Thursday, February 01, 2007

what is a speakers proposition?

Tzu – Wei Hung gave a talk last night that utterly baffled me. My failure to understand I think was to do with the word “proposition” and perhaps certain assumptions about what cognitive states it is possible to be in. The puzzle that the talk drew out was that given a contextualist infinite regress argument it seems that there is an unbridgeable gap between the speakers proposition and the sentence proposition. I fear to say exactly what the regress argument is because I’m fairly certain I haven’t got it right, but it seems that if you try and remove the context sensitivity out of a sentence you end up with another context sensitive sentence. This sentence in turn can be clarified by another sentence. Presumably there will be an infinite number of these clarifications, therefore there will be no sentence that expresses the proposition that the speaker was trying to express. Although I don’t see that there is necessarily going to be an infinite number, I am a great fan of language and am willing to accept that this part of the regress argument is true. So when I write “John is ready” I mean “John is ready for the written exam” which means “John is ready for the Metaphysics written exam on Tuesday the tenth of April, 2007” by which I mean “John has written the exam time down in his personal organiser and has a clear idea of which topics he is going to answer.” And so forth.
I guess the puzzle is that it seems clear that none of these sentences is going to express exactly the proposition I had in mind. Some are too vague, others too precise. They can’t all express it, since this would mean that the speaker's proposition is infinite and we are assuming this is impossible. So there is an unbridgeable gap between the speaker proposition and the sentence proposition.
Now I wonder what the speaker's proposition is. Gabe’s suggestion was that it is the propositional content of what it is that the speaker is trying to express by the sentence. Tzu-Wei said that the speaker grasps the truth (conditions,) of the proposition when he grasps a proposition. Propositions are it seems at least minimally true or false. If we take some kind of belief norm of assertion, then we can say that the speaker proposition is the propositional content of the belief that he is asserting. Then here comes the analytic philosophy assumption that I think is too blatantly false to have been assumed for so long.
1. In order to believe a proposition one must grasp its truth conditions.
There are of course a number of ways of expressing this thought.
1* In order to believe a proposition one must know in which possible worlds the proposition is true.
This is certainly unrealistic since even to individuate a possible world would require an infinite cognitive effort.
1** In order to believe a proposition one must have a criteria for assessing the truth of p.
The problem with 1** is that if you believe p it seems that you have already assessed that p is true using some criteria.
1*** In order to believe a proposition one must have a translation schema such that the proposition is true if and only if some sentence in one’s language of thought is true.
Does anyone believe 1***? Or does everyone apart from me believe 1***? The problem with 1*** is that is really is infinitely regressive.
For me the problem with the whole notion of grasping a proposition by grasping its truth conditions is that it is possible that the world could be such that it is not determined that a particular belief is true or false, or true and false or neither.
Examples. I believe that John feels cold. John doesn’t know whether he feels cold or not. Is my belief true or false?
I believe that there are twice as many numbers as even numbers? Do I grasp the truth conditions of this proposition?
I believe that President Bush is unpopular. Is it not possible that the world is such a way that were I omniscient I still wouldn’t know whether Bush was unpopular or not?
I believe that space time is infinite and that the laws of physics are constant throughout. Do I really grasp the conditions under which this proposition is true and under which it is false? If we are to assume that I cannot grasp the infinite then I certainly don’t grasp the truth conditions of this proposition. But I believe it none the less.


Anonymous Anonymous said...

We can distinguish the basic intuition behind 'grasping truth conditions' from some of the technical ways in which this is modelled. The intuition, I take it, is that when we express a content (and we use language to enable us to do this) we have the ability (in principle, and often in practice) to distinguish conditions under which what we said either holds or does not hold. If we did NOT have that ability, there would be no connection between the content of what we say and our life in the world.
We can distinguish this from what ought to be a separate question: what sort of info does a sentential form encode.

9:16 AM  
Anonymous jonny said...

"distinguish conditions under which what we said either holds or does not hold"
This is what interests me. Is this a verificationist principle? One grasps that p when when can tell when one would know that p. A proposition is something that is knowable.

4:03 PM  
Anonymous Johann said...

