Abstract

This chapter shows that the embedded view has explanatory power. Inferential accounts are often preferred to competing accounts — for example, to Gibsonian alternatives — because they explain a variety of visual phenomena that are otherwise mysterious. The chapter provides embedded models of visual cases that pose difficulties for direct views. These include cases where vision appears to be inflexible, as in stability, constancy, misperception, and illusions, as well as cases where the visual system displays a measure of flexibility, as in multi-stability. The chapter shows that the embedded view is also compatible with cognitive penetrability. One of the crucial moves is to recognize that the embedded view does not hold that no processing happens between retinal stimulation and visual percepts. It only holds that such processing does not consist of an inference.