Thanks for asking. There are two things to address: Searle and the inconsistency bit.

First, Searle. Searle is saying, "X is Y, not Z" ("The proposition is the content of the belief, not the object of the belief"). Such a statement implies that it is in some way sensible or plausible, albeit incorrect, to say that X is Z. That is where I want to part ways with Searle. There is nothing sensible or plausible about reifying a proposition such that it would itself be the object of belief. That's not the sort of thing a proposition is. Hence Searle's conclusion seems to be invalid, namely his conclusion that beliefs are (usually) not propositional. In other words, I agree that reified propositions are not the object of belief, but I maintain that propositions are the object of belief. (It's hard to convey because the crucial difference is between two different conceptions of propositions.)*

But again, if Searle is merely rejecting an error of his contemporaries, then what he is doing is understandable. My critique here is admittedly a bit subtle. If someone reifies a proposition and makes it the object of belief, then we should not say, "Beliefs are not about propositions." Instead we should say, "You do not understand what a proposition is. When we talk about a proposition we are talking about a representation of 'objects and states of affairs in the world'." This is merely that "single puzzle piece" that nags at me in Searle's account. Whether it affects our larger conversation, I do not know.

Second, the inconsistency bit. Sorry - I had meant to edit that post because it is sloppy, but I forgot. The idea was that if logic cannot capture belief, and yet belief is subject to canons of consistency, then logic is not coextensive with consistent subjects.

Maybe a more formal example would be more helpful:

Doesn’t logic just set out what it is we can say, consistently? — Banno

It's been awhile since I studied mathematical logic, but I believe that Gödel, in his 1931 paper which contains his incompleteness theorems, showed in effect that there are arithmetic truths that logic (in the Principia) cannot prove, such that arithmetic is not a branch of logic as Russell had hoped. I can try to dig out some sources for this if you'd like. If that is right, then arithmetic is a consistent discourse that logic cannot fully model or contain.

In particular, what the belief is about is not shown by this analysis... — Banno

Do you think there is an analysis which shows what the belief is about? Do you think you would be able to draw up a deeper analysis that avoids the recursion problem I noted above?


* Edit: A proposition is like a mirror. Its whole purpose is to reflect something. When I stand in front of the mirror I am not looking at the mirror; I am looking at myself. More precisely, I am looking at a reflection of myself in the mirror. The mirror itself is not the object that my act of sight terminates in. So if I say, "I saw a bump on my eye when I was shaving," the listener will assume that a mirror (or something like it) was being used. There is no other way to look at my eye. If someone comes to believe that mirrors are things to be looked at, apart from their reflection, then they don't understand mirrors.

I would suggest that before settling on a formal representation we need a good understanding of the use of the word in the wild. — Ludwig V

We are supported in understanding the "use of the word in the wild" by formal analysis.

"X is Y, not Z" — Leontiskos

That does not strike me as an adequate parsing of what Searle is pointing out. Indeed, it would appear that you are making much the same claim as Searle, the one I expressed as that beliefs range over propositions but that the object of the belief is (usually) not the proposition, but the object of the proposition.

There are, of course, various formal ways to represent beliefs. As with any use of logic, the formalism you choose is dependent on what you wish to do. The point here is not that beliefs cannot be formalised, but that there are inadequacies in just using first order logic to do so.

arithmetic is a consistent discourse that logic cannot fully model or contain. — Leontiskos

Gödel showed that for any sufficiently advanced system (roughly, one that includes counting) there are true well-formed formulae of that logic that cannot be proved within that logic. Note the bolding. The unproven true well-formed formulae are part of the system. So I'm not sure that "arithmetic is a consistent discourse that logic cannot fully model or contain" is quite right.

And the mirror analogy just does not seem to work. Of course one can take a mirror and look at it, rather than what it reflects. One does so in order to clean it, or to check it for scratches. So with language, one can look to the logic of propositions in order to understand their structure.

