Sunday, October 19, 2008

Better Russell Bit

It occurred to me last night that I should do a better job of giving Russell's position. Now, certainly this isn't perfect; but it should hopefully be better!

Russell assumes:

1. Propositions meaningful are non-linguistic entities.
2. Meanings are non-linguistic entities.

I happen to not agree with this so much, but I presume it is obvious to everyone else and stands in no need of defense.

But there is a tradition of British philosophy which denies this as well. They would argue:

3. Language stands for ideas having meanings.

This is a concept Empiricism: the basic words and concepts, basic ideas, are extracted from experience. So a private sensory language exists which is meaningful because it is about our ideas. (So when a baby sees a red patch, his private language gives meaning to his sensations and gives him ideas.)

This is a most queer view, but nonetheless was historically popular. It is further held that:

4. In every judgement there is something, the true subject of the term, which is not an idea and does not have meaning.

The motivation here is clear: we don't want to be idealists, with only ideas and meanings existing in the mind. So there are some transcendentally real objects, even if we can only idealize them. Kant or Locke would want this. There are 'unknowables' or 'I-know-not-whats' to deal with.

But this line of thinking brings the conclusions:

5. So meaning is linguistic (or experiential).
6. So propositions, to have meaning, must be linguistic (experiential).

I add "(or experiential)" since a good Empiricist will argue for all sorts of non-linguistic knowledge, awareness, etc. grounded in the magical power of experience... as you can tell I do not accept this view. But an empiricist like CI Lewis will argue in defense of (5) that sense-meanings are essential for any meanings to exist at all, and in defense of (6) that propositions are only meaningful if they phenomenologically reduce into statements about immediate experience.

Yes, Lewis is a phenomenalist. It seems that (3) and (4) commit us to Kantian transcendental idealism if we want to be 'realists' and phenomenalism if we are happy to be solipsists. We can argue for 'realism' is we are unhappy with being solipsists.

But now we see that:

7. ~( (1) & (6) )
8. ~( (2) & (5) )

So something has to give. Even thought I don't like (1) or (2) so much, I don't like (3) or (4) either! What is wrong with (3) and (4)?

3. Language stands for ideas having meanings.

4. In every judgement there is something, the true subject of the term, which is not an idea and does not have meaning.

I don't like (3) because it give meanings primarily to ideas, which later get hooked up to language. So there are pre-linuistic means and concepts acquired directly through experience. This violates the Myth of the Given. This view isn't in favor of innate ideas, but has the same basic picture in mind.

I don't like (4) because subjects have meanings. Lewis would even argue that they have sense-meanings. Something like direct reference will show that I can refer to something directly, even if I cannot access it with sensations or if I only know some contingent facts about it. I mean Jones when I say 'Jones' or point at him. Does it make sense to say 'the real Jones' is hidden, meaningless, etc.? That seems queer. That is Jones there, damn it!

Russell makes the point that:

5. Words have meaning, in the simple sense that they are symbols which stand for something other than themselves.

So there is a psychological and logical element to meaning. The defender of (3) and (4) are getting these messed up. Ideas seem to relate to psychological meaning, which is distinct from denotation which relates to a logical meaning.

This confusion is evident if we consider (6):

6. Propositions, unless they are linguistic, do not themselves contain words, only containing entities indicated by words.

Propositions do have meanings, but are not word-things. But this isn't a contradiction!

Agents are talkers. Words are talkings and entities are talk-eds. We shouldn't confuse the talkings and talk-eds! A proposition is not a talking at all, it is a talk-ed. A talking simply represents it.

To motivate (5) and (6) we can appeal to denotation as a kind of meaning, distinct form psychological or empirical notions of meaning. A poor Empiricist like Kant may wonder how I can have an idea of a thing-in-itself, the real deal object behind my sensations and ideas. Russell sees that I still denote that thing, even though I cannot experience it, such that I have no idea of it and it is 'meaningless' to me. This lets us conclude:

7. Meaning, in the sense in which words have meaning, is irrelevant to logic.


8. Meaning, in the sense in which propositions have meaning, is relevant to logic.

And this non-psychological sense of meaning is denotation!

So Russell seems to accept the limitations of language and of ideas, of empirical notions of meaning. But this isn't a defect, since there are more robust systems of representation in logic and propositions, which are non-empirical and non-psychological notions of meaning.

A criticism here that I won't fully cash out is this: the traditional Empiricist appeals to the Givenness of sensation to do the heavy lifting. The traditional Rationalist appeals to the Givenness of intellect to do the heavy lifting. It strikes me that Russell wants to accept the Givenness of acquaintance, coupling the traditional Empiricist and the traditional Rationalist together.

I hope this is better than the last formulation of Russell's argument.

No comments: