Thursday, July 21, 2011

Indispensability of proper names

To: Julian P.
From: Geoffrey Klempner
Subject: Indispensability of proper names
Date: 23 December 2003 10:41

Dear Julian,

Thank you for your e-mail of 14 December with your second University of London essay, 'On the Indispensability of Proper Names'.

I did not find your argument psychologistic, nor did I think the style too personal. This is clear, well organized and easy to read.

The Julian

If we construe 'Julian' strictly as a proper name, then 'the Julian' makes no sense. (It makes as little sense as 'the this'.) However, there does seem to be a way to construe 'the Julian' as a description, namely, 'the person called "Julian"'. What's wrong with this? (Assume that we can add surnames etc. if necessary.)

Imagine that I am a philosopher of language who believes he has worked out the definitive theory of proper names. So, when I say, 'The Julian', meaning, 'the person called "Julian"' I am able to give necessary and sufficient conditions for any object to be 'the person called "Julian"' and moreover I am confident that these conditions do indeed apply to just one object. This would be a case where proper names are used by the populace while the philosopher kings (such as myself) translate proper name discourse into non-proper name discourse.

This would be a way for proper names to be useful, but ultimately dispensable. But it assumes the possibility of a 'definitive theory of proper names' as used by the populace, which is one of the points at issue.

Knowing about Bonnie

You have uniquely identifying knowledge of Bonnie but are unable to convey this knowledge to anyone else. Anyone who picks up knowledge about Bonnie from you relies on you for the 'route to reference', e.g. 'the woman Julian calls "Bonnie" who gave him helpful hints'. In a world where people did not have names we would say, 'the woman with blonde hair and a red sweater who gave Julian helpful hints'.

Even if you know Bonnie's full name, that's not going to help others distinguish her from other women with the same name. You will still need to add descriptive knowledge. This descriptive knowledge will either rely on you for the route to reference (as above), or it will be sufficient for anyone to identify Bonnie without relying on you to make the identification.


The description, 'the xth person born' clearly won't work, for the reasons you give. There are two issues here. The first concerns the nature of descriptions of the form, 'the xth...'. E.g. 'the first person to drown in the sea'. We assume that someone must have been the first person to drown in the sea, or the 223455th person born, but these kinds of description are parasitic on descriptions which enable us to locate particulars in space and time. It would be impossible for all descriptions to be like the parasitic case.

The second concerns Donnellan's point about referential and attributive descriptions. Even if philosopher kings succeeded getting a law passed banning proper names, people would naturally use definite descriptions in just the way we now use proper names. Proper names are just so useful, that it seems impossible to imagine that a race of beings could develop a language which lacked anything that functioned like a proper name. However, even if we accept this point, it falls short of a demonstration that proper names are logically indispensable.

Telly Tubbies

The difference between using a language and communicating through pictures is that pictures cannot communicate thoughts about other times, or about generality - unless you turn them into a language (i.e. a system of signs whose 'pictorial form' is 'logical form', in the sense of the 'Tractatus'). You are right that in the Telly Tubbies pictorial communication there is nothing that functions like a proper name. But nothing follows from that with regard to our actual language. (Hence your admission, 'that is not where I expected to end up!'?)

A Theory of Recognition

Here is the most important bit. I read it as point about Frege, that sense cannot be reduced to descriptions. Our knowledge of objects essentially involves a capacity to encounter things and become acquainted with them. (I would argue - some might disagree - that this encounter necessarily involves not just perception but physical agency. So there could not be a race of intelligent trees who, unlike the Ents, merely looked around and talked to one another but did not act or move in any other way.) There has to be a level at which we know objects which is below that of descriptions, however one construes this 'knowledge'.

This is really just Russell's point, but I am stretching 'acquaintance' to objects in the physical world.

However, this does not establish the necessity for proper names. What it does is make a place for proper names which can also be filled in by demonstratives or by the referential use of descriptions.

You say, 'any trigger that activates the web of connections associated with a referent object is equivalent to a proper name of that referent'. I don't think you meant this literally. A smell is just a smell, surely? What is true that we physically relate to the world in a way that involves non-descriptive encounter with objects.

Happy holidays!

All the best,