Peception Involves Bottom and Top Down Processing Essay Example

94); how this information is then interpreted is what we term perception. Whilst science can implicitly explicate the visual system, visual perception is individual and thus subjective and susceptible to alteration; inferences about the processes involved in visual perception are therefore derived from theory, research and experiments.

Ideas put forward about the processes involved are influenced by whether the primary goal of perception is considered to be action or recognition, and whether the direction information is seen as flowing up or down through perceptual system. As we engage with the environment, information is received via the senses; in visual perception, light arrives at the retina and is subsequently processed via the visual system whereby a description of the world is generated to form an interpretation of what is being perceived.

This process is termed Bottom-up as perception is achieved from sensory data flowing up through the perceptual system. This sensory data cannot however merely enter and exit the perceptual system; once sensory information is analysed and a description of the environment is formed, the internal representation of what is being sensed must be stored so as the memory of that experience can inform and guide subsequent interactions with the environment.

Utilizing stored knowledge to aid interpretation of future interactions with visual stimuli is referred to as top-down processing (Pike & Edgar pp. 68-69). These concepts suggest that perception can arise through bottom-up or top-down processes; whilst one may assume that it is an interaction of these processes that result in perception, some approaches advocate that only a singular process is necessary to formulate perception. One such advocate is J.

J Gibson

(1950, 1966) whom proposes that the principle goal of perception is action. Gibson puts forward that to achieve action it is not necessary for sensory information to endure complex cognitive processing; rather action arises as a result of the perceptual system’s active engagement with a dynamic, visually rich environment, containing sufficient information to lead to direct perception and immediate interaction, without the need for further analysis or inferences (as citied in Pike & Edgar,2010 p. 3).

The vibrant environment to which Gibson refers is composed of images and surfaces with varying densities, textures, sizes, distances and shapes, and the interaction or movement of the observer and the movement of the object provide ample information to generate perception. Gibson emphasises the significance of what he terms the ‘optic array’ whereby ight reflected from a multitude of textured surfaces unites in the visual field occupied by the observer, thus altering in conjunction with the position of the observer, hence providing information about where the object is in space.

Furthermore, Gibson purposed that in addition to the optic array, invariants or higher-order features, such as ‘horizon ration relation’ and ‘texture gradient’ offer explicit insight into the nature of the environment (as cited in Pike & Edgar, 2010, pp 74-76). Horizon ration relation’ explored by Sedgwick (1973) enables us to judge relative heights and distances of objects; the distance of objects can be judged as the ratio of the proportion of the horizon above to below the object remains constant, however if objects viewed at the same distance differ in height the overall height of the horizon correspondingly increases (as cited in Pike & Edgar, 2010 pp

74-76).

Likewise, texture gradient provides information about the density, height, and width of individual elements contained within and varying from surface to surface. Additionally Gibson posits motion and movement of the observer as a critical part of perception, whereby unambiguous information about shape, distance, and size of objects are revealed in the changing nature of flow patterns in the optic array.

Gibson proposes that the wealth of invariant information and cues derived from features of objects give rise to their affordances in that the ensuing perception enables us to climb stairs, place cups on tables, throw balls into hoops, in general, to act within our environment. It cannot be disputed that the surrounding world is indeed rich and dynamic, and in support of Gibson’s claim, one might concede that our hunter-gatherer ancestors had to rely upon their surroundings to inform and guide their actions.

However for Gibson to imply that there is no need for cognitive processes or memory in perception is intransigent, in that civilisation would never have evolved if humanity did not learn and acquire knowledge as a result of their experience with their surroundings. Indeed Gibson’s theory presents as somewhat of paradoxical concept in that he accents the need for action and interaction, yet in denying the need for cognitive processes it seems to suggest that humans are passive by merely ’picking up’ information from the senses but never using that experience to uide future behaviour.

At the premise of Gibson’s theory is the notion that the environment is abound with rich, sensory information, however this notion is starkly contested, with many alternatively viewing sensory stimuli as impoverished

and incomplete, thus necessitating prior experience and stored knowledge to aid interpretation of the information flowing up from the senses. Richard Gregory’s (1980) theory lies within the ‘constructivist approach’ and proffers that visual perception involves top-down processing.

