Abstract
We investigated whether attention is useful for binding shape to color by studying the acquisition of novel shape/color combinations.
Methods. We created pictures of 60 familiar objects depicted in novel colors, such as a green dog, a blue frog, etc. Each picture had a strong outlined shape and a vivid color (red, green, yellow, blue, or purple) which was fixed for each picture. Two such pictures were flashed simultaneously, left and right of fixation, on each trial. On half the trials the pictures were the same, and on half they were different. On every trial, all the subjects reported ‘same’ or ‘different’, an easy task requiring little attention. Half the subjects also named the pictures, while the other half of the subjects identified a masked letter at fixation. Both of these tasks were attention-demanding. After each block of 30 trials, all subjects received a Yes/No recognition test for the shapes of ten of the pictures portrayed in monochrome and then reported the color of any recognized shapes.
Results. For subjects who were required to name the pictures on each trial, and who presumably attended them closely, recognition of both shapes and shape/color combinations improved rapidly and was highly accurate within 4 blocks (120 trials). For the remaining subjects, who were required to attend to the masked letters, shape recognition increased at the same rapid rate, but shape/color combinations were still reported poorly after 6 blocks.
Conclusion. Attention benefits the perceptual acquisition of novel shape/color combinations in the periphery, even when the shapes themselves do not benefit from attention.