Wednesday, April 04, 2012

Are semantic rules true or false?

Over at Daniel's place the other day there was an interesting discussion about McDowell's views of what he calls the "endogenous given." I'm not going to talk about McDowell here though; instead I want to back up and look at the fault lines in the ensuing discussion. Basically I see a conflict between two readings of, or ways to follow, Wittgenstein. I'll say more about that later, but in this post let me jump right into the way it came up in the thread. The hope is not so much (yet, if ever) to show that my team is right, whatever that may turn out to mean, as to see where the differences lie (which may in fact turn out to be all we want to do). That itself says something about my team, or at least about me; but let's press on.

Whether or not it amounts to a theory of truth or even a definition of "truth", my team customarily refers to the disquotational schema "'[P]' is true iff [P]" (where "[" and "]" are Quinean corner brackets; you know the drill) as a platitude (to wit, the "disquotational platitude", hereinafter DP) – i.e., something whose truth is not in dispute.

For example, I am inclined to say that this English sentence
(1) This is red.
is true iff the object indicated is indeed red. Note that this does not solve our problem (below, concerning the status of (1) as an empirical statement or a semantic rule); but nobody said it would. All the schema does is disquote. After that you're on your own. If after being dumped into the object language we still don't know what to say about the reason we brought it up, that's not (1)'s fault, nor that of the schema. This may become clearer as we continue.

In particular, it seems to me not to matter to the truth value of (1) if it was meant as an ostensive definition of "red" rather than "just" an empirical observation about the thing indicated. If the thing isn't red then how can it function as part of an ostensive definition of "red"? And if it's red then (1) is true. When you ask me to bring you a red thing, the thing pointed to in (1) is not exempt simply because it's paradigmatic of redness. So it's red, and (1) is true, like I said.

Compare the meter stick. Sometimes people arguing in what they take to be a Wittgensteinian spirit deny that the meter stick is a meter long. As I recall, this is because (or at least if) to be "a meter long" just means that it is the same length as the meter stick, and what that means (on this view) is that if you hold it up to the meter stick they match; and of course you can't hold the meter stick itself up to the meter stick, so the idea that it is itself a meter long collapses into incoherence, making "the meter stick is a meter long" nonsensical, or at least sinnlos, and thus neither true nor false.

I get the point of saying this, which I take to be a healthy resistance to platonistic reification of things like "lengths"; but I don't think taking this line is worth it. Of course the meter stick is a meter long; that's why I can use it to measure off meter-long pieces of cloth (or wood, to make more meter sticks). There are better ways of resisting platonism than by this sort of semantic sleight-of-hand.

Note the form of that statement. Naturally most critics of this "Wittgensteinian" position are defenders of platonism in the relevant sense. If I just reject it in favor of "common sense" I look like one of these critics, or at best depriving our side of a potent weapon. Instead I say that weapon isn't worth it: it costs too much and we have better ones anyway. I'm still just as much a critic of platonism, and a follower of Wittgenstein in this respect, as anyone else.

Still, we need to account for the seeming imperviousness to empirical refutation of definitional statements; and that something is red is in most cases both contingent and empirically revisable, as are most uses of (1). This is what motivates the idea that (1) considered as a definition is neither true nor false, but instead a semantic rule for use of "red": revisable, if I want to change the way the "game" is played; but not empirically falsifiable – and thus, on this view, strictly nonsensical as an assertion of fact.

This is what I get from N.N.'s remarks on the Soh-Dan thread. The idea is also batted about by Wittgenstein in On Certainty: again, the idea is that such statements are neither true nor false, truth-values being reserved for statements which are, as we might say, moves in the game rather than rules setting out how the game (of indicating how things are by making true statements about them) is supposed to be played. But as I said, (1) doesn't seem to work for this. Let's try another sentence:
2) This color is red.
I don't think this helps any. Its only function, as opposed to (1), would be to make clear that it is the color of the indicated object that you are calling "red", but it most likely already is – "red" is after all a common English color-word – and in our context it certainly is (unless you're pointing to a photograph of Red Auerbach, or indicating which sections of "The Communist Manifesto" are objectionable to your free-market sensibilites, or something). How about this one?
3) This color is called "red."
This is the same as (2). Here the speaker is emphasizing the linguistic aspect of his statement, which of course you're always allowed to do; but it's still true if the thing is indeed red (that is, the color which English speakers call "red") and not otherwise, just like (1). Finally, in an effort to wrest control of the matter, we could try:
4) I will call this color "red."
After trying to retreat from empirical statement (about the object and its color) to grammatical rule with (1) - (3), we finally come to a first-person statement – a statement not about the object, or even its color, but about me and my (subsequent) verbal behavior. This is what Daniel and N.N. go back and forth about on the thread.