That's what logic is for.

This conversation does not appear to be progressing.

Seems to be a formalism along the lines of the logic of belief revision?

That approach might be quite informative, although it bothers me that the set of beliefs is closed. And that it is dichotomous. It's different to human belief.

But interesting stuff.

While @Leontiskos and I quibble about whether belief can be formalised, folk such as yourself just go ahead and formalise it. Nice.

That does not strike me as an adequate parsing of what Searle is pointing out. Indeed, it would appear that you are making much the same claim as Searle... — Banno

It is quite possible that I am misunderstanding Searle or saying the same thing in different words. Thing is, I don't know whether this question about the nature of propositions is relevant. If you think it's relevant and want to look at it, we can. If not, we can let it slip back into the water.

To help get us back on track, would you be able to clarify this post? Could you draw out your claim that we cannot substitute salve veritate using your Superman example? Specifically, what is the substitution that would fail to save truth? I don't disagree with your claim, but I want to make sure I am clear on what you are saying.

Secondly, if you do not think beliefs can be represented by first order relations, then what alternative would you turn to?

Gödel showed that for any sufficiently advanced system (roughly, on that includes counting) there are theorems of that logic that cannot be proved within that logic. Note the bolding. The unproven theorems are part of the system. So I'm not sure that "arithmetic is a consistent discourse that logic cannot fully model or contain" is quite right. — Banno

Yes, but the reason we know the unproven logical sentences are true is because they correspond to (provable) truths in arithmetic via "Gödel numbers". Is the unproven sentence "part of the system"? The unproven sentence is represented in that system, but is not provable in that system. (Maybe my logic is rusty, but I am wary of calling an unprovable sentence a theorem.)

And the mirror analogy just does not seem to work. Of course one can take a mirror and look at it, rather than what it reflects. One does so in order to clean it, or to check it for scratches. So with language, one can look to the logic of propositions in order to understand their structure. — Banno

I don't disagree that we can examine propositions qua propositions, but if we think that is what propositions are, or are for, then I think we have made a mistake. (Very likely the hangups around propositions have to do with their ontology.)

So when Searle says, "Most [intentional states] are directed at objects and states of affairs in the world independent of any proposition," it seems to me that he is saying something like, "Most of the time when we use our rear-view mirror to see what is behind our car, we are looking at objects and states of affairs in the world independent of any mirror." Just as I don't think using a rear-view mirror can be done independent of a mirror, so I also don't think intentional states can be directed at the world independent of propositions. We direct our beliefs at the world through propositions in the same way that we view the world through a mirror. Searle isn't a direct realist, is he?

already alluded to the idea that if beliefs are not propositional, then it is hard to understand how they could be wrong. If my belief is directed at the world independent of any proposition, then how could I ever be wrong about what I believe?

P.S. Thanks for the SEP article. I didn't realize this topic was so well-explored.

I am wary of calling an unprovable sentence a theorem. — Leontiskos

Yeah, I have a bad habit of calling true well-formed formulae "theorems", which is incorrect. @sime has picked me up on it before.

Later.

Could you draw out your claim that we cannot substitute salve veritate using your Superman example? — Leontiskos

It's a standard example. Lous believes (knows, loves, is worried that...) Kent wears glasses. Kent is Superman. But Lous does not believe (know, love, worry that...)Superman wears glasses.

You must have seen it?

Right, I get that. But does Lois believe that Clark Kent is Superman? That's where the ambiguity arises for me. The falsity of the substitution is overdetermined. For example:

1. Lois believes that Clark Kent can type at 140 words per minute; Clark Kent is Superman; But Lois does not believe that Superman can type at 140 words per minute (because Lois does not believe that Clark Kent is Superman).
2. Lois believes that Clark Kent wears glasses; Clark Kent is Superman; But Lois does not believe that Superman wears glasses (and yet Lois does believe that Clark Kent is Superman).

