In art it is hard to say anything as good as: saying nothing

(Wittgenstein 1980 p. 23e)

The problem that I examine in the thesis is this: what part does the artist play in determining the meaning of his paintings? There are so many plausible theories in circulation that between them they allow too see almost anything in any given painting. How do we, and do we need to decide on a correct account? In this dissertation I will examine Wollheim's account of pictorial meaning as adumbrated in Painting as an Art in relation to writers who have commented either directly or indirectly on the central concepts operative in his account of pictorial meaning. Norman Bryson's semiological account of meaning will be the main contrasting view I examine. The limitations of the study are that the discussion is held within the boundaries of painting and philosophical aesthetics.

I hope to show that there is a strong connection between Wollheim's philosophy of mind and moral philosophy, both informed by Freudian and Kleinian psychoanalysis, and his aesthetics. To understand what it is to be a person we have to distinguish between the person, the person's life, and the living of that life. An analogy can be made between a painter a painting and a process that culminates with the painting, the painter and the painting are understood in terms of the process. The discussion is contextualised by reference to the semiotic, relativist view of meaning in painting typified by Norman Bryson. This involves two different starting points for the explanation of meaning. Wollheim's explanation is in terms of the first person which implies an artist expressing himself and whose expression can be understood in terms of our sharing a universal human nature. Bryson's account is in terms of the third person. This implies meaning by intersubjective agreement based upon a linguistic account of meaning subject to rules codes and conventions whose communication that can only be understood through a public language. Wollheim argues that what a painting means rests on triad of factors. First, the mental state of the artist second, the way this causes him to mark the surface and third, the mental state that the marked surface sets up in a sensitive and informed spectator.

The aim of this dissertation is to examine and evaluate this account of pictorial meaning. The arguments examined here have important consequences for understanding the relation between interpretation and evaluation of painting.

Painting, here and throughout, is to be understood as "painting practised as an art". The concept 'art' is assumed to apply to these paintings, but how some become 'art' and others do not will not be examined. The argument is presented in four sections:

I. Painting a picture that carries meaning is a singularly human activity. The analogy between understanding human beings, by looking at their exhibited behaviour, and of looking at and understanding paintings is introduced. In this first section I examine Wollheim's view that both the artist painter and spectator are roles fulfilled by a person who shares in a universal human nature. This notion is contrasted to the socially constructed 'self' posited by poststructuralists.

II. In this section I consider the relevance of intention in determining pictorial meaning. This is explored against the claim that the work and artist are separate entities and therefore reference to the artist is irrelevant.

III. To understand paintings must we look at them? Wollheim argues that the painting reveals the artist's intention. Here I examine the claim that the meaning of a painting is determined by what can be seen in it in conjunction with visual and non-visual information. This is set against two arguments: that pictorial seeing is not a matter of experience but social consensus and that visual perception may not be a reliable foundation in which to ground pictorial meaning.

IV. Here I examine intended meaning and subsequent interpretation. Is meaning dependent on public agreement or on the individual artist's intention? Is meaning absolute or relative? My discussion focuses on the challenge posed by the semiotic view to the notion of pictorial meaning