Introduction

In his paper, “On Referring,” P.F. Strawson argues against Bertrand Russell’s theory of descriptions, citing that Russell’s theory fails to make the distinction between sentences (or expressions) and the uses of sentences. Consequently, Russell’s theory fails to describe adequately how meaning operates in ordinary language and Strawson argues that we should instead adopt his theory, one that does not suffer from the fatal consequences of Russell’s.

Russell’s Theory of Descriptions

Strawson begins with an account of Russell’s motivation for crafting the theory of descriptions in order to show precisely where Russell goes wrong (41-3). The question motivating Russell’s theory is, by now, a familiar one. Consider the following the sentence:

     
          “The king of France is wise.”

It is clear that it is a meaningful and significant sentence: we all understand it. However, we also know that there is, at present, no king of France. But how can this be? How can this sentence be meaningful and significant if the subject, “the king of France,” fails to describe, or refer to, anything in the real world? Russell’s primary motivation for his theory of descriptions is to avoid Meinong’s response to the problem—-that is, the “postulation of a world of strange entities” that are non-existent (42). According to this account, “the king of France” does refer to an object, but it is a non-existent object that belongs to a special category. Russell wanted nothing to do with such things and rejected this conclusion in favor of his own.

But in the process of rejecting this conclusion, Strawson claims, Russell concedes a far more important principle. Russell believes that only “logically proper names” can occur as subjects of sentences and that these names will lack meaning unless there is, in fact, an individual object to which they refer. In other words, the meaning of a logically proper name “just is the individual object which the expression designates” (43). In our example, “the king of France” is not, by definition, a logically proper name—it fails to designate an individual object in the world. So what is it? Russell argues that it is a complex existential proposition that asserts two things:

          (1)   Existential Condition: “there is a king of France”
          (2)   Uniqueness Condition: “there is only one king of France.”

When combined with the predicate “is wise,” the proposition also asserts a third thing: “there is nothing which is the king of France that is not wise.” After translating the entire sentence, the proposition it asserts may be written:

          (P)  “There is an x, such that x is a present king of France, and nothing other than x is a king of France.”

However, even though our original sentence lacks a logically proper name and, hence, a subject, the sentence is still significant. Accordingly, there seem to be only two different ways that a sentence can be significant for Russell. Either it can contain a logically proper name as its subject or it is not a subject/predicate sentence at all, but rather can be analyzed as a complex kind of existential proposition (43).

Russell’s Big Mistake: the Distinction between a Sentence and its Use

Strawson argues that this dichotomy is false: neither of these conditions confers “significance” upon a sentence. According to Strawson, Russell fails to make the distinction between sentences and expressions, on the one hand, and the _uses_ of sentences and expressions on the other (45). Let us make the distinction now. Consider our sentence: “the king of France is wise.” One can imagine this very same sentence being uttered across different centuries. Despite the fact that the sentence would be voiced by radically different speakers in radically different contexts, something remains relatively stable through time and this something is what Strawson means by a “sentence” (44). Contrast the “sentence,” with the “use of a sentence.” Imagine our speakers again, using the sentence in different time periods. As you might imagine, this very same sentence could be used to refer to different persons—different kings in different time periods. Now, despite the fact the fact that these speakers use this sentence to refer to very different individuals, we should not confuse these different uses of the sentence with the sentence itself.

By recognizing this distinction, we can see how Strawson thinks Russell has fallen prey to a false dichotomy. We cannot say that the sentence, “the king of France is wise,” is about a particular person because it can be used to refer to different persons at different times. The sentence does not designate any individual at all—-it can only be used to designate someone. Similarly, we cannot say that the sentence itself is true or false. Only particular uses of this sentence will express true or false propositions—the sentence itself makes none (44). As Strawson argues:

          The question of whether a sentence is significant has nothing to do with the question of whether the sentence, 
          uttered on a particular occasion, is on that occasion, being used to make a true or false assertion or not, 
          or whether it is being used to refer to, or mention, anything at all. (45)

In other words, whether a sentence is significant or not has nothing to do with whether it expresses an existential proposition or contains a logically proper name.

Strawson: Meaning as General Conventions of Use

So, according to Strawson, why is the sentence “the king of France is wise” meaningful and significant? Strawson states it rather nicely: meaning is a function of the sentence or expression, whereas reference and truth value are functions of the use of the sentence or expression (45). Consequently, when we give the meaning of a sentence, we do not consider whether the sentence itself is true or false, nor do we consider what the sentence refers to. Rather, we give the meaning by simply stating the general rules we follow in order to use the sentence, either to refer to a particular object or to make a true or false assertion. The meaning of a sentence consists in the “rules, habits, [and] conventions governing its correct use on all occasions,” and not merely one usage of it (45).

In order to see clearly how Strawson’s distinction between a sentence and its use differs from Russell’s own account, let us return to our famous example:

          “The king of France is wise.” 

Now, recall that both Strawson and Russell find this sentence significant, but for very different reasons. Russell finds it meaningful in so far as the expression “the king of France” can be translated into an existential proposition, thus rendering the sentence either true or false:

          (P) “There is an x, such that x is a present king of France, nothing other than x is a king of France, and x is wise.”  