One worry about this seems to be that if (4) is simply predictive of my future behavior, it doesn't mean what we want it to mean (this is N.N.'s point in certain comments). I myself wonder about the force of that italicized adverb.

Another reading can be paraphrased as "I intend to call this color 'red'". According to the DP, this is true iff you do indeed intend to call that color 'red'. (Again, the DP is remorselessly unhelpful here; but that's no reason to reject it as false. Instead, again, we just need to recognize that you get out what you put in, semantically speaking.) But then we get into the performative aspects of stating that you intend to do something, and what effect that has on the truth value of (4) – see esp. Daniel's remarks about Roedl there.

In any case, the trick here, for N.N. and his Hackensteinian friends, is to try to pack into the statement itself as an interpretation accomplie the idea that it is asserting a grammatical rule rather than making an empirical claim, and thus to indicate that responses of "no it isn't" will be met by "what do you mean, it isn't? I'm telling you how to use the word."

That's what leads us, finally, to N.N.'s example (replacing similar statements about bachelors, which he takes to be too baggage-laden for our purposes):
5) A blork is a purple flower.
or
6) A blork is by definition a purple flower.
Chew on that for now; I'll come back to talk about blorks later on.

22 comments:

N. N. said...

If the thing isn't red then how can it function as part of an ostensive definition of "red"? And if it's red then (1) is true.

This is a brilliantly clear statement of something I want to oppose. But I think it's best to wait for the discussion to turn to "blork." Just as a teaser, consider your statement but with "blork" in place of "red":

(1) This is a blork.

If the thing isn't a blork then how can it function as part of an ostensive definition of "blork"? And if it's a blork then (1) is true.

I especially like your account of my progressive attempts, (1)-(4), to distinguish definitions from descriptions. However, you left out the last (and perhaps best) one: (5) Call this red; or let us call this red. This way of putting the definition (viz., as a command) is particularly resistant to its being understood as a description.

I look forward to the sequel.

Duck said...

Me: If the thing isn't red then how can it function as part of an ostensive definition of "red"? And if it's red then (1) is true.

You: This is a brilliantly clear statement of something I want to oppose.

Me again: No it isn't. It's ambiguous (not my intent, of course, but it seems that you insist) and you are reading it (perversely to my mind) the other way.

I will try to save my main points about the blork example for another post; but you brought it up here, so let me address this in the context of the claim I just made.

You ask me to consider:

(1a) This is a blork.

and the claim, analogous to one I made about the red thing indicated in (1):

(*) "If the thing isn't a blork then how can it function as part of an ostensive definition of "blork"? And if it's a blork then (1a) is true."

I see what you did there. You are, it seems to me, reading the analogue to (*) (i.e., the one about "red") in a platonistic way – as if red things were (or redness is) sitting out there independently, waiting for us to refer to them/it in the correct way, which is what it seems to you that the ascription of (contingent) truth to (1a) confirms has happened – that is, which it might not have done (whew! glad we got it right!). This sort of platonism is the only thing that the oddness of (1a)/(*) can reveal as nonsensical – because of course there was no such thing as "blork"-ness before you coined the term.

And I do agree that that sort of platonism is nonsensical. Yet I would also say – using the term as (1a) tells me to, and not thereby endorsing platonism – that blorks existed before you coined the term. You didn't create them, did you, by uttering (1a)? That would be idealism!

Look at it this way: are you suggesting the following exchange?

[Note to readers: we are not insane; we are merely doing philosophy.]

