Abstract
The interaction of the On and Off pathways in motion processing with motion-defined-form (common-fate) signals was investigated. A modified version of the global dot-motion stimulus was used. Signal dot number was fixed at 4 and the number of noise dots was varied to establish the threshold S/N ratio required to determine the direction of the signal dots (up/down). The same dots remained signal dots over the 3-frame motion sequence and two spatial arrangements for them were used: a square pattern and a random pattern that changed from trial to trial. To minimise the possibility of attentional tracking of the signal dots, the first motion frame contained 12 distracter patterns, i.e. noise dots arranged into the same pattern as the signal dots. Three luminance-polarity conditions were used: 1) same polarity for all the dots; 2) mixed polarity for the signal and noise dots (half light, half dark); and 3) same polarity for the signal and mixed for the noise dots. In the same-polarity condition, performance for the square pattern was better than for the random pattern, e.g. threshold of 1.7% versus 5.3%. This advantage was lost when the polarity of the signal and noise dots were mixed. However, when only the noise dots had mixed polarity, performance for the square pattern was markedly better, 0.9%. To further rule out attentional tracking, the number of motion frames was reduced to two, and the same pattern of results was obtained. CONCLUSION: These results show that the On and Off pathways are kept separate in the motion processing of motion-defined-form signals. This is different to the situation with standard global-motion processing, in which they are pooled, suggesting that this form-based motion task is processed in a different area, possibly in the ventral pathway.