Well, let's not kill the messenger. Davidson approaches the discussion in this way both because it is his area of interest, and because it presents a common ground for disparate branches of conceptual relativism. — Banno

Yes, I'm aware of that claim. That's the reason I brought up Kuhn and Quine as Davidson's targets. They certainly didn't think that way. Kuhn actually says "translation, if pursued, allows the participants in a communication breakdown to experience vicariously something of the merits and defects of each other’s points of view". Basically, I'm saying none of Davidson's targets equate non-translatability with incommensurablility, so I think Davidson needs to make a clearer case as to why he thinks that are the same if he is to attack Quine, Kuhn etc. If he's not opposing their positions, but instead merely presenting an alternative framework for looking at apparent differences, then he need make no such connection, but that's not what he's doing here, and that's not the place you're trying give his conclusions either.

Well, do you believe that the cat is on the mat? then you will hold "the cat is on the mat" to be true. And if not, then I suppose you will not. — Banno

As I've been trying to explain, I don't see belief in binary terms, belief is a probability statement. I have a degree of certainty that the cat is on the mat. Not believing it then simply doesn't make sense, Not believing it is just believing some alternative, contradictory thing, with a greater degree of certainty.

The point of this line of questioning though, is to find how Davidson makes the leap from Tarskian truth theories in formal languages, to truth theories in the real world (the content of Fit-Reality type conceptual schemes). He seems to me to do this in a rather 'hand-waivy' kind of way claiming that we all have fundamentally similar beliefs and we have to assume as much to talk to each other, such that the meaning of the meta-language's 'the cat is on the mat' is a statement about the way the world is (which is where he breaks from Tarski). It is this 'way the world is' that Davidson seems to want to invoke as a sufficiently general level of agreement that all conceptual schemes are fitting/organising.

The problem is, I don't think that description fits the two important areas of conceptual schemes that are being attacked here - complex matters of science, and fundamental level schemes of cognition/perception. I don't see any evidence fo the similarity of beliefs here that would bring Tarski out of formal languages into the real world.

A side issue - I think this is wrong, in that it gives primacy to strings over cats. Both are perhaps real. — Banno

Perhaps both are, but one is built from the other. If we are realists (and I'm presuming we are) then the only thing we're committed to axiomatically is an external reality. That reality has to be heterogeneous, otherwise we would be making up the variations of cats, people, trees, etc..., and we're back to idealism. But that's it. That's all that realism alone commits us to. To say that some collection of reality is a 'cat' and some other collection is 'not cat' requires some justification for putting the boundaries there and not elsewhere. Some justification for saying objects are determined by their form, not say, their function, or their energy fluxes, or any other aspect of reality we could have chosen. Those justifications are conceptual schemes (in the Quinean sense) and they cause the objects to be, they're not about cats and mats, they about sensory inputs, and such inputs are not shared, they vary according to the conceptual scheme, they also are often pre-linguistic, thus one cannot apply Tarskian truth theories to them.

Where? It's just that the you use notation may differ from that in the article. Is a, for instance, an individual? — Banno

I see where we've got crossed wires. The first presentation is from Ian Rumfitt's "Ramsey on truth and meaning", it's a n expression Rumfitt uses, I though might be more succinct a way of expressing Ramsey's later thought from Ramsey unpublished work "On Truth". The actual version is

"B is true :=: (∃p). B is a belief that p & p."

My later comment was in reference to the description of belief in terms of certainty which is indeed in 'Truth and Probability'. Two different parts of Ramsey's thinking which I'm bring together here. Apologies for any confusion there.

