To: Bogdan P.
From: Geoffrey Klempner
Subject: Meaning of 'Zeus was a Greek god'
Date: 9th February 2010 12:49

Dear Bogdan,

Thank you for your email of 1 February, with your essay for the University of London BA Logic module, in response to the question, ''Zeus' does not refer to anything. So the claim 'Zeus was a Greek God' is neither true nor meaningful.' Discuss.

You have taken a very definite line on this case, arguing that objects have different kinds of 'existence'. Bill Clinton is a real human being, an inhabitant of the physical world, and his properties are the kinds of properties that you expect to find in physical objects and/or human beings. Sherlock Holmes is a character in the novels of Arthur Conan Doyle. Holmes as the properties of a physical object and/or human being *in the novel*, but what these amount to in reality is information which can be gathered from the pages of Conan Doyle's collected works. 'Sherlock Holmes played the violin' is true if and only if there is some reference to Sherlock Holmes paying the violin in at least one of Conan Doyle's novels or short stories about Holmes. (From recollection, I believe that Holmes does play the violin -- badly.)

Actually, what I've just said is not strictly correct, because novelists sketch their ideas for characters which don't always end up on the printed page. So, for example, if asked, Conan Doyle might explain why Holmes plays the violin by the fact that Holmes' father was a noted violinist. In that case, 'Holmes' father was a noted violinist' is true by virtue of Conan Doyle's character sketch of Holmes, even though it is false if we restrict the domain of investigation to the published works.

I don't know if you can see a problem with this -- because I can. Let's look at Zeus. There is a sizeable body of literature about Zeus, from various sources, and it is not at all clear how we should rate the authenticity of these various sources. A scholar studying the myth of Zeus, or the history of the Greek gods, will take a balanced view, putting more reliance on one source, and less on another.

Unlike the case of Holmes -- at least if we agree to restrict our inquiries to what Conan Doyle wrote about Holmes in his published works -- there is room for genuine disagreement between scholars over whether a particular property belongs to the mythical being Zeus, or not. One way around this is to qualify the statement, 'On one tradition, Zeus has a golden chariot, on another tradition he does not have a chariot.' (I don't know if this is true, I'm just inventing this for the sake of an example.) So what is the truth about Zeus, or Holmes for that matter?

Real objects, objects in the real world have more properties than we will ever discover. It is possible for a large number of our beliefs about a real object to be false, because we have failed to gain a sufficiently good look at it. By contrast, a fictional or mythical character has all and only those properties which it is stated -- on a page, or in an oral tradition -- that it has. A fictional or mythical object is a construct of the things that are said about it.

In that case, statements about fictional or mythical objects lack at least one of the important hallmarks of truth. I think this is what Russell objected to in Meinong's theory. You need a robust sense of reality. You have to understand that 'Bill Clinton was President of the USA' is 'really' true, while 'Holmes played the violin' or 'Zeus rode a chariot' are merely 'true' because someone has made a statement to that effect.

Of course, if we're doing semantics, and looking for a truth conditional explanation for the combinatorial powers of language, then at one level it doesn't really matter that statements about Holmes or Zeus are not 'really' true. You just stick to a minimal account of truth according to which the term 'is true' is the correct substitution for 'is T' in Tarski's schema.

However, if you go that way, you had better make sure that you give some account of your domain of reference which is not completely trivial. At some point, you are going to have to flag the difference between 'real truth' and 'ersatz truth', 'real existence' and 'ersatz existence'. In the case of real truth, you go to the object and compare the object with the predicate that has been applied to it; in the case of ersatz truth, you collect together the various statements that have been made about the object and determine whether or not these are consistent with the predicate which has been applied to it. That's surely a big enough difference to warrant saying that there are no *truths* about Zeus, only things that, at one time or another, have been 'held to be true'.

All the best,

Geoffrey