I don't know whether its a 'verificationist' position or not (the label has become a bit of a bogey shibboleth). But it seems reasonable to me to think that in so far as we understand something, we have (ipso facto) an understanding of how things might be for that something to be true.

10:41 AM  
Anonymous jonny blamey said...

had to look up "shibboleth", now I understand it. Does this mean that I understand how things must be for it to be true that vericationism has become a bogey shibboleth? It is true that I have encountered arguments from philosophy graduates which use as a premise that verificationism is false. Also people who say things that seem to be verificationist in essence will vigourously deny that they are verificationists.
knowing the truth conditions seems indistinguishable to me from being able to say what constitutes evidence arguments or reasons for a proposition. Understanding the meaning of a sentence on the other hand can't be the same as understanding the truth conditions since many sentences don't express propositions.

12:17 PM  
Anonymous Johann said...

'shibboleth' is a good word!

Do you think sentences have meanings? There's no need to suppose that they do, which is not to deny that we can utter sentences and thereby express something with content (and some forms of content will have truth conditions, others perhaps wont).

A fortiori [just to show how learned I am] understanding the meaning of a sentence cannot be the same as grasping truth conditions.

12:58 PM  
Anonymous jonny said...

Yes, I think the test for whether something has a meaning is that there is an answer to the question what does that mean?. Or what do you mean by that? So most sentences have a meaning, but not all. only assertions have truth values, assertions express beliefs. You can't believe something without grasping its truth conditions.

2:01 PM  
Anonymous johann said...

I can point to something and mean that thing, but my pointing itself need not mean that thing. Likewise, it is a coherent position to say that I can mean something by uttering a sentence without that sentence having the meaning that I express, or even without there being a coherent notion of sentence meaning at all. When we ask about what someone has meant we ask precisely that - what did you mean 'by saying that' - but this does not entail that the reason why you meant something was that the sentence you uttered had that meaning and expressed it for you (so to speak). It is coherent to suggest that we use sentences as tools for us to express meanings, without the 'tool' itself containing the meaning that we express. That might be a nontraditional notion but it might also be correct.

9:09 AM  
Anonymous jonny said...

Johann, yes you are right. Now "sentence" is confusing me. There are tokens and types. One might say it is not the token sentence that has the meaning but the type. But take "the stairwell needs painting". the same sentence has different meanings on different occasions. This makes your tool analogy seem good. The sentence doesn't have a meaning, you mean something by a sentence. What it is one means by a sentence I guess is an important philosophical entity. People might want to say its a proposition, but enter Wittgenstein and we find many things meant by linguistic utterances aren't true or false, (questions, commands, jokes). The plot thickens.
Meaning is related to importance, relevance, use and value. Something of no importance doesn't mean anything. I'm exploring the idea that true and false are red herrings, and the real binary system for meaning is good and bad. There is a deep link between meaning and intention, meaning and function and meaning and causation. So "he meant well","Thats meant to switch off when the water boils" "flowers mean spring is on its way". I don't believe that this is an arbitrary ambiguity. But I don't know other languages.

9:51 AM  
Anonymous johann said...

Glad you like the tool analogy! - though its not my idea, i've pinched it from someone else. (though isn't the token/type issue irrelevant here?).
But maybe you are combining different nuances that 'meaning' can have - there is a technical lexical/semantic question, and there is meaning as relates to importance, relevance etc. Even if truth conditions sometimes seem to give us no grasp on an utterance meaning, might we need some extra arguments before plunging into good/bad as the replacement for truth/falsity? You might be jumping into just a different tub of herrings.

12:34 PM  
Anonymous Jonny said...

The type/ token distinction is interesting for this reason. The sentence "I am 18" has the same meaning whether it is typed in red blue, written on paper or black board. But it has a different meaning depending on who says it and also what being 18 amounts to in the context. when two people write down "I am 18" is this two tokens of one type? Or two types?
The advantage of good/bad is that it subsumes true/false. If I know on what occasions it would be good to say "I am 18", then I get the truth conditions for free. This way I can "grasp" what you mean when you say "Are you 18?" without having to worry about truth conditions when there don't seem to be any. Generally speaking it is good to believe what is true. It doesn't even make sense to say that it is true to believe what is good.

4:42 PM  

Post a Comment

<< Home