Quine's Argument

Quine’s overall argument is something like the following:

Important Terms:

Empiricism: minimally, the claim that all ideas (concepts, thoughts, etc) originate in sensory experience.

Analytic: a statement is analytic if, and only if (iff) it is true in virtue of meaning alone

Analytic truths are independent of wordly facts (facts in the world outside of language and our minds)

Synthetic: a statement is synthetic iff its truth depends on the world outside our heads and language

Reductionism: for the empiricist, reductionism is the thesis that all statements of fact can be reduced to statements about sensory experience

Example:the statement ‘electrons are negatively charged’ can be (must be) reduced to a statement involving sensory experiences.

Language: A language is set of terms (singular terms—terms that refer to an individual object like ‘Socrates’; general terms—terms that refer to collections of objects like ‘red’ and logical terms like ‘or’, ‘and’, ‘not’, etc.), rules on how to combine the terms to form simple sentences, and rules on how to combine the sentences to perform logical operations (like deductive and inductive inferences, etc). 

Fact: These are the things in the world that make sentences (or utterances or thoughts) true. For example, the sentence “snow is white” is true because of the fact that snow is white.

Value:These are things that are the objects (the targets) of desire; they can be thought of as either dependent upon desires or independent of desires. For example, my desiring dogs more than cats implies that dogs are more valuable than cats for me (this is value as dependent on desire). Alternatively, love is more desirable than indifference, suggests that love is worthier of being desired than indifference regardless of whether anyone actually desires love (this is value as independent of desire).

Here’s the Main Argument

1. The standard empiricist position includes (a) the analytic/synthetic distinction and (b) reductionism

2. Both (a) and (b) are false.

3. If (a) and (b) are false, then the standard empiricist position is false.

4. Hence, the standard empiricist position is false.

5. If (a) and (b) are false, then the claim that there is a separation between language and facts is false.

6. Hence, the claim that there is a separation between language and facts is false.

7. Hence, there is no separation between language and facts.

8. If there is no separation between language and facts, then changes in a language implies changes in facts.

9. Hence, changes in a language implies changes in facts. 

10. If changes in a language implies changes in facts, then science is determined by all sorts of values.

11. Hence, science is determined by all sorts of values.

While the above argument is valid (if the premises are true, the conclusion must be true as well) we need some defense of some of the premises. I will ignore defending premise 2. Premises 5, 8, and 10 are what I will focus on here. They are the most interesting to me. In another post, I will explain how someone might resist premise 2. So, let’s take a look at premise 5.

5. If (a) and (b) are false, then the claim that there is a separation between language and fact is false.

Premise 5 states that if there is no analytic/synthetic (a/s) distinction and reductionism is false then the fact vs language distinction is false as well. Let’s first look at why the denial of the a/s distinction seems to imply the denial of the fact vs language distinction. 

We think of truth in terms of language and non-linguistic facts. If a sentence S agrees with a non-linguistic fact F, then the statement S is true. According to the a/s distinction some sentences of the language are such that they are true without agreement with a non-linguistic fact—they are true simply in virtue of themselves, independently of the non-linguistic facts. This seems like it has to be the case otherwise we could never specify meanings independent of their relationship to non-linguistic facts. In other words, if this is not true—if there are no analytically true statements—, then the relation between language and non-linguistic fact has to be reconceived. And this is because there will be no such thing as meanings or even truth that is completely separate from experience or the so-called non-linguistic facts. It is this that Quine is ultimately arguing for. The idea that there is a language, there are non-linguistic facts, and there is a relation between them, is precisely what Quine is arguing against. 

Think of it like this: it is quite normal for us think that there is a world outside of our heads and outside of our language and concepts. In order to have true thoughts or beliefs we have to get our thoughts to match the world outside our heads. But this seems to assume that we can think of language (and thought more generally) as independent of the world outside our heads. So, we have thoughts and sentences and utterances and these are supposed to correspond to the world (in order for our thoughts to be true). But if Quine is right and there is really no such thing as analyticity, then there is not really such thing as a language inside our heads that has to match the world in order for it to be true. We can adjust the language anyway we want (due to varying desires/values) or we can adjust the world anyway we want (due to varying desires/values). Indeed, an adjustment to language just is an adjustment of the world, and an adjustment of the world just in an adjustment to language. 

Let’s now look at why the denial of reductionism seems to imply a denial of the language vs fact distinction. Recall that reductionism (in this context) is the claim that all factually meaningful statements can be reduced to observation statements. So, the statement ‘protons are made up of quarks’ is either reducible to a statement about observations (i.e. sensory experiences) or it is factually meaningless. Quine’s point here is that this is all false because no single statement all by itself can be confirmed or disconfirmed by some observation. It is always possible to make adjustments elsewhere in one’s language to accommodate the observation. If I think there are no cats, and you present me with what appears to be a cat, I can (a) change my belief that there are no cats to the belief that there are cats, (b) deny that the appearance is a reason to believe that there are cats, (c) deny that I am observing a cat, or more radically (d) claim that there is no inconsistency in believing both that there are no cats and that what you are holding is a cat. (d) will require me to make adjustments to logic, while (a) through (c) will require me to make adjustments at other, less central places, in my beliefs/language. But the point is that no single statement is ever refuted by some experience. The experience itself is dependent upon the system of thought (the system of language and concepts and logic, etc) that is itself dependent upon various values that such a system of thought embodies. 

Given premise 5, premise 8 seems obvious. 

8. If there is no separation between language and facts, then changes in a language implies changes in facts.

We have just seen that the so-called facts are not just given to us nakedly. They are determined by the language (the system of thought, concepts, logic, etc) that we all employ in our day-to-day interactions with the so-called world. Facts and language co-mingle; they are inseparable once you ditch the a/s distinction and reductionism. Thus, a change in language is a change in facts. 

Given premise 5 and premise 8, premise 10 seems obvious or at least plausible.

10. If changes in a language implies changes in facts, then science is determined by all sorts of values.

The language that we employ is not answerable to the facts. It is not determined by the facts. We normally think that our language should conform to the facts. You say “snow is not white” and we say “adjust your belief/language so that it fits the facts.” But this way of thinking assumes the very things that Quine has argued against, namely, the a/s distinction and reductionism. Once we get rid of those and get rid of the language vs. fact distinction it is easy to see that the conceptual scheme or language that we adopt to describe the world is a product of our desires or values. We adopt the language of the sciences because it helps us in ways that the language of Homeric myths does not. And since we value or desire the way that the language of science helps us and no longer value the way that the language of Homeric myths helps us, we adopt the language of science. By calling the language of science the truth or factual or whatever all that we are really doing is expressing our values and desires. 

So, there you have it. Science is dependent upon value. What is value dependent on? Quine never really says. But notice that we could actually embrace just about everything Quine says and deny the major implication of what he says. Here’s how: suppose that values are objective. In other words, suppose there really are good and bad things, right and wrong ways of thinking and acting and so on and that these are not determined by us. If that is right, then while science may depend on values, values themselves are objective and hence science and facts and whatnot are objective too.