find groping attempts at linguistics, or physics or other sciences throughout our strange history), and since these philosophers weren't clear on the two frames they're mashing together, by and large neither are we.

Yes, well, I drive a car by sitting like so, and moving my arms and legs thus. But moving my arms and legs thus is not driving a car.

Well, on pain of being unable ourselves to say what our own utterances are about, it had better be something we have access to during the construction of those utterances, and that isn't the state of the world, only our inferences of it. — Isaac

I'm just not following this. Do we make inferences and form beliefs about the world and its state, even though we don't have access to it?

If you want to tell a causal story about why we say what we do, you should probably also have a story that gets you from facts to meanings, or you never get "aboutness" at all. (Quine and Grice, to name two, both have such stories.)

Supposing that you can get aboutness here, how do you pick which cause in your chain is the one the utterance was about?

Suppose you've conditioned your dog to say "sausages" given a certain stimulus. The pathways are there, just waiting for it. (Making a hash of the neuroscience.) When that stimulus shows up and your dog says "sausages", isn't it more natural to single out that stimulus, or the event of its occurring, as of special importance, rather than whatever happened in your dog's brain? (Still not clear that cause would be what an utterance is about.)

As for whether your dog is talking about your dinner, even Ryle would have little trouble with that. A single conditioned response it's just not how we judge competence at understanding or producing meaningful speech.

Maths Tutor: "Williams! What is three times seven?"
Williams: "Sir! I believe three times seven is twenty-one, sir."
Tutor: "Don't be irrelevant, Williams. I asked you about the product of three and seven, not your state of mind, which I assure you is of no interest to me or to anyone else."
Williams: "Sir?"
Tutor: "I have no interest in what you believe, Williams, which is I why did not ask you, 'What do you believe the product of three and seven is?'"
Williams: "Yes, sir. It's twenty-one, sir."
Tutor: "Andrews! Is Williams correct?"
Andrews: "I thin-- yes, sir. It's twenty-one."
Tutor: "Good! Now, we'll have no more of this 'believe' business in my classroom, is that clear?"
All: "Yes, sir!"

Oh, but young Williams's maths tutor is concerned precisely with Williams's state of mind, you might complain; his one job is to make sure Williams holds the right beliefs. Is that the only interpretation? Couldn't we also say the tutor's job is to ensure that Williams gives the correct answer when asked a direct question? I can just hear Wittgenstein describing this scenario as "training".

Wittgenstein's remark, by refusing to acknowledge the de dicto/de re distinction, has another little oddity:

Wittgenstein: "Is there a fire in the next room, Williams?"
Williams: "No, sir."
Wittgenstein: "Show it to me."
Williams: "Sir?"
Wittgenstein: "The fire I asked you about, Williams. Where is it?"
Williams: "There is no fire, sir."
Wittgenstein: "Don't be absurd, Williams. Would I have asked you about something that doesn't exist?"
Russell [ entering ]: "I believe, dear Wittgenstein, that young Williams here [ pats Williams on the head and winks at him ] is trying to say that the next room is such that if something is in it, it is not a fire."
Wittgenstein: "Oh shut up, Russell."

The "aboutness" of a sentence is not always a simple matter. What one can and can't say is almost never a simple matter. Why then should we expect to reach simple conclusions about what one can and can't say about what?

(( I couldn't find it online either. Curious bit of history in the SEP article that Church seemed to have recently seen Moore's thing when he anonymously gave Fitch the horrifying paradox that bears his name. ))

@Isaac thinks it overthrows the correspondence theory of truth with a single blow. That's a bit more vehement than anything I've posted.

Why does Moore say/think it would be said? — Ciceronianus the White

Fair enough.

The thing is, Moore sentences screw with what, to a budding philosopher, might seem like natural answers to the general question, "Why do people say the stuff they do?"

There is, sad to say, a gulf between the fact that p and someone believing that p. Thus, even if people always and only said what they sincerely believe, we could not deduce p from someone asserting it, or from p that people would assert p and not assert ~p. So much we all know well enough.

We also all know that from someone asserting p we cannot conclude that they believe p. They may be insincere; they may have misspoken; they may not have meant p by saying what they said, even though we might take them to have meant p, and there are subcases of this last that are particularly noteworthy, such as irony.