Strawson, on the other hand, thinks that Russell has missed the boat on the question of whether the sentence is meaningful because Russell considers the meaning of a particular use of this sentence and thinks that he has given the meaning of the sentence itself. In response, Strawson has us imagine a person walk up to us and sincerely say: “The king of France is wise.” What would our response to such a statement? Would we say that she had uttered a false proposition, as Russell maintains? Not likely. We might say to her, “you’re mistaken,” but the question of whether the sentence is true or false does not arise because there is no king of France. Nevertheless, this does not mean that the sentence is not significant. The sentence “the king of France is wise” is significant, not because the sentence itself refers to or asserts anything, but rather because there are rules and conventions we follow that enable us to use this particular sentence on different occasions. Thus, despite the fact that there is no present king of France, the sentence is still meaningful because it could have been used, at a time in the past perhaps, to refer to a particular person and to assert a true or false assertion.

Strawson and Frege

We can also contrast Strawson’s view with Frege’s theory, found in “On Sense and Reference.” How would Frege treat the sentence: “the king of France is wise”? Recall that Frege distinguished between the sense of a sentence and the reference of a sentence. For space considerations, I will not go through his reasons for believing so, but suffice it to say that Frege regarded the sense of a sentence as the thought it expresses, and the reference of the sentence as its truth value. According to Frege, the sentence “the king of France” expresses a thought and, therefore, has a sense, which is precisely why the sentence is significant. But Frege also held that the sentence has no reference. Why? Due to the fact that the sentence contains an expression, “the king of France,” which itself has no reference (given that there is no present king of France to which the expression refers), the sentence in which it appears fails to have a reference—it fails to have a truth value. It is, for Frege, an imperfection of language that we can have such non-referring expressions.

Strawson’s viewpoint is similar, since he also recognizes that the sentence has no truth value. However, Strawson’s view is significantly opposed to Frege’s view, for Frege regards reference to be a function of the sentence itself and its constituent parts, rather than a particular use of the sentence. Under Strawson’s view, sentences themselves do not refer—only particular uses of that sentence refer. Thus, “the king of France is wise” fails to have a reference because when it is used in modern times, there is no king to which the sentence refers—it is not because of any improper “nonreferring expressions.” It is precisely this “defect” of ordinary language that allows us to use the same sentence across time, to refer to different individuals, and to assert true or false propositions. Consequently, it is no defect of ordinary language that this occurs. Far from it, it is not an imperfection of language, but rather a more accurate characterization of how it operates.

The Significance of Identity Statements

Using Strawson’s theory, we can also answer the question that motivated both Frege and Russell: the cognitive difference between a=a and a=b. Russell claims that his theory of descriptions provides us with the best answer to this question. Strawson, as we might expect, disagrees for all of the reasons we have been exploring. Strawson has us imagine a pair of sentences:

          (1)   “That is the man who swam the channel twice on one day.”
          (2)   “That man swam the channel twice in one day.”

In sentence (1), it does not seem like the predicate of the sentence (“the man who swam the channel twice on one day”) is being used in the same way as the predicate of sentence (2) (“swam the channel twice in one day”). Thus, in order to distinguish the two sentences, we might say that the predicate of sentence (1) is being used referringly_—-that is, that it refers to an individual object (53). But if we do this, it seems that _both the subject and predicate of sentence (1) are referring to the same object. The problem is that we are either making a trivial identity statement in which we identify the man with himself (a=a) or, or we are making reference to the same man twice and saying nothing about him.

Strawson believes this puzzle only arises for those (like Russell) who think that the meaning of an expression is the object to which it refers. How does Strawson answer the puzzle? Strawson argues that the best way to understand the difference between these two different sentences can “be best understood by considering the differences between the circumstances in which you would say [(1)] and the circumstances in which you would say [(2)]” (53). You would only utter sentence (1) if you believed that your audience knew that someone had, in fact, swum the channel twice in one day:

Nathan: “Hey Daniel…you know the man who swum the channel twice in one day that we read about in the paper.”

Daniel: “Yeah…what about him?”

Nathan: “That’s him over there!”

Daniel: “Where? You’re pulling my leg…”

Nathan: “Over there! That is the man who swam the channel twice on one day.”

Daniel: “Well, I’ll be!”

Consequently, you would only utter sentence (2) if you believed that your audience did not know that someone had, in fact, swum the channel twice in one day:

Nathan: “Hey Daniel…guess what?”

Daniel: “If this is another one of your crazy stories…”

Nathan: “No…it’s nothing like that. You see that man over there.”

Daniel: “Yeah.”

Nathan: “That man swam the channel twice in one day.”

Daniel: “You don’t say? He looks pretty small to be a swimmer…”

So the difference in cognitive significance between these two sentences is simply that the sentences have different meanings—that is, they each have corresponding rules of usage and conventional uses. Due to the uses that govern sentences like these, Strawson thinks it is more appropriate to call these “identification-statements” instead of “identity statements” (54). It seems more appropriate to say that we are identifying individuals and their properties, rather than drawing identity statements between objects.

-- DanielS - 19 Sep 2005