My aim is to show that the accounts of depiction offered by Christopher Peacocke and Robert Hopkins assume rather than explain one of the central features of depiction. This feature is pictorial realism. It is a constraint upon any adequate theory of depiction that it be able to explain pictorial realism; however, Peacocke and Hopkins seek to meet this constraint by employing the notion of resemblance. I raise three problems with Peacocke's account and point out an error in Hopkins's use (...) of solid angles (upon which his notion of resemblance rests). It is suggested that while these theories must be rejected, there are various non-resemblance theories, including that proposed by Gombrich, which might prove adequate. (shrink)

There is not one but many ways to picture the world--Australian "x-ray" pictures, cubish collages, Amerindian split-style figures, and pictures in two-point perspective each draw attention to different features of what they represent. Understanding Pictures argues that this diversity is the central fact with which a theory of figurative pictures must reckon. Lopes advances the theory that identifying pictures' subjects is akin to recognizing objects whose appearances have changed over time. He develops a schema for categorizing the different ways pictures (...) represent--the different kinds of meaning they have--and argues that that depiction's epistemic value lies in its representational diversity. He also offers a novel account of the phenomenology of pictorial experience, comparing pictures to visual prostheses like mirrors and binoculars. (shrink)

This paper examines a form of pictorial realism that has epistemic import. Gombrich and Schier claim that some pictures are realistic because they convey accurate information. The difficulty is that judgments of realism vary across cultural and historical contexts. Goodman counters that pictures belong to different systems and realistic pictures belong to familiar systems. However, this does not explain the revelatory realism' of pictures in novel systems. I propose that two views can be combined: a realistic picture is one which (...) belongs to a system that conveys the kind of information that suits the needs of users in a context. (shrink)

Pluralism—the incommensurability and, at times, incompatibility of objective ends—is not relativism, nor, a fortiori, subjectivism, nor the allegedly unbridgeable differences of emotional attitude on which some modern positivists, emotivists, existentialists, nationalists and, indeed, relativistic sociologists and anthropologists found their accounts.

The A.W. Mellon lectures in the fine arts 1956, National Gallery of Art, Washington.

This revised edition includes a new Preface outlining developments in Relevance Theory since 1986, discussing the more serious criticisms of the theory, and ...

In this paper, I discuss the influential view that depiction, like language, depends on arbitrary conventions. I argue that this view, however it is elaborated, is false. Any adequate account of depiction must be consistent with the distinctive features of depiction. One such feature is depictive generativity. I argue that, to be consistent with depictive generativity, conventionalism must hold that depiction depends on conventions for the depiction of basic properties of a picture’s object. I then argue that two considerations jointly (...) preclude depiction from being governed by such conventions. Firstly, conventions must be salient to those who employ them. Secondly, those parts of pictures that depict basic properties of objects are not salient to the makers and interpreters of pictures. (shrink)