To: Shan A.
From: Geoffrey Klempner
Subject: What is the best account of proper names?
Date: 9th July 2008 14:43
Thank you for your email of 30 June, with your essay in response to the University of London exam question, 'Defend what you think is the best account of proper names.'
This is not a bad piece of work. You have done enough research to clarify the most obvious differences between a causal theory of names and a description theory of names, in relation to the problems which prompted the search for a theory of reference in the first place.
Your conclusion, that a 'hybrid' theory works best, is consistent with the problems you raise with each theory. It seems common sense that if there are two rival accounts, and each account has shortcomings which are addressed by the other account, then the easiest option is just to combine them.
However, I have to say that I wasn't really convinced. In fact, I don't get the feeling that you were gripped by this problem. You are reporting on what philosophers have said, for and against the different theories, but I didn't find the arguments very persuasive.
Let's start at the very beginning. Suppose you told a colleague that you are studying philosophy and, as it happens, currently working on an essay. 'What's the essay about?' 'Proper names.' 'What's philosophical about that??'
What IS philosophical about the problem of proper names? What is at stake? Why do we need a theory of names? or for that matter a theory of meaning?
Mill's concern is not really with a theory of meaning at all, but simply with a distinction, obvious to anyone, between terms which convey a meaning, because they have a descriptive content, and terms which we use as labels, where any meaning is accidental and irrelevant. 'A rose by any other name would smell as sweet.'
Let's take the three alleged 'problems', and try to mount a defence of Mill:
i. Whether an identity statement is informative or not depends on the information you already possess. Names are just labels of objects, but you can meet an 'object' in different circumstances. In your knowledge base, you have a label A, which you associate with a list of beliefs about a certain object, and you also have a label B which you associate with another list of beliefs about a certain object. What you don't know is that the two objects are, in fact, the same. For example, two criminals known as 'A' and 'B' in a police database.
ii. 'Calvin' is a fictional label for a fictional character, which we have taken up and use as a real label of that fictional character. A truly 'empty' label, on the other hand, has no use. If I say, 'Bishbashbosh is clever,' you will want to know who Bishbashbosh is. 'No-one, I just made up the name.' 'In that case, you didn't *say* anything.'
iii. How can you have labels for things that don't exist? One response would be to say that existence isn't a property which we attach to things. It is a 'second-order' property which is predicated of a description. But that's a rather fancy explanation, which presupposes Frege's account of quantification. So let's stick with Mill. Harry Potter is a fictional character, like Calvin. So we can use the name as a label. However, we can also say that no actual human being is identical to Harry Potter, and in this sense he 'doesn't exist'.
What is wrong with the label theory (if you can call it a theory)? The key to understanding the problem of proper names is grasping why there is anything for philosophy to grapple with here. What is it that we need to know about the 'basic nature of reference'?
Understanding is a species of knowledge and, as such, requires an explanation. If I asked, how a person can know that there is a tomato on the table, you would tell me some story about perception, and how human knowledge is built up from things which we see, hear, touch etc. etc.
If I asked, how a person can know that Aristotle taught Alexander, then one of the things that needs to be explained is how a person can gain knowledge from a statement in which the terms 'Aristotle' and 'Alexander' appear. Proper names have to be *understood*. Saying that a 'name is a label' doesn't explain anything. The task for a theory of proper names is to explain what a hearer understands, when they understand any statement in which that name appears.
The key issue here is what we require of 'understanding'. This is something which you do not say anything about. Dummett, in his original response to Kripke's 'Naming and Necessity' (added as an appendix in his book 'Frege: Philosophy of Language) makes the point that there is a crucial difference between understanding a sentence containing a name and 'acting like a tape recorder'. If your wife tells you that 'Mr Brown's secretary called about the meeting next week,' she has succeeded in conveying information to you, even though for her 'Mr Brown' is just a label.
What more is needed for understanding? Interestingly, Gareth Evans changed his mind about this point when he came to write the series of lectures (which I attended) which were eventually published, posthumously, as 'Varieties of Reference'. Originally, as you explain, Evans had given an account close to Kripke's. At the time, there was much talk of the 'division of linguistic labour' (Hilary Putnam's term) as an explanation of how we are able to use words that we don't fully understand. But this just puts the problem back. We still need an account of understanding in the full sense, when you are not just repeating something you've heard but really know what you are saying.
But now we've got the problem that with proper names you can never know 'enough'. No-one has ever met Aristotle or Alexander. Are we all just repeating something we've heard about these historical figures?
I hope you can see from this that the question of meaning and understanding is fundamental to epistemology. Our knowledge of things outside our immediate vicinity -- outside our experience -- cannot be merely 'magical'. An account is needed which does justice to the way language works in extending human knowledge throughout space and time.
All the best,