Let Jenny be out 2-year-old. there are four possible beliefs she might have:


Jenny believes that other people have minds
Jenny does not believe that other people have minds
Jenny believes that other people do not have minds
Jenny does not believe that other people do not have minds


some of these can be paired up consistently, others, not without contradiction. You perhaps have taken be as asserting the second option; but I wish to assert both the second and the last - that is, that Jenny has no beliefs about the minds of other people. — Banno

I agree, Jenny has no beliefs about the minds of other people, she doesn't even have the concept 'minds of other people' her world is one of freely accessible knowledge about her world. So 'teaching' doesn't make any sense to Jenny (how can someone give her information, where would it have come from?), different emotions in other doesn't make sense to Jenny (how could others feel differently about things?)...

Whole fields of knowledge simply have no correlates in Jenny, no way of making her see what the mind-theorist sees. Jenny's difference in beliefs means that she has no 'other minds' in her world view for anything to even be about.

In the language of the simple world (using Davidson's example), she has no shoes in her closet to re-arrange. The problem Davidson is trying to dissolve is that we seemingly cannot say to her "Oh, I arrange shoes this way", by way of 'translating' (and therefore dissolving) our different conceptual scheme, because Jenny will just reply "What on earth are 'shoes'?".

But, in the simple world, we can explain what shoes are in terms that Jenny does understand and in this respect Davidson is exactly right - we'd been duped into thinking Jenny had a different conceptual scheme becasue she didn't have 'shoes' in her closet, but really all that had been the case was that we'd though we were arranging 'shoes' when actually we were arranging 'bits of leather' and 'feet'. Now Jenny knows what shoes are in those terms - perfect.

Except... going back to minds, it doesn't work. What are the underlying constituent parts common to both Jenny's world and ours? Jenny just has to feel that there are other minds, she has to develop the actual neural networks which alter her actions to behave as if there were other minds. It's not a linguistic matter. The same is true of all fundamental cognitive models, all complex scientific models, anything where we have not got good cause to assume shared beliefs are constituting the models.

Thanks for the clarification.

"B is true :=: (∃p). B is a belief that p & p." — Isaac

This seems to still fall to my counterexample, that there can be truths that are not believed.

. Basically, I'm saying none of Davidson's targets equate non-translatability with incommensurablility... — Isaac

Hmm. I think you are going out on a limb here.

See for instance the Stanford article on Kuhn, which treats translation at length.

I'll grant that Kuhn emphasises the psychology of science while Davidson emphasis the language.

As I've been trying to explain, I don't see belief in binary terms, belief is a probability statement. I have a degree of certainty that the cat is on the mat. Not believing it then simply doesn't make sense, Not believing it is just believing some alternative, contradictory thing, with a greater degree of certainty. — Isaac

And do you claim the same fro truth? Is it subject to degree?

So conceptual schemes are for you psychological as against linguistic, and hence you take it that translation is irrelevant.

Well, that strikes me as confused. I cannot see how you can sensibly divorce one from the other. Conceptual schemes are as much about cats and mats as they are about sensory inputs. There need be no justification between what counts as cat and what as mat.

What are the underlying constituent parts common to both Jenny's world and ours? — Isaac

Directly perceptible things... common referents(says Davidson).

Except... going back to minds, it doesn't work. What are the underlying constituent parts common to both Jenny's world and ours? Jenny just has to feel that there are other minds, she has to develop the actual neural networks which alter her actions to behave as if there were other minds. It's not a linguistic matter. The same is true of all fundamental cognitive models, all complex scientific models, anything where we have not got good cause to assume shared beliefs are constituting the models. — Isaac

Another good post.

Kuhn, as I understand him, would deny the commonplace that Newtonian physics is a subset of relativistic physics, accurate only within certain limits; Kuhn would say that the difference is so great that terms such as mass and velocity stand for different and incommensurate concepts. (Checking, this is the case presented in the Stanford article, so I'm not on my own in this view).

Now Davidson as I read him is saying that such an approach will not work, since there is common ground between, say, mass-for-newtonians and mass-for-relativitics.

Understanding relativistics gives us an insight into mass; we have a better grasp of how it behaves. Something is gained by adopting the relativistic description.