You: This is a blork.
Me: So it's a blork then?
You: Well, no, I didn't say that exactly; I'm just telling you what "blork" means.

That makes NO SENSE.

Let me stop here; I'll discuss "Call this red" next. I do think it will help, but maybe not for the reason you think.

Duck said...

Now: you also suggest

(4a) Call this red, or (4b) Let us call this red.

This might help, in this sense. Now that we have commands, or at any rate, non-assertions, their satisfaction-conditions and their truth-conditions come apart (that is, they don't have truth-conditions at all, but they are satisfied if I do as you ask). But now there is no theoretical bar to the thing's being red. You can now freely admit that it is, as its redness does not entail the truth of (4a) or (4b), as it does of (1) - (4). But now since the whole point of (1) - (4b) was to do exactly the same thing – to instruct me in your use of "red" – the cases are not that different after all. Admitting the truth of (1) - (4) does not interfere with their purpose of semantic instruction, any more than it does in the case of (4a) and (4b), which do not have truth values. In either case you are committed to calling the thing red, and thus the truth of (1) - (4).

I (we?) are just saying that in addition to doing this (the semantic task), it (i.e. (1) - (4)) also tells me – how could it not?? – that the thing is red. Are you in a position to deny it (= that the thing is red)?? You just said it was! [Also: if relevant, tell me how you feel about the meter stick. If you simply want to deny that that's what it means for s.t. to be a meter long, then the example is orthogonal and won't help us here.)

So I see it this way: when (4a) and (4b) relieve the theoretical pressure on saying that the thing is red – which, again, is what (4a) and (4b) tell me to say! – then we see that we should have been allowing that all along, in the cases of (1) - (4). That is, (1) - (4) are true statements (by your lights anyway; I may disagree, even while taking instruction about your use; this is hard to imagine in cases of color, but not impossible, and easier for other things). Their point is not simply to inform me about the redness of this or that object, but also to instruct me in your use of "red"; and I may be perfectly well able to interpret you that way.

Duck said...

How about this one: isn't

(7) I now pronounce you man and wife.

said under the appropriate circumstances by the appropriate person, true? That is, isn't it the case that its performative aspect does not interfere with its truth? If so, then why can't that be true of (1) - (4) (but not (4a/b), which can't have truth values no matter what)?

N. N. said...

I'd apologize for my perversity, but it's still not clear to me how you're statement can, in the end, be read in a non-Platonic way. I did not think that you were intentionally suggesting that words carve nature at the joints, but (unless I'm missing something important, and I may well be), that is the unavoidable pitfall here.

Yet I would also say – using the term as (1a) tells me to, and not thereby endorsing platonism – that blorks existed before you coined the term. You didn't create them, did you, by uttering (1a)? That would be idealism!

Not thereby endorsing platonism? I don't see how the claim that blorks existed before "blork" was coined can avoid the endorsement. It seems to me that categorial distinctions are not "out there" in the world independently of our adoption of them. We differentiate Bengal tigers from other subspecies of tigers. If we didn't would there be Bengal tigers? That is, would there be a kind of thing different than other kinds of thing such as Siberian tigers (or lions, for that matter). We are tempted to answer: Of course, there would be those things even if we had not decided to call them by the expression "Bengal tiger." But which things are those things as opposed to other things? Don't we need some standards of same and different kinds in order to even talk intelligibly about "those things" as opposed to others? Sure, this is a kind of idealism, but that doesn't worry me.

I'll respond to your other comments soon, but opening day is right now!

N. N. said...

But now there is no theoretical bar to the thing's being red.

But there is. It's the one that you think costs too much. In the case of (1), it's as if I made the color a sample and then compared it to itself: That color is that color (pointing both times to the same thing).

Of course the meter stick is a meter long; that's why I can use it to measure off meter-long pieces of cloth (or wood, to make more meter sticks).

Does the fact that I can use it to measure off meter-long pieces prove that it's a meter long? Why can't I say: the reason why I can can use it to measure off meter-long pieces is because it's the standard for being a meter long. This doesn't commit me to holding that the standard is itself a meter long. To claim otherwise is to beg the question.

