In lieu of an abstract, here is a brief excerpt of the content:

68 Books whilst still widely separated, can at last be seen to be convergent . Ullman's difficult but rewarding book left me with the feeling that, whilst much remains mysterious in the darkness of unexplained phenomena, a new beacon illuminates an alternative route towards the understanding of perceptual mechanisms. Most of the previous literature on movement perception, on which Ullman draws widely, adopts the psychophysical approach in which visual input (the stimulus for motion) is compared with visual performance (reported sensations of motion). Such studies are essential and have documented many fascinating, if enigmatic, phenomena. The limitation to this approach is that it is extremely difficult to deduce from these experiments how the visual system, treated as a 'black box', achieves its performance. The main reason for this is that, as with other aspects of perception, motion sensations, which may be either genuine or illusory, are frequently clear and unequivocal, even though the visual input consists of insufficient or of ambiguous information. What are the rules, assumptions or procedures that the brain utilizes to interpret an ambiguous visual input as unambiguous motion? Usually this type of question is answered by the supposition that the input is modified and interpreted by relatively high level cognitive processes variously termed unconscious inference (Helmholtz), perceptual set and perceptual schemata (Abercrombie ). Although these concepts have proved useful, they have an embarrassingly Cartesian tenor and offer little insight into the mechanism. Recently, two new approaches have become possible. One is the direct study of the visual mechanism through neural physiology, which is still in its infancy, although it has already achieved notable results. These are not the subject of this book, although I was surprised that Hubel and Wiesel's discovery of unidirectional motion-responsive neurons receives no mention. The other is the computational approach adopted by Ullman, in which is studied tasks accomplished by the visual system in order to investigate the computations it performs. The assumption behind this methodology is that visual performance is the result of computations from the visual input that can be studied independently from the mechanism that accomplishes them. For example, consider the following familiar illusion to which Ullman gives detailed attention: When a simple static image is succeeded after the correct interval by a second static image in a new position, an illusion of motion may result. How does the visual system compute which parts of the first image correspond to the same object or same pattern when they reappear in a different position in the second image? A fascinating solution to this problem of correspondance is offered in Part I. Part II, which deals with 3-dimensional interpretive problems, is equally interesting. The subject matter of this book is directly relevant to animation, to neon-tube displays and to kinetic art. I suspect that it also has a bearing on the illusions of motion utilized in the Op art of numerous painters. Unfortunately, the book is a progress report written for those who can follow arguments in mathematical form. However, artists interested in motion will find that it contains much useful information that is not as difficult to follow as at first appears. Ullman writes succinctly but perspicuously, though he is over-fond of initial letter abbreviations. This is a difficult but absorbing book of interest to a wider readership than that to which it is addressed. Biology and Ethics: Reflections Inspired by a Unesco Symposium . Bruno Ribes, Unesco Press, Paris, 1978. 202 pp. Reviewed by Patrick Shaw* As the subtitle suggests, the book consists of personal reflections following a conference of biologists and moralists. It embodies a tentative approach towards an ethic for biologists, guidelines for research. The approach is described by the 'Dept. of Logic, University of Glasgow, Glasgow G12 8QQ, Scotland. author as a 'functional' one. Broadly, it is summarized in the following quotation: 'Accordingly, the ethical approach to the tasks and purposes of biology consists fundamentally in our view, in urging scientists to discern and continue in the direction in which life has progressed, in other words to promote the further development of its "logic" (p. 55). However, there is not a clear statement of programme. Earlier, for example, Ribes states: 'Our intention, therefore...

pdf

Share