In (1) the substitution is false because of an ontological fact of identity that Lois does not hold as a belief. In (2) the substitution is false merely because Superman dons a disguise, and not because of any lack of knowledge or belief on Lois' part. Your example could be interpreted either way, and yet in each case the substitution fails for a different reason.

Of course the substitution would remain true in 3:

3. Lois believes that Clark Kent can type at 140 words per minute; Clark Kent is Superman; Therefore Lois believes that Superman can type at 140 words per minute (because Lois does believe that Clark Kent is Superman).

We are supported in understanding the "use of the word in the wild" by formal analysis. — Banno

Mutual support? Interdependence? But perhaps not on topic here.

If my belief is directed at the world independent of any proposition, then how could I ever be wrong about what I believe? — Leontiskos

The standard way is to post an "intentional object" (in this case a proposition). But then, what's a proposition? The standard "meaning of a sentence" doesn't help much. I believe that Frege thinks it is something like thinking of a state of affairs without affirming or denying it. Are there no other proposals around.

The other way might be to push your metaphor a little harder and say that if a is directed at b, it can still miss.

Statements are combinations of nouns and verbs and such like; Some statements are either true or false, and we can call these propositions. So, "The present king of France is bald" is a statement, but perhaps not a proposition.

Beliefs range over propositions.

Beliefs are stated as an association between an agent and a proposition. This superficial structure serves to show that a belief is always both about a proposition and about some agent. It might be misleading as the proposition is not the object of the belief but constitutes the belief.

This association is such that if the agent acts in some way then there is a belief and a desire that together are sufficient to explain the agent's action. Banno wants water; he believes he can pour a glass from the tap; so he goes to the tap to pour a glass of water.

Notice that I am using the word "association" and not "relation", for reasons explained above.

Some folk hereabouts think something like that there are beliefs which are not propositional. It remains unclear to me how that could work. It's supposed that there are hinge beliefs that are in some way not propositional, but that is quite problematic, since hinge beliefs are also supposed to ground other beliefs by implication, and implication relies on propositions. If a hinge belief cannot be expressed as a proposition, it cannot be used to ground other beliefs.

Far better to adopt the grammar described above, with "beliefs" restricted to range over propositions.

I don't see a salient issue with (2) and (3). Roughly,

2. B(Lois, Kent (wears glasses)); Kent=superman; but  ◇~B(Lois, Superman(wears glasses)

3. B(Lois, Kent (types fast)); B(Lois,Kent=superman); ⊢B(Lois, Superman(types fast))

To my eye this sets out pretty clearly why substitution fails to preserve truth in (2) but not in (3): in (2) it need not be true that Lois believes superman is Kent. The various modal Doxastic Logics might be used to capture this more formally.

I must be misunderstanding you, since it seems you are saying that logic is inadequate to the task of dealing with beliefs, when it sets stuff out quite clearly.

Unless your point is that Lois might have inconsistent beliefs?

I dunno. Perhaps if we drag this back to your opening post. You proposed that to believe is to think with assent; I guess one might ask: to think what? If what you are thinking cannot be expressed as a proposition, is it a thought? Or is it better thought of as a sensation, a feeling, an impression, an intuition?

It seems worth making the point that parsing natural languages into formal languages is not a game of finding the one, correct, interpretation. Rather one chooses a formalisation that suits one's purpose.

Beliefs are stated as an association between an agent and a proposition. This superficial structure serves to show that a belief is always both about a proposition and about some agent. It might be misleading as the proposition is not the object of the belief but constitutes the belief. — Banno

The last sentence, in particular, nicely reconciles the issue about the object of the belief. There might be more to be said, but it is a good starting-point.

This association is such that if the agent acts in some way then there is a belief and a desire that together are sufficient to explain the agent's action. Banno wants water; he believes he can pour a glass from the tap; so he goes to the tap to pour a glass of water. — Banno

Quite so. "Association" is vague enough to enable it to stand up even when he cannot pour a glass from the tap (because the water has been turned off).