SO would one say that in learning relativistic physics, one just has to feel that mass changes with velocity? In a sense, yes, but the feeling would be worthless without the equations. The feeling won't do on its own.

Ans so with Jenny. The feeling will not suffice. She demonstrates her understanding of other minds by changes in language and behaviour.

SO I think I can agree with what you are saying while maintaining that it's the language and associated behaviour that really count.

It's the word "underlying" that I object to; it suggests some sort of fundament were there is none.

I'm not angry. — Janus

Well I am. I'm pissed at you for levelling personal insults at me. It demeans the discussion.

So 'teaching' doesn't make any sense to Jenny (how can someone give her information, where would it have come from?), different emotions in other doesn't make sense to Jenny (how could others feel differently about things?)... — Isaac

The teaching would be to set up situations that challenge Jenny's existing behaviour, situations in which it is beneficial for her to acknowledge that someone else may have a different perspective. Perhaps the strongest challenge to Jenny is deception. Here we are teaching her that other folk may have beliefs that differ from her own.

The reason theory of mind is of interest is that it is absent in some folk, and as a result they cannot behave as expected.

Davidson, in discussing partial incommensurability, describes it as differences of belief, not of conception. I gather that you do not think this applies in Jenny's case, since she has no beliefs about other minds. I'm suggesting that the situation is better described as a difference in belief, since that allows us to challenge the erroneous beliefs and hence to help Jenny build a theory of mind; in a way that simply saying "she lacks the concept..." does not.

I am. I'm pissed at you for levelling personal insults at me. — Banno

:halo: :halo: :halo:

I've thought long and hard about whether or not there are such things as incommensurate schema as compared to non translatable ones. Ultimately, I think it comes down to whether or not the referents within any given framework are directly perceptible or not. I'm leaning towards the conclusion that some cannot be effectively translated one into the other without losing crucial meaning...

It demeans the discussion. — Banno

The level at which you discuss (with me at least) is already demeaned. But don't worry too much about it, since you won't answer straight questions I've lost interest.

It's the word "underlying" that I object to; it suggests some sort of fundament were there is none. — Banno

Yes, and in line with Davidson's rejection of an uninterpreted world(and yours too, if I understand you correctly). That is - it seems to me - a major underlying difference between our views.

:wink:

Convention T, redundancy, and deflationary accounts focus upon "is true". Truth as predicate. Correspondence theory holds that truth is correspondence.

"B is true :=: (∃p). B is a belief that p & p." — Isaac

I'm skeptical that beliefs can play any sensible part in a definition of truth simpliciter. Though I can read what you have written as a definition of a true belief but not necessarily a truth. Specifically, a true belief is a belief in a truth.

One reason for my skepticism is that beliefs are a modal and truths are not. For an example.

Let's take the definition that "B is true" iff "B is a belief that P and that P":

(A) There are less than 3 bodies in orbit around Saturn at the moment.
(B) There are exactly 3 bodies in orbit around Saturn at the moment.
(C) There are more than 3 bodies in orbit around Saturn at the moment.

Let P = (A) or (B) or (C), the disjunction of all three of them. P is true since it exhausts all possible cases of the number of bodies in orbit around Saturn. I will believe that P.

This entails that (A) is true, or that (B) is true, or that (C) is true. By the above definition, this entails that (A) is a belief or that (B) is a belief or that (C) is a belief. But I don't believe in any of them, I simply believe in the disjunction. Surely, then, there must be three different people who exist in order for the disjunct P to be true. If not, it would be strange that (P) could be believed but none of its elements need be believed in order for (P) to be believed (IE, one of the disjuncts needs to be true, but it doesn't seem to need to be believed in order for it to be true).

The situation gets a bit worse:

(1) There are 0 bodies in orbit around Saturn.
(2) There is 1 body in orbit around Saturn.
(3) There are 2 bodies in orbit around Saturn
...
(n ) There are n bodies in orbit around Saturn.
...

