Kernis (this issue) notes that, although researchers have traditionally been concerned with the relevance of self-esteem level, stability of self-esteem also serves a critical role in healthy psychological functioning. Kemis proceeds to synthesize research on self-esteem level and stability to arrive at a description of optimal self-esteem. He defines optimal self-esteem as a secure (rather than fragile or dysfunctional) form of high self-esteem. What are the causal mechanisms that underlie secure self-esteem? We believe that the search for causal and developmental antecedents of self-esteem would benefit greatly from the use of genetically informed designs. Such designs allow for the specification of both genetic and environmental influences on individual differences. Moreover, genetically informed designs provide one method of ensuring that putative environmental effects are indeed environmental rather than heritable (Plomin, 1994). A small body of literature has investigated the genetic and environmental architectures that underlie self-esteem level. Studies of both adults (Kendler, Gardner, & Prescott, 1998; Roy, Neale, & Kendler, 1995) and children (McGuire, Neiderhiser, Reiss, Hetherington, & Plomin, 1994) suggest that genetic influences on self-esteem level explain approximately 30% to 40% of the variance. These studies also suggest that shared environment does not account for variability in self-esteem level. That is, familial variables (e.g., parental warmth, parenting styles), which contribute to similarity of siblings raised in the same household, have rather minimal impact on self-esteem level. On the other hand, nonshared environments (such as differential treatment of siblings by parents or individual self-esteem moments), which contribute to between-sibling differences, do have a substantial influence on self-esteem level (for a review, see Neiss, Sedikides, & Stevenson, 2002). In many ways, the observations we offered in the previous paragraph repeat a standard behavior genetic argument: Genetic and nonshared environmental influences on behavioral traits are often substantial, but shared environmental influences are minimal (Plomin & Daniels, 1987; Turkheimer, 2000). We reignite the argument, however, because it has implications for the origins of optimal self-esteem. Furthermore, behavior genetic findings complement existing hypotheses about causal mechanisms assumed to underlie optimal self-esteem. For example, Kemis suggests that a match between children's intemal states and adult interpretation of children's inner experiences facilitate authenticity (and presumably optimal self-esteem). Stated somewhat differently, a gene (child's disposition) x environment interaction (parental feedback) effect may be a critical causal mechanism underlying optimal self-esteem. Such a hypothesis can be tested within a genetically informed design. In the absence, as yet, of known genetic markers for genes predisposing to high self-esteem level, such a study would have to be undertaken with adopted children. Specifically, the sample would consist of children (a) whose biological parents have high or low self-esteem level, and (b) who are reared by adoptive parents fostering a high or low self-esteem level. Unfortunately, it has been difficult to identify such gene-environment interactions within adoption studies (Plomin, DeFries, & Fulker, 1988). Indeed, the prospect of genetic markers of risk becoming available in the near future is so close (Plomin & Crabbe, 2000) that it might be more sensible for researchers to wait for the availability of such markers rather than attempt to identify possible gene-environment interactions through quantitative genetic methods (e.g., the adoption design). We do not wish, and we are not in a position to, overemphasize the genetic underpinnings of self-esteem level: Research on the heritability of self-esteem level shows that the largest portion of variability in self-esteem is related to nonshared environment-clearly, then, genetic main effects do not tell the whole story. To complicate matters, the pathways linking genes to self-esteem level are unknown. Nevertheless, Cloninger, Svrakic, and Przybeck's (1993) psychobiological model describes one possible pathway between innate temperament and self-concept. These researchers proposed that heritable biases in information processing reflect four basic dimensions of temperament: novelty seeking, harm avoidance, reward dependence, and persistence. Temperament influences self-concept and self-esteem via information processing biases. This model also allows for reciprocal influences between information processing and self-esteem. For example, self-esteem may shape attention to and interpretation of environmental stimuli (Gaertner, Sedikides, Vevea, & Iuzzini, 2002; Gramzow, Gaertner, & Sedikides, 2001; Green & Sedikides, 2001; Sedikides & Skowronski, 1993). Specific genes might be linked to neurochemical substrates associated with the basic temperament dimensions, although attempts to identify such linkages have met with limited success (Herbst, Zonderman, McCrae, & Costa, 2000).
Read full abstract