Personality Structure

The modern study of personality structure arguably began with Jung's (1921) introduction of the concept of introversion—extraversion (see Watson & Clark, 1997, for a more detailed history of the development of this concept). In the mid1930s, using the then newly developed technique of factor analysis, Guilford and Guilford (1934, 1936) demonstrated that Jung's concept was multidimensional. This research was followed by more than a decade of scale development work to assess the identified traits, culminating in the Guilford-Zimmerman.

Temperament Survey (Guilford & Zimmerman, 1949). This instrument contained a number of specific scales that defined the three higher order factors of neuroticism, sociability, and restraint versus impulsivity. Based on this work, Hans Eysenck began a series of construct-validation studies that ultimately resulted in the Eysenck Personality Questionnaire (EPQ; Eysenck & Eysenck, 1975). The EPQ contains scales to assess the dimensions of neuroticism, extra-version—introversion, and a third scale called psychoticism, which is generally agreed to assess a dimension more akin to psychopathy.

A second line of research into personality structure was proceeding independently of this work, based on Galton's lexical hypothesis that important psychological dimensions will be embedded in language. Beginning with a list of 4,504 trait-disposition terms compiled by Allport and Odbert (1936), Cattell (1945, 1946; Cattell, Eber, & Tatsuoka, 1980) used rational content sorting, cluster analyses, and factor analysis to construct a set of 16 primary traits, which again can be arranged hierarchically to form a smaller number of higher order factors. Based on this same set of terms, a number of other researchers have contributed to the identification of a consensual set of five personality dimensions, now widely known as the Big Five (see Goldberg, 1993, for a delightful rendering of its developmental history). Among both Cattell's higher order factors and the Big Five traits, two dimensions are easily identifiable as neuroticism and extraversion-introversion; two more-agreeableness and conscientiousness—

In a series of studies designed to elucidate the structure of questionnaire-based personality, Zuckerman and colleagues have demonstrated that only a few (at least three and no more than seven) broad dimensions are sufficient to describe personality assessed in this manner (Zuckerman, Kuhlman, & Camac, 1988; Zuckerman, Kuhlman, Thornquist, & Kiers, 1991, Zuckerman, Kuhlman, Joireman, Teta, & Kraft, 1993). Their analyses also indicate that it may not be particularly fruitful to argue about whether there are “really” three, five, or seven dimensions, because the different solutions exist largely in hierarchical relation to each other. That is, with the exception of the Big Five's fifth dimension (variously called openness, intellect, or culture) whose existence or importance is controversial,¹ the three-factor solutions generally incorporate the five-factor solutions, which in turn incorporate the seven-factor solutions.

The important point is that there now exists a broad—though by no means universal—consensus that self-reported personality can be described by a small number of broad, phenotypic dimensions (see Block, 1995; McAdams, 1992, for dissenting views). Two of these dimensions—neuroticism and introversion-extraversion—appear in remarkably similar form in all current major models (Zuckerman et al., 1993). Two more dimensions—agreeableness versus aggres-sion-hostility and conscientiousness versus impulsivity (more fully labeled “impulsive unsocialized sensation seeking” by Zuckerman et al., 1991)—ap-pear in most models either as separate traits or combined into a broad, higher order dimension such as Eysenck's psychoticism, Guilford's impulsivity versus restraint, or disinhibition versus constraint (Watson & Clark, 1993).

A Unified Mood-Personality Structure

Although, as mentioned, most of the structural work in the domains of mood and personality was carried on independently, Tellegen (1985) had the insight to realize that the two higher order mood dimensions lay in a one-to-one relation with the “Big Two” broad dimensions of personality. Tellegen hypothesized that mood constituted the basic core of these personality traits, and he reconceptualized them as primarily dimensions of affective temperament. He also renamed them negative emotionality (for neuroticism) and positive emotionality (for extraversion) to underscore their fundamentally affective character.² Furthermore, he developed the theoretical foundation of the domain by elaborating on the connections (previously noted by Eysenck) between these dimensions and the biologically based dimensions of motivation described by Gray. I will discuss these connections in more detail in a later section.

Supportive evidence quickly accrued for the mood-personality connections, and again I illustrate them primarily using my research with Watson. In table 91, data are presented from nine samples who completed mood rating forms and personality questionnaires. For Samples 1 through 4, the instructions were for short-term, “state” mood, ranging from “currently” to “over the past few weeks.” For Samples 5 through 9, the instructions were for general “trait” mood; that is, the research participants rated how they “generally felt.” A clear convergent-discriminant pattern emerges for both types of instructions, although the pattern is somewhat stronger for trait than for state mood. Specifically, NA relates more