Abstract
Purpose. We have found that mislocalization due to stimulus motion (De Valois & De Valois, 1991; Ramachandran & Anstis, 1990) is more pronounced when the observers pointed the target by action than when they visually judged its location using a ruler (Yamagishi, et al., ARVO 2000). This result is considered to support the idea of separate brain pathways for vision and action (Milner & Goodale, 1995). To gain additional support for this argument, we repeated the experiment for luminance and colour stimuli. If the activation of the dorsal stream (action pathway) is weakened by equilluminant stimuli, target localization would depend more on the ventral stream (perception pathway). It was therefore predicted that difference in localisation bias for perception and action would be less pronounced for colour stimuli than for luminance stimuli. Method. Observers viewed the screen through a mirror. After a drifting Gabor patch (vertical carrier of 0.5 c/deg), modulated either in luminance or in red-green colour, was presented at around 10 deg in the right of the fixation mark, the observers either pointed the target position using a pen without seeing their hand (action task) or reported the location using the visual ruler presented later (perceptual task). Modulation depth was four times the detection threshold. Three speeds (1.5 – 6.3 Hz) were tested. Results. The action task yielded larger positional biases for luminance stimuli with a prominent effect of speed, but less biases for colour stimuli without a distinctive effect of speed. The perceptual task yielded similar biases for luminance and colour stimuli without a large effect of speed. Conclusion. The prediction was correct in essence. Pointing action is subject to anticipatory correction only for luminance motion, supporting the idea that the dorsal pathway is used for action control.
Supported by ATR Human Information Processing Research Laboratories.