Borders, contours, and mechanism

Kogo and Wagemans claim that subjective contours are assigned from the earliest processing stages. I argue that in making this claim, Kogo and Wagemans are mistaking subjective experience with the perceptual mechanism. There is ample evidence that before figure assignment occurs object properties on opposite sides of unassigned borders compete for perception as figures. In order for these properties to compete, these must be a point in processing at which a border exists before it is assigned to one side.