Routledge Companion to Philosophy of Language

Gillian Russell, Delia Graff Fara, Gillian Russell, Delia Graff Fara

• Publisher: Routledge
• Year: 2013
• Print ISBN: 9781138776180
• eBook ISBN: 9781136594076
• Topic: Filología
• Subtopic: Lingüística

Philosophy of language is the branch of philosophy that examines the nature of meaning, the relationship of language to reality, and the ways in which we use, learn, and understand language.

The Routledge Companion to Philosophy of Language provides a comprehensive and up-to-date survey of the field, charting its key ideas and movements, and addressing contemporary research and enduring questions in the philosophy of language. Unique to this Companion is clear coverage of research from the related disciplines of formal logic and linguistics, and discussion of the applications in metaphysics, epistemology, ethics and philosophy of mind.

Organized thematically, the Companion is divided into seven sections: Core Topics; Foundations of Semantics; Parts of Speech; Methodology; Logic for Philosophers of Language; Philosophy of Language for the Rest of Philosophy; and Historical Perspectives.

Comprised of 70 never-before-published essays from leading scholars--including Sally Haslanger, Jeffrey King, Sally McConnell-Ginet, Rae Langton, Kit Fine, John MacFarlane, Jeff Pelletier, Scott Soames, Jason Stanley, Stephen Stich and Zoltan Gendler Szabo--the Routledge Companion to Philosophy of Language promises to be the most comprehensive and authoritative resource for students and scholars alike.

Section III Foundations of Semantics

3.1 Names

Barbara Abbott

DOI: 10.4324/9780203206966-26

John Stuart Mill regarded all definite noun phrases (NPs) as names, giving the place which the wisdom or policy of antiquity had destined for the residence of the Abyssinian princes as an example of a complex name (1843: 16). However, nowadays most people take the term “names” as equivalent to “proper names”—just NPs like Madonna and Barack Obama. Proper names have caused a lot of headaches for semanticians and philosophers of language, prompting David Kaplan to remark “if it weren't for the problem of how to get the kids to come in for dinner, I'd be inclined to just junk them” (1978: 224). This chapter will try to make clear what the main problem with proper names is, what solutions have been proposed, and why none of those solutions has been universally adopted as the correct one.

The Problem

The main problem with proper names concerns their contribution to the meaning of sentences (or utterances) in which they occur. (There are other problems, many of them related to this one, but we won't have space to go into them here.) Saying more about this main problem requires a bit of background. We'll start by looking at a different kind of definite NP—definite descriptions—which we will compare and contrast with proper names. Definite descriptions are NPs which (in English) begin with the determiner the. They are linked to their referents by expressing properties of those referents: the star of Evita refers to Madonna because she has the property of having starred in the movie Evita, while the mother of Lourdes refers to Madonna because she has the property of being the mother of Lourdes. Even though those two definite descriptions (the star of Evita and the mother of Lourdes) refer to the same person (Madonna), they do it by expressing different properties that Madonna has. These referent-determining properties are called “senses,” after Gottlob Frege (1892).

The senses of definite descriptions obviously contribute to the meanings of sentences in which they occur—the propositions those sentences express. They do not, however, usually matter to the truth value of simple sentences. (1a) and (1b) below differ only in that one uses the NP the star of Evita and the other uses the mother of Lourdes.

(1) a. The star of Evita will be at the party.b. The mother of Lourdes will be at the party.

These two sentences obviously mean different things—they express different propositions. But it is clear that they have the same truth value; if one is true, then the other must be also. Indeed this is exactly what we would expect, since how we refer to some entity shouldn't matter to the truth of what we say about it. This generalization is the law of substitutivity (also called “Leibniz's Law”), indicating that we would expect to be able to substitute one NP for a coreferential one without affecting the truth value of the containing sentence.

However, there are significant groups of sentences in which one can't exchange one definite description for another with the same referent and be sure of maintaining the same truth value. We focus here on one such group, which consists of sentences about mental attitudes such as knowledge, belief, desire, hope, anticipation, etc. So, it seems that (2a) might be true while (2b) is not.

(2) a. Frank knows that the star of Evita will be at the party.

b. Frank knows that the mother of Lourdes will be at the party.

The standard view is that belief, knowledge, desire, etc. are attitudes towards propositions— hence the term “propositional attitudes” for them. In the case of (2) the propositions in question are the ones expressed by the embedded sentences following knows that (these are the same as the sentences in (1)). And given that the properties expressed by a definite description contribute to the proposition expressed by the sentences it occurs in, it is not surprising that we can have a change of truth value when we exchange coreferential definite descriptions. The proposition which Frank is said to know in (2a) is different from the one he is being said to know in (2b).

We turn now to proper names. Unlike definite descriptions, proper names do not seem to achieve their reference by expressing properties of the things they denote. In other words they seem to be nondescriptional, lacking a Fregean sense. This is true even for names that originally expressed properties. Mill pointed out that, although the English town of Dartmouth originally got its name because it lay at the mouth of the Dart River, if an earthquake should cause the river to change its course the town would still be called by that name (Mill 1843: 20).

If proper names do not have any associated descriptional sense, then exchanging coreferential ones in a sentence should not change the proposition being expressed. The two sentences in (3) should express the same proposition.

(3) a. Mark Twain was a writer of humor.

b. Samuel Clemens was a writer of humor.

And we expect to be able to substitute coreferential proper names for each other in propositional attitude sentences without affecting the truth value of those sentences. However, it's not clear that this is the case; it seems that (4a) and (4b) could differ in truth value.

