PERSONALITY PROCESSES AND INDIVIDUAL DIFFERENCES Was Gordon Allport a Trait Theorist?

Contemporary discussions have generally defined a trait as a pervasive, cross-situational consistency in behavior. A careful examination of Allport's (1937, 1961, 1966) writings reveals that he did not believe in such pervasive consistencies. In fact, he maintained that behavior in different situations is frequently inconsistent, even contradictory, because different traits are aroused to different degrees in different situations. Allport was, therefore, an interactionist in the sense that he recognized that behavior is determined by the person and the situation. However, he failed to develop an interactionist process theory that could predict and explain the situational variability that he observed and described. Allport's concept of a trait as well as those of several contemporary theorists are compared and found to differ in terms of (a) the ontological status they ascribe to traits and (b) the type of consistency over situation they require. Distinctions among types of trait concepts are then used to clarify the debate about the existence of traits. It is argued that the evidence for traits as pervasive, cross-situational consistencies is quite negative but that there is considerable evidence for the existence of traits as average levels of responses and as consistent patterns within delimited ranges of situations. Mischel's (1984) concept of a context-bound consistency is shown to be surprisingly similar to Allport's concept of a trait. The future of trait constructs is discussed in relation to the descriptive, predictive, and explanatory functions of traits.