But having gotten to irony, it looks like we've made a mistake.
A: "How was work?"
B: "Just peachy."
B is most likely not asserting that work was just peachy; she is asserting that work was not peachy at all by saying that it was. She is still asserting what she sincerely believes though, right?
A: "That bad, huh?"
B: "No, it was okay. Long stupid meeting this afternoon, that's all."
So B was also exaggerating, and didn't actually believe that the work day had been the opposite of peachy, although that is what she meant by what she said.

We can mostly infer what someone believes from what they say, and we probably have to, because reasons. We can even do this taking some extra steps between what they said and and what they meant by what they said. (Or "what they meant by saying what they said," if that's better.)

The connection between what someone means and what they believe is clearly not just (logical) implication, as we all know. It looks a lot of the time like what Grice called 'implicature': this is a slightly weaker connection than implication in that the audience is encouraged or expected to make an inference, but not only does the speaker not require their audience to infer their belief, the speaker might actually block that inference. Implicature is cancelable.

In the example above, we get one of each: in the first exchange, B encourages A to take her to mean work was the opposite of peachy; in the second exchange, B then cancels A's inference that B believes work was the opposite of peachy. And this is all perfectly ordinary. I've put some of Grice's terminology on it, but we all do this stuff everyday, and it seems barely even deserving the word 'theory'.

And it's also wrong. Moore sentences show that clearly, and Grice himself makes this very point. The inference of belief is not cancelable and is not implicature. I attempted a little sleight-of-hand here, which I'm guessing most readers caught: exaggeration is, like irony, a sort of insincerity, whether or not it is intended to deceive. Grice was aware of Moore's paradox. He treats irony as a violation of the maxim of quality*, and the assumption that the speaker believes what they (in the indicative mood) say as just assuming they are following the maxim. (And further, if you make an indicative mood utterance you intend that the audience think you believe what you are saying.)

According to this story, I'm not inferring that you believe p when you assert p, but assuming you do. (Leaving aside whatever interpretive hoops we jump through, for the moment.) When I assert p, I don't intend you to infer that I believe p, I intend you to believe that I believe it. Just the conclusion without the inference to get you there.

But are we really done with inference? Doesn't it seem like I'm actually reasoning something like this:
IF you assert that p AND IF you are observing the maxim of quality, THEN you believe that p.
Or we might put it this way:
IF you assert that p, THEN EITHER you believe that p OR you are not observing the maxim of quality.
Sure. But we could rephrase: when you say something either you believe it or you're lying, or exaggerating, or speaking ironically, or any of the other ways of violating the maxim. It's the violations that give rise to implicature, and the default is just non-inferential belief ascription.

I think this should strike most philosophers (of an analytic bent, anyway) as a somewhat bitter pill to swallow, rather like Hume showing you can't justify your reliance on induction. I think we want to say that ascribing beliefs to others based on their indicative mood utterances is rational -- and I believe it is -- but if it is, it is not because we infer what they believe from what they say. I mean to say: it is a question of rationality, within the sphere of reasoning, evidence and so on, not that "the rational thing to do" is believe people are always honest or something. Not the conclusion, but the process.

So if we are to find a place for rationality it's going to be somewhere else, which is a little surprising because the natural hook to hang rationality on would surely have been right around here somewhere, right? Language use, propositional attitudes, belief formation -- this looks like the place.

But we need something more. Since I've more than nodded at Grice, I'd like to be able to give his answer, but I'm not sure we get one. (And I'm no Grice scholar.) The next thing he reaches for should look familiar by now:

I think that this consequence is intuitively acceptable; it is not a natural use of language to describe one who has said that p as having, for example, "implied," "indicated," or "suggested" that he believes that p; the natural thing to say is that he has expressed (or purported to express) the belief that p. He has of course committed himself, in a certain way, to its being the case that he believes that p, and while this commitment is not a case of saying that he believes that p, it is bound up, in a special way, with saying that p.

But this ends up not being much of an account, because all Grice is going to claim is that when you make an indicative mood utterance you intend the audience to think you believe it. That might get us to treating the commitments of indicative mood utterances as the same as belief reports, but nothing more. It's not an account of what those commitments are or of the sense in which the management of such commitments is a rational matter.

(No one take this as any kind of final word on Grice, please, because I think it likely he addresses these issues in stuff I haven't read. If anyone knows, speak up.)

* Try to make your contribution one that is true; do not say what you believe to be false; do not say what you do not have adequate evidence for.

I'm assuming that nobody would say "It's raining" if they thought it wasn't raining, unless they wanted to lie for one reason or another — Ciceronianus the White

Of course nobody would say it. The question is, why not?