By the way, why is the price too high? What do you need the money for? What is at stake for you here? That is, what is the undesirable result of this view of the standard meter, etc.?

Are you in a position to deny it (= that the thing is red)?? You just said it was!

Isn't whether I said that precisely what's in dispute? Granted, I've used a propositional sign that in different circumstances can be used to describe the color of the thing, but that hardly suffices to establish that it's use as an ostensive definition is also a description.

Concerning (7), my answer to your question is no. Think of two cases: First, a priest says (7) to an engaged couple during a wedding ceremony, etc. so as to marry them. Second, a priest says (7) to an engaged couple, months before in his office, so as to describe for them what he will say after they say their vows: "You say: I do. I pronounce you man and wife." In the second case, he is not marrying them, but describing his future behavior. In the first case, he is not describing his present behavior, but marrying them. Rephrasing what he says makes this clear (rephrasing to bring out a difference in grammar is what W. calls a grammatical investigation, which is what you're post and our discussion boils down to). In the first case, we (our priests, magistrates, etc.) don't say "Let them be married," but we could. I wouldn't call this a command, but it's similar to one. But in the second, we couldn't. We'd have to mention, rather than use that phrase, e.g., we'd have to preface it with "I say:"

Duck said...

I'm not sure where to say what, but let me address a few things here.

As I was right to suspect, you endorse the story of the meter stick: that as a matter of "grammar", the meter stick cannot be a meter long, and "the meter stick is a meter long" isn't even false but rather nonsensical (as opposed to "that ruler is a meter long," which is simply false); and you also, as I suspected in bringing it up, assimilate this example to those of blorks and red things, where it seems especially problematic.

Considering the inference from "(1) ["This is red"] is a grammatical remark indicating how "red" is to be used" to "(1) so construed is neither true nor false" to "the thing indicated is not red": this seems to indicate that you accept the disquotational platitude. If not, that the thing is red would not even seem to be a threat (that is, making "this is red" true). So I will take the DP to be platitudinous and thus not an issue for us. Otoh it still looks funny to see the redness of the thing to threaten "Call this red."

About that supposedly not-red thing, let me press you a bit more. When you picked out something to hold up while you uttered (1), why did you pick that thing rather than this other one? It seems to me that you have to say that

(9) Because that other one's not red, and this one is.

is nonsensical. No doubt that looks no more painful than anything else so far; but imagine that as the exchange continues, it turns out that when you said (1), I didn't really hear what you said. Why did you want me to hand you this book rather than that one? So you explain further:

(10) No, I wanted to show him what "red" means; that's why I needed the red one.

In order to show someone what "red" means (as opposed to telling them to think of ripe McIntosh apples and stop lights and fire engines), you need a red thing to show them while uttering (1) - (4b). That's why you wanted me to hand you Reading Rorty rather than Rorty and his Critics. The latter is a better book, but the former (i.e. its cover) is red while the latter is not. This makes perfect sense to me; while to deny that Reading Rorty is red on the grounds that you used it (or wanted to use it – what if you never got to utter (1) because your target wandered off while you were looking for something red, uh, I mean, let's see, something you could use in this way? Would you get to admit that it's red then?) to demonstrate your use of "red" seems like a fairly thorny nettle.

However, the reason I don't want to (have to) say that is less that the pain is too great – everyone wants to adjust what others say in some respects, if that difference seems to be causing confusion – but that the supposed benefit is unimpressive. So when you ask "why is the price too high?", my answer is less that I need those particular resources (here, that the nettle is simply too painful) than that what you are offering in exchange doesn't have the features I'm looking for anyway (as you suggest by asking "what do you need the money for?"). In any case that's more the form my answer will take.

I think one thing this does is point up a difference between "red" and "blork", which I will also save for later.

Duck said...

Next, you ask:

"Why can't I say: the reason why I can can use it to measure off meter-long pieces is because it's the standard for being a meter long. This doesn't commit me to holding that the standard is itself a meter long."

