What is the difference between learning the meaning of a word and learning to use the word? — Banno

Meaning of a word is in conceptual level. Using a word is in application level. They are different.

So it should be no trouble to set out that difference with a bit more than a mere label.

How do you tell that someone has the "concept" red?

By seeing how they use the word, and what they do in the world with red things.

So again, what more is there to understanding the concept "red" than being able do stuff with red things...


Welcome to the wonderful world of Wittgenstein.

Not everyone agrees with Wittgenstein. If you don't know what red means, how could you use the word "red"? If you are a colour blind, how could you tell red objects? For you being able to use the word red means that you know what you mean by red from your experience of seeing red via your perception, and folks describing red objects as red.

Not everyone agrees with Wittgenstein. — Corvus

So what.

Think on it some more. Colour blind folk do use the word "red" correctly - how can that be?

So what. — Banno

Wittgenstein was wrong.

Colour blind folk do use the word "red" correctly - how can that be? — Banno

They could be using the word red metaphorically, rhetorically or idiomatically to mean something other than the colour red such as red tape, redline, red-light district.

Think on it some more. — Banno

... and Red herring, red meat of course.

So now you are making use of a private rule... this (indicating the qual) is red...

And how can we say that what you pick out by "this" is the same as what I pick out?

I don't see that question as having any significance. — Banno

This seems to conflate several issues. Why is my description of my red quale a private rule? What would be the (correct, presumably) use of a public rule to describe the quale? I'm not seeing the alternative.

As for whether you and I are naming the same quale, wouldn't the answer be: Conceivably we aren't, but it's unlikely, given how color names are learned. And again, even if our meanings turned out to be different, it would have no bearing on whether we intended rather than extended (so to speak). When I point to the red beads and call them "red", this has nothing to do with their extension. Because, as above, the same beads might be green, with no difference in extension. If the extension were all that mattered, how would I know if they were red or green? So what then is the difference?

What I'd really like -- what I think would help most for me to see this picture -- would be to hear your alternative account of how, for instance, we can label "red" without allowing that "red" means that color (or quale). And, anticipating you, if that account involves learning how a word is used, what is the "extensional version" of that? What do we teach a child when we teach them color names? "When you point to that, say 'red'?" And if the child replies, "Why?" what do we say?

Not everyone agrees with Wittgenstein. If you don't know what red means, how could you use the word "red"? If you are a colour blind, how could you tell red objects? For you being able to use the word red means that you know what you mean by red from your experience of seeing red via your perception, and folks describing red objects as red. — Corvus

Quine, in Word and Object, addresses how folk who are color blind use the word correctly.

“Uniformity comes where it matters socially; hence rather in point of intersubjectively conspicuous circumstances of utterances than in point of privately conspicuous ones. For an extreme illustration of the point, consider two men one of whom has normal color vision and the other of whom is color-blind as between red and green. Society has trained both men by the method noted earlier: rewarding the utterance of ‘red’ when the speaker is seen fixating something red, and penalizing it in the contrary case. Moreover the gross socially observable results are about alike: both men are pretty good about attributing ‘red’ to just the red things. But the private mechanisms by which the two men achieve these similar results are very different. The one man has learned ‘red’ in associate with the regulation photochemical effect. The other man has learned ‘red’ by light in various wavelengths (red and green) in company with elaborate special combination of supplementary conditions of intensity, saturation, shape, and setting, calculated e.g. to admit fire and sunsets and exclude grass;…”

Wittgenstein was wrong. — Corvus

Well if your are to convince me of this I'd first have to be convinced that you understood Wittgenstein.

They could be using the word red metaphorically... — Corvus

... and so on. If I ask for the red pen, and they hand me the red pen, that's not metaphorical, nor is it merely rhetorically, and it certainly isn't idiomatic. It's pretty much literal and extensional.

This seems to conflate several issues. Why is my description of my red quale a private rule? What would be the (correct, presumably) use of a public rule to describe the quale? I'm not seeing the alternative. — J

Quite so; if someone referred to "the sensation S" for themselves alone, then the qual is private; and if they do it for others, isn't it just the colour red? Here's the problem with qualia: if they are private, then they are outside of our discourse, and if they are public, they are just our common words for this or that.

? What do we teach a child when we teach them color names? "When you point to that, say 'red'?" And if the child replies, "Why?" what do we say? — J

Rather we play with them, ask for the red block, offer them a lolly - but only the red one, and so on. We teach them to use the word. Then there is no "why?" as the task is of forthright interest.

As for whether you and I are naming the same quale, wouldn't the answer be: Conceivably we aren't, — J

...and it doesn't matter!. Becasue what counts here is the use!

That is, and here I'm grossly overgeneralising, the extension.

Well if your are to convince me of this I'd first have to be convinced that you understood Wittgenstein. — Banno

If one says one can use words without knowing its meanings, then he is wrong, whoever he is.

... and so on. If I ask for the red pen, and they hand me the red pen, that's not metaphorical, nor is it merely rhetorically, and it certainly isn't idiomatic. It's pretty much literal and extensional. — Banno

They must have been acquainted with something other than "red" to be able to do that by habit or guessing. That doesn't mean they know what "red" is. Their use of "red" could be based on the high chance of fluke guessing.