Some folk hereabouts think something like that there are beliefs which are not propositional. It remains unclear to me how that could work. It's supposed that there are hinge beliefs that are in some way not propositional, but that is quite problematic, since hinge beliefs are also supposed to ground other beliefs by implication, and implication relies on propositions. — Banno

That's at least partly about different uses of propositional forms. If I point to something red and announce "That's red.", I may be making an empirical observation, or teaching someone what "red" means, which case the object is a sample and to be used in a different way. That's only an example. There would be other cases to consider.

Unless your point is that Lois might have inconsistent beliefs? — Banno

Isn't that a problem, though? Any proposition will have a cloud of implications around it. There's no guarantee that I can see that Lois will have drawn all those implication or that she would instantly agree to all of them if they were presented to her, or that she will not draw any false conclusions from what she does believe.

It seems worth making the point that parsing natural languages into formal languages is not a game of finding the one, correct, interpretation. Rather one chooses a formalisation that suits one's purpose. — Banno

Your analysis of Lois' beliefs illustrates how useful formalization can be.

Or is it better thought of as a sensation, a feeling, an impression, an intuition? — Banno

I would rather say that sensations and feelings (when expressed in the grammatical form of a proposition) and impressions (ditto) can give rise to beliefs, rather than being beliefs. Not sure about intuitions. One can intuit that...

To my eye this sets out pretty clearly why substitution fails to preserve truth in (2) but not in (3): in (2) it need not be true that Lois believes superman is Kent. — Banno

Oh, that's quite right. You say that (2) fails to preserve truth, and this is undeniable. My question is whether you are saying (2) fails to preserve truth because Lois does not believe that Kent is Superman, or only because Superman uses a disguise?

This is why I am curious: When you say that first order relations fail to model belief because your example does not admit of substitution salva veritate, there are two different ways to understand your claim, because there are two different ways that your example does not admit of substitution salva veritate. I am wondering which of them you have in mind.

I must be misunderstanding you, since it seems you are saying that logic is inadequate to the task of dealing with beliefs, when it sets stuff out quite clearly. — Banno

Well, does (3) solve your salve veritate problem? Could we "substitute salva veritate if the equality relation was redefined to take into account belief"? Perhaps the belief relation is only referentially opaque when it prescinds from belief in the equality of the substituted term. Namely, once Lois believes that the substituted term is equal (B(Lois, Kent = Superman)), the substitution saves truth.

Unless your point is that Lois might have inconsistent beliefs? — Banno

No, I don't want to say that or go there. :grin:

I dunno. Perhaps if we drag this back to your opening post. You proposed that to believe is to think with assent; I guess one might ask: to think what? If what you are thinking cannot be expressed as a proposition, is it a thought? — Banno

To think a proposition. I agree with you that beliefs are about propositions ().

To clarify, I have been disagreeing with Searle when he says, "Most [intentional states] are directed at objects and states of affairs in the world independent of any proposition," and this goes back to my posts which compare propositions to mirrors. I have been positing a stronger notion of the propositionality of belief than Searle, not a weaker one.

It might be misleading as the proposition is not the object of the belief but constitutes the belief. — Banno

At this point I am inclined to think we are just talking past one another. When you or Searle say that a proposition is not the object of a belief, presumably you are not flirting with direct realism. Instead, you are saying that the object of a belief is not a proposition qua proposition, just as when I look through a mirror to see a reflected object my act of sight does not terminate in the mirror itself. Yes?

The standard way is to post an "intentional object" (in this case a proposition). But then, what's a proposition? The standard "meaning of a sentence" doesn't help much. I believe that Frege thinks it is something like thinking of a state of affairs without affirming or denying it. Are there no other proposals around. — Ludwig V

I think I understand what Searle is saying now. The clause, "...independent of any proposition," felt strange at first, but probably he is saying that the subjective act of belief prescinds from notions of propositionality or representation. ...And I have no problem with Frege's account of a proposition.

