To: Chris M.
From: Geoffrey Klempner
Subject: On what modes of existence there are
Date: 24th February 2009 12:02

Dear Christian,

Thank you for your email of 15 February, with your 'free' essay towards the University of London Logic paper, written to your own topic, 'On What Modes of Existence There Are'.

At 6300 words, this is more than double the length of pieces of work that I am normally prepared to accept. Please try to keep your work closer to the 2000-2500 word target length! There is a good reason for this restriction: work which my students send me is queued and allocated to time slots. There is a certain amount of leeway (for example, if I find a particular problem or issue difficult to grapple with and need to think about it). However, if it takes me twice or three times as long to read and digest a submission, the time available for a thoughtful response is inevitably reduced.

I wasn't convinced by your theory of A-existence, B-existence and C-existence. It is a brave try. However, I feel that you haven't sufficiently appreciated the importance of distinctions between the different kinds of entity that exist: spatio-temporal particulars, events, as well as the various grades of 'impure' abstract object, and 'pure' abstract objects. (The set of British Prime ministers since 1800 would be an impure abstract object. The set containing the null set together with the set containing the null set would be a pure abstract object.)

Let's start with the example you give of your dog. You don't give your dog's name, but let us call him Charley. Charley is a spatio-temporal particular. As you say, the criterion for the existence of entities like Charley is observability. However, this is not true of all entities that exist. The criterion for the existence of the null set is something like 'logical definability'.

According to your theory, Charley A-exists by virtue of the fact that we have available the term 'Charley' which might or might not denote anything, or indeed (in the case of self-contradictory descriptions) might not be capable of denoting anything. Charley B-exists by virtue of the fact that you have a concept of Charley, as an object of mental reference concerning which Charley's C-existence of full-blooded existence may either be asserted or denied.

However, we need to look more closely at the question under what circumstances Charley's C-existence might be denied.

Let's say you have a neighbour, Alice, who suffers from paranoid-schizophrenic delusions. Alice believes that she 'owns' a dog called Bruce who sits at the table and eats breakfast with her every morning. Bruce 'exists' for Alice but we would say that Bruce does not exist. What exactly are *we* referring to when we use the term 'Bruce'?

Russell and Quine (along with the vast majority of formal logicians) would say that we are not referring to any entity called 'Bruce'. We are denying the existence of an entity satisfying the description given by Alice: 'The dog who eats breakfast with me every morning.'

Why would there by any motivation to say that 'Bruce' refers to an entity which 'exists' in some weaker sense? To Alice's psychotherapist, Bruce is something very real, a presence in Alice's life which affects practically everything she says and does. 'Bruce wouldn't like me to do that,' Alice might say. 'I did it because Bruce told me to.' For the psychotherapist, 'Bruce' is a term which refers to an existing entity. But this entity is not the kind of entity which the name 'Bruce' purports to refer to, for Alice. Exactly what kind of existing entity this is, is something which one might debate in the pages of psychoanalytic journals.

Say, if you like, that when the psychotherapist talks to Alice about 'Bruce' they are talking at cross-purposes. That is the nature of much psychoanalytic discourse. There is a process of bracketing and interpreting through which the patient's statements are filtered, in order to enable something resembling dialogue.

Arguably, we need to expand our list of kinds of entity even further to include 'theoretical entities'. It seems somehow wrong to describe the psychotherapist's 'Bruce' as merely an 'impure abstract object'. 'Bruce' only exists if the theory in question is true. ('Phlogiston' is an example of an entity which doesn't exist -- when, knowing this, we refer to 'Phlogiston' we are referring to the Phlogiston theory, as an intellectual construct.)

The greatest mistake, however -- which I think you are very close to making -- is to think that you can somehow 'add' C-existence to B-existence.

Charley (your dog) is not a Bruce-type entity to which full-blooded existence has been added. Charley is not, and never was anything but a spatio-temporal particular.

Within the 'univocal' view of existence there is plenty of room for dispute about the ontological status of various kinds of entity. Quine is famously inclined to parsimony. According to Strawson, there is 'no entity without identity'. Gareth Evans has written sceptically about the ontological status of vague objects. Do possible worlds, or in general, possibilia, exist? We don't have to decide for or against parsimony in order to embrace the univocal view. So it is misleading at best to imply that the univocal view requires Quinian parsimony. These questions are up for grabs. What about the ontological status of events? Does the null set exist? Why insist that entities be countable? and so on.

One thing which I would go along with in your account is the idea of 'simulation'. Alice would be an example. When the psychotherapist asks Alice, 'What did you discuss with Bruce at breakfast this morning?' he is engaging in an act of mental simulation, putting himself in some sense in Alice's shoes, drawing conclusions from what she says on the basis of a continually refined and updated mental model. However, to repeat, this process is fully accounted for by allowing 'Bruce' to a theoretical entity, a psychoanalytic term of art. When he says (to a colleague) 'Bruce does not exist' he is not referring to this theoretical entity but rather denying that anything satisfies the description offered by Alice.

So I think there is room for developing the 'simulation' idea, along lines which are consistent with, rather than opposed to, a univocal account of existence.

All the best,

Geoffrey