OTHER TITLES IN THE SERIES IN EXPERIMENTAL PSYCHOLOGY Vol. 1 GRAY Pavlov's Typology Vol. 2 HOLLAND The Spiral After-effect Vol. 3 LYNN Attention^ Arousal and the Orientation Reaction Vol. 4 CLARIDGE Personality and Arousal Vol. 5 FELLOWS The Discrimination Process and Development Vol. 6 BEECH and FRANSELLA Research and Experiment in Stuttering Vol. 7 JOFFE Prenatal Determinants of Behaviour Vol. 8 MARTIN and LEVEY The Genesis of the Classical Conditioned Response Vol. 9 BAIRD Psychophysical Analysis of Visual Space Vol. 10 MELDMAN Diseases of Attention and Perception Vol. 11 SHAW and SICHEL Accident Proneness Vol. 12 LYNN Personality and National Character Vol. 13 STROH Vigilance: The Problem of Sustained Attention Vol. 14 FELDMAN and MACCULLOCH Homosexual Behaviour: Therapy and Assessment Vol. 15 RACHMAN: The Effects of Psychotherapy ASPECTS OF MOTION PERCEPTION BY PAUL A. KOLERS P E R G A M ON PRESS O X F O RD • N EW Y O RK • T O R O N TO S Y D N EY • B R A U N S C H W E IG Pergamon Press Ltd., Headington Hill Hall, Oxford Pergamon Press Inc., Maxwell House, Fairview Park, Elmsford, New York 10523 Pergamon of Canada Ltd., 207 Queen's Quay West, Toronto 1 Pergamon Press (Aust.) Pty. Ltd., 19a Boundary Street, Rushcutters Bay, N.S.W. 2011, Australia Vieweg & Sohn GmbH, Burgplatz 1, Braunschweig Copyright © 1972 Paul A. Kolers All Rights Reserved. No part of this publication may be re- produced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of Pergamon Press Ltd. First edition 1972 Library of Congress Catalog Card No. 73-188746 Printed in Great Britain by A. Wheaton & Co., Exeter 08 016843 4 FOR JULES V. COLEMAN AND GEORGE A. MILLER WHO MADE LIGHT OF DARKNESS Human vision is such an enormously rich complex of experiences, and human beings are so diversified in habits and interest, that no two of us value our eyes for quite the same set of reasons. If asked what aspect of vision means the most to them, a watchmaker may answer "acuity" a night flier, "sensitivity," and an artist, "color" But to the animals which invented the vertebrate eye, and hold the patents on most of the features of the human model, the visual registration of movement was of the greatest importance. G. L. WALLS (1963, p. 342) FOREWORD IN 1964 I published a paper that pointed to certain differences in the way the visual system implemented the perception of motion derived from physically moving objects, and illusory or apparent motion derived from flashing lights. The paper came to the notice of H. J. Eysenck, editor of the International Monograph Series in Experimental Psychol- ogy that is published by Pergamon Press. I accepted his invitation to prepare a monograph on the subject of apparent motion, but between that time and 1969 my attention was occupied by other things. At the latter date, at Bell Telephone Laboratories, I undertook the study of some aspects of motion and form perception, particularly as they relate to the quality and fidelity of electronically transmitted images. This monograph grew out of that study. A number of problems confront the engineer interested in trans- mitting pictures electronically, because pictures tend to be quite detailed but the transmission medium is both limited and expensive. Hence a number of engineers actively explore the topic of bandwidth compres- sion, seeking a means by which maximum amounts of visual information can be transmitted at minimum cost. It seemed to me that a contribu- tion to this topic that a psychologist might make would be to study the kinds of information that people actually use in their perception of continuing sequences of briefly presented pictures, such as characterize television transmission of real scenes. Not everything in a picture is perceived, nor do people typically have to look at great amounts of detail in order to come away with an adequate representation of what they are being shown. The reason is that much of our perceiving is anticipatory and inferential; it is based on an active sampling of clues rather than on passive reception of details. What is not known, how- ever, is whether the sampling is haphazard or whether it is rule- governed; whether, that is to say, any aspect of any object is as good a clue to its identity and is as likely to be sampled as any other. The ix X FOREWORD intuitive answer to this uncertainty is a negative one; and in this regard concepts of salient or distinguishing features that characterize objects and facilitate their identification have come to the fore in recent years. Hence we may ask, what aspects of pictures best aid their recog- nition; what features most readily aid in distinguishing them? What are, in fact, the salient or distinctive features of pictures? Can one formalize these characteristics? Questions of this kind motivated some of the research I shall report later in this monograph. The direction followed in seeking an answer was the classical one of dealing with simple, stripped down instances of stimuli. Although the pictures most likely to be transmitted in an elec- tronic system