It is curious, though, that 'belief' insofar as it is distinguished from knowledge really is propositional in the way that Searle is talking about. If I say, "I believe X but I do not know X," then apparently there is an intentional propositionality, and one which is much more common than Searle's example of Bernoulli's principle.

the object of a belief is not a proposition qua proposition, just as when I look through a mirror to see a reflected object my act of sight does not terminate in the mirror itself. — Leontiskos

Another metaphor, but still, it works for me.

the subjective act of belief prescinds from notions of propositionality or representation. — Leontiskos

I think this is OK. Where do we go from here?

It is curious, though, that 'belief' insofar as it is distinguished from knowledge really is propositional in the way that Searle is talking about. If I say, "I believe X but I do not know X," then apparently there is an intentional propositionality, and one which is much more common than Searle's example of Bernoulli's principle. — Leontiskos

We are saying roughly that believing uses a proposition rather than mentioning it. But we are also saying that it is possible to believe something of a proposition - second order belief, presumably based on a proposition about a proposition.

"I believe X but I do not know X" seems to take back or modify in the second half ("..but I do not know X") something that is asserted in the first half ("I believe X"). But it isn't a flat contradiction. So I think you are saying that "I believe X but I do not know X" expresses a view about the certainty of, or evidence for, X - that certainty is less than complete, or that evidence is less than conclusive.

Is that something like what you meant?

The scope of the belief statements surely makes explicit your quibble? The difference is between "Kent is Superman" and "Lois believes that Kent is Superman".

I'll try one more time. Suppose we have some agent L a nd some possible state of affairs which we might present as either p or as f(a); that is, p ≡ f(a). Then that L believes this state of affairs as B(L,p); and for some purposes this will suffice. But the problem here is that B(L,p) looks like a first-order relation between the agent and the state of affairs, and this is not so. "p" is not the thing the belief is about. We might get to that thing by parsing the believe as B(L, f(a)). The belief is that f(a), a proposition, that is about the individual a.

And the quote form Searle seems to me to be making that point; that B(L,p) is somewhat inadequate, and that the belief is about the individual named "a"

@Sam26, @creativesoul, Searle is not saying that beliefs are not propositional. Beliefs range over propositions.

We are saying roughly that believing uses a proposition rather than mentioning it. — Ludwig V

Yep.

But nup. Davidson's analysis.

Yep.

But nup. Davidson's analysis. — Banno

That's extremely interesting. But I don't understand it. Could you give me a reference for Davidson's analysis?

Sam26, creativesoul, Searle is not saying that beliefs are not propositional. Beliefs range over propositions. — Banno

Where did I say that?

I think the main source is On Saying That.

Edit: The idea is something like that we sometimes both use and mention; SO "Galileo said that the Earth moves" might be analysed as a conjunct of "The Earth moves" and "Galileo said that", where the demonstrative "that" points to "The Earth moves", or even to Galileo's utterance of "The Earth moves".

But that'd be it's own thread.

You may not have. But at some stage you have left me with the impression that you think there are beliefs that cannot be put into the form "X believes that P" where P is some statement. Whereas I, and I think most folk who have given the issue some consideration, might say that this is part of the grammar of belief.

So to believe something is to believe that something to be true; and what is true is this or that state of affairs, this or that statement.

I hope I have misunderstood.

Edit: Something to do with hinge beliefs not being propositional?

What I do say, is that beliefs are not necessarily a product of statements such as, "X believes that P." Those are only beliefs that are part of the language of statements. The fact is that beliefs can exist quite apart from any linguistic expression of that belief. You are starting with language and working backwards. I start prior to language and work forward. So, given my understanding, prior to language there were still beliefs. These beliefs were shown in the actions of those who had the belief. The act is prior to language, then comes language, where we are able to express the belief. In my estimation you and others are putting the cart before the horse.

