Notes on Russell “Descriptions”

One of the questions of focus for this course has been “are truth and reference the only semantic properties?” To put it in other words, are truth and reference all there is to meaning? For some philosophers, the postulation of senses, in addition to reference and truth, has seemed necessary.

One consideration discussed already is the informativeness of identity statements.

Another is the problem of statements about things that don’t exist, as in: James Bond has a license to kill Note that since James Bond doesn’t exist, the noun phrase ‘James Bond’ has no reference. If reference were the only semantic property of subsentential parts, then ‘James Bond’ would be strictly meaningless (like ‘googooblahblah’) and the sentence ‘James Bond has a licesnse to kill’ is neither true nor false (which seems odd) and also strictly meaningless (which seems ludicrous).

One way to solve this ‘problem of non-existing referents’ is to postulate senses: ‘James Bond’, while lacking a referent, nonetheless has a sense. This is the Fregean sort of thing to do.

Another way to solve this problem, and one that is a more explicit target of Russell’s article, is to enrich our ontology to include not only the things that exist, but also things which merely ‘subsist’ which is the funny kind of existence shared by James Bond, Santa Clause, The Easter Bunny, and four sided triangles. This kind of solution was suggested by Meinong (someone mentioned in the article). On this solution, the only semantic property of noun phrases is still refernce, but the number of things that can genuinely be refered to includes not just the things that actually exist.

Russell can be read as showing how a certain kind of noun phrase—definite descriptions—can be treated without having to admit senses or any other spooky things like subsiting non-existing referents into our ontology.

Russell’s critique involves recognizing multiple categories in which names may fall into (which, allegedly, Frege does not recognize). Some names are logically proper names: “simple indefinable words whose meanings do not depend on the meanings of any constituent expressions” (Moore, p. 3). For them, their meaning just is their reference. (Thus, there is no part of their meaning that is constituted by sense.)

What are examples of logically proper names? Not anything that could be defined. Russell thinks that even ‘Plato’ could be defined (perhaps as the student of Socrates and aouthor of The Republic).

What about definite descriptions (singular noun-phrases of the form ‘the so-and-so’)? Can their meaning be identified with their reference? Well, they function very differently from logically proper names. Consider a definite description that doesn’t have a referent, like ‘the present king of France’ (this phrase has no referent because France, at present, has no king). If we were to treat it as a logically proper name by saying that it’s meaning was identical to its reference, we would have to say that it is meaningless, since it has no reference. But obviously it is a meaningful phrase. When you hear or read it, yhou understand it, unlike genuine meaninglessness like ‘ooggah booggah kazooey!!’. Frege would want to say that these sorts of considerations point to the existence of sense.

Russell has another way out. To understand Russell’s analysis, consider putting the definite description together with a predicate (like ‘is bald’) to create a sentence: ‘The present king of france is bald’. If we were treating ‘the present king of france’ as a name, then since it has no referent, the sentence itself would have to be regarded as meaningless (which seems bizzare) and thus as being neither true nor false. But Russell’s inovation was to treat such a sentence as analyzeable (decomposable) into several other sentences, each of which could be seen to be true or false.

For example,

The present king of france is bald

Is analyzed as the conjunction of three sentences:

1) There is at least one person who is king of France

2) There is no more than one person who is king of France

3) Any person who is the king of France is bald

Since the first sentence is false (there is no king of France), the 3 part conjunction is false, and thus the original sentence that is analyzed as the three-way conjunction is false.

“What are the implications for sense? Well, those Fregean names for which an identification of meaning with reference is ruled out (definite descriptions) are not, no this conception, proper constituents of sentences at all, with any kind of meaning of their own. It is a mistake even to ascribe references to them. And a sentnece of the form ‘the so-and-so is the same as the such-and-such’ [an identity statement with definite descriptions as the noun phrases] can be both true and informative, not because there are two different modes of presentation [or senses] of the same thing [referent] but simply because it is both true and informative that at least one thing satisfies both descriptions. . . . The arguments for sense simply lapse” Moore, p. 4.

The key idea here is that a sentence like 'the present king of france is bald' can be regarded as meaningful without having to include any semantic properties other than reference and truth value. No mention of sense, then, is required.

Note the problem left over: what about identity statements made with geniuine logically proper names? Couldn’t they be informative? If so, then there still may be a case for sense. Another problem: what about the predicates? Can their meanings be accounted for only in terms of reference?