Prototype categorization of emotion

Psychologists have yet to agree on a definition of emotion. Attempts at a classical definition, whereby a concept is defined by a necessary and sufficient set of criterial attributes have not met with success. The purpose of this research, therefore, was to test the feasibility of an alternative to the classical view, namely prototype theory. According to the prototype view, concepts are organized in terms of prototypes, which are the clearest cases or best examples of the category. Category members are nonequivalent and can be ordered in terms of their degree of resemblance to the prototypical cases. Boundaries between categories are therefore ill-defined. In this research, the feasibility of conceptualizing the everyday concept of emotion as structured in terms of prototypes was tested using Rosch's approach. Rosch and her associates have recently demonstrated that many natural language categories such as fruit, furniture, and vehicle can be conceptualized as prototypically organized. Rosch has also demonstrated that many natural language categories are organized hierarchically. For example, the set fruit, apple, Granny Smith apple illustrates a hierarchy with a superordinate, middle, and subordinate level. The first two studies examined the hierarchical structure of emotion categories. In Study One, "emotion" was taken to be the highest, or superordinate level. Subjects were asked to list members of the category emotion. As predicted, prototypical category members like "happiness", "anger", and "sadness", were listed first and with greater frequency than less typical members like "respect", "awe", and "boredom". The purpose of Study Two was to explore the possibility of a subordinate level of the hierarchy. Subjects were asked to list types of emotion categories generated in Study One. It was discovered that unlike the Study One results where all responses were single, words, emotion categories at this level of the hierarchy are not coded monolexemically. Subjects had to "invent" subordinate categories. Consequently there was little agreement in their responses. Internal structure refers to that general class of conceptions of categories in which categories are composed of a core meaning and in which items within the category may be considered differentially representative of the meaning of the category term. In this research, representativeness was operationally defined by means of subjects' ratings of how good each item is as an example of its category. In Study Three, prototypicality ratings were obtained for 20 emotion terms (generated in Study One) as a measure of category representativeness. As predicted, subjects found it meaningful to rate the extent to which each instance was a good example of the category emotion. Moreover, subjects agreed with one another in their responses. Representativeness of items within a category was then shown to affect certain dependent variables important in psychological research. Study Four concerned speed of processing. Subjects were asked to verify statements of the form "An exemplar is a category name". As predicted, response times were shorter for verification of the category membership of highly prototypical than less typical exemplars. In Study Five, subjects were given the 20 target emotions and were asked to give the general category to which each belonged. As predicted, "emotion" was given as the superordinate category name more often for prototypical than nonprototypical exemplars. In Study Six, subjects generated attributes of the 20 target emotions. A family resemblance score was computed for each emotion based on the attributes each had in common with the other category members. As predicted, prototypical category members resembled the entire family to a greater degree (i. e. had a higher family resemblance score) than nonprototypical members. Overall, the results suggested how people may organize their concept of emotion. People need not be able to define "emotion" in order to use the concept in an orderly and comprehensive way.