Neuroscience Research, 18 (1993) 11-18 11 © 1993 Elsevier Scientific Publishers Ireland, Ltd. All rights reserved 0168-0102/93/$06.00 NSR 00679 Update article Visual attention revealed by an illusion of motion Okihide Hikosaka a,., Satoru Miyauchi a and Shinsuke Shimojo b a Laboratory of Neural Control, National Institute for Physiological Sciences, Okazaki 444, Japan, and b Department of Psychology, University of Tokyo, Tokyo 153, Japan (Received 16 July 1993; accepted 23 July 1993) Key words: Visual attention; Motion illusion; Motion perception; Illusion; Psychophysics; Human Summary Attention is a mechanism to select sensory information. It is a modulatory process which normally cannot be observed as overt responses. A new psychophysical method using an illusion of motion perception allowed us to visualize the field of the magnitude of attention and its dynamic changes. Based on our experiments using this method we suggest that (1) both passive (bottom-up) and active (top-down) attention exert their effects on the early stages of visual processing, (2) active attention can quickly and briefly be replaced by passive attention induced by an external event, but can be restored in about 400 ms, and (3) attention is directed to an object, not space, and follows the object as it moves. Introduction Attention is a mechanism to select sensory informa- tion. Based on the selected information we make ac- tions and store memories. The actions and memories then are used to select sensory information. Through such a cyclic process the brain will adapt itself to environment and acquire automaticity. However, we know little of attention. This is largely because we cannot observe or measure attention di- rectly. When we fail to perceive a particular object among many that are in our view, we realize that attention was not directed to the particular object; we realize that sensory information has been selected for some reason; we realize that the brain is not like a mirror. Many psychologists and physiologists have at- tempted to describe the nature of attention and to * Corresponding author. Present address: Department of Physi- ology, Juntendo University, School of Medicine, Hongo 2-1-1, Bunkyo-ku, Tokyo 113, Japan. Tel.: 03-5802-1026. discover its mechanism, but many questions remain to be resolved. What aspect of sensory information is modified by attention? - Is it intensity (Bashinski and Bacharach, 1980; Egly and Homa, 1984) or speed (Eriksen and Hoffman, 1972; Posner, 1980; Sternberg and Knoll, 1973) that is modified? At what level of sensory processing does attention act? - Does atten- tion act before perception (early selection; Broadbent, 1958; Hillyard, 1985) or after perception (late selec- tion; Deutch and Deutch, 1963)? We can direct our attention to an object voluntarily; or our attention can be drawn to an event in our environment. Can these phenomena be called attention (Kahneman and Treis- man, 1984; Schneider and Shiffrin, 1977; Yantis and Jonides, 1984)? Do they share common mechanisms? What is the objective of attention? - Is it space, object, or its attribute (Kanwisher and Driver, 1992)? Our experiments on attention started when we dis- covered an illusion of motion (Hikosaka et al., 1993a). We have shown in the following experiments that the illusion represents the intensity (or its slope) of atten- tion. We then would like to answer the questions raised above.