Note the form of this question, i.e.: "why am I required not to talk this way?" That seems significant. Because my response is: well, you could if you want; I just don't see the point (or: I do see what the point is supposed to be, but again: no sale). You, on the other hand, seem to think that I must speak the way you do instead of the way everyone else does (which you condemn as irredeemably platonistic no matter who says it). This concerns not so much a burden of proof as the role that requirements (that we say this or that) play in what we are doing, as well as the role of context of utterance in interpretation.

I'll say more about this later, but for now remember my gripe with Kripkeans in my 3QD essay about Errol Morris (which I know you've seen already; rereading it, I bet our readings of PI §241 differ too, so we may want to take another look at what I say there). Of two ways of speaking about names, Kripkeans demand that we speak one way rather than the other (in response to an "internalist" or "descriptivist" doctrine that we must speak the other way rather than the one); where it seems to me we could speak either way if we wished. Neither is required, either by the nature of things OR by pre-established semantic rules. Yet to say this is not to deny the normativity involved in linguistic behavior, but instead to look again at what we really want to say about it.

N. N. said...

e need to discuss something that I'll call "role reversal."

The standard meter in Paris, so long as it is the standard for something's being a meter long, is neither a meter long nor not a meter long. This, I think, should be uncontroversial.

However, we can change what we use as the standard meter, and we can do so in such a way that what was previously described as being a meter long is now to be used as the standard for something's being a meter long.

Suppose we measure the time it takes the speed of light to travel from one mark on the standard meter bar to the other. We can then describe that distance (viz., the distance that light travels in that amount of time; hereafter L) as one meter in length. That is, we have compared L to the standard meter, and seen that they match. Hence, the sentence "L is a meter long" is true.

However, we can then decide to use L as the standard for something's being a meter long, e.g., instead of using the standard meter in Paris to help us make wooden "meter sticks," we use L. That is, we compare the sticks to L, and if they match, then it is true that they are a meter in length (and we can ship them out).

The reason that we are now using L as the standard meter is because it was previously determined to be the same length as what was the standard meter. But now that L is the standard of something's being a meter long, what would it mean to say that it is a meter long? It cannot be compared to itself. To call it a meter long requires using something else as the standard. Therefore, if it is presently being used as the standard, it makes no sense to describe it as being a meter long.

But now consider the wooden "meter sticks" that were previously described as being a meter long. These, of course, can themselves be used as standards of something's being a meter long, but when they are, they cannot be described as being a meter long.

In other words, to describe something as being a meter long involves comparing it to some standard or other. Anything that is presently the standard, can be measured by some other standard, in which case it is not presently the standard. On the other hand, it if it presently the standard, then there is no such thing as its being a meter long (or not) because there is no such thing as comparing it to itself.

The fluidity, relatedness, and multiplicity here can cause confusion. But if we keep our eye on the central point -- viz., that something is a meter long in virtue of being compared to a standard, and nothing can be compared to itself -- then the distinctions sketched above fall into place.

Duck said...

Yes, I already knew you could say all that if you wanted – sorry to make you spell it out. We can stop (for) now about the length of the meter stick.

Is the same thing happening for colors too? Are we all like Wittgenstein's shopkeeper?

N. N. said...

We don't usually operate with samples as the shopkeeper does, but we could. And we all initially learned the meaning of "red" by an ostensive definition.

My point in going through the meter stick example in more detail was to emphasize that, even if something can be truthfully described as a meter long, or as red, on one occasion, it can be used as a standard of something's being red on another. But it can't be both at the same time.

If that's right, then if I am using the sentence "This is red" as an ostensive definition, and hence the thing pointed to as a sample of red, then this sentence does not describe the color of the thing, and hence this object, used as a sample, is neither red nor not red but a standard of something's being red.

Again, consider the sentence "That color is that color" pointing both times at the same thing. If it were possible to simultaneously use the thing pointed to as a sample and to describe it as the same color as the sample, then this sentence would make sense. But it doesn't.

N. N. said...

Just in case my point is still not clear, let me add the following.

I agree with you that, in explaining the meaning of red to someone with an ostensive definition, the reason I picked this object (as opposed to that one) to use as a sample of red is because it is red. But I don't think that admitting this jeopardizes my contention that, used as a sample of red, the object is neither red nor not red.

