CONTENTS

1   INTRODUCTION

2   EXPOSITION

3   THE HIERARCHY OF REFERENCE

4   THE NEUTRALITY OF LANGUAGE

5   CONCLUSIONS

3 THE HIERARCHY OF REFERENCE

Having looked at the structure of language in general, let us now consider each of the types of reference in turn, at each stage considering the ontological implications of having such a category. At each stage we are trying to demonstrate two points:

(a)	that it makes sense to treat these expressions as having an objective reference;
(b)	that the Fregean account of the semantics of the expressions is unique, being determined by the structure of saying.

The first of these points relates back to Assertion B of Section 2.1, while the second relates to Assertion A.

3.1 Objects

The first hurdle to be overcome before one can understand the Fregean notion of objects is to grasp the extreme generality of this notion. One must abandon all narrow preconceptions, derived from experience of the actual, physical world, of what constitutes an object. One must strip away all ideas of time, space and matter. The Fregean, logical concept of object is prior to all these. Objects are whatever we talk about, and it is the world, through our experience of it, which determines what things we are able to talk referentially about. Language provides the bare categories and then the world fills them up. Physical objects, spirits of the dead, gods and demons, numbers, points and lines, these are all putative objects in the Fregean sense.

This agnosticism about what object exist should not be taken as a licence for ontological extravagance. The existence claims of any putative class of objects have to be examined with care and the due degree of scepticism. But what Fregean ontology provides for us is an initial liberation, a liberation from parochial preconceptions about what objects have to be like. We need not, for example, reject out of hand a Platonic interpretation of arithmetic, in which numbers are said to be real objects, merely on the grounds that numbers are utterly unlike physical objects and that therefore our means of knowing about their existence and nature must be quite different from the knowledge we gain through our physical senses.

But can we do without objects altogether? The object by itself appears to be a formless substance, a ground of being to which properties and relations are added to give it form. All that the object supplies is the being. This picture then leads to the suspicion, especially when one has become convinced of the doubtfulness of the status of existence as a predicate, that the object is nothing more than a conventional peg on which to hang properties and relations, and can be eliminated in favour of collections of properties, and relations between these collections.

The basic intuition here derives, at least in part, from a materialist view of the world, namely that all objects ultimately owe their existence to being made of matter, and all their properties derive from the way this matter is configured in space. But this is a contingent theory of the actual world, and such preconceptions have no place at the level of logical ontology. Another part of the motivation for the elimination of objects is the following, perhaps unconscious, identifications:

properties and relations   =   experience, phenomena

objects   =   underlying being, noumena.

If these identifications are coupled to a positivistic prejudice that experience is all that we can meaningfully talk about, then there is a perfect motivation for the elimination of objects.

In resisting the elimination of objects from ontology in this way, we have to challenge the presuppositions on which the whole picture is built. Objects are not physically separate from their properties. This is part of the content of Frege's notion of the incompleteness of predicate reference. We can not speak separately of an object and its properties, and then link them together with some relation like, "... is attached to ...". This is to reduce the original object and its properties to a collection of related objects, and as we saw above this reduction always fails to eliminate all incomplete reference. We can, by logical analysis separate object and property, but the fact that this is not a "physical" or "objectual" separation is permanently recorded by the incompleteness of the property. Note that this distinction between objectual and logical separation is one of the things which can not be expressed in the "sayable model", but can only be pointed to by saying, "it's not like this".

As an actual example of the claim that objects can be eliminated in favour of their features, consider the following quotation from a physics textbook, Goldstein's "Classical Mechanics". Talking about the ontology of field theory, he says,

"... the electromagnetic field was long thought of in terms of the elastic vibrations of a mysterious ether. Only in recent times was it realised (to quote an apt phrase of Professor S L Quimby) that the ether had no other role than being the subject of the verb 'to undulate'."

Following this logic, and a thorough, though mistaken, indoctrination in the belief that Special Relativity eliminated the possibility of an underlying extended object which does the undulating, physicists have persisted in the belief that they can talk of field excitations without an ontological commitment to the underlying field. This is incoherent. Pace Professor Quimby, verbs need objects if we are to make coherent sense. This is not just an arbitrary convention of language as human invention. It is an imperfect natural-language reflection of the underlying necessary structure of Fregean language. If one does not want to talk about fields, then one is required not to speak of undulations.

Another temptation we must resist saying that the properties are what can be known, with the objects either relegated to the status of the inaccessible underlying reality or eliminated altogether. When I see a tree, I do not see greenness and tree-shapedness. Instead it is the object I see, as well as the truth of certain statements about the tree. I see:

the tree, that it is green, that it is tree-shaped.

The claim that I see the tree is not an ontological or metaphysical one. It is a statement about language and how it refers. If one harboured specific doubts about the ontology of trees, then one might wish to introduce objects of some other sort, and express our sensations in terms of seeing them. But we would not have stopped talking about seeing objects.

A belief in objects corresponding to the names, atomic or compound, one uses, is a matter of ontological commitment. This virtue, so named by Quine (though in a somewhat different context), is fundamental to a realist approach to language and the world. We do not use language without using names (and I argue here for the stronger conclusion that we can not do so), so we are committed to the existence of objects. This is a linguistic point and says nothing substantive about the world. What objects happen to exist is then the business of substantive ontology.

Click here to go to the next part of Section 3