Perceptual Categorization

Perceptual categorization involves the formation of prototypes on the basis of perceptual features. It occurs effortlessly and automatically and does not involve abstract representation. Development of this type of category appears to go from the concrete to the abstract.

Definition and Overview

Perceptual categorization is the process by which our nervous systems automatically divide the world into discrete “things” on the basis of perceptual features. Individual objects and events are perceptual categories in this sense, because categorization Opens in new window is a way of grouping things together, and perceptual categorization groups individual features together into unified objects and events

In perceptual categorization, objects are computed based on their appearance; and are used for purposes of recognition and identification. Unlike conceptual categories Opens in new window, perceptual categories are implicit and are therefore inaccessible to conscious awareness.

Perceptual categories operate directly on our sensory input at the earliest stages of perception, a phenomenon previously described as feature extraction Opens in new window.

Basic features such as a sound’s pitch and loudness are perceived at the most primitive level of auditory categorization, which combines the stimulation from separate auditory nerves into features.

The next level of perceptual categorization is the binding together of different features that have already been extracted. At this stage of perception, individual features are bound together to form unified entities, often based on invariances in the environment.

Thus features such as edges, areas of color, surface texture, and other visual properties that remain essentially the same when viewed from different angles are combined to produce the visual impression of an object.

Certain features such as the acoustical resonance and timing characteristics of a particular musical instrument would be examples of auditory invariances.

Perceptual categorization functions to create boundaries in our perception of the world, and boundaries help create units for memory Opens in new window. Indeed, in this very basic sense, the edges of a visual object, or the beginning and ending of a pitch are category boundaries.

That we recognize a musical event as being separate from its background is a kind of rudimentary perceptual categorization. In this sense a single musical event is a perceptual category. Proximity, similarity and continuity are some of the operating principles of perceptual categorization.

Perceptual categorization is also the process that unifies our different and changing views of an object into a constant “thing.”

That we perceive the sounds of a musical instrument or a person’s voice as coming from the same source across multiple events and despite changes of pitch, timbre, and timing is an example of this kind of perceptual categorization.

This is a higher level of perceptual categorization where multiple events are categorized as coming from a single source. Thus perceptual categorization is a basis for the constancy of objects as they are perceived.

Basic to the idea of a perceptual category is the idea that, given a continuous range of values of some stimulus, we will tend to perceive only a limited number of discrete categories within that range. A classic example of this is the color spectrum of visible light.

We know that the visible spectrum is a continuum of frequencies of light energy, and yet when presented with the full spectrum of these frequencies, as in a rainbow, we do not see a continuum—a smooth or continous change of color—but a small number of color categories (bands) with fuzzy edges that shade into each other.

This particular categorization is neither learned nor related to language distinctions about colors (Hardin, 1988). Rather, it is built into the structure of our mechanisms of vision.

The division of continuous auditory experience into the three levels (event fusion, melodic and rhythmic, and formal) is another example of a broad type of categorization that is not learned.

Image