DISCUSSION*

Is an Individual Nothing But a Packet of Properties?

Concerning individual objects such as the Charles Bridge or the Etna the question may be raised whether individuals do form a reducible ontological category. The answer to this question is a clear "no" for those philosophers who adhere to a substance theory, or, as it is called these days, a theory of the bare particular. According to these philosophers, what makes a bundle of properties an individual is the bare particular which is a pure individuator (and particularizer).¹ This theory may be criticized on the following grounds. Because the bare individuator is stripped of its properties, it is something we cannot know, which means that we are never able to say what makes the Etna the individual it is. Postulating a "something I know not what" is a solution to which we should not incline unless there is no other way out. Besides, on this theory the bare particular consist merely of its being related to its properties. But I do not see how an entity a can stand in relation to something else, b, if a consists merely of being related to b. In order that something may be a term in relation it must have other characteristics besides being related in this way. And further , the regress argument against relation, which can be found below , holds for the relation between the bare particular and its properties.

As a reaction to the substance-theory, it is said that an individual is nothing but a complex of properties. Russell is best known defender of this theory.² But the theory had already been defended by Moore in 1899.³ Moore's universe consists of nothing but concepts: an existing individual is nothing but a complex concept standing in

unique relation to the concept of existence. Moore defends a one - category ontology: the world consists of nothing but universals. At first, this theory does not appear relevant for philosophy today, but I will show that certain version of this theory is defended these days, and that this version may be criticized along with the same line as Moore's theory.⁴ In this theory no account is given of the fact that certain properties may exist together, making up an individual, whereas others may not: whiteness and grammaticality cannot be properties of one individual, whereas an individual which has redness as property must also have colouredness and extendedness as properties.

Secondly, how can it be accounted for that there is more than one individual? Because existence is a universal, it is the same existence-concept which is related to all properties which make up an individual. This means that there is only one very complex individual. And, thirdly, how is it possible that concepts become related to each other: Bradley's regress argument applies in a strong atomistic conception of the universe. In order that concept A and concept B may be related by concept R1 there has to be a concept R2 to relate A and R1 and a concept R3 to relate B and R1. Because we can always ask the question 'But what relates Rn to R n+1?' we never get to a really relating concept. This means that there can be no individuals, for they are complexes of concepts standing in a certain relation to the concept of existence. Suppose this theory can account for relations, even then it cannot account for relational order in asymmetrical relations. (I only mention this

contents

MARIETJE VAN DER SCHAAR/ DISCUSSION*IS AN INDIVIDUAL . . . ?

59

criticism because it has no direct bearing on the problem of individuals.) Roberto Casati, one of the participants of the Workshop, criticized the bundle theory of individuals because it cannot account for change in individuals: a different property implies a different bundle. This criticism certainly applies to Moore's theory. the first and the third of these arguments against Moore's theory are connected with the atomistic position which is implied in one-category ontology. In such an ontology no account can be given of any connection between the atomic entities, that is, the concepts. In order to avoid confusion about the term atomism I make a distinction between strong atomism and weak atomism. Weak atomism is the position that there are last unanalysable elements, but that these elements are not all there is. This position is exemplified in a theory of the proposition which says that the proposition can be analyzed into ultimate components, but that it is not reducible to these components: the proposition is more than the sum of its components (Bolzano's position). Strong atomism, on the other hand, says that there is nothing but atomistic entities. This is Moore's position. From now on I will use the term 'atomism' as short for 'strong atomism'.

Another argument against this theory was put forward by G.F. Stout in 1921.⁵ If this is true that an individual is nothing but a bundle of universal properties, two individuals with exactly the same properties must be identical, which means that according to this theory the principle of identity of indiscernables is necessary. But because two individuals may be exactly alike

the principle is not necessary. From this contradiction the conclusion is drawn that it is not the case that the individual is nothing but a bundle of universals. Whether this argument is valid is dependent on the statement that two different individuals may be exactly alike, which is not accepted by everyone, and on what you take as an assumption in your reductio ad absurdum argument, for it is possible to take a different proposition as refuted assumption, for example , that assumption that there is a plurality of individuals.