This is an instance of role reversal. When the object is not being used as a sample of red, it can be described as red. But when it is being used as a sample of red, i.e., as a standard for something's being red, then it cannot be described as red because that would involve comparing it to itself which can't be done.

noiln said...

Empirical rule
Empirical rule, empirical
http://www.infoaw.com/article.php?articleId=926

nogre said...

Yes, my duckrabbit is more stylish. But I was also trying to make a quick point.

I deal with the Wittgensteinian definition/revision problem differently. I have not once, in my life, ever used the statement, "This color is red." This is not to say I haven't tried to color match; in fact it is a current concern because modern monitors/graphics programs come with color chips to make sure that the monitors display color correctly. We use these color chips as standards, not because "They are red" but because they do not variate over time in a way that causes problems. Since they do not variate, they can be used as a standard. I am reminded of PI80

I say "There is a chair". What if I go up to it, meaning to fetch it, and it suddenly disappears from sight?---"So it wasn't a chair, but some kind of illusion:.------But in a few moments we see it again and are able to touch it and so on.------"So the chair was there after all and its disappearance was some kind of illusion".------But suppose that after a time it disappears again---or seems to disappear. What are we to say now? Have you rules ready for such cases--rules saying whether one may use the word "chair" to include this kind of thing? But do we miss them when we use the word "chair"; and are we to say that we do not really attach any meaning to this word, because we are not equipped with rules for every possible application of it?

What is important about the non-variation? I say that we are able to have standards like "This is a meter" because they appeal to a greater level of certainty than ordinary facts which we could be mistaken about. I like On Certainty 594

My name is "L.W." And if someone were to dispute it, I should straightaway make connexions with innumerable things which make it certain.
He then continues to give examples (brainstorm?) of things that have such connections with our everyday life that if they were mistaken, then huge amounts of our experience would have to be considered mistaken as well. The fundamental interconncection with much of our experience gives greater certainty to those things.

So I get myself out of this definition dilemma by reductio ad absurdum. The things I consider to be standard are standards because they are stable in ways that are so connected with other parts of my world that if they were to change radically (such as chairs pulling cheschire cat acts), I might as well throw in the towel. If I am trying to give someone a definition about something fundamental, it is the weight of my life experience that provides the force. This is neither true nor false; it is the ground.

Duck said...

Hi Noah - thanks for stopping by. I did suspect you were making some kind of point with your remarks on the d/r, but it was not clear to me what that point was, so I didn't mention it.

I actually don't see this "problem" as much of a problem, or I wouldn't if people didn't make it one. I have no previous philosophical reason not to say, with the folk, that when I say "this is red", indicating the color, that I believe, and indeed am claiming, that the thing is red. I recognize and endorse the anti-platonistic motivation for (thinking about) denying this, in Wittgenstein and elsewhere, but I just don't think it's necessary to follow through on it (which I actually think he doesn't).

I naturally agree that things we use as standards shouldn't vary in the relevant respect – no meter sticks made of rubber – but that seems orthogonal to the semantic point here.

I like §180 too, but I'm not sure I see how it requires or even suggests that standard red things aren't red. In fact it suggests that we have some flexibility here – that semantic rules don't need to dictate what to say in extreme cases in order for them to do their everyday work – so that when asked pointedly about the meter stick itself, I may simply shrug. It's the platonist who requires what I merely allow, and N.N. (read the other comments if you haven't) who forbids it (and is concerned that I allow it). That's why I say it isn't really a problem for me. Insist (I have been saying to N. N) on saying this rather than that in your extreme case if you want; but that he sees it as required strikes me as non-Wittgensteinian (and independent of that, as not worth insisting on). And of course he says the same about me – that I open the door to platonism, or worse – in spurning the Hackensteinian line.

Epistemologically I don't see the point of drawing a categorical line between ordinary beliefs and those with "greater certainty" – esp. if that means that I am to take those with greater certainty as thereby not true (!?). That's too much philosophical dogma for too little bang; and doesn't make sense of what even On Certainty, where the idea is ruminated over at the most length, says about the contextual variability of "hinge propositions" (think of the riverbank image). As I've mentioned elsewhere, and should probably say more about now, the best sense of "hinge proposition" that I can use (in my pragmatist context) is as synonymous with "belief" – which is of course belief that something is the case, and thus something with a truth value.

