Philosophy of Language

Curtis Brown

Midterm Review

We have looked at theories regarding two main (related) topics: reference and meaning. Here is a summary of some of the highlights. You should be familiar with, and able to write about, the issues, terminology, and arguments mentioned or briefly summarized below:

I. Theories of Reference

1. Definite Descriptions

A. referential theory: the meaning of a definite description is simply its reference

problem 1: apparent reference to nonexistents (if the referential theory were true, a sentence with a nonreferring definite description would seem to be meaningless: if meaning = reference, then a definite description could only have a meaning if it had a reference. But in fact nonreferring definite descriptions like "the present king of France" seem perfectly meaningful.

problem 2: negative existentials. A special case of problem 1. If the referential theory were true, then a sentence like "the present king of France does not exist," if meaningful, would be self-contradictory (or at least necessarily false). But in fact it seems both meaningful and true.

problem 3: Frege's puzzle (how can identity statements be informative and contingent?) If meaning = reference, then "the morning star = the evening star" is just a fancy way of saying Venus = Venus, which seems to be both necessary and uninformative. However, "the morning star = the evening star" in fact seems to be informative and contingent.

problem 4: Substitutivity. If the referential theory were true, then we would expect that substitution of one definite description for another coreferential definite description would be guaranteed to preserve truth. But in some contexts, e.g. belief contexts, this does not seem to be the case.

B. Russell's theory of definite descriptions. (You should be able to describe the theory, and also to explain how it attempts to deal with the four objections above.)

C. Miscellaneous stuff to know:

context-relativity of definite descriptions

referential vs. attributive uses of definite descriptions

2. Proper names

A. referential theory: meaning of a proper name is simply its reference. Note that this faces exactly the same four objections as the referential theory of definite descriptions (although the examples of course need to change to involve names rather than descriptions).

B. the description theory of proper names (simple version): every proper name is synonymous with a particular definite description. So we can use Russell's theory of definite descriptions to explain the meaning of names as well.

problem 1: the same person may associate different definite descriptions with a name at different times. And even at a single time, there may be no single description that the person would associate with the name.

problem 2: different people associate different descriptions with the same names.

C. the cluster theory: the meaning of a name is a "cluster" of descriptions, not a single description

problem 1: (metaphysical objection) if a name is synonymous with a cluster of descriptions, then it must be a necessary truth that all or most of the descriptions are true of the bearer of the name. But in fact this doesn't seem to be necessary. Even if the descriptions are true, this fact seems contingent, not necessary.

problem 2: (epistemological objection) if a name is synonymous with a cluster of descriptions, then it must be a priori that the descriptions are true of the bearer of the name. But even if the descriptions are true, they seem to be a posteriori rather than a priori.

problem 3: (semantic objection 1) if a name is synonymous with a cluster of descriptions, then it must be true that all or most of the descriptions are true of the bearer of the name. But it seems that this could turn out to be false. (Thales, Gödel/Schmidt: maybe Thales and Gödel don't have any of the properties we associate with them.)

problem 4: (semantic objection 2) if a name is synonymous with a cluster of descriptions, then it must be true that whatever has all or most of the descriptions is the bearer of the name. (The converse of the previous point.) But examples like Gödel/Schmidt seem to show that this isn't correct. (Even if Schmidt had all the characteristics we associate with Gödel, we wouldn't be referring to Schmidt when we use the name 'Gödel'.)

D. The causal-historical theory of how the reference of names is determined. (Note that this is not a theory of the "meaning" of names if meaning has to do with what a competent user understands. Kripke's view is that reference is determined in part by factors that the user may not know about: it's the chain of transmission from the original dubbing to my use that determines what I'm referring to, but referring to, say, Socrates when I use the name 'Socrates' does not require that I know much about the chain of transmission.)

problem 1: empty names.

problem 2: reference change.

problem 3: multiple dubbings.

E. miscellaneous things to know about Kripke, causal-historical theory, and related matters:

rigid designator vs. flaccid designator

Twin Earth

II Theories of Meaning

1. Ideational Theories
(Lockean version: ideas are like images)

problem 1: images are too specific (as Berkeley pointed out, you can't have an image of a triangle in general: it will be isosceles or scalene or . . . )
problem 2: lots of words have no associated images
problem 3: images are too idiosyncratic (i.e. they vary from one speaker to another; thus, if meanings were images, words would all be multiply ambiguous)
problem 4: it's not at all clear how to extend the image theory of meaning to meanings of sentences.

