On P. F. Strawson's "On Referring"

On Strawson’s Functionalism

In the 20^th century, many philosophers of language look at view words, not in isolation, but as parts that interact with other things to generate meaning. For Russell, denoting phrases interacted with various parts of speech in order to generate meaning for whole sentences. In this sense, words by themselves have no meaning, but they contribute meaning to sentences. Strawson takes this functionalism a step further: words, expressions, and sentences alone have no meaning and have no truth-value by themselves. Meaning and truth-value are absurd concepts until a person uses a sentence.

Strawson rebuts Russell’s analysis of proper names. Russell does not merely build his own theory, as Strawson does not merely build his own theory. Russell constructed his position by arguing against Meinong’s theory of names. According to Russell, Meinong’s arguments entail seemingly contradictory conclusions. In particular, Meinong’s philosophy holds that if a meaningful sentence has a subject as one of its parts, then the subject must have meaning. Furthermore, if the subject has meaning, it must refer to something that subsists. In Russell’s example, when we examine “The king of France is bald,” we notice that Meinong’s philosophy leads us to the conclusion that the denoting phrase ‘king of France’ refers to a subsisting king of France. Russell rejects this absurd conclusion, and advances his own theory of descriptions to solve the problem of meaningful denoting phrases that fail to refer.

Strawson argues that Russell advances an unsuccessful solution to this puzzle. Specifically, Strawson agrees with Russell in rejecting the whole of Meinong’s arguments. However, Strawson points out that Russell accepts a key premise from Meinong. This premise, that logically proper names and descriptions exist for each sentence, according to Strawson, is a blunder. It is possible for a sentence to be significant, and “begin with an expression used in the uniquely referring way” (p. 230) and fall outside of Russell’s categorical analysis. In fact, according to Strawson, this is the rule, rather than mere possibility, since he says, “there are no logically proper names and there are no descriptions” (p. 230).

Strawson’s first step in demonstrating this conclusion is by creating three distinctions of sentence and expression formation. According to Strawson, the first category is the sentence or expression. The second category is a particular use of a sentence or expression. The third category is an utterance of a sentence or expression. To see how this distinction operates, consider this example. Last semester I took several exams. Suppose I talked about my performance on one of these exams, and said shortly after I received my graded exam, “I did well on the exam.” Now, consider an exam that I took two weeks ago. Suppose, also, that I talked about how I did on the more recent exam, and said, following my receipt of the graded exam, “I did well on the exam.” According to Strawson’s distinctions, the sentence and its component expressions are identical in the two occasions. In addition, my usage of the sentence on the two occasions was the same: I used that sentence in the same way. However, Strawson’s distinction leads me to conclude that even though I used the sentence in the same way on two occasions, the utterances of the sentence were different. This difference allows room for the two sentences to differ in meaning, refer to different objects, and have independent truth-values.

The distinction of utterances means, for Strawson, that sentences and expressions alone have no meaning. In fact, Strawson says that unused sentences and expressions have no truth-value either. This argument that forbids unused words and sentences from having meaning or truth-value seems to solve the puzzle of self-contradictory denoting phrases. However, it is less clear how Strawson solves the problem of when a person utters the sentence “The present king of France is wise.”

Strawson’s answer is that the utterance is not genuine (p. 233). According to Strawson:

If, when he utters it, he is not talking about anything, then his use is not a genuine one, but a spurious or pseudo-use: he is not making either a true or a false assertion, though he may think he is. (p. 233)

In particular, when a person utters the sentence, “The present king of France is wise,” he or she does not concurrently assert that the present king of France exists (p. 234). Although such a sentence implies that the present king of France exists, this implying does not mean that the example logically entails that the present king of France exists. This leaves room for Strawson to claim that no contradiction arises, since we are “giving a reason for saying that that the question of whether it is true or false simply does not arise” (p. 234).

Strawson connects this analysis to Russell by arguing that Russell conflates sentences with sentence use and sentence utterance. When Russell argues that sentences are true or false and that denoting phrases have meaning, he puts sentences to use, rather than showing that unused sentences have a truth-value or have meaning. Russell utilizes the context and conventions relevant to a particular utterance. Russell’s error, according to Strawson, is that he uses this information to analyze a non-uttered sentence. For example, in considering the sentence “Scott is the author of Waverley,” Russell discusses the occasion where a nobleman uses the phrase to ask who wrote Waverley, at the end of this analysis, Russell takes the facts specific to the utterance of a particular sentence and applies those details to that sentence as if the sentence was never uttered. Strawson thinks that this is a mistake; an utterance’s context and corresponding conventions are only relevant when a person utters a sentence. If we analyze a sentence aside from its utterance, we are not entitled to appeal to information contained in a context or a set of conventions.

The only way for Russell’s analysis to be fruitful in light of Strawson’s objections is to consider the logical possibilities of a sentence when a person utters it. It may be possible that a certain sentence will always produce a stable meaning and/or truth-value in every possible utterance. Russell produces examples in his response to Strawson: “the square-root of minus one is half the square-root of minus four” and “the cube of three is the integer immediately preceding the second perfect number” (p. 243). Russell argues that these are examples where “egocentricity of wholly absent” (p. 243), but even the above examples feature some egocentricity. For both examples, the sentence still depends on a context and conventions of interpretation to reveal the meaning and the truth-value. Without this information, our refusal to examine utterances robs us of the essential information needed to determine the truth-value and meaning of particular utterances. Russell would have us believe that matters of mathematical analysis are free of human interpretation (i.e., egocentricity), but this claim asks language to function independently of human use.