According to Stout, an individual is certainly nothing apart from its properties, but properties are not to be identified with repeatable universals. Properties are as particular as the individuals to which they belong, and most of them are entities in space and time. The blueness of this ball is not identical with the blueness of another ball, even if these blues are of exactly the same shade. According to this theory we ought to distinguish between qualitative identity and numerical identity. Although the colour-properties of the two balls are qualitatively identical, they are numerically different, for there are two blues. In order to distinguish particular properties from universals Stout calls them abstract particulars. These entities can only be separated by the mind, because they necessarily depend on their individuals for their existence. Therefore they are called abstract. This means that the difference between abstract and concrete is not the same as the difference between general and particular. A universal is abstract and general, whereas a property is abstract and

contents

MARIETJE VAN DER SCHAAR/ DISCUSSION*IS AN INDIVIDUAL . . . ?

60

particular, for a property is non-repeatable. A theory of abstract particulars can be very useful to account for acts, processes, events, boundaries, etc.

This idea of properties as abstract particulars is defended at the present time by some Australian philosophers,⁶ most explicitly by Keith Campbell, as for Stout, an individual is nothing apart from its abstract particulars. But there is also an important difference between the two theories. Whereas Stout recognizes besides abstract particulars individuals as separate ontological category, for Campbell there is only one category, that of abstract particulars. Connected with this position is Campbell's idea that certain abstract particulars, such as the blue of the sky, may exist independently. Campbell's one-category theory strongly resembles Moore's. except that the category is not that of concepts but of abstract particulars, or tropes as he sometimes calls them following D.C. Williams. Campbell's theory is open to the same criticism as Moore's except that some arguments have to be reformulated.⁷ Universals and individuals may be seen as two different dimensions in which tropes may be grouped. In this sense there is a correspondence with Mach's Theory of 'Empfindungen' in his book Analyse der Empfindungen (Leipzig, 1886). Mach conceives of the I and the world as constructed out of neutral group of elements ('Elemente' or 'Empfindungen'). The common core enables the two theories to bridge the gap between universals and individuals, or between I and world. The theory of abstract particulars is historically related to the psychology of perception, of which Mach

is a representative. The red as perceived is a particular. When this entity is interpreted as existing independently of a perceiving act, as for example Stout interpreted it, we easily get a theory of abstract particulars.

This theory is a variant of logical realism which tries to overcome the gap between logical entities on the one hand and physical objects and the knowing mind on the other hand. I may directly perceive the blueness of this ball, and that blueness is part of the universal blueness. I will not concentrate on the epistemological problems with which a trope-theory is confronted, but these problems arise immediately because of the impossibility of relations in a one-category theory. How can my mind, which is nothing but a cluster of tropes, ever become connected with a trope which is not compresent with it?

Whether a group must be conceived of as a collection, bundle, or set, or rather as a sum, is not so clearly stated by Campbell. The individual is not a sum of tropes, for Campbell denies that the relation between an individual and its tropes is a relation between a complex whole and its simpler parts.⁸ This means that the relation between tropes and the individual is probably a membership-relation, and that the individual is a set or bundle of tropes. But a set or bundle is certainly not a trope, it belongs rather to the category of concrete particulars, that is, of real individuals. On the other hand, if the individual is nothing but tropes with a relation of compresence, this relation must itself be a trope. But how can this special trope hold the other tropes together. How can it connect? As in Moore'ws theory, the infinitive regress argument against relation holds.

contents

MARIETJE VAN DER SCHAAR/ DISCUSSION*IS AN INDIVIDUAL . . . ?

61

But suppose the relation of compresence does relate, which tropes are related by this relation? Does it relate trope 1 and trope 2, or trope 1,2 and 3, or trope 1 and 3? Where does one individual end and another begin?

The relation of compresence is neither necessary nor sufficient to account for individuals. Apart from the problem that not all individuals are in space, some individuals cannot be described as a group of tropes with a relation of compresence because its tropes are not compresent at all, such as the United Kingdom as it existed in the nineteenth century. And not every group of tropes together with a relation of compresence forms an individual: not every arbitrary part of the group of tropes which does make up an individual is itself an individual, although these tropes are compresent. Or a compresent group makes up more than one individual: Jonah still existed as independent individual after he was eaten by the whale.

The problem of relational order does not arise for a symmetrical relation like of compresence. For asymmetrical relations Campbell introduces the concept of filling the initial place in a relational fact aRb.⁹ But such a structured fact is not merely a sum of tropes, but a structured whole of a, R and b, something which many less strict minds are not inclined to accept in their ontology.

