Abstract
First- and second-order systems have been proposed to explain visual processing (e.g., Lu & Sperling, 1995). Most studies reported asymmetric transfer of perceptual learning between contrast-defined motion and luminance-defined motion (Chen, et al., 2009; Petrov & Hayes, 2010; Zanker, 1999). Others, however, showed either both-way transfer of adaptation and tilt after-effects (Cruickshank & Schofield, 2004), or no transfer at all (Vaina & Chubb, 2012). Here, we examined transfer of perceptual learning between the first- and second-order mechanisms in a fine, static orientation discrimination task. Four experienced observers were first trained to make orientation discriminations (9 deg. peripheral, monocular viewing) with luminance-modulated (LM) gratings, then trained with contrast-modulated (CM) gratings. In CM condition, observers were tested with gratings of approximately ±0.5 deg (d' = 1.5 in LM), but all failed in the task. We reverted to ± 45 deg. foveal viewing and gradually shaped down to a final difference threshold of about ±5.0 deg. Another four observers were initially trained with the CM grating of ±45 deg. Thresholds were also about 10 times higher than found with LM gratings. In addition, the learning curves of the two groups were similar, indicating no perceptual learning transfer from the first-order system to the second-order system. We are currently investigating potential perceptual learning transfer from second-order to the first-order. Preliminary evidence indicates that there might be some savings, but more data are needed.