Continue on for all natural numbers. Then the disjunction (1) or (2) or ... is true and believed, but the beliefs are mutually exclusive. Then there are arbitrarily many believers? Or do all these believers necessarily not believe that the list elements (1)...(n) are mutually exclusive?

I think the root of these pathologies is that belief doesn't distribute over disjunctions (a consequence of being a modality). If you make belief a component of truth, then belief in some sense must distribute over a disjunction.

As for beliefs necessarily being probability distributions assigned to sets of statements, this is also quite contentious, there's no probability distribution that assigns indifference to the list (1)...(n) even when there is no information about (1) to (n ) (equal mass assignments either are all 0 and so the measure doesn't sum to 1 or equal mass assignments sum to infinity).

Yes, but the idea behind those accounts is to indicate why the "is true" holds. And the answer is that the "is true" holds if the proposition describes the facts. "The cat is on the mat" is true iff the cat is on the mat. This formulation which is known as the T-sentence is showing us that the statement "The cat is on the mat" is true if and only if it describes the fact that the cat is on the mat. And it does describe this fact if the cat is on the mat.

Now I've used the word "describes", but the alternative "corresponds to" means exactly the same as far as I can tell.

I don't think of correspondence as theory, but merely as a descriptive account of what we mean when we say a statement or proposition is true; the very same thing that the T-sentence is meant to show in formal language.

"B is true :=: (∃p). B is a belief that p & p." — Isaac


This seems to still fall to my counterexample, that there can be truths that are not believed. — Banno

If you don't mind, I've tried to deal with the truth/belief issues in my reply to @fdrake below, so hopefully I respond to this there are it will save me a bit of typing if you read that first and see if it answers your concern (or makes it worse!).

I'll grant that Kuhn emphasises the psychology of science while Davidson emphasis the language. — Banno

Yes, again this touches on my approach throughout this (although it's Ramsey's approach originally, I'm not claiming original thoughts here). We have only the human mind, which must have primacy over language. There's no doubt the one arises out of the other, there's no mutual standing in causation. Learning German doesn't alter one's thinking (much), but both physical and mental trauma to the mind can affect language. I know it's a bit territorial coming from a psychologist, but I'm in no doubt that the human mind is as far back as we can go in our understanding of the world (everything we know/understand is undoubtedly formed by, and contained within, it, and we can go no further than it because we cannot get outside of it). So I have a lot of sympathy for a linguistic approach, especially when it comes to philosophical problems, but psychology takes precedent, for me.

And do you claim the same fro truth? Is it subject to degree? — Banno

Yes, in some contexts I think we'd have to say that. Is it true that I like whiskey? - well, partly. I like some whiskeys and not others, there are occasions where I don't fancy a whiskey (can't think of any right now, maybe a bad example), I didn't like whiskey when I was seven...I think we'd have to say either that "I like whiskey" is simply not well formed enough to admit of truth vales (which gets us back to Tarski's limitation), or we'd have to admit that it is partially true.

I cannot see how you can sensibly divorce one from the other. Conceptual schemes are as much about cats and mats as they are about sensory inputs. There need be no justification between what counts as cat and what as mat. — Banno

Why do you say there need be no justification? I mean, with cats and mats, there's nothing to dispute, but that's not the target here. It's 'gravity, 'phlogiston', 'ether', 'strings', 'atoms', 'ecosystems', minds', 'selves', ...I don't really understand how you can say that there's no need for any justification as to what counts as any of those things. The justification seems to be entirely the point of any theory about them, and opponents of those theories simply deny their existence. I understand Davidson would point to translatability as an indicator of some more atomic objects which are being 'organised' into these 'conceptual objects', or he might point to a difference in belief about them within a network of shared beliefs (the key term being 'about them'), but I don't see that this makes them commensurable.