Naturally these things become harder to revise the closer we get to the center of my "web of belief"; but that doesn't mean they're not beliefs. It just means they're hard to revise. And again, that we don't know what to say in such cases shows that we don't take ourselves to be required to say this rather than that.

nogre said...

My problem with the 'this is red' is that I can't think of a single time I have ever had to use the sentence (or been in the presence of anyone using it). Which makes me very worried that it only gets mentioned in philosophical talk and hence never has anything to actually do with being red. Maybe if I were teaching a baby to speak I would say things like that, but otherwise, never. So I think it is much more likely to be the refuse of old metaphysics.

I guess my reading of §180 is a bit different. I always took it to mean that a chair that wasn't quite there might be better served by being called something other than chair. If I found an object that disappeared and reappeared at random, I would think it significantly different to give it a new name. Likewise a meter stick that varied in size shouldn't be called a meter stick.

Mu reading of On Certainty seems different too. My reading is based in his discussion of things we can't be making a mistake about, and that these are different from things we could be mistaken about. If we get too close to the center of the "web of belief" we might collapse the entire web (and fundamental definitions live in this area). It is not just a question of being more difficult to revise, it is that revision and doubt have stopped at this point. But LW died that week, so there isn't much textually for me to go on- this is just my favorite reading of these passages.

N. N. said...

At this point (and our discussion has helped me to see this), the question "Are standard red things red" seems to be, well, a red herring. I admit that they are, but deny that this is relevant. The relevant question, I think, is "Does an ostensive definition of 'red' describe the thing pointed to as red," and so far as I can tell, the answer is clearly "No." You want to say "Yes" because, if the thing pointed to is red, then the sentence "This is red" (pointing at the sample of red) is true. But the sentence true only if it, as it is used in this context, is the sort of thing that can be true or false. And that is what I (now) deny. In this context, the sentence is not a description, and therefore, the fact that the sample is red does not make it, the sentence used as a definition, true.

In fact it suggests that we have some flexibility here – that semantic rules don't need to dictate what to say in extreme cases in order for them to do their everyday work – so that when asked pointedly about the meter stick itself, I may simply shrug.

You may shrug, but not because shrugging is an adequate response. You need a coherent story to tell about how the "extreme" case differs from everyday ones. Presumably, you don't think that, if the standard meter were the only measure of a meter, that it would make sense to say that it is a meter long. Or if you do, I'd like to hear why. But if you don't, then you must agree (it seems to me) that no standard can be compared to itself. And if that is agreed upon, then the rest of my position follows.

On last point about forbidding things. I'm not forbidding anything. That makes it sound like there is something that can be done (e.g., measuring the meter stick with itself) that I am, for some reason or other, prohibiting. That's not how I see it at all. As I understand what I am up to, I'm merely pointing out that there is no such thing as what you take me to be forbidding. You say, the definition is true. And I say: I don't know what you mean. If, at the end of the day, it turns out that you didn't mean anything, then there is no need of a prohibition because there is nothing to prohibit.

N. N. said...

Another way to bring out my point.

What sorts of things are true or false? Not sentences simply. "What time is it?" is a sentence, but it's not true or false. It is sentences used to make statements that are true or false.

If the sentence "This is red" is used to define the word "red," then it is not a statement. That, I think, is where our disagreement really takes place. For argument's sake, suppose that I'm right in this – definitions are not statements. Then definitions are not true or false.

Now, if the definition "This is red" is not true or false, then the fact that the thing pointed to is red (a fact we agree on) does not make the sentence, used as a definition, true. So, that fact can drop out of the discussion.