The Moore sentence is clearly pathological. But how do you know that? What rule, principle or maxim does it violate? Or is it defective in some other way?

I can't explain it, and that bothers me.

(I think @Snakes Alive had a promising approach.)

The fact is that birds are subjectively experiencing [...] So not just biochemical and physiological drive [...] — Enrique

This is just not the tree I was barking up, but that's on me, I could have been clearer.

If I put together a jigsaw puzzle by selecting a piece at random and then performing a brute-force search for pieces that connect to it properly -- matching shapes and colors, the usual rules -- and then repeat this process with the new edges of my work in progress until the puzzle is complete, I can be successful without having any idea what the final form will be. (I do this consciously, sure, and we can say all sorts of stuff about how I do this and how I come to be able to do this.)

It was this distinction I was trying to draw attention to: between a more-or-less top-down approach, guided by knowledge of what the final result is supposed to be, and a more-or-less bottom-up approach that focuses only on what the permissible next steps are.

Formal logic -- which you allude to at the top -- aspires to the mechanical, to rules only concerned with what the permissible next steps are. This is a little odd.

Like humans, songbirds have facility with structure concepts, for they erect nests that are intricate masses of sticks and brush, clearly envisioning how parts fit together as a whole [my emphasis]. — Enrique

Is there research that establishes this? I think I can almost imagine experiment designs that might get near questions like this -- or at least showing sensitivity to the in-progress shape. (Do they correct for mistakes? Do they adjust for the effects of an obstacle? But that would be the barest beginning.)

It's not the word "envisioning" which bothers me.

It's specifically the question of whether the behavior is guided toward a specific end result, for two reasons:

• rationality is, in some vague sense, a matter of the role an ideal plays in behavior (the ideal in this case being the shape of the not-yet-completed nest); the ideal is something not part of the physical environment but is still "acted on", "striven towards", "abided by", that sort of thing; (a)
• logic in particular is concerned precisely with the reduction of informal methods of reasoning -- which are often guided by the desired end result in ways we consider fallacious -- to a mechanical procedure, in deliberant disregard for what the procedure will produce.

A nest-building bird that followed a procedure mechanically -- add a piece to what we have so far by entwining it in a certain way, leaving ways to use it for the next bit, and preserving a local curvature of such-and-such -- could consistently produce nests with no knowledge of the overall shape its procedure leads to. (b)


(a) Fine with me if someone claims it is "physically present" in the builder's brain -- at the level of behavior, materials used, artifacts constructed, it's not there, but maybe I shouldn't shrug this off so quickly.

(b) And we would presumably look to natural selection to explain why this is the procedure they follow, which is whole 'nother thing.

There's no particular use here for P1. Just use P2-P4 to derive C1.

The general form here is sometimes referred to as "argument by cases".

Better still do this:

1. x & y (premise)
2. ~y (premise)
3. y (from 1)
contradiction.

Is the following a fallacy ?
x ∧ y
~y ;
hence ~x — Philarete

Yes it is a fallacy. Once you know either of the conjuncts is false, you know the conjunction is false without ever looking at the other conjunct or knowing its truth value. (Disjunction behaves similarly with truth rather than falsehood.)

In particular, here you have learned nothing at all about the truth or falsehood of R, though you have both P and Q being false.

No. C2 and P2 are inconsistent, but R could still be true for all that. C3 doesn't follow.

Indeed.

Good luck with your work.

[Foucault's] concept is neither linguistic, nor psychological — Number2018

Then you'll want to say something like this:

By 'speak' I don't mean [[speak]], by 'proposition' I don't mean [[proposition]], by 'true' I don't mean [[true]], and by 'assert' I don't mean [[assert]].


Which I would see as directly related to my point that in speaking you accept certain responsibilities to your linguistic community. See Humpty-Dumptyism.

We often take a common practice and formalize it, more or less abstractly. Often there are options for how to carry out such a formalization, and it's even possible to screw up, have a formal process that doesn't match up well with the original. It's natural to think of mathematics beginning some such way, and many people have thought just that.

That's tricky though, right? Because the sort of abstraction and structure building we associate with mathematics seems to be what we use to formalize existing informal practices. There's some chicken and egg trouble here.

