Philosophy of Language:
Verificationism

Curtis Brown

If we abstract away from certain points which are important in other contexts, we can represent the progression from the 17th-century empiricists to the positivists to Quine's "Two Dogmas of Empiricism" as a gradual development of verificationism. It may be that, as Ian Hacking argues (in Why Does Language Matter to Philosophy?), the views of the classical empiricists about language should not really be called a theory of meaning, and it is certainly true that the positivists were at least as interested in providing a theory about what makes language meaningful (as Carl Hempel puts it, a "criterion of cognitive significance") as they were in providing a theory of meaning. And Quine himself would prefer to simply drop the term "meaning" from our vocabulary. Nevertheless, it can be quite illuminating to think of these views as stages in the development of certain root ideas about meaning.

The basic idea they all share is that meaning at bottom must be a result of links between language and experience. It may be that to some extent we can explain linguistic expressions in terms of other linguistic expressions (by, for example, giving definitions). But ultimately we need some way of breaking out of the circle of words, and connecting language with the world rather than just with other words; empiricism claims that language connects with the world via experience. But each of the three views explains the relation between language and the world differently. I’ll consider classical empiricism and positivism in this handout (with occasional allusions to Quine), and will discuss Quine’s views in a later handout.

Classical Empiricism

The classical empiricists held, essentially, that the meaning of a word was an idea; complex ideas were compounds of more basic ideas, and the most basic ideas, simple ideas, were essentially copies of parts of our experience. So the meaning of any word was, ultimately, something like a set of experiences.

Criticisms

We have already considered some of the details of this view, and some criticisms of it. (My handout on the 17th-century background to the philosophy of language explores both matters in considerably more detail than this one.) For instance, although the view has some plausibility for concrete nouns, it becomes extremely implausible for more abstract nouns, and even more implausible for other parts of speech, such as prepositions or logical connectives. This seems to be symptomatic of a more general difficulty, which is that the most direct bearing of experience on language seems to be as evidence: our experience gives us reason to think claims about the world are true or false. You can't have evidence that, for example, dog. That is, you can't have evidence for words; the notion does not even make sense. What we do have evidence for is sentences: for instance, although it is nonsense to speak of evidence for "dog", it is quite reasonable to speak of evidence for "There is a dog in the room." If experience bears on language primarily as evidence, then it seems that it bears more directly on complete sentences than on individual words. The positivists made use of this insight, and so we reach our second view.

Logical Positivism

Positivism came in many varieties; the following sketch is a rather crude version of some positivistic ideas. But I hope it accurately conveys the flavor of positivistic views.

The relation between language and the world

Just as the empiricists thought that complex ideas had meaning because they were compounded out of simple ideas, and that simple ideas had meaning because they had a direct connection with experience (namely being copies of it), so the positivists thought that some sentences had meaning because they were definable in terms of other sentences, and that at the "bottom" one would find basic sentences, sentences which had their meaning because of their direct connection with experience (in this case being reports of it rather than copies of it). For the positivists these basic sentences were observation sentences. The connection between the world and language thus boils down to a connection between observation sentences, on the one hand, and experiences--the observations reported by observation sentences--on the other. Of course, there is more to the world than experience, and more to language than observation sentences, but the idea is that the world is connected to language only via experience, and experience is connected to the rest of language only via observation sentences. We thus might have the following picture:

Image4.gif (4637 bytes)

In this picture, theoretical sentences are related to observation sentences by chains of definition, while observation sentences stand in a one-to-one connection with particular experiences. (Compare Ayer’s claim that "it is the philosopher’s business to give a correct definition of material things in terms of sensations" (50-51).)

(Empiricism and idealism)

(For our purposes, we needn't worry about the relation between experience and the rest of the world. But I would like to mention one point about this relation. It is sometimes held (e.g. by Michael Devitt and Kim Sterelny in Language and Reality, Cambridge: MIT Press, 1987) that positivism tends to lead to idealism, and our chart helps to show why. Theoretical sentences are about the rest of the world--the parts that we experience only indirectly. But theoretical sentences are held to be definable in terms of observation sentences, which simply report experiences. Since talk about the rest of the world is definable in terms of talk about experiences, it seems to follow that the rest of the world is in some sense constructed out of experiences. Thus, just as Berkeley's empiricism led to idealism, so positivism sometimes led to a version of idealism known as phenomenalism. This seems implicit in Ayer’s claim that Berkeley "discovered . . . that material things must be definable in terms of sense-contents" (53).)

Philosophy as analysis; the analytic/synthetic distinction

A. J. Ayer writes that philosophy is concerned only with "analysis," with analyzing the meanings of words. "The philosopher, as an analyst, is not directly concerned with the physical properties of things. He is concerned only with the way in which we speak about them." In terms of our chart, this means that the philosopher is concerned with explaining the connections between the various items on the chart, and in particular with the definitions which link theoretical sentences to observation sentences. In a natural extension of meaning, analytic philosophy was construed by the positivists as concerned with producing analytic sentences, that is, sentences which are true by virtue of the meanings of the words they contain. These are contrasted with synthetic sentences, which are true because of the way the world is, and can only be established empirically.

So far, I have said nothing about the analytic/synthetic distinction. It is time now to see how it fits into our picture. In the chart above, no analytic sentences are explicitly represented: I intend both the observation sentences and the theoretical sentences to be synthetic, to make empirical claims about the world. Observation statements directly make a claim about experience, and theoretical statements indirectly make a claim about experience, since they are definable in terms of observation sentences. Nevertheless, there is a sense in which analytic sentences are implicitly represented in our chart. They are implicitly represented because they provide the framework for the chart, or to switch metaphors, they provide the rules of the game the chart summarizes. Less metaphorically, analytic sentences describe the links between other sentences. I have said that the theoretical sentences are supposed to be definable in terms of observation sentences: it is the analytic sentences which provide these definitions. Analytic sentences are in effect about the relations between other sentences, not about the world.

Since analytic sentences give definitions, or provide the rules of our language game, they are not empirical. We can define our language however we please, or so it seems; similarly, we can invent the games we will play, and thus decide what the rules of the game will be. So the analytic sentences must be true regardless of what experiences we do or do not have: while empirical sentences are confirmed by some experiences and disconfirmed by others, analytic sentences are, as Quine puts it, "confirmed no matter what." Seeing how the analytic/synthetic distinction is implicit in our chart helps in seeing how fundamental it is to the positivist project. If, as Quine claims, there is no sharp analytic/synthetic distinction, then our nice neat chart breaks down, since there is no longer a clear set of rules relating the theoretical sentences to the observational ones.

Semantics and Epistemology

Another point to notice about the positivistic picture is how closely it intertwines semantics with epistemology. The meaning of a sentence is, ultimately, a certain set of experiences; it is this very same set of experiences which would, if we had them, confirm the sentence, that is, give reason to think it true. So knowing the meaning of a sentence is precisely knowing under what conditions one would be justified in thinking it true. (This is why, to jump ahead a bit, Quine's force-field metaphor simultaneously gives an account of empirical confirmation and an account of the nature of meaning.)

Foundationalism

A final point to notice about the positivist picture just sketched is that it is foundationalist. That is, it takes certain sentences as basic, namely the observation sentences, and builds up the content of the rest of language out of these basic building blocks. The observation sentences are the foundation of language, and the rest of the edifice of language rests on this foundation. This is closely related to a point Quine will make, namely that the positivist picture is reductionist because it thinks all theoretical sentences of the language can be reduced to observational sentences.



Last update: September 24, 2014
Curtis Brown | Philosophy of LanguagePhilosophy Department | Trinity University
cbrown@trinity.edu