Puzzling Over the Paradoxes of Caspi and Moffitt

Abstract

It has been difficult for me to know what to make of Caspi and Moffitt's target article. It brings together an impressive array of experimental studies from diverse areas of psychology, suggests mechanisms underlying the coherence of personality, and makes interesting references to anthropological ideas and evolutionary theory. On the other hand, the reasoning underlying the structuring of all this evidence seems to me arbitrary and unsound. The first confusing premise is that individual differences, after early childhood, are very hard to detect because they are masked by social roles. Therefore, a pressing problem is to devise a strategy for removal of masks. It is true that, after early childhood, most of us do not sing songs to our physician or try to take turns with the stethoscope. We conform to the role of patient. But, we do not hide our hypochondriacal concerns, tension and irritability, intellectual curiosity, or stoic acceptance. Caspi and Moffitt quote Block (1981), but they do not listen to his full message, to McCrae and Costa (1990), or to many others. There is massive evidence for the validity and longitudinal stability of individual differences in personality. Individual differences (as assessed by inventories and other personality measures) are indeed affected by our work roles (Kohn & Schooler, 1978) and by other role configurations over time (Helson & Picano, 1990), but they are not masked, and they continue to influence behavior in important ways. The next step in Caspi and Moffitt's argument is to show what personality psychologists already knowthat differences in personality are more apparent in situations that do not pull for the same behavior from everybody. Caspi and Moffitt cite many pages of evidence that individual differences in dogs, monkeys, babies, and children are conspicuous in novel, ambiguous, and threatening situations. They do not mention that personality differences might also be reduced in such situations or that personality differences are apparent without such emergency. Rather, they fix on the idea that, to find personality differences when they are clear-cut, we should study them in life transitions that are novel, ambiguous, and threatening. This idea represents a fundamentally different perspective that Caspi and Moffitt describe as paradoxical. Although transitions have hitherto been regarded as times when people change, the authors say, transitions actually are times when people show their stripes. In my view, this apparent paradox is maintained by shifting meanings of terms. When it is suggested that people change in connection with the to parenthood, for example, one is referring primarily to mean level change in personality characteristics (Feldman & Aschenbrenner, 1983) or perhaps to general processes of adaptation to change (Stewart & Healy, 1985). The question of rank-order change in personality over the course of transitions, which is what Caspi and Moffitt studied in girls before and after menarche (if behavior problems count as an adequate assessment of personality), has received less attention, perhaps because, as these authors say, rank-order stability tends to be high. I know of no consensus or major theory holding that rank-order change is pronounced in transitions, so I have trouble seeing the paradox. It is true that some studies report less rank-order stability from adolescence to young adulthood than over other periods of the life span (Finn, 1986; Haan, Millsap, & Hartka, 1986; Helson & Moane, 1987). The change from the structured and sheltered environments of home and school to the heterogeneous wide world has often been an unprotected transition in which some young people do better and others do worse than previous adjustment leads one to expect. This evidence goes against Caspi and Moffitt's argument. Although there was less stability from ages 21 to 27 than from ages 27 to 43 in the Mills Longitudinal Study (Helson & Moane, 1987), stability was still high, and individual differences among the women in the senior year of college showed highly significant relations to subsequent role patterns. We have pointed out that our findings might be particularly striking because the participants were members of a transitional cohort for whom there were strong pressures to marry and have children along with new opportunities in the labor force and the gathering storm of the women's movement (Helson, Elliott, & Leigh, 1989; Helson & Picano, 1990). However, it was the social change in the whole country that made it possible for differences in personality among these women at their time of transition to the adult world to be expressed in different life-paths. Thus, the consequences (importance) of individual differences in a life transition depend on the social and historical context. The experimental model adopted by Caspi and Moffitt throughout most of their target article does not seem to me sufficient for the study of personality in