Gregory proposes that that information flowing up from the senses is incomplete, and to make sense of it, it is necessary to use stored knowledge to make inferences about incoming sensory information in order to build representations and construct perception. In contrast with Gibson, Gregory identifies the goal of perception as recognition as opposed to action and proposes that visual perception derives from using existing knowledge in conjunction with sensory information to generate a series of perceptual hypotheses about what the object may be (as cited in Pike & Edgar,2010, p. 91).

Gregory recognized however that by integrating prior knowledge, a false perception can sometimes arise, in that some knowledge about objects are so deeply ingrained, that if presented with a similar, yet novel object we are likely to reject the correct perception and generate an incorrect perceptual hypothesis in favour of how we understand and know the object to usually be. Evidence to support this notion derives from analysis of visual illusions; visual illusions are often perceived differently by individuals, thus demonstrating that the brain uses individual, past experience to interpret what the eyes see, albeit sometimes incorrect.

Gregory’s constructivist theory of perception offers an explanation of how we integrate bottom-up sensory information with top-down knowledge; and whilst Gregory provides limited evidence to support his theory, there is abundant evidence to demonstrate that humans are accomplished learners with a vast capacity for knowledge, which

is undoubtedly employed to inform and guide perception. However, Gregory’s theory, like Gibson’s presents as, to some extent conflicting, in that the purpose of constructing a somewhat haphazard, best guess perception would surely be better utilized in an nvironment that was rich and ever changing, whereby it would be impossible to attend to such a wealth of information resulting in best guesses being advantageous.

If however the environment were impoverished as Gibson suggests, one might expect a more accurate method of processing to have evolved. Gibson’s and Gregory’s theories have presented two alternative ways in which visual perception may be processed, and to some extent arguments for both are supported by evidence from the physiology of the visual system.

Schneider (1967, 1969) gathered evidence to suggest that within the visual system there are two distinct, yet interconnected visual pathways. The ventral stream, termed the ‘what pathway’ projects to the infer temporal cortex and is thought to be involved with object recognition, whereas the dorsal stream or ’where’ pathway projects to the parietal cortex, a region considered to be specialized in analysing information about the position and movement of objects (as cited in Pike & Edgar,2010,p. 95).

Norman (2002) purposes that the ventral stream, which analysis’s and process information at a slower rate and draws on existing knowledge to arrive at a description of an object, is utilized in Gregory’s, perception for recognition, top-down approach. Alternatively, the dorsal stream, due to its short-term storage, speedily receives and processes information regarding movement and position, lending support to Gibson’s notion of perception for action, bottom-up approach. Norman suggests that the streams act in synergy so

as perception for recognition and action are both achieved (as cited in Pike & Edgar, 2010, p. 6).

Norman’s argument for synergistic bottom-up and top-down processing is supported by research carried out by Enns and Di Lollo. Enns and Di Lollo (1997) propose that the re-entrant pathway, which communicates information back and forth from different brain regions may be responsible for combining top-down, higher level perceptual hypothesis with lower level bottom-up descriptions. Enns and Di Lollo, in a backward masking task, used non standard mask that did not contain contours which overlapped or coincided with the target, to examine whether, as standard masks do render the target nvisible.

They presented the mask either simultaneously or very briefly after the target and found that masking still occurred. Enns and Di Lollo suggest that whilst bottom-up processes generate a description of the target by the time the top-down perceptual hypothesis achieved and communicated back via the re-entrant pathway, the target is replaced with the mask, subsequently rendering the target invisible(as cited in Pike & Edgar,2010, pp. 98-100) .

Enns and Di Lollo infer that it is bi directional communication between bottom-up and top-down processes, operating in unison that result in perception. In conclusion, theories, such as Gibson’s and Gregory’s that lie within opposing frameworks enable us to consider competing and distinct aspects and depending on whether we need to quickly respond to something, like averting danger, fast visually guided action may indeed triumph without need for consulting prior knowledge.

Alternatively, in order to learn that something in the environment may be a threat, this experience must be stored and employed to guide future

action. So whilst bottom-up and top-down processes may be used independently, Enn’s and Di Lollo’s suggestion that visual perception involves communication and interaction of both processes offers the most likely explanation, in that whilst some systems within the central nervous system may be specialised for different functions, it is the interaction of these processes that result in our highly sophisticated ability to perceive. ?