In my last post on the topic of names, I examined the referential theory and explained some of the classic difficulties that it has faced. The descriptivist theory of names attempts to avoid these difficulties by denying that the semantic content of a name is the object to which it refers. Instead the theory claims that some kind of description of the object is included in the semantic content of the name. The first person to propose this kind of theory was Frege – and the descriptivist theory was to remain the dominant view up until the second half of the twentieth century. It was displaced as the dominant theory and replaced once again by the referentialist view, thanks largely to the criticisms of Saul Kripke. Given the failure of the descriptivist theory, we are left in a somewhat baffling position, with no theory being entirely satisfactory. We are left in the position where the process of reference, something as familiar to us as eating, remains an unsolved part of larger mystery concerning how we use language.

Frege proposed that the meaning of a name is not just the object it refers to, but also what he called its ‘sense’. But what is sense? Consider the examples of Hesperus and Phosphorus from the previous post. It’s often referred to as the evening star. Part of our understanding about Hesperus is that it appears in the sky at a certain position in the evening. Frege would claim that this information is included in the sense of the name, and therefore as part of the meaning of the name. In Frege’s words, he would say that the sense encodes the mode of presentation of the object. Intuitively, this is supposed to capture the idea that objects are presented to us in different ways. We never see the whole object, only different aspects of it. Sometimes those aspects are so different from one another that we mistakenly construe them as belonging to different objects – when in fact they belong to the same object. The notion of sense explains this phenomenon.

For Frege, the sense of a name is what we understand when we understand its meaning. Hence we can understand the meaning of a name even when it doesn’t refer to anything at all – like “Santa Claus”. This provides us with an immediate solution to the problem of bearerless names.

We can see how Frege’s theory solves the other two problems as well. Frege’s Puzzle is solved because we can now explain why one of the identity statements is informative, and the other one is not. “Hesperus is Hesperus” is not informative for not only do they both refer to the same things, they also have the same sense. “Hesperus is Phosphorus”, on the other hand, is informative because one name contains the sense of “the morning star” and the other contains the sense of “the evening star”. Hence when we discover the second identity statement is true, we learn that two separate phenomena are actually different modes of presentation of the same object.

The problem of belief is solved by the fact that, acccording to Frege, names occuring in the scope of a belief operator (that’s just a technical way of saying that the name occurs after the words “believes that” in a sentence) refer to their customary sense as opposed to their referent. So in the sentence:

1) Lois believes Superman is strong

‘Superman’ doesn’t refer to the object, but the sense. This prevents substitution of names with the same referent unless they also have the same sense associated.

These are all convenient solutions to the traditional problems associated with the referential theory of names. As a solution it was well regarded for quite some time until Saul Kripke came along and demolished it.

Before we can look at his objections we must first outline one aspect of Frege’s theory that I haven’t yet touched upon – and this is the idea that the sense of a name determines its referent. As an example, consider Hesperus. If the sense of this name is: “The point of light that appears in the sky at a certain location and at a certain time in the evening” then this description determines the referent because only one object answers to this description. Part of the reason why Frege included this condition in his theory was to provide some explanation of how it is that names hook onto objects in the way that they do. This is something that the direct reference theory of names struggles to do and has to appeal to other theories in order to provide some explanation for this. The descriptivist theory of names does not have this problem.

Keep this in mind as we proceed, because it is this aspect of the theory that has caused most of the problems for the descriptivist theory of names.

Consider the sentence:

2) George Bush is the President of the United States

This sentence is true. But consider, it is reasonable to assume that the sense of ‘George Bush’ is equivalent to: ‘the President of the United States’. And for many people, this might be all they know about George Bush, and as such wouldn’t associate any greater sense than that. But if this is correct then the sentence is equivalent to:

3) The President of the United States is the President of the United States.

But sentence 3 is trivially true, while sentence 2 is not. Another way to state this difference is to say that sentence 2 is contingently true (i.e. it could have been false), while sentence 3 is necessarily true (it could not have been false). We call these the modal properties of the sentence.

This argument is due to John Searle, but there were ways out for the descriptivist theory. Kripke’s attack was much stronger – though related in theme. Consider the next sentence:

4) If George Bush exists, then George Bush is the President of the United States

Kripke’s argument was to point out that while we would consider this a contingent statement (after all, he might not have been the president, and very nearly wasn’t) – but the descriptivist theory implies that it is in fact necessary. Why is this so? Well consider the part of the descriptivist theory that I mentioned above, that sense determines reference. If the sense of George Bush is ‘The President of the United States’ – then the object it determines has to be the one that satisfies that description. Whatever object that turns out to be, it must be the president. So it can’t be the case that it’s not the president – hence sentence 4 must be necessarily true.

A related argument from Kripke – is called the epistemic argument. Consider against sentence 4. Kripke points out that we could only know sentence 4 from experience. We have to look out into the world, look at the news headlines and whatnot, to discover whether or not it is true. The descriptivist theory, however, implies that we don’t have to do this. Under the descriptivist theory, all we need do is understand what the meaning of the name is, to know whether sentence 4 is true. In other words, the descriptivist theory would claim this sentence to be known apriori – however its clear that its known empirically.

The arguments turned the tide against the descriptivist theory of names. While philosophers turned wholesale back toward the referentialist theory of names, still they didn’t have adequate solutions to the original three problems. In dismay, some philosophers gave the descriptivist theory another shot and built highly complex and technical theories. The referentialists similarly had to build in complex, theoretical machinary in order to defend their position. We’re left in a position where the seemingly simple act of reference, almost as banal and familiar to us as anything could be, remains without a satisfactory explanation.