A Discussion of Pictured Impossibles with Reference to Nelson Goodman’s Analysis of Fictional Objects

Leonardo, Vol. 11, pp. 129-130. Pergamon Press 1978. Printed in Great Britain A DISCUSSION OF PICTURED IMPOSSIBLES WITH REFERENCE TO NELSON GOODMAN’S ANALYSIS OF FICTIONAL OBJECTS Sheldon Richmond* 1. Do sentencesaboutfictional objects refer to anything? Do pictures of fictional objects refer to anything? The first question has been discussed since Plato’s Sophist. The second, though analogous in form, has been considered recently by Nelson Goodman [I]. Nominalist philosophers prefer to find ways of explicatingsentences about fictional ‘objects’ that do not confuse fictional ‘objects’ with real objects. They take their cue from Bertrand Russell, who stated that sentences apparently naming fictional objects are actually descriptions with nulldenotation , (that is, not about any object). Goodman adapts Russell’s view to his explication of representational pictures of fictional objects. Can his explication of such pictures be extended to depictions of impossible objects? If not, do they denote objects of a special fictional kind? In my analysis I shall use the following terms: Picture refers to an object consisting of one or more characters, each made of marks placed on a 2-dimensional surface, that are part of symbolic systems. Symbolic system refers to a system of characters (or, classes of marks) that may or may not refer to objects in the real world. A picture or sentence may or may not have reference; a picture or sentence that has no reference is a null-denoting sjmbol. According to semiotics, a representational picture is an iconicsymhol;it representsthrough resemblancethe visual appearance of real objects [2]. According to Goodman, a representational picture is a symbol in a disjoint and nondifferentiated symbolic system. A symbol system is disjoint when its marks can be classified into one or another character; a symbol system is non-differentiated when its marks cannot be classified uniquely. For instance, sentences (unlike pictures) have marks that can be classified uniquely; and so sentences are part of differentiated symbolic systems. 2. A picture of a real person is a representation of that person when it denotes the person. A picture of Pegasus is a representation of Pegasus, but it does not denote Pegasus, because Pegasusis a fictional winged horse. How then can one say that a picture of Pegasus represents Pegasus? Goodman answers that by convention a picture of Pegasus is a representation of Pegasus, it is a ’Pegasus- *Philosopher. 2 Inman St., Apt. No. 2, Cambridge. MA 02138, U.S.A. (Received 24 Aug. 1977) picture’. A picture of a real person can function as a representation because it denotes a unique real object. A picture of Pegasus can only function as a definite description of a fictional animal; that is, the picture has null-denotation as regards the real world. Goodman’s explication appears trivial in the light of Russell’s view of fictional sentences. Goodman avoids using the semioticdefinition for a representational picture as an iconic symbol by saying that representational pictures do not represent through resemblance but through membership in a symbolic system. According to him, pictures are symbols that represent objects by their role in a symbolic system. However, unlike verbal descriptions, pictorial descriptions are syntactically nondifferentiated ; their marks are not uniquely classifiable into characters [I, pp. 135-1361. This is the point that allows Goodman to transfer Russell’s explication of fictional sentences to pictures of fictional objects: both fictional sentencesand pictures of fictional ‘objects’are in symbolic systems that are syntactically disjoint; that is, have marks classifiable into one or another character [1, p. 1341.Note that Goodman does not explicitly assert that syntactic disjointness must be a minimum condition for pictures. However, his remarks in Reference 1 (pp. 225 ff.) might be taken to imply that this is true. (Pictures, unlike sentences, are syntactically non-differentiated; that is, pictures do not have marks uniquely classifiable into characters, such as do letters of an alphabet [1. p. 2261.) 3. Does Goodman’s explication of representational pictures of fictional objects apply to pictures of impossible objects? The latter, apparently executed according to the rules of linear geometric perspective for depicting 3-dimensional objects in pictorial space,cannot exist in real space. M. C. Escher made pictures ofcomplex objects that are impossible in the real...