(4) a. Mary knows that Mark Twain was a writer of humor.

b. Mary knows that Samuel Clemens was a writer of humor.

So our problem is to explain this apparent difference.

The Solutions

The problem sketched above has been a worry to philosophers of language for a long time. First, below, we outline some early proposals to solve it. We then review a seminal work of Saul Kripke's which changed the landscape, and by and large forced people to look in other directions for a solution. Following that, we take up some more recent proposals, separated into three groups. We'll pause after the first group to look at another of Kripke's contributions, which presented new challenges. We finish with the final two groups of approaches. We will not be able here to review all the analyses which have been put forward; important contributions from Mark Richard (1990), Graeme Forbes (1990), and others have had to be omitted. Nevertheless it is hoped that the reader will get some idea of the range of possibilities.

Early Solutions

Both Frege and Bertrand Russell concluded from data like those given above that (contrary to our intuitions and Mill's argument about Dartmouth) proper names do express properties of their referents, similar to those expressed by definite descriptions. Frege suggested that the name Aristotle might express a property like being the student of Plato who taught Alexander the Great (1892, 58 n.), and Russell suggested that the name Bismarck might be equivalent to the first Chancellor of the German Empire (1918: 171). Given this approach we might propose that Mark Twain means something like ‘the most famous American humorist’, while Samuel Clemens expresses a different sense (although it's a little difficult to say what that one would be).

One problem with such a theory is that it does not explain the difference in impact between examples such as those in (5).

(5) a. The student of Plato who taught Alexander the Great was a student of Plato.

b. Aristotle was a student of Plato.

The sentence in (5a) sounds analytic—that is, true just in virtue of its meaning. However, the sentence in (5b) does not sound analytic. But if the name Aristotle expressed the description suggested by Frege, then (5a) and (5b) should have exactly the same meaning, and if one sounds analytic then the other should too.

Others have solved this problem by proposing a looser kind of descriptiveness for proper names. This approach is sometimes called a “cluster” theory. On this kind of view, we associate a number of descriptions with a proper name, not all of which actually need to hold of the denoted entity. John Searle is one follower of the cluster theory. (Cf. also Wittgenstein 1953: ξ79.)

Suppose we ask the users of the name ‘Aristotle’ to state what they regard as certain essential and established facts about him. Their answers would be a set of uniquely referring descriptive statements. Now what I am arguing is that the descriptive force of ‘This is Aristotle’ is to assert that a sufficient but so far unspecified number of these statements are true of this object

(Searle 1958: 171)

Ultimately Searle concluded that only one of the descriptions need hold true for a name to refer. That being the case, we no longer have a problem with the examples in (5). On Searle's theory, the logical form of (5b) is something like (6).

(6) The entity who had at least one of the following properties: X₁, X_p, . . ., X_n, had property X_m, where 1≤ m ≤ n.

And sentences of the form of (6) are not analytic.

Kripke (1972) on Proper Names

In a major series of lectures titled “Naming and necessity” (later published as a book), Kripke (1972) gave a number of arguments against both single description theories and cluster theories, and in favor of Mill's conclusion that proper names are nondescriptional. He pointed out that people can know and use names for individuals when they don't know any identifying properties of those individuals. I've heard of Anaximander, for example, and can use his name to ask questions about him, but at present I don't know anything more about him than that he was an ancient Greek philosopher. Contrary to what the quote from Searle suggests, I could not give any uniquely referring descriptive statements.

Kripke also pointed out that when we speak of alternative possibilities—different possible worlds—definite descriptions differ from proper names in potentially referring to different entities. In the actual world the NP the star of Evita denotes Madonna, but in different circumstances the movie maker could have picked a different person for the starring role, and in such a possible world the star of Evita would refer to that person and not Madonna. Thus (7) is true.

(7) Madonna might not have been the star of Evita.

However, proper names seem to pick the same individual out of any possible world. Because they don't express a property of that entity, they have no way to change their referent under different possible circumstances. Thus (8) does not seem to be true.

(8) Madonna might not have been Madonna.

(In thinking about alternative possible worlds, and how expressions of our language apply to them, it is important to be clear that we are keeping our language fixed. The purpose of these exercises is to become clearer on meaning in our language as we speak it. Thus, it is not relevant that Madonna might not have been named that in a different possible world, or that the word star might not have been used for leading movie actors.)

Kripke called proper names “rigid designators,” meaning that (as we use them) they refer to the same entity in all possible worlds (1972: 48). He proposed that instead of being connected to their referents by expressing properties of those referents, proper names are connected by a historical or causal chain—a chain of uses of the name going back to the original baptism or dubbing.

These facts present a problem for both the single description theory and the cluster theory of proper names. No matter which single description we pick, as long as it is a description of the type suggested by Frege, Russell, or Searle, it will not hold of the same entity in all possible worlds. (There are what we might call “necessary” descriptions, like the sum of 5 and 7, which apply to the same entity (12, in thi...

Table of contents

• Cover Page
• Half Title Page
• Series
• Title Page
• Copyright Page
• Contents
• Notes on Contributors
• Acknowledgements
• Introduction—Gillian Russell
• Section I Core Topics
• Section II Foundations of Semantics
• Section III Foundations of Semantics
• Section IV Methodology
• Section V Logic for Philosophers of Language
• Section VI Philosophy Of Language For The Rest Of Philosophy
• Section VII Historical Perspectives
• Index