Analytic and Synthetic: A Quinean Approach

Rob Bass

[The following is based on correspondence with two different people, though only one is quoted here. Since neither of them saw the exact version of my responses and comments that appears here, this is more my work than theirs. Certainly, neither of them deserves even anonymous discredit for having missed points that may not have been made clearly in the actual correspondence.]

On Sun, 24 Feb 2002, X wrote:

[....]

> I do think that the "analytic-synthetic" distinction
> is a valid one (after reading Laurence Bonjour's "In
> Defence of Pure Reason"),
> and I think that confusion
> stems from people confusing what the major issue is,
> and the fact that the philosophers rejecting the
> distinction have resorted to misrepresentation and
> multiple interpretations.

I'm not sure what you have in mind. The distinctions I'd use run approximately like this:

The tautological is what can be certified by logical operations alone, e.g., 'all trees are trees,' 'all trees are either trees or bachelors,' etc. You don't need to know anything about the meanings of the non-logical terms to show that those are true. I use "tautological" for items that are truths of logic as they stand or else are substitution instances of truths of logic. Thus,

"(x) ~(Fx & ~Fx)"

is a truth of logic, and

"There is no man who is (at the same time, in the same respect, etc.) both married and not married"

is a substitution instance of a truth of logic.

I don't think you get within shouting distance of the serious questions about the analytic-synthetic distinction until you start talking about something true in virtue of its meaning -- and not just by virtue of performing logical operations upon it. So, once the tautologies are segregated out, then, of the remainder, analytic statements are, if true, true in virtue of their meaning alone, or, if false, false in virtue of their meaning alone. A standard example of the first is 'all bachelors are unmarried,' and of the second, 'some bachelors are married.'

Then, everything left is synthetic. That is, any statement that is neither tautological nor analytic, whether true or false, does not owe its truth or falsehood to its meaning alone. An example which might be true is 'all bachelors are more than four feet tall.' No inspection of meanings, without paying attention to the heights of bachelors will tell you whether that's true or not.

Now, denying that there is a sharp analytic-synthetic distinction is not the same as denying that some claims are more analytic than others. That is, there may be a property of analyticity that is present or absent to varying degrees, so 'all bachelors are unmarried' is near the analytic end of a spectrum, while 'there is some red-haired bachelor who is 4-feet, 6 inches tall' is much nearer the synthetic end.

Objectivists, as you know, dispose of the analytic-synthetic distinction by identifying meaning with reference. Since one way of expressing that distinction is by way of the claim that some truths, the ones that are analytic, are true by virtue of their meaning alone, while others, the synthetic truths, are not, but depend also upon the referents of synthetic claims (compare 'all bachelors are unmarried' and 'some bachelors are lonely'), identifying meaning with reference surely rules out the possibility of an analytic-synthetic distinction. If the meaning of a term or phrase is the same as what it refers to, no wedge can be driven between what is true in virtue of its meaning and what is true in virtue of something else.

But that is only one possible way of drawing the lines. Broadly speaking, there are five ways that meaning and reference (or, better, the sets of meaningful and referential terms or phrases) could be related. First, they may be identified, so that the meaning of a term is the same as its reference and vice versa. Second, they may be disjoint, so that a term is meaningful if and only if it is not referential. Third, they may be intersecting, so that some terms are both meaningful and referential, some meaningful without being referential, and some referential without being meaningful. Fourth, meaning may be a subset of reference, so that all meaningful terms are referential, but not vice versa. And fifth, reference may be a subset of meaning, so that all referential terms are meaningful, but not vice versa.

Since all parties to the dispute agree that all referential terms are meaningful, two, three and four are all out, because each allows for at least some terms to be referential but meaningless. Only the first and the fifth possibilities are serious contenders. And the first is not a very serious contender. There are obvious problems with identifying meaning and reference. As I've written elsewhere,

[One] problem is with concepts that have no referents. We do not, for example, mean the same thing by "centaur" and "mermaid." However, if meaning were the same thing as reference, we should conclude either that they have the same meaning (because they have the same referents, i.e., none) -- which is absurd -- or else that they have no meaning (because they have no referents) -- which is likewise absurd.

