It is widely accepted that motion and color are processed in separate brain areas of primates. Numerous studies on monkeys suggest that neural mechanisms responsible for motion processing respond faster than those for color. Recent studies on humans, however, provide contradictory evidence. Is this discrepancy due to a gap between species (animal vs. human), or between measures (neurophysiological vs. behavioral)? To help resolve this issue, event-related potentials were acquired as human participants viewed motion and color stimuli. Results indicated that the physiological response evoked by motion arose earlier than that by color, which is consistent with previous ¢ndings in animals.This temporal precedence of motion signal processing over color was corroborated in a parallel behavioral experiment.
IntroductionPhysiological research in primates suggests that different attributes, especially motion and color, are processed by different parts of the visual system [1,2]. Motion and color travel from the retina to the cortex primarily through the magnocellular and parvocellular pathway, respectively. In the cortex, motion receives specialized processing in area MT whereas color processing occurs in area V4 (V8) [3][4][5].Given the fact that motion and color information processing have different and relatively independent pathways, it would not be surprising if they also demonstrate different temporal properties. Indeed, single-unit studies in animals suggested that neurons specialized for motion processing respond earlier and faster than those for color in both the retina and lateral geniculate nucleus [6,7]. At the cortical level, it was found that neurons specialized for motion signals in V2 activate earlier than those for color [8]. It was also reported that area V4 activates later than MT and MST [9].This motion-led-color asynchrony can, however, be reversed or eliminated when human observers perceive joint occurrences of motion and color. A striking phenomenon, motion-color asynchrony, was recently demonstrated by Moutoussis and Zeki [10]. Participants viewed moving squares with oscillating changes in both the color and the direction of motion, the temporal phase of which varied in different trials. Participants had to indicate the combined perception of color and direction (e.g. green up, red down or green down, red up). Results showed that observers tended to associate the change in direction of motion together with the color change that occurred about 60-80 milliseconds (ms) later. This kind of 'miss-binding' was interpreted as color being processed faster than motion. Similar phenomena were observed in other paradigms [11,12]. Nishida and Johnston [13], however, used a different response mode and showed that motion and color were perceived without asynchrony. Bedell and colleagues [14] also proposed a twostage model to account for the different observations on color-motion asynchrony, with the first stage dependent on the sensory processing and the second stage dependent more on task.Clearly, there is a...