Nice.

...and to admit lobsters only after boiling.

Different persons growing up in the same language are like different bushes trimmed and trained to take the shape of identical elephants. The anatomical details of twigs and branches will fulfil the elephantine form differently from bush to bush, but the overall outward results are alike. — Word and Object

If one says one can use words without knowing its meanings, then he is wrong, whoever he is. — Corvus

I agree. But to know a word is to use it, and to use it is to know it.

That doesn't mean they know what "red" is. — Corvus

...but nor does it mean that they do not!

I agree. But to know a word is to use it, and to us either is to know it. — Banno

We have a part agreement here, which is a rare event.

...but nor does it mean that they do not! — Banno

Actually it is difficult for me to imagine what colour blind would be like without being one myself, hence the point was purely from inference. You could be right. Please carry on.

Fixed error from self correct: "to use it"

Thanks.

So to the rather odd paragraph about physical necessity. There's a bit of having one's cake and eating it, too, going on here.

Upon the contrary-to-fact conditional depends in turn, for example, this definition of solubility in water: To say that an object is soluble in water is to say that it would dissolve if it were in water. In discussions of physics, naturally, we need quantifications containing the clause ‘x is soluble in water’, or the equivalent in words; but, according to the definition suggested, we should then have to admit within quantifications the expression ‘if x were in water then x would dissolve’, that is, ‘necessarily if x is in water then x dissolves’. Yet we do not know whether there is a suitable sense of ‘necessarily’ into which we can so quantify? — p 158-9

So we need necessity in order to do physics; but we must debar it from logic. A difficult path to tread.

That is, and here I'm grossly overgeneralising, the extension. — Banno

I think you are. I was "seeing the picture" up to this point, but you'll have to work harder to explain how use reduces to extension. I believe you still need to respond to the bead question: How do we make coherent a situation where the extension remains the same but the color changes?

It's not that use reduces to extension, but that the use of a (proper) name is it's extension - what it refers to. This in contrast to Quine rejecting the use of proper names.

How do we make coherent a situation where the extension remains the same but the color changes? — J

I'm not seeing a problem with that. It might have been that beads 4,5, and 6 were the red beads. In which case, in that domain, "...is red" would be extensionally equivalent to {4,5,6} instead of {1,2,3}. And an extensional sentence about the red beads would have the same truth value as an extensional sentence about the beads {4,5,6}, and passes the test of substitution.

We want to say that there is more to being red than being {1,2,3}; but note that that "more" is intensional rather than extensional.

There need be no "intermediary step" of the sort you suggested,

So going back to this,

If planets and planètes have the same extension, then "The number of planets is greater than 7" means the same thing as "The number of planètes is greater than 7". Is there any intermediary step that would show this to be true? — J

Extensionally,

Planets = Planètes = {Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, Neptune} — Banno

The list of planets just is the "meaning" of both Planets and Planètes, and so since their number is greater than seven, both the English and French sentences are true.

There is a seperate issue, why Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, and Neptune count as planets, while Pluto, amongst other things, does not. But given that we accept the list of planets, then

The number of the planets > 7 = "Le nombre de planètes > 7

without further explanation. That is, if planets and planètes have the same extension, then "The number of planets is greater than 7" means the same thing as "The number of planètes is greater than 7" without further ado.

Missposted.

It's not that use reduces to extension, but that the use of a (proper) name is it's extension - what it refers to. — Banno

I agree, because what else could a proper name refer to? Some "essence of Banno"? But I'm suggesting that a color name is different, because in addition to referring to the various extensions of "is red," it also (appears to) refer to something intensional, namely the quale we can each call to mind. I know how much you value ordinary usage, and I would maintain that this is a clearcut case: No one scratches their head and says, "Yes, but what is Banno? What does 'Banno' mean? How can I use 'Banno' intensionally?" whereas we surely do talk this way about most terms that have definitions in addition to an extension. A proper name merely signifies without defining. So I still think the burden of argument is on you to explain why this way of talking has to be mistaken.

We want to say that there is more to being red than being {1,2,3}; but note that that "more" is intensional rather than extensional — Banno

Yes. Does this mean you countenance an intensional use that is not extensional -- or by "we want to say," did you mean that we'd like to but we can't justify doing so?

The "we want to say" is a nod to what cannot be said, but is instead done or shown.

After all, can you give a reason for saying that {1,2,3} are red, that does not involve showing us or at least looking at the beads?

can you give a reason for saying that {1,2,3} are red, that does not involve showing us or at least looking at the beads? — Banno

No. So take me to the next step -- what does that tell us about "red"?

what does that tell us about "red"? — J

:smile: Isn't it your bed time...?

In one sense it tells us that there is nothing more to say about red; given the domain is only the beads, red just is {1,2,3}.

I agree that there is something annoying here, but I suspect that it cannot be well articulated.

:smile: Isn't it your bed time...? — Banno

Ha! Argumentum ad tempus requiescendi.