Nor is meaning the same as reference when there are referents. "The animal that, species-typically, has the highest brain-to-body-mass ratio" has the same reference as "human being" but not the same meaning. We could discover, contrary to current scientific opinion, that some other animal had the highest, species-typical, brain-to-body-mass ratio, but that discovery, if made, would not be a discovery that those animals were human beings or that we are not.

Identifying meaning with reference is absurd, so we are left with some version of number five: all referential terms are meaningful, but not all meaningful terms refer. That still leaves open questions as to exactly how they are related. If the boundary were sharp and fixed, that would leave us with a robust version of the analytic-synthetic distinction. My view is that the boundary between the two (a) is not sharp, (b) its non-sharpness is such that reference can make a difference to meaning, and (c) it is not determinable in advance when this will happen. Or, in the terminology of Gilbert Harman, the analytic-synthetic distinction amounts to the belief that there is a principled distinction between dictionaries (which record meanings) and encyclopedias (which give you facts in addition to meanings). But there is no reason to suppose that. There are different things that we believe or stipulate about the terms we use and what they refer to, the various things may be held more or less firmly, but none of them are things that cannot be affected by future evidence. Meaning and reference are (relatively slowly) shifting and interpenetrating fields, not features of terms or judgments that remain fixed independently of what happens elsewhere.

More than this can be said, though. To go back a bit, if we define the analytic as what is true in virtue of its meaning alone, then recognizing that some claims are analytic will depend upon knowing when terms, phrases or statements have the same meaning. That is, analyticity (as Quine pointed out) is going to depend upon synonymy. But how are we to determine that? How do we determine that "bachelor" means the same as (say) "unmarried male"?

Here are three things to be kept distinct in formulating an answer:

1) "I stipulate that 'bachelor' and 'unmarried male' are intersubstitutable."
2) "I mean 'unmarried male' by 'bachelor.'"
3) " 'Unmarried male' has the same meaning as 'bachelor.'"

The first you can do, but that just turns the combination of "all bachelors are unmarried" and the stipulation into a tautology (in my sense) and does not leave it as true in virtue of its meaning.

The second is sometimes confused with the first, but it can be shown to be different by cases in which a person is presented with and discomfitted by a case that matches his stipulation but not his meaning. Example:

A: "Oh, so Max, an unmarried male dog, is a bachelor, then ...?"
B: "No, I meant it had to be human, too, of course."
A: "So, my 13-year old nephew would qualify ...?"
B: "No, adult human."

And so on. You may or may not get something that exactly matches your meaning, but my present point is that even when it's your meaning in question, you don't necessarily have instant access to a set of necessary and sufficient conditions that captures your meaning. You can think you've got it -- and still be wrong.

Given that, a statement about the meaning of a term, even your meaning, is a kind of hypothesis about whether you will find any persuasive counter-examples.

All of that becomes even more clearly important when you switch from talking about your meaning to the meaning. It is at least initially plausible (though not, I think, ultimately quite right) that your meaning is solely a matter of something that goes on or has gone on in your mind. It is not in the least plausible that the meaning is solely a matter of something in your mind. What the meaning of a term is depends on how people generally use the term, when they are using it properly. (The fact that people do not always speak correctly enormously complicates the lexicographer's task, but the complication is unavoidable.) That is something you cannot hope to have assured access to just by considering the contents of your own mind.

Accordingly, if analytic truths are the ones that are true in virtue of their meaning (not in virtue of being tautologous, or stipulated equivalences, or your meaning) and if they are also a priori (in the sense that their truth can be determined by thinking), then there just are not any.

If, you relax the requirement that truth-values for analytic statements be determinable a priori, then you can treat the analyticity of some claim as an hypothesis to be borne out -- or not -- by later investigation.

[....]

Rob
_____
Rob Bass
rhbass@gmail.com
http://oocities.com/amosapient