As with Moore's theory no explanation can be given of the fact that certain tropes are necessarily part of one individual, whereas others necessarily

are not. For tropes are as simple as Moore's concepts. In one place Campbell suggests that certain tropes are dependent on others.¹⁰ This means either that they are connected by a relation of dependence -- which is not possible because of the impossibility of any relation: or that there are both independent and dependent entities -- which amounts to a two-category ontology.

It is clear that tropes fulfil both the nature-providing role and the particularizing role. But do they also fulfil the individuating role as is suggested by Campbell?¹¹ Does each trope individuate itself and the individual to which it belongs? Campbell clearly says that the individuation of the trope itself is basic and unanalysable.¹² But what makes what we normally call an individual this individual? How is it possible that its tropes, including the relation of compresence, make an individual this individual as Campbell suggest? Do these tropes have one aspect in common, so that each makes the individual the individual it is, or can they fulfil the individuating role only together?

Because I do not see how tropes can individuate an individual, I cannot conceive of a group of tropes as being something different from a complex property. This is not meant as an argument against Campbell's theory, but rather as an attempt to understand. I cannot see that what makes an individual this individual is something different from the individual itself. It seem that we are in need of an extra ontological category besides the trope. All the tropes of an individual determine its nature, which means that a trope forms part of nature

contents

MARIETJE VAN DER SCHAAR/ DISCUSSION*IS AN INDIVIDUAL . . . ?

62

of an individual. Tropes are necessarily dependent on the individual to which they belong: they are dependent parts. There are also parts of the individual as existent, such as my hand or my nose: these are not necessarily dependent parts. A theory which acknowledges dependent entities (particular properties and relations) and independent entities (individuals, and propositions or states of affairs) is not committed to an atomistic position in the way a one-category ontology is. Bradle's regress does not hold if a relation is dependent entity which is part of an independently existing complex consisting of the relation and the related terms.

According to Campbell, "any two-category theory is going to be bedeviled by a non-relational tie problem"¹³, that is the problem of inherence of a property in an individual. Is this also the case for the two-category theory here proposed? I think not. There is no problem about how an individual may be connected to its properties, if properties are conceived of as dependent or incomplete entities. Only if we start with two complete and independent entities does the problem arise as to how these entities can ever be connected. Campbell's point makes sense only for a two-category theory which is as atomistic as his own.

NOTES

1. This position was defended by Pavel Materna during the Workshop: see also Pavel Tichy, The Foundations of Frege's Logic, Berlin, New York, 1988, p. 210 ff.

2. Bertrand Russell, An Inquiry into Meaning and Truth, London 1980 (first published in 1940), Ch. VI.

3. G.E.Moore, 'The Nature of Judgement', Mind, 8, 1899,pp. 176-193, esp. 182, 183.

4. Armstrong gives four different arguments against the bundle theory of individuals, but although these arguments are valid against Moore's theory, they do not apply to the version of the bundle theory, which I will argue against in this paper, cf. D.M.Armstrong, Universals and Scientific Realism, Cambridge,1978, I, Ch. 9.

5. G.F.Stout, 'The Nature of Universals and Propositions', Proceedings of the British Academy, 10, 1921, pp. 157-172, esp. pp. 161-163: see D.M.Armstrong, Universals, I, Ch. 8 for a clear exposition of this argument.

6. It is no accident that this idea has been spread over Australia, for Stout spent the last years of his life in Sydney.

7. Other arguments against Campbell's theory of tropes are given by J.P. Moreland, 'Keith Campbell and the Trope View of Predication', Australasian Journal of Philosophy, 67, 1989, pp. 379-393. My arguments against Campbell concentrate on the atomistic position which is implied by this theory.

8. Keith Campbell, 'The Metaphysic of Abstract Particulars', Midwest Studies in Philosophy, VI, pp. 477-488, Minneapolis, 1981, p. 482.

9. Keith Campbell, Abstract Particulars, Cambridge, 1990, 3,5.

10. idem, pp. 98, 99.

11. idem, p. 10.

12. idem, 3,7.

13. idem, p. 15.

Marietje van der Schaar

University of Leiden

contents