Abstract
Introduction. Multiple cues are often processed separately in different streams or modules, then integrated together. Integration theorists debate whether integration occurs in a “master map” (a dedicated brain area) or via connections between modules encoding separate attributes.
This experiment was a critical test of master area theories. Poirier & Frost (2001) previously found that orientation after-effects (OAE) transfer equally well to the same or different attribute pairs from adaptation pairs. Here, we measured the ability of the system to maintain opposite OAEs contingent on attribute pairs. Most master map theories do not allow for multiple OAEs within the master map.
Methods. We generated two vertically oriented gratings, each of the same spatial frequency, defined by changes in different attributes (luminance, color, texture, or motion). Then they were cut in horizontal strips, and interleaved with a ¼ phase difference (alternate strips contained a piece of grating defined by a different attribute). Global orientation of each single attribute-defined grating became ambiguous because it had ½ phase difference, but with integration across the two attributes, an unambiguous orientation was apparent.
A second similar display used the other two available attributes, and was oriented in the opposite direction. Participants adapted to both displays in temporal alternation within a trial, and were tested randomly with either attribute pair aligned vertically to see if they resulted in the OAE.
Results. Preliminary data shows that some participants could maintain simultaneously two OAEs in opposite directions, contingent on attribute pairs.
Discussion. These results are suggestive of adaptation within and between selected attribute maps rather than in a master map. We suggest a model where modules higher in the hierarchy receive already-integrated information from lower levels, and further processing can elaborate the pattern of integration.