are of people in movement, we began with simpler instances; in fact, with such classically simple ones as plane geometric figures. Moreover, we studied their perceptibility not when they were presented one at a time, measured for their specific geometry and luminance, but when they were presented in brief sequences. Therefore we were immediately dealing with illusory motion of the configurations, for it is a characteristic of the visual system that it perceives brief, properly sequenced presentations as objects undergoing spatial trans- lation. It is just this characteristic, in fact, that the motion-picture pro- jector and the television screen take advantage of, albeit in substantially different ways, in triggering those perceptions of objects in motion that we do see with those media. As we shall see below, the technological successes with those media do not depend upon a correct understanding of their effects on the nervous system; the technology works, but no currently available account of the perceptibility of forms in motion is correct. Theory lags practice rather extensively in this domain. One outcome of our research was the discovery that the visual system actively constructs, implements, and fills in features of objects derived from a sampling of the physical presentation. The visual system does not merely record what has been presented to it (if it records that at all); it actively creates its own pictorial reality. Moreover, the mechanisms underlying this creation seem to be substantially different from the mechanisms utilized in feature analysis. What role, if any, that feature analysis plays in normal perception is not entirely clear. These facts emerged from a study I carried out with James R. Pomerantz in the fall and winter of 1969 and published in the Journal xi FOREWORD of Experimental Psychology in 1971. Intrigued and puzzled by them, but under a deadline created by other forces, we followed through with this investigation during the early spring of 1970. Reminded by Pergamon Press of my committment to do a monograph on apparent motion, I thought that these new investigations might be embedded in the context of this old problem; for the fact is that people have been looking at motion pictures for more than seventy years, and at television screens for more than twenty-five, but the mechanisms governing these perceptions are still unknown. Regrettably, our work does not answer all of the questions it set out to answer. Indeed, we were blocked for some considerable time, for in our thinking we continued to seek positive answers to certain questions whereas it becomes increasingly likely that some of them cannot be answered positively. Thus for conceptual and administrative reasons it was not possible to round out this work, and it is incomplete. I present it at least for the questions it raises if not for the answers it provides. On and off between 1964 and 1968 I discussed various aspects of the perception of illusory motion at Project Zero at the Harvard Graduate School of Education. I think it was in response to those beneficent challenges that some concern with pictorial notation crept into my own thinking, although whatever distortions it may have there are not to be laid at the door of Project Zero or its director, Professor Nelson Goodman. It was during that period also that I carried out some preliminary experiments on form and motion perception at the Research Laboratory of Electronics, in connection with a course on perception and cognition I taught in the Department of Electrical Engineering at Massachusetts Institute of Technology. Moreover, as my own command of German is poor, I am indebted to several friends and former col- leagues for aid with various aspects of the older literature. In particular I thank the former Misses Katharine Gilbert and Jean Mechlowitz, and Dr. D. C. Milne in this connection. I also thank Drs. David N. Perkins, Ronald Cohen and Ivan Bodis-Wollner for their comments on the manuscript. My greatest obligation, however, is to James R. Pomerantz for his conscientious, creative, and energetic contributions to our short-lived collegiality. I am not sure that I would have had the patience or subtlety to pursue these elusive realities alone; indeed he always shared and sometimes even led in the formulation of our xii FOREWORD experiments. And of course I shall forever remember my stay at Bell Telephone Laboratories, and here record my thanks to that institution for enabling me to carry out this research. Toronto, Ontario ACKNOWLEDGEMENT I thank the authors and the American Psychological Association for permission to reproduce the following figures: 3.6, 4.3, 4.4, 4.5, 6.2. xii FOREWORD experiments. And of course I shall forever remember my stay at Bell Telephone Laboratories, and here record my thanks to that institution for enabling me to carry out this research. Toronto, Ontario ACKNOWLEDGEMENT I thank the authors and the American Psychological Association for permission to reproduce the following figures: 3.6, 4.3, 4.4, 4.5, 6.2.