Letters

Letters Readers’ comments on texts published in Leonordo are welcomed. ’Ihe Editors reserve the right toshorten letters. Letters should be written in English and sent to the Main Editorial OfTice. Comments on Generating Caricatures Since well before the brothers Carracci drew their caricatures at the turn ofthe 17thcentury, exaggerated drawings have captured artists and their audience. Susan Brennan’s Caricature Generator can now fire imaginations of computer-literate artists and students ofvisual perception. As a member ofthe latter group let me say why this is so. Brennan (Leonardo 18. 170- I78 [ 19851) suggested that “Caricature.. .seeks, paradoxically , to be more like a face than the face itself.” A caricature captures the essence of its subject better than many other possible representations. But be on guard about this point. Indeed, Ryan and Schwartz [I], Fraisse and Elkin [2] and Dwyer [3] all reported that stylized, accentuated drawings were more easily identified, or aided learning more, than photographs of the same objects. But none studied faces. When Perkins and Hagen [4] compared caricatures and photographs of people, they found no evidencethat the former were better than the latter. Yet I have always felt this to be the wrong comparison. As employed by Brennan, bandwidth compression is the key idea for guiding such comparisons. Notice that her caricatures need only 200 points to start with but that even a poor quality photograph needs a million. What is psychologically important, then, is that caricatures seem much better than rotoscoped (traced) line drawings from a photograph. Brennan’s is the first available system to allow systematic tests ofthis idea. As examples, her Fig. 7 tracings of Fay Dunaway, Diane Feinstein, John F. Kennedy and Elizabeth Taylor look less like their subjects than the computer-composed caricatures. Why caricature works so well is not clear, but it seems related to a phenomenon in ethology known as supernormality. For example, a whiteeggspeckled with dark brown spots is preferred by a nesting plover over its own light brown speckled egg [5]. In this case, in many others, and in caricature, some information to which the perceiver is acutely attuned seems enhanced and hence preferred. Following Gibson [6], some authors [7,8] have supposed that invariants in faces are preserved in photographs, and perhaps even better in caricatures; but since these are unnamed and unmeasured this argument, while it may eventually prove true, is currently little more than explanation by incantation. What is most exciting about Brennan’s system is that it allows exploration ofthespace of depiction, investigating the preferred amounts of exaggeration, the best norms, and the types of facial information to which we are most sensitive. There is no real need for an expert caricaturist; the theory of caricature has been made sufficiently concrete so that it can be implemented on a machine. References I. T.A. Ryan and C. Schwartz, “Speed of Perception as a Function of Mode of 2. 3. 4. 5. 6. 7. 8. Presentation”, American Journal o f Psychology 69, 60-69 (1956). P. Fraisse et E.H. Elkin, “Etude GCnCtique de I’Influences des Modes de PrCsentation sur le Seuil de Reconnaissance d’Objets Familiers”, L’AnnPe Psychologique 63, 1- 12 (1963). F. M. Dwyer, Jr, “Adapting Visual Illustrations for Effective Learning”, Harvard Educational Review 37,250-263 (1967). D.N. Perkins and M. Hagen, “Convention , Context, and Caricature”, in M. Hagen, ed., The Perception o f Pictures, Vol. I (New York: Academic Press, 1980.) N. Tinbergen, The Study o f Instinct (Oxford: Clarendon Press, 195I). J.J.Gibson, “The Information Available in Pictures”, Leonardo 4, 27-35 (1972). E.H. Gombrich, “The Mask and the Face: The Perception of Physiognomic Likeness in Life and in Art”, in E.H. Gomrich, J. Hochberg and M. Black, Art. Perception. and Reality (Baltimore, MD: Johns Hopkins University Press, 1972). D.N. Perkins, “Caricature and Recognition ”, Studies in the Anthropology of Visual Communication 2, 1-23 (1974). James E. Cutting Department o f Psychology Uris Hall Cornell University Ithaca, N Y 14853. U.S.A. Comments on “Electronic Thinking Cap: Microcomputer-Enhanced Creative ProblemSolving ” From my vantage point as a psychologist, the Metros article (Leonardo 18, 100-104 119851) raises a number of provocative questions...