But there are further puzzles. It's also quite natural to think that formalization is possible in the first place because the underlying structure was there and operative all along. Formalization would then be not an invention we superimpose on a practice but the discovery of the true structure, the essence of what we were doing, in our imperfect way, the whole time. That puzzle becomes particularly acute in the cases of mathematics and logic.

I have some sympathy for the idea that logic is a formalization of, say, the pre-logical practice of inference, or of the cognitive virtue of consistency, or of the social requirement of predictability -- I'm not even sure what to fill in there! Maybe all of the above and more. (What I'd really like to put here is linguistic dispute.) But that leaves untouched lots of questions about how that formalization is even possible or why the needs it meets are needs in the first place.

Any thoughts on what we might portentously call "the origins of logic"?

But I don't think you can build this fortress yourself.

There's an episode of "Barney Miller," an old sitcom, in which an old man is about to be taken off to an asylum because he seems to be babbling gibberish, but Sgt. Dietrich finds another old man who speaks the dialect of Ukrainian the supposed crazy man has been speaking all along.

Is it possible to find a link between "I lie" and "I speak"? Both can be split into two interdependent propositions. — Number2018

Not sure what you mean here. Are you thinking of versions where these are taken to be self-referential? (You'd have to say, because to my ear these both sound more like "habitual present tense" or whatever the right term for that is.)

One thing I was wondering about was whether we're to take your "I speak" as being spoken in a specific language. As I suggested above, I'd be tempted to see any utterance in a living language as also carrying a commitment to a linguistic community, which carries with it certain rights that can be claimed and certain responsibilities that ought to be met. If you want to abstract away the specificity of the language so that no linguistic community is implicated, not even an abstract one, you either want a private language -- and someone will be along shortly to tell you you can't have one -- or maybe a Whitman-like "barbaric yawp". Is a barbaric yawp speech? Unlikely. Is it a signal? I'd like to say unlikely as well, because once signaling became voluntary (on the way to becoming speech) it also became possible to make a noise for the sheer pleasure (or at least sensation) of noise-making. This is plain in small children. But of course we've drifted away from the linguistic now...

I suppose I'll have to have another look at the McDonnell & Abbott paper, because I think that's kind of what I'm talking about.

To put it relatively starkly: are there cases in which you can know your performance is improving not because you can test that directly but because you know the procedure you're following to improve is theoretically sound? Imagine an iterated version of our game in which your winnings each round go into an account you don't have access to. You do want to increase your winnings, but you'll only know how much they've grown when you stick. In the meantime, you gather data from each round. You could say that it's just a different perspective on what the experiment is, and maybe that's all there is to it.

I thought of one analogy: suppose you're firing artillery at an enemy position but you have no spotters. (Sometime in the past, no aerial surveillance or anything.) You could use your knowledge of how the enemy troops are usually disposed on this type of terrain, and also how they're likely to respond to bombardment. You could, for instance, attempt to fire at the near side of their line first -- using only physics and an estimate of how far away they are -- and then fire slightly farther, if you expect them to retreat a bit to get out of range. If you had some such general knowledge, you might be able to systematically increase the effectiveness of your strikes without ever seeing where they hit and how effective you were on target.

It strikes me as a deep problem because so much of what we do is guided by faith in a process. Many of those processes have been tested and honed over time, but induction is still a leap of faith. Some things we do get to test directly, but some of the very big things we don't. I'm thinking not just of science here, but also our moral, social, and political decisions. I'd like to think certain acts of kindness or courage, small though they may be, might improve a society over time, but I don't expect to know that by experiment. Anyhow that's part of what's been in the back of my mind here.

Right, that was the idea. Whether it might be possible to make a better second (or later) choice without knowing how good the previous choice (or choices) was.

I accidentally addressed this before, I think, when I noted that any arbitrary cutoff that helps you by falling in the [x, 2x] interval, works as a criterion of success (for having chosen 2x).

I'm going to mull it over some more. The natural answer is of course not! but I'm not so sure. Any information you might use to improve future efforts is also information you could use to evaluate past ones... [That's an argument against the idea.]

I'm just wondering if there's an alternative to the usual predict-test-revise cycle in cases where you can't get confirmation or disconfirmation, or at least don't get it straight off. In our case, that might be, for the iterated case, using a function or a cutoff value you revise as you go, without ever finding out how well your earlier picks went. Within the single game, finding out the value of one envelope is not enough to be confirmation of your guess, but you might still use it to make your next guess better.

Is this any clearer? Apo and I have talked about the desirability of getting binary answers. I'm just thinking about how you might proceed without them. I expect such issues have been thoroughly chewed over already by people who actually do science!

