Categorization Exemplar and Prototypical Categorization

Categorization

Exemplar and prototypical categorization and their relationship to typicality

Prototype theory is a concept of learning that suggests certain types of objects are more central to forming mental categories than other types of objects. For example, although people may come to acknowledge there are many types of dogs, a Golden Retriever may seem to be more prototypical of the category of 'dog' than, say, a Chow or a Chihuahua. Prototype theory suggests that first the human mind develops a concept, such as a 'sport,' from more typical examples, like baseball and football, and gradually that conceptual group assimilates less typical examples, like fishing and marathon running, which could also be classified as hobbies or fitness activities. "To cope with an incessant flow of information, humans and other organisms must group their experiences into meaningful categories…the traditional view, dating from the time of Aristotle, is that categories are mentally represented as sets of necessary and sufficient conditions. According to this view, an exemplar is assigned to a category if it satisfies the category's membership conditions" (Dopkins & Gleason, 1997, p.1)

But according to prototype theory, boundaries of categories are more fuzzy and pliable by virtue of necessity (Green, 2000, p.301). It seems an "almost common-sense notion that, as an organism, what one wishes to gain from one's categories is a great deal of information about the environment while conserving finite resources as much as possible," and to be cognitively efficient, categories must be both specific enough to reflect all essential information and general enough not to overwhelm the classificatory consumer with irrelevancies" (Green, 2000, p.302). Too many categories results in a waste of needless cognitive space that is why "a category's mental representation is based on a prototypical exemplar or prototype" that does not strive to subsume all of the aspects of every conceivable example, only the most salient aspects of the categorical components (Dopkins & Gleason, 1997, p.1). Another example might be how, "the representation for the BIRD category might be based on ROBIN. To decide whether an entity is a member of category, one compares it to the category's prototype," and asks if an ostrich sufficiently resembles a robin (Dopkins & Gleason, 1997, p.1). The ostrich's similarities confirm the ostrich as a member of the typified 'bird' categorization in that it has feathers, lays eggs, has a beak vs. its dissimilarly large size to the robin. After all, taken to an extreme, if one defines all birds as robin-sized, that would require different categories for sparrows and ducks as well as ostriches.

The fact that not all examples need to conform to the prototype is one of the strengths of the prototype theory: better examples better resemble the categorical prototype but there is a continuum of better (Dopkins & Gleason, 1997, p.1). Cognitively, according to prototype theories, "people represent categories in terms of some central tendency computed over the category training instances and classify objects on the basis of how similar they are to the prototypes of the alternative categories. By contrast, according to exemplar models, people represent categories by storing the individual training instances themselves" without prototypes (Zaki et al. 2003, p.1160).

Thus, if contrast to prototype theories, so-called exemplar theory suggests that learning proceeds from specific examples, and if prototypes are formed, this is only much later on in the subject's cognitive development, and is not universally the case. "According to the exemplar theory, a category's mental representation encodes the exemplars that compose the category. To decide whether an entity is a member of a category, one compares it to the category's exemplars" (Dopkins & Gleason, 1997, p.1). These exemplars are empirically rather than deductively conceived. Exemplar theory can also account for the problematic phenomena of exceptions: "If all of the exemplars of a category are encoded, all exemplars need not share the same properties. Stimuli will be 'better' exemplars of a category the more closely related they are to the category's exemplars (Dopkins & Gleason, 1997, p.1). Categories are "mentally represented simply as collections of exemplars; categories themselves have no independent representation" and there is no prototype (Dopkins & Gleason, 1997, p.3).

In prototype theory, the prototype of a category does not have to correspond to an actually perceived exemplar but can exist as an "abstract, idealized prototype" (Mobius 2004, p.11). This is a "controversial assumption" given that it assumes a level of abstract, one could higher-order thinking that is arguably only be present in the minds of individuals with certain types of education, past a particular developmental stage, or of particular cultural mindsets (Mobius 2004, p.11).

Exemplar theory has also been called 'cloud' theory in that each category of exemplars is represented in memory by a "cloud of remembered exemplars" (Mobius 2004, p.3). "The exemplar cloud is continuously updated by new speech events" (Mobius 2004, p.3). Random selection of matching, however, "causes increasingly broader, fuzzy-edged categories" that are, in their own way, just as flexible as the typicality of prototype theories (Mobius 2004, p.5). Exemplar theory very explicitly 'maps' the brain, stating that each new stimulus encoded as a location on the organism's cognitive map, and while areas of map are associated with specific functions, stimuli are "categorized on the basis of most similarity to the most exemplars near that location" not according to prototypical types (Pierrehumbert, 2003, p.8)