To: Craig S.
From: Geoffrey Klempner
Subject: In what sense is it true to say that Pegasus is a winged horse?
Date: 9th November 2009 13:42

Dear Craig,

Thank you for your email of 3 November, with your essay for the University of London BA Logic paper, in response to the question, 'In what sense, if any, is it true to say that Pegasus is a winged horse?'

This is a decent summary of the main positions which have been taken with regard to the truth conditions of statements about non-existent objects. Your conclusion seems to be, 'in a variety of senses depending on your favourite theory'. Apart from the account given by modal realism which you righly criticize, according to you all the other contenders do the job reasonably well. So you can pick the one you like.

In the background here is a view which has become increasingly prominent amongst analytic philosophers, deriving originally from Quine's attack on the analytic-synthetic distinction. According to this view, when analysing a given concept -- or logical form -- the question is not, 'What do we mean by XYZ?' but rather, 'What concept or analysis of XYZ do we want?' The answer is to be evaluated not just by whether it 'works well enough' but also how well it integrates into our general semantic theory. In other words, there is no 'fact' of the matter here, but it is also not the case that you can 'say what you like'. The question is rather which of the various not-incoherent alternatives is the most fruitful/ useful. For what we are engaged is not so much 'analysis' in the traditional sense but what Quine calls 'regimentation'.

That said, you have thrown down the gauntlet with your defence of Meinong. I am prepared to accept that Meinong has to some extent been unfairly maligned. As it happens, my own views about ontology are fairly relaxed. I don't feel the compulsion that philosophers like Quine feel, to prune things down to bare essentials. However, that said, there seems something worrying about the Meinongian approach.

Consider the following scenario. Inspired by the film, 'The Da Vinci code', I have gone on a quest for the Holy Grail. However, after watching the film my belief that the Holy Grail is some kind of goblet remains unshaken. There are some things which are true of the Holy Grail, whether it exists or not. Jesus drank from it at the last supper. It contained wine on at least one occasion. I also believe that it was made of gold.

The combination of these three statements might or might not be true of the Holy Grail, depending on the truth of my belief. Whereas the first two would be said to be (pace Quine) definitive of 'the Holy Grail'.

As luck would have it, I find the Grail, in an old churchyard in Sheffield. We could of course get into the question of how I *know* that it is the Grail, but let's assume the question of truth is separable from the question of how good my evidence is. Assuming that Jesus was an existing historical figure, it is either true or false that *this* is the goblet he drank from at the Last Supper.

But have I found what I was 'searching' for, on Meinong's theory? I didn't know this, but what I was searching for was indeed an existing object. So far as I knew for sure, however, the Grail was a merely 'subsisting' object like Pegasus.

The problem is that we are tempted to view existence as a property added to subsistence. Some subsisting objects exist while others do not. Or, maybe, some merely subsist until they are brought into existence, while other objects cease to exist but continue to subsist.

If you are going to set out to defend Meinong, then I think it is crucial to show that his theory is not committed to this incoherent view of existence as a property which subsisting entities possess or do not possess. That is not to say that the idea that existence being a first-order property is incoherent in itself. (By definition, not-(Ex)(not-eggists(x)), where (Ex) is the existential quantifier or Fregean 'second-order 'concept of existence and 'eggsists' is the first-order property of existence which every object in the domain of reference has by definition.)

Given this proviso, the question then becomes how easy it would be to construct a formal semantics embodying Meinong's theory. Here, I am tempted towards Quine's view that if we try to 'regiment' talk of subsisting entities then we will find them to be extremely disorderly. How many subsisting Holy Grails are there? My 'Holy Grail' may be more or less idiosyncratic compared to the 'objects' of the propositional attitudes of others engaged in a similar search. The 'rhinoceros in the doorway' objection also creates serious difficulties with the identity conditions for subsisting entities, which might force one to give up the important principle, 'no entity without identity'.

Have you looked at the literature on Meinong? It would give added weight to your case if you can cite a philosopher who has successfully incorporated Meinong's insights into a semantics for statements about non-existent entities. (J.N. Findlay's book, 'Meinong's Theory of Objects and Values' would be a possible starting point for understanding the intricacies of Meinong's theory. I just did a Google search for 'rehabilitating Meinong' which gave a surprisingly large number of entries.)

All the best,

Geoffrey