“God” as a proper name

Most people, even those who don't believe in God, think that there could have been a God. So they will concede that there are some possible worlds in which something like the Jewish, Christian, or Moslem God exists.

This does not just apply to discussions of God's existence. We made a parallel observation about genes in 4.4: if you can't observe them directly, you need some way of associating the term “gene” with things you can observe. And a similar point applies to proper names other than “God” as well. If I say, “Dorothy exists,” there's a sense in which you don't understand what I've said until you have some idea whom it is I'm talking about. And there are two ways in which we are normally introduced to a personal name, such as “Dorothy”—two ways, that is, in which we normally know who it is a name names.

One way of learning a name is by being introduced to the person, Dorothy herself.

A second way of learning a name is by being told some facts about a person along with the name. So if I say, “Dorothy is coming,” and you say, “Who's Dorothy?” I might reply, “She's the woman who wrote that very good book on metaphysics.” Here you come to know of someone not by acquaintance but, as Russell put it, “by descrip­tion.” Knowledge by description is knowledge of a person or thing acquired without direct perception of them. Now, the descrip­tion I just gave you of Dorothy doesn't look like it is enough to iden­tify her uniquely. There might have been several women who have written very good books on metaphysics. (There are!) And though you know this one is called “Dorothy,” there might be several Dorothys who were excellent metaphysical authors. (There are!) But I said that this Dorothy was “the woman who wrote that very good book on metaphysics.” And this implies that I think you know which book I am talking about and that there is only one woman I could mean. If that book has only one author, then you do indeed have a piece of information that is uniquely true of the Dorothy I'm talking about.

If you have an individuating description of a thing, then you asso­ciate something like a Fregean sense with the name of the thing. For, as I said in 3.4, a sense is a way of identifying the referent. And if you have an identifying description, you have a way of identifying the ref­erent. The reason such an individuating description isn't a Fregean sense is that senses are shared among all speakers of the language. But each person can associate a different individuating description with the same name. This is why we don't speak of the “meaning” of a proper name: they don't have shared meanings in this sense, only shared references. When we discussed Frege's “On Sense and Reference,” in 3.4, I went along with Frege's idea that “the Morning Star” had a sense. That was easy to do because this is a rather unusual name in that it has, so to speak, a conception of the referent built into it. You can tell from the name that the object in question is supposed to appear near the horizon in the morning. So this is an example of a name that has a public, shared conception associated with it, which is why it was a good example for Frege to use. But, as I say, for names more generally, we don't require that every user have the same con­ception. Still, everybody needs some conception of a person about whom they are thinking or talking, even if each of us can have a dif­ferent conception of the same person.

One reason it is a good idea to have an individuating description is that since many people can have the same name, there is the possi­bility of confusion unless you know which Dorothy (or John or Mary) I'm talking about.

There are thus three major kinds of possible confusion about names:

1.      You can file information about two different people in the same file.

2.      You can file information about the same person in different files.

3.      You can mistakenly open a file when there isn't a person at all.

But if you have a piece of (true) individuating information in the file, you have a way in which, at least in principle, you can sort these con­fusions out.

Take the first kind of confusion: mixing information about two distinct people. This happens quite often, just because names are shared. I hear you saying something about someone called “Michael” and I file it away in the file for Michael Jordan; but in fact you were talking about Michael Jackson. If I have an individuating description of Michael Jordan—”the world's best basketball player,” say—then all I need to do to avoid mixing these two people up is to find out whether the person you're talking about fits the description. Of course, I may not be able to find this out: but if I have no indi­viduating description of a person, then there's no way I can keep my files from getting muddled up even in principle.

There's a similar reason for wanting individuating descriptions: to avoid filing information about the same person in two files.

Finally, if you have a piece of information that is uniquely satis­fied by one person, then you know that you aren't opening a file for someone who doesn't exist. It's because names work like file labels in this sort of way that we don't often say things like “Dorothy exists.” I wouldn't have opened a file for information if I didn't think the person existed; and until I have a file, I won't really understand whom you're talking about. And you couldn't introduce me to some­one by acquaintance or by an individuating description unless they existed. So the very use of a name ordinarily commits you to the existence of the person named.

Bertrand Russell's analysis suggests that when people say, “So- and-so exists,” what they really mean is that there's something that satisfies a certain individuating description. So, for example, I might tell you a very sad story about someone called “Mary,” as if it were fictional. Suppose at the end of the tale I looked extremely glum. You might seek to cheer me up by saying, “Come on, it's only a story. Mary doesn't exist.” And if I replied, “Oh yes she does,” you would take me to be saying that there was an actual person about whom the story I had just told was true. If we took all the references to Mary out of my story and replaced them with a variable, “X,” and then wrote in front of the story “There is something, X, such that...

This consideration of how we use ordinary proper names, such as “Jane” and “Dorothy,” suggests a way to proceed with thinking about the name “God.” When somebody says, “God exists,” we need to ask what conception of God, what individuating description (or descrip­tions) of God we should rely on in evaluating this claim. And we can understand the person to be saying that there exists something that satisfies that individuating description.

This is a point that David Hume puts in the mouth of Cleanthes, a character in his famous Dialogues Concerning Natural Religion, first published in 1779. Cleanthes says very near the beginning of Part IV of the Dialogues.

The Deity, I can readily allow, possesses many powers and attributes of which we can have no comprehension; but, if our ideas, so far as they go, be not just and adequate and correspondent to his real nature, I know not what there is in this subject worth insisting on.

Cleanthes is arguing that unless we have some conception of God, it is hard to see what point there is in saying we believe in him.

Because “God” is a proper name, it doesn't have a fixed sense associated with it, so different people may identify God in different ways. As we saw with Dorothy, that need not lead to trouble as long as everybody is in fact talking about the same person. But as we also saw, with the mixed-up Michaels and Professor Moriarty just now, we should be on the lookout for two possible confusions. One is that different people are using the word “God” to talk about different persons. The other is that God is known to us in many different ways, but we have not recognized that he is, in fact, only one person.

And, of course, there is always the possibility that we opened a file for God in error, and there is no such being at all.

8.4