Version: v2, Published online: 2010
Retrieved December 11, 2019, from https://www.rep.routledge.com/articles/thematic/vision/v-2
3. Direct theories of vision
The ‘Gestalt’ movement of the early twentieth century rejected the view, prevalent since Berkeley, that complex percepts can be analysed into simple sensory components (see Gestalt psychology). According to the Gestalt theorists, perception is holistic: perceptual wholes are not built up out of more basic sensory elements, in the way, for example, that a painting is just the combination of all the paint-covered segments of the canvas. Gestalt theorists claimed further that perception is direct – perceptual processing is not correctly described in terms of psychological or mental processes. The structure of a visual experience is to be explicated in terms of the structure of the underlying brain states, that is, in neurophysiological terms. The Gestalt psychologist Wolfgang Köhler characterized as a physical gestalt any dynamic system that settles into an equilibrium state of minimal energy. A soap bubble forming a perfect sphere is an example of a physical gestalt, as is, Köhler argued, the brain producing an organized percept. Köhler proposed a theory that appealed to electrical fields within the brain to account for perception (and all other mental processes). Gestalt speculative physiology was not borne out by subsequent brain research, which failed to discover evidence of Gestalt mechanisms implicated in perceptual processing. That said, however, work by Gestalt theorists to characterize perception in terms of very general organizational principles (such as proximity, the idea that nearby elements in the image tend to be grouped together, or similarity, the idea that visually similar elements in the image tend to be grouped together) has proved useful in motivating the search for computational mechanisms that realize and explain these principles.
The psychologist James J. Gibson shared with the Gestalt theorists the belief that visual perception is not mediated by processes characterizable in psychological terms. Gibson argued that indirect theorists have mischaracterized the information in the optical array. If the effective stimulus for the visual system is taken to be retinal flow (the smooth transformations of the optic array as we move about), then, according to Gibson, there are important constancies in the stimulus that indirect theorists have typically missed. There is therefore no need to posit inferences, calculations, memories, association of ideas, or any other intervening psychological process, to explain our perception of size and shape constancy. In addition to brightness and colour, properties directly picked up in the stimulus include, according to Gibson, higher-order properties that remain invariant through movement and changes in orientation. These higher-order invariants specify not only structural properties such as ‘being a cube’, but also what Gibson called ‘affordances’, which are functionally significant aspects of the distal scene, like the fact that an object is edible or could be used for cutting.
Two fundamental assumptions underlie Gibson’s ‘ecological optics’: (1) that functionally significant aspects of the environment structure the ambient light in characteristic ways; and (2) that the organism’s visual system has evolved to detect these characteristic structures in the light. Both assumptions are controversial. With respect to (2), indirect theorists have complained that Gibson provides no account of the mechanism that allegedly detects salient higher-order invariants in the optical array. His claim that the visual system ‘resonates’, like a tuning fork, to these properties is little more than a metaphor. But it should be noted that in claiming that perception of higher-order invariants is direct, Gibson is simply advocating that the mechanism be treated as a black box, from the point of view of psychology, because no inferences, calculations, memories or beliefs mediate the processing. (The physiological account of the mechanism’s operation will no doubt be very complex.) This claim might be plausible if assumption (1) is true – if there is a physically specifiable property of the light corresponding to every affordance. But for all but the simplest organisms it seems unlikely that the light is structured in accordance with the organism’s goals and purposes. More likely, the things that appear to afford eating or cutting or fleeing behaviour structure the light in all kinds of different ways. This likelihood has led indirect theorists to claim that something like categorization – specifically, the bringing of an object identified initially by its shape, colour or texture under a further concept – is at work when an organism sees an object as food, as a cutting implement, or as a predator.
Egan, Frances and Nico Orlandi. Direct theories of vision. Vision, 2010, doi:10.4324/9780415249126-W047-2. Routledge Encyclopedia of Philosophy, Taylor and Francis, https://www.rep.routledge.com/articles/thematic/vision/v-2/sections/direct-theories-of-vision.
Copyright © 1998-2019 Routledge.