[ small clarification ]

I keep thinking about how the two rounds of the game compare.

The general problem would be something like this: can you improve your performance even in situations where you are unable to evaluate your past performance?

I think the answer to this turns out to be yes, and I would consider that a result of some importance.

***

There is more information in the second round -- it's unclear.

I only meant that terms like "performative utterance" and "illocutionary force" originated in a philosophical tradition I know a bit about (Austin, Strawson, Grice) and found a home later in ("as" might be more accurate) the linguistic field of pragmatics, about which I know less.

Since you don't seem to be talking about either of those, I doubt I can be any help.

Not my area then, and any comment I could make would be uncharitable.

If you haven't read Austin yet, I'd recommend doing so.

The conditional probabilities also change when you know the value of one of the envelopes, but you don't know what they are anyway. — Srap Tasmaner

Clarifying a little. Suppose X=5. This is the sort of thing you don't know, but choice being what it is, you can still be confident that P(Y=X | X=?) = P(Y=2X | X=?).

P(Y=X | X=5, Y=10) = 0, but for you that's P(Y=X | X=?, Y=10) = ??. 0 or 1. Changing that to a pair of terms multiplied by P(X=...) just moves your ignorance around, though it's better formally.

You can recognize that conditioning on the value of your pick changes the probabilities, even though knowing exactly how they change amounts to knowing the value of X. And you don't.

I still look at the problem this way:

First you're presented with two sealed envelopes; you can't tell which is bigger.

You open one and observe its value; you still can't tell which is bigger.

If you had an estimate for how much money would be in play, observing a given value may or may not lead you to revise that estimate. Depends.

The conditional probabilities also change when you know the value of one of the envelopes, but you don't know what they are anyway.

The value you observe is no help making the one decision you have to make, so it's reasonable to treat the second choice you face as a repeat of the first, meaning it's just a coin flip: you're either going to gain or lose, relative to the first choice, but you don't have any idea which, not even probabilistically, and the amount you might gain is the same as the amount you might lose.

If it would be your own money at stake here, you shouldn't be playing at all.

" I speak" has a special privileged status, different from any other statement; it can support or destroy the whole theory. — Number2018

What whole theory is that?

Actually, " I speak" is a kind of artificial construct and for better
understanding we need to relate this utterance iwith a concrete situation. — Number2018

I know you didn't mean it this way, but isn't that true of all sentences?

(I have no idea what Lazzaroto is talking about.)

If the players wants to think that means 5 or 20, then that is not the fault of the game master. — Jeremiah

Of course. Someone about to buy such an envelope on the street ought to hope a friendly and helpful philosopher would be walking by to point out that

⊢ p ∨ q p ⊻ q ↔ P(p) + P(q) = 1

but that you cannot infer further any of these:

• P(p) = P(q)
• P(p) > 0
• P(q) > 0

edit: meant "exclusive or"

Could you imagine somebody speaking without involving a kind of illocutionary commitment? — Number2018

Specific commitments can be negated by the speaker, at least in many cases. (Moore's paradox an apparent exception.) But some commitment?

I doubt you can say anything that couldn't be taken as making some kind of commitment. And that's not irrelevant since having your words taken in a particular way is a key element of commitment.

I would lean toward speaking at all indicating a commitment to a shared framework of communication using language. The trouble though would seem to be defining speech here -- there are uses of word-sounds that aren't exactly speech, and not just among young children, the mentally ill, and parrots. Competent rational speakers use words for signaling too. So my idea appears to be circular. Hmmm.

Bingo. — JeffJo

Saw what you did there.

If we skipped all the preliminaries and just offered for sale, at the price of £10, envelopes advertised as containing "either £5 or £20" -- well, I'm guessing that would be illegal lots of places. There might be no envelope on offer containing £20, which would make this a simple scam.

If you add a guarantee that there's at least one envelope worth more than the price of the envelope, you're on your way to reinventing either the lottery or the raffle.

This is why I think talking about average returns over repeated games is a red herring. — Michael

Repetition is actually built into the game. You choose between a pair of envelopes, then you choose again between that same pair of envelopes. More repetitions could readily be added.

Your view is precisely that your chances of gain are better on the second choice than the first.

If you know only one value, you don't have enough information to prefer either sticking or switching. Flip a coin. That's what you did the first time you chose an envelope, and that's what you're doing now.