I will try to articulate the annoyance better, but probably not tonight, as the moon rises slowly over the Gulf of America . . .

Gulf of America — J

Not a location I recognise. :meh:

Good night.

So embarrassing. Maybe I'll sleep for 20 years.

So we need necessity in order to do physics; but we must debar it from logic. A difficult path to tread. — Banno

Compare Quine to Martin Lof, the inventor of intuitionistic type theory. According to Lof, analytic sentences, at least in the context of intuitionistic type theory, are de-dicto definitions that are regarded to consist of perfect information, as in a complete table.

So in terms of your beads example, Lof would regard your proposed function mapping numbers to colors as analytic. But it is important to note the utility of calling this function definition "analytic" is only in relation to existentially quantified propositions about the analytic definition, which Lof classifies as "synthetic". E.g the theorem "there exists three red beads" is synthetic for Lof in relation to your bead function definition, because to determine the truth of the theorem requires checking.

In general, Lof regards a theorem in relation to intuitionistic logic to be 'synthetic' if the theorem contains an existential quantifier whose existence requires a proof in relation to the analytic definitions provided. Lof regards a synthetic theorem to be 'a priori' if the theorem can be proved de dicto via a process of deduction using the supplied analytic definitions that makes no recourse to facts about the external world. This is of course the case with intuitionistic logic, since its deductive system is constructive, i.e. de dicto. Hence for Lof, most of the theorems of intuitionistic mathematics are synthetic a priori (with the exception being postulated mathematical axioms). Generally, synthetic a priori propositions are undecidable.

Of course one might question whether the rules of the deductive system are correctly applied or whether one's analytic definitions are correct, in which case one's definitions are treated as being truth apt synthetic propositions in relation to some other underlying analytic definitions. So the analytic-synthetic distinction Lof intended is pragmatic without implying an absolute metaphysical distinction.

I think that Lof's reasoning is very much in line with Quine, whose notion of "physical necessity" I understand to be synthetic a posteriori, being in relation to the external world, but nevertheless also in relation to an analytic definition of physical terminology that undergoes constant revision on the basis of a posteriori evidence.

For example I imagine that Quine would consider the theorem "All swans are white" to be an analytic definition in the sense that Lof referred to, namely that the theorem doesn't contain a non-negated existential quantifier and so cannot be regarded as "true" except in the de dicto sense. This of course doesn't imply that the theorem's negation is analytic, which consists of a non-negated existential quantifier that answers to de re evidence. To me, such examples suggest that when counter-examples cause theory change, the falsified older theory is often not even wrong, in that the older theory cannot express the counter-example that it is wrong about.

Not something with which I am familiar. But in intuitionistic type theory, isn't a theorem synthetic if its truth depends on constructive proof rather than mere definitions? That is, not all synthetic theorems contain existential quantifiers. Consider "Every red bead appears before every blue bead on the string", which is not analytic, which must be determined by inspecting the arrangement of beads, and which uses universal quantification only. I may be misunderstanding your point, but being synthetic is not dependent on existential quantification only. However if your point is just that theorems containing an existential quantification are always synthetic because they require constructive proof, then yep.

In practice, we allow empirical counterexamples to revise our concepts, meaning that the statement was never purely analytic to begin with. Quine might say that "All swans are white" was part of a revisable web of belief rather than something analytic. It's not that we stipulate that "all and only white waterfowl are swans", an unfalsifiable, analytic and false proposal.

In any case, the idea of using an intuitionistic logic here is interesting.

Not something with which I am familiar. But in intuitionistic type theory, isn't a theorem synthetic if its truth depends on constructive proof rather than mere definitions? That is, not all synthetic theorems contain existential quantifiers. Consider "Every red bead appears before every blue bead on the string", which is not analytic, which must be determined by inspecting the arrangement of beads, and which uses universal quantification only. I may be misunderstanding your point, but being synthetic is not dependent on existential quantification only. However if your point is just that theorems containing an existential quantification are always synthetic because they require constructive proof, then yep. — Banno

Apologies for any misleading. To clarify, in type theory synthetic judgments can be identified with existential quantification due to the fact all propositions are types: having a proof that proposition A is true is equivalent to constructing a term a of type A, written a : A.

When referring to existential quantification, Lof wasn't referring to an existential quantifier within a proposition, but to an existential quantifier over terms representing a proof of a proposition type. Furthermore, the terms of a proposition type are definitionally equal by fiat, i.e a proposition type is the equivalence class of all proofs of that proposition.

My example referring to the swans was potentially misleading for conflating the two sorts of existential quantification, but nevertheless valid. A term cannot be constructed for the proposition type "All swans are white", indeed for any proposition containing a universal quantifier over an infinite domain, unless the proposition is interpreted intuitionistically such that the proposition can be proved by mathematical induction.

Perhaps a better example is the proposition "Nothing can accelerate beyond the speed of light". In relativity, a proof of that proposition implies contradiction. Hence presumably, the negation of the proposition is analytic in the theory special relativity, meaning that the proposition doesn't imply the physical impossibility of faster than light travel.