On J. J. Gibson’s New Perspective

Leonardo, Vol. 4, pp. 359-360. Pergamon Press 1971. Printed in Great Britain ON 3. 3 . Nelson Goodman* GIBSON’S PERSPECTIVE In ‘The Information Available in Pictures’ [I], James J. Gibson has abandoned the main position I attacked in Languages of Art [2]. No longer does he regard realistic representation as defined by geometrical optics-as measured by how nearly the bundle of light rays delivered from a picture duplicates a bundle delivered from what is represented. Furthermore, we agreeinrejectingcriteria of realism based on the deceptiveness of pictures or on their similarity to their subjects. And I am more in agreement with him than he seems to realize in insisting upon a significant distinction between paragraphs and pictures-between linguistic and pictorial symbols. Although I wrote that we must read a picture-in other words, interpret rather than merely register it-an appreciable part of my book is devoted to defining the distinctions between linguistic, pictorial, and other symbol systems; and I entirely agree that there is no vocabulary of picturing as there is of saying. Still, our meeting of minds is not complete. In the first place, I do not agreethat reverseperspective ‘cannot be consistently applied’. A long opaque rectangular box, viewed obliquely, may be drawn in * DepartmentofPhilosophy,HarvardUniversity,Emerson Hall, Cambridge, Mass. 02138, U.S.A. (Received 19 April 1971.) standard perspective as in Fig. 1 (a) or in reverse perspective as in Fig. 1 (b). Drawings of the box viewed from the end will be as in Fig. 2. If the box is transparent, then Fig. 2 (b) will serve as a drawing in standard perspective (with the outer square for the near end) or in reverse perspective (with the inner square for the near end). In the second place, I am wary of Gibson’s proposal for a new theory of realistic representation. His outline is somewhat vacillating and vague. In what he calls a formal definition, he says that the optic array from a picture provides information of the same kind as does an optic array from the object but afewsentenceslaterherequires,morereasonably, that the same information be provided. Again, he says that optical information ‘consistsof inuariants, in the mathematical sense, of the structure of an optic array’ but he does not specify, mathematically or otherwise, what these relevant structural invariants are. Later, he seems to identify these invariants with the ‘invariant distinctive features of objects’. But this seems to complete a circle: in saying that the information provided by optic arrays consists of the invariant distinctive features of the object represented, are we saying anything more than that what is constant among the several views or pictures of an object is the object they represent -a statement that may be true enough but that is notably lacking in information. ( a ) Standard Fig. 1. 359 ( b ) Reverse 360 Nelson Goodman ( a ) Standard Fig. 2. ( b ) Reverse If these difficulties are resolved by a fuller statement of the theory, I have other worries. Is it not clear that the same information about the opaque box is provided by Fig. 1 (b), as a drawing in reverse perspective, and by Fig. 1 (a), as a drawing in standard perspective; and that more information about the box is provided by Fig. 2 (b), as a drawing inreverseperspective,than by Fig. 2 (a),asa drawing in standard perspective? Moreover, the information yielded by these drawings obviously depends not only upon them but also upon the system according to which they are interpreted. And this points to my most seriousmisgivingabout Gibson’sproposal : that the notion of ‘the information provided by an opticarray’ ismuch tooellipticalto servethepurpose at hand. The amount and kind of information derived from an optic array or anything else is usually no constant function of what is encountered but varieswiththeprocessing. A symbolmayinform in as many different ways as there are contexts and systems of interpretation. REFERENCES 1. J. J. Gibson,The Information Available in Pictures,Leonard0 4,27 (1971). 2. N. Goodman,Languages of Art (Indianapolisand New York: Bobbs-Merrill, 1968) pp. 10-19. ...