Directly perceptible things... common referents(says Davidson). — creativesoul

This presumes that the creation of a referent is never axiomatic, and I don't see where Davidson would be getting such a assumption from. The creation of objects of perception out of the raw sense data is not only entirely possible, but, given modern neuroscience, seem highly probable. No pre-existing - more atomic - objects of perception that they are built from, just constructed from raw data to which we (consciously, and so linguistically) have no direct access.

Ans so with Jenny. The feeling will not suffice. She demonstrates her understanding of other minds by changes in language and behaviour.

SO I think I can agree with what you are saying while maintaining that it's the language and associated behaviour that really count. — Banno

Really count for what, though? I completely understand the pragmatic issue with trying to get at non-linguistic concepts. 'Other minds' is a public concept and Jenny can only show she's 'got it' publicly - language and behaviour. I get that (though, as above, I'd place a lot more emphasis on behaviour - in fact I personally see language as just a type of behaviour, but that's another thread). But if we then remove the notion that these public concepts have themselves been derived, constructed, by some process, where does that leave cognitive science? Similarly, if we treat all public concepts this way, even those which are widely disputed, how can we frame such disputes? Jenny might have to show she's 'got the rule' publicly, but that does not constrain us in talking about 'the rule' in other terms. We needn't reify it simply because we become necessarily bound to it in each context.

Davidson, in discussing partial incommensurability, describes it as differences of belief, not of conception. I gather that you do not think this applies in Jenny's case, since she has no beliefs about other minds. I'm suggesting that the situation is better described as a difference in belief, since that allows us to challenge the erroneous beliefs and hence to help Jenny build a theory of mind; in a way that simply saying "she lacks the concept..." does not. — Banno

This is a very persuasive argument. A pragmatic one, I'll insist, not an ontological argument for other minds. But I think I'd have to concede that talking this way is more useful as an approach to helping people change erroneous beliefs, and the same would be true of your previous example of relativistic physics. I'm not sure a pragmatic way of improving people's beliefs is the sole remit of Davidson's paper, but I certainly find myself agreeing with your line of thought here. Nicely put.

OK - here's my interpretation of Ramsey's answer to that problem.

Firstly, it's important to note the way Ramsey opens the section on logical consistency(what we're dealing with here) in Truth and Probability - He says...

the ideally best thing is that we should have beliefs of degree 1 in all true propositions and beliefs of degree 0 in all false propositions. But this is too high a standard to expect of mortal men, and we must agree that some degree of doubt or even of error may be humanly speaking justified.

What he's saying here is that the description of beliefs must be first and foremost psychological. To put it another way - we cannot, no matter how hard we squeeze, get any more out of 'the cat is on the mat' than that I believe the cat is on the mat. Ramsey is not saying that truth's are beliefs because of some logical deduction about truths. He's saying so because the notion is contained within a mind and there are limits that places on what it can possibly be.

So, to the objections...

Let's take the definition that "B is true" iff "B is a belief that P and that P":

(A) There are less than 3 bodies in orbit around Saturn at the moment.
(B) There are exactly 3 bodies in orbit around Saturn at the moment.
(C) There are more than 3 bodies in orbit around Saturn at the moment.

Let P = (A) or (B) or (C), the disjunction of all three of them. P is true since it exhausts all possible cases of the number of bodies in orbit around Saturn. I will believe that P.

This entails that (A) is true, or that (B) is true, or that (C) is true. By the above definition, this entails that (A) is a belief or that (B) is a belief or that (C) is a belief. But I don't believe in any of them, I simply believe in the disjunction — fdrake

In order to believe in the disjunction, it is necessary that you believe in all of them. The disjunction is a tautology of your belief in the three parts.

For Ramsey, a belief that p is a disposition to act as if p, and here speech is taken as an act. So if you hold the disjunction to be true, if someone were to ask "is it possible that there are less than 3 bodies in orbit around Saturn at the moment?", holding the disjunction to be true compels you (all other matters being equal) to answer "yes". So you do have a belief, in some degree, in each of the three options. Without such beliefs you could only justifiably believe the abstract logical truth of disjunction, not the specific one regarding Saturn and its moons.