The other point is, even where language exists, many beliefs (those expressed in our nonlinguistic acts) are never put into statements, but that doesn't mean the belief doesn't exist as part of the act that shows the belief (the act of opening a door shows my belief that a door is there, apart from whether it's stated or not). The belief doesn't pop into existence just because the belief is stated. It can, but not necessarily.

What I do say, is that beliefs are not necessarily a product of statements such as, "X believes that P." — Sam26

Sure. It's not a product of such statements. The statement sets out an aspect of the grammar of belief, as between an agent and a proposition.

Those are only beliefs that are part of the language of statements. — Sam26

I maintain beliefs can be stated.

The fact is that beliefs can exist quite apart from any linguistic expression of that belief. — Sam26

Sure. Beliefs can be shown as well as stated. But they can also be stated. Note also the word "exist' here, and the implicit hypostatisation. When one says that a belief exists, what more is one saying, apart from that thinking the world is such-and-so accounts for this behaviour... the beetle is in this box, but you still cannot see it, yet you can talk of it existing. Nothing is brought into existence here.

so, given my understanding, prior to language there were still beliefs. These beliefs were shown in the actions of those who had the belief. The act is prior to language, then comes language, where we are able to express the belief. In my estimation you and others are putting the cart before the horse. — Sam26

Yes, I understand that you see it this way. But in the end all you have are the actions - both verbal and non-verbal; never the belief. You infer the belief from the act, beetle from the box.

the belief doesn't pop into existence — Sam26

Yep. Stop there.

Sure. It's not a product of such statements. The statement sets out an aspect of the grammar of belief, as between an agent and a proposition. — Banno

Yes, but that's not my only point. I'm pointing out that there's the agent and the proposition, but there is also those acts that show the belief. You keep reiterating that beliefs can be stated, no one is disputing this, that's obvious. And it's just as obvious that beliefs can be shown in animals, pre-linguistic man, and in modern man, apart from language. You seem to agree with this. However, then you say...

Note also the word "exist' here, and the implicit hypostatisation. When one says that a belief exists, what more is one saying, apart from that thinking the world is such-and-so accounts for this behaviour... the beetle is in this box, but you still cannot see it, yet you can talk of it existing. Nothing is brought into existence here. — Banno

I've addressed this before, and you keep trying to put my account of belief into Witt's beetle in the box. I would agree if my account was limited to the subjective, viz., pointing to the thing in the mind as if that's the belief. Nowhere do I do that, that's simply your interpretation of what I'm saying, but it's not accurate. My account of belief is based in nonlinguistic actions (showing the belief), and in the statements that an agent makes about their belief or the beliefs of others. So there's no beetle in the box. The belief exists as a function of those two kinds of acts. So minds do bring into existence the proposition and the nonlinguistic acts that show one's belief. Exists, as I'm using the concept, refers to those things we do, both linguistic and nonlinguistic, that can be said to be beliefs. Again, so the ontology of belief refers to those things minds do in the world that can be said to be beliefs.

The meaning of the concept belief is a function of what we do in language (language-games and forms of life), it's not a function of something we can point to in the head/mind. Wittgenstein was pointing out that the meaning of a concept is not something internal to us. My account of what we mean by belief nowhere suggests that it's the thing in the mind that gives meaning to belief. Thus, the beetle in the box doesn't apply to my account. Even the way I'm using the word existence depends on the external, not the internal.

:rofl:

Around and around. Seems to me you talk as if the belief is something more than the behaviour, existing beyond that, until I push the point, then you agree with me that it isn't.

Oh, well.

It seems you're pushing me in that direction. Maybe it's that I put more importance on the mind, but that doesn't mean that meaning resides in the mind. My beliefs about consciousness might push you in that direction, I'm not sure.

:wink: Cheers.

:up: