ABSTRACT
Theories of perception, especially of early vision and of late visual cognition, have burgeoned in the last three decades, but I think they are almost entirely unrelated to each other. I think that the case can be made (and was made by Helmholtz) that mental structures, opaque to the processes of early vision, intervene between sensory input and our reportable knowledge of the visual world. Nakayama (1990) used a title that I wish were free for me to use here: “The Iconic Bottleneck and the Tenuous Link Between Early Visual Processing and Perception” (although actually I discuss a somewhat higher level bottleneck).