My grasp of modal logic is not that great, so it may well be that I'm missing something crucial in what you're saying, but if not, then it seems that what we're talking about is a difference in what constitutes a 'belief' rather than anything else.

As for beliefs necessarily being probability distributions assigned to sets of statements, this is also quite contentious, there's no probability distribution that assigns indifference to the list (1)...(n) even when there is no information about (1) to (n ) — fdrake

Ramsey says...

the Principle of Indifference can now be altogether dispensed with; we do not regard it belonging to formal logic to say what should be a man's expectation of drawing a white or a black
ball from an urn; his original expectations may within the limits of consistency be any he likes; all we have to point out is that if he has certain expectations he is bound in consistency to have certain
others. This is simply bringing probability into line with ordinary formal logic, which does not
criticize premisses but merely declares that certain conclusions are the only ones consistent with
them.

I take this to mean that indifference is acceptable within bounds, but that your belief in the disjunction would act as just such a bound.

↪creativesoul Yes, but the idea behind those accounts is to indicate why the "is true" holds... — Janus

Is it?

Seems to me that some argue that it's superfluous/redundant, and others show/argue what it means.

....it seems that what we're talking about is a difference in what constitutes a 'belief' rather than anything else. — Isaac

I think that this is far more enlightening an observation that may seem at first glance, or at least, it could be...

Directly perceptible things... common referents(says Davidson).
— creativesoul

This presumes that the creation of a referent is never axiomatic, — Isaac

I don't think so. Rather, it acknowledges that some are not. Whereas, you seem to be taking a hard line stance that we have no direct access to any referents at all; Have I misunderstood?


----------------------------------------------------------------------------------------------------------------------


Truth is existentially dependent upon thought and belief. Our awareness of the role that truth plays in our thought and belief requires language. Belief does not. Truth does not. Meaning does not.

The problems - self-imposed - arise when we completely divorce truth and meaning from thought and belief.

My criticism remains the same. When we say that some statement is true, we do not mean the same thing as when we say that it is believed. The view you espouse denies this, and hence is wrong.

Specifically, there are things that are not believed and yet true. You are obliged to deny this in order to maintain your view.

Is it?

Seems to me that some argue that it's superfluous/redundant, and others show/argue what it means. — creativesoul

If it was "superfluous/ redundant" then there would be no need to mention it in the first place. The T-sentence is designed to show the logical conditions for the "is true". Despite my repeated questions, no one has managed to show how the logic of correspondence is any different than the logic of the T-sentence. If there is a difference, how come no one, apparently, can explain what the purported difference is. All I've been getting is vague objections.

Despite my repeated questions, no one has managed to show how the logic of correspondence is any different than the logic of the T-sentence. I — Janus

What is it in the T-sentence that you think corresponds, and what does it correspond to?

If it was "superfluous/ redundant" then there would be no need to mention it in the first place. — Janus

That is some weird logic Janus. Becoming aware that some language use is superfluous/redundant requires it's use in the first place. "Is true" is one such use. It adds nothing meaningful to a belief statement.

Try rewriting the T-sentnce without the "is true" or some analogue, then.

'P' IFF P

The "Snow is white" corresponds to snow being white if snow is white. You could also say it successfully describes snow being white if snow is white. A true statement successfully describes actuality which is the same as to say that it corresponds with or to actuality. That's the logic. Does this mean that there will only be the statement 'P' if there is the actuality P? Or?

I'm more than sympathetic to the need for any theory of truth to take proper account of correspondence. I do not find it(convention T) compelling if correspondence is left in the wind. However, there are significant problems with attempting to combine the two. It's all beside the point of the thread however... which is about translation of seemingly incommensurate conceptual schema.

I feel separating philosophy form systems analysis and design is intellectual suicide. I'll read Davidson's article later. At this point i would say i agree with Davidson.