Graphic Languages ed. by F. Nake, A. Rosenfeld (review)

278 Books Graphic Languages. F. Nake and A. Rosenfeld, eds. North-Holland Pub. Co., Amsterdam, 1972. 442 pp.. illus. 519.30. Reviewed by Michael Thompson* This is a collection of over 20 papers and transcriptions of three panel discussions; the proceedings of the IFIP Working Conference on GraphicLanguages held in 1972. About 50 specialists were invited and proceeded to amaze each other with their enormous diversity of approaches and intentions. Indeed, the innocent title ‘Graphic Languages’ can be interpreted in many ways representing areas of research that sometimes have little in common except that they deal in some way with two-dimensional material intended for viewing and the use of computers. Rosenfeld attempts to establish some order with a threeway classification that I mention here to give the reader an idea of the contents of the book: (1) Conip~irer graphics. The user gives instructions to the machine and in return gets pictorial output, usually in the form of a line-drawing. Here the book can refer to the language used by the man sitting at the display console, that is, the ‘command language’, or ‘interactive language’ or it could mean the language in which the computer programs are written. (2) Inioge riianipularion. Here the objective is usually to improve an image in some way, by enhancing contrast or removing irrelevent features. There is an artificial gulf created between this topic and the previous one, because computer graphics tend to deal with line drawings whereas image manipulation works with greyscale pictures. The picture is divided into an array of say 1000 x 1000 squares and the intensity and maybe colour is measured for every one of them. In this case, the book is not too obviously applicable, since the image manipulation is done by large specially written systems and the language is just FORTRAN. Image manipulation appears in the book only as the subject of one of the panel discussions. (3) Imnge Arinlysis. The process here starts with either a grey-scale picture or a line-drawing and ends with an analysis that may be regarded as ‘recognition’, ‘description ’ or ‘perception’, depending on the use to which it is to be put. There are two main approaches here and, although further research may unite them, at present they are two sides of a controversy. One side being concerned with the ‘linguistic’ aspects of the visual (loosely this means hierarchical, organizational, structural and non-numerical aspects). Their approach tends to be mathematical and general, with little application forthcoming. The other approach seeks application, is problem oriented and at this early stage of research suffers from specialization, e.g. to images of blood cells, or to images of clouds or to line drawings of rectangular solid blocks. The first approach tends to stress syntax in a desire for generality, whereas workers seeking application stress semantics. How does all this relate to art? Well, few answers are available in terms of the work of artists and I have speculated on this subject in this issue of Leonardo (Leonardo 7 , 227 (1974)). Certainly an increasing knowledge of visual matters might be relevant to the artist. Also each new technique invented offers a challenge to creativity. There are two papers that explicitly deal with this. Firstly, G. Courtieux describes a system to output animated pictures on a cathode ray tube with the idea of having ‘students from the Fine Arts School use the language and give advice on how it should be changed‘. He stresses that the language has no facilities for calculations at all and here the brief discussion is well worth reading, for the specialist audience cannot take this and are sure that very complicated equations of motion must be needed. Courtieux replies that his system is not devised for teaching physics but for visual creation. The other paper is by Ken Knowlton who describes his work with Stan Vanderbeek and Lillian Schwartz. He makes the point that *1 Remez Road, Kadimah, Israel. although artists would like to explore a wide variety of possible ways of describing and producing graphics, ‘any one of these ways can become an artistic medium only if it developed iir dcprh to the point where artists and viewers...