2. "Use" theories (Wittgenstein)

Wittgensteinian ideas to be familiar with include:

language game

rules of language

family resemblances

language as abstraction vs. as concrete social practice

3. Psychological Theories (Grice)

Grice's definition of speaker meaning
roughly: S speaker-means P by uttering x if and only if:
(1) S uttered x intending that A (S's audience) form the belief that P;
(2) S intended that A recognize that (1);
(3) S intended that A form the belief that P because A recognizes that (1)

problem 1: soliloquizing (no audience; possible responses: you can be your own audience, or the audience can be hypothetical)

problem 2: speaker may not intend audience to acquire belief (preaching to the choir; the examinee)

Grice's account of sentence meaning. You don't need to know much about this, as it is only vaguely specified in Lycan, but very roughly the idea is that a sentence S means P if people usually, or under ideal circumstances, speaker-mean that P when they utter a sentence of the same type as S. (By "same type" I mean another instance of the same abstract sentence -- the same words in the same order.) We have a somewhat more developed version of a related account in Lewis's definition of what it is for a population to use a language.

5. Verificationism (Ayer)

verifiability criterion of cognitive significance
verificationist theory of meaning
strong vs. weak verifiability
verifiability in principle vs. in practice
relation between verificationism and classical empiricism (as in e.g. Locke)
Ayer's conception of the role of philosophy (not "metaphysics"! Rather, "analysis.")
analytic/synthetic distinction (related to "analysis" and "analytic philosophy")

Objections to verificationism:

problem 1: examples of propositions that aren't verifiable but seem meaningful (the world came into existence five minutes ago; everything doubled in size overnight)

problem 2: does the verificationist criterion of cognitive significance apply to itself? If so, is it cognitively significant?

problem 3: the Duhem-Quine objection. Individual sentences do not have observational consequences except in conjunction with auxiliary hypotheses. As a result, they are not verifiable or disconfirmable one at a time. Moreover, if meaning = observational consequences, then sentences don't have meanings one by one either! (One route to Quine's rejection of the analytic/synthetic distinction.)

6. Truth-Condition Theories 1: Davidson

T-sentence

7. Truth-Condition Theories 2: Intensional (Possible-Worlds) theories

extension vs. intension (of singular terms, predicates, sentences)
intensions as meanings allow us to distinguish between coextensive but nonsynonymous terms or sentences

8. Relations between (some of) the theories

Lewis's "Languages and Language" as a way of connecting intensional truth-condition theories (which regard a language as an abstract object that we can discuss using set-theoretical apparatus) with Gricean theories of languages as social practices.
The connection, as Lewis sees it: a language (in the abstract sense) is used by a population if and only if there is a convention of truthfulness and trust in the language in that population. (Truthfulness in a language means attempting, by and large, to utter only sentences to which the language assigns, as intensions, a function whose value at the world of the utterance is "True." Trust in a language means that when you hear someone else make an utterance, you believe (by and large) the intension that the language assigns to the sentence.)
So: for a population P to use a language L is for there to be, in P, a convention of truthfulness and trust in L. But what is a convention? In short, "Conventions are regularities in action, or in action and belief, which are arbitrary but perpetuate themselves because they serve some sort of common interest" (p. 164). In more detail:

There is a regularity R of the appropriate kind in P. (In our case, a regularity of truthfulness and trust in L.)
Everyone believes that everyone else acts in accordance with the regularity.
There is a general preference for general conformity to R.
The fact that this regularity obtains gives individuals a reason (in conjunction with their other goals) to continue acting in accordance with the regularity.
There are other possible regularities that could in principle work as well as the actual one, if they were followed as widely.
All the above facts are "common knowledge," in the sense that everyone knows them, and everyone knows that everyone knows them, etc.

You should be able to explain Lewis's account of convention, and explain how it applies (a) to nonlinguistic conventions, e.g. driving on the right, and also (b) to the linguistic conventions that are Lewis's chief interest in the essay.

Last update: March 3, 2004.
Curtis Brown | Philosophy of Language | Philosophy Department | Trinity University
cbrown@trinity.edu