Back to whether definitions are statements. There are, I think, two reasons for denying that definitions are (also) statements. First, they can be rephrased as (something like) commands: "Let us call this 'red'" or "Call this 'red.'" Clearly, the latter are not statements. Second, if the definition "This is red" states something, then what does it state? It would have to be that the object pointed to is red. So, the sentence would have two wear two hats at the same time: definition and description. These, it seems to me, are incompatible. Think of how someone who didn't know the meaning of "red" would have to understand this sentence as simultaneously a definition and description. He would have to somehow apply the rule he's just received to the thing pointed to. As far as I can tell, he would have to understand the sentence as saying "This color is this color," pointing both times to the same thing. And that, I think, is incoherent.

Here is where the "blork" example would be useful. The person learning the meaning of "red" is in a position that is (for present purposes) the same as that which the coiner of a word is in.

Charles T. Wolverton said...

the question "Are standard red things red" seems ... a red herring.

Cute! But what does "standard" mean here? Well, here's what it might mean. A subject S is in a controlled environment in which an object can be illuminated so as to produce reflected light at S's eyes with wavelength in any of the half-dozen or so basic pure spectral color bands. S has been found to have normal vision in all respects and to be a reliable reporter of visual phenomenal experiences. S was brought up by parents who - being scientists themselves - were careful during S's prelingual development to use a similarly controlled environment in teaching S to associate the experience of seeing objects similarly illuminated with the experience of hearing spoken color names - eg, "red" - and later with uttering color names. An experimenter now asks S "Which of these objects is red?" while pointing at one of several test objects, and S answers "This looks red" while pointing at one of the objects. The tester then says "Correct."

Why did S give that answer rather than "This is red?" Because S - knowing how the environment works - realizes that there is no way to tell what color the object "is". It could be a white object (or even a mirrored one) illuminated by RED light or an object with reflective properties such that incident light with a certain spectral power distribution results in reflected light with spectral power concentrated in the RED band. Such environments might be described as "standard" for testing the ability to identify colors (somewhat as Sellars suggested in EPM), but no "standard red things" are involved. In short, I don't think the DP helps because I consider the truth condition resulting from disquotation of "This (object) is red" to be meaningless.

Does an ostensive definition of 'red' describe the thing pointed to as red?

The setup described above can be used to "ostensively define red" in the following sense. After S leaves the environment, all objects other than the one S identified as "looking red" are removed. Another subject is brought in and told that if asked in English "What color is this?" (pointing at the object) one should answer "This is red". But this defines the use of "red" (ie, it becomes part of its "description" per PI §10 if I understand that section correctly), and I'm inclined to agree with NN that while an example of word usage is a sentence, it is not a proposition. However, I see "This is red" as a comment about one's phenomenal experience, not a description of an object.

The problem with the proposition "the standard meter is 1m long" seems to be that defining the meter as the distance between marks on the standard meter bar was an instance of the distinction made in PI §50 between something represented and a means of representation. Thus, to see if something was 1m long required comparing its length with the distance between the marks on the bar - a representation - which as noted earlier in the thread makes no sense if the something is the distance between the marks themselves. However, the current definition as the distance traveled by light in a given time interval avoids that ambiguity since it defines the meter as a true distance rather than as a representation of one. To check the length of something one need only see how long it takes light to travel from one end of it to the other. The problematic proposition now has no meaning since "1m long" is a property of an object, not a distance. The proposition must therefore be changed to "the standard meter is a 1m distance" which is trivially true. (This general idea was broached in NN's comment of 4/9 11:21 AM, but this way of putting it seems a little more straightforward.)

Duck said...

Hi Charles, thanks for your comment.

I think I should back up a bit and say something about the similarities and differences of the underlying motivations here. The bit about standards makes for an interesting connection, as it happens; but I will probably start a new thread for those things. I may have a few more comments here too, but they might make more sense in a broader context. Next week is busy but I should have some time after that, so stay tuned.

N. N. said...

Did you lose interest in the discussion, or are you presently too busy to take it up again? I am very interested to find out where it ends because if I'm wrong, I need to change my view!

Duck said...

Sorry N, I haven't lost interest - It's just that at this point I need to back up and take the bird's eye view (not God's eye though, don't worry). And yes, other stuff keeps coming up. For instance I have a new 3QD post: http://www.3quarksdaily.com/3quarksdaily/2012/07/whats-wrong-with-being-a-sophist.html