Genuine, Defensive, Unconscious and Authentic: How do They All Fit Together?

A couple of months ago, a student of mine approached me with an idea for her dissertation. She was interested in variables that might mediate prejudice. One such variable, she thought, might be self-esteem. To my estonishment, she argued that individuals with high self-esteem would be more prejudiced than would those with low self-esteem. She justified this by pointing to the literature that indicated that those with high self-esteem tend to value their own groups and values more highly than do those with low self-esteem (Baumeister, Tice, & Hutton, 1989; Greenwald et al., 2001). In at least one study (Crocker, Thompson, McGraw, & Ingermane, 1987), high self-esteem individuals also devalued outgroups and outgroup values more than did low self-esteem individuals. Hence, my student said, high self-esteem whites should value other whites and traditional white values more than would low self-esteem individuals. It is further possible that they would also value blacks and nonwhite values less than would low self-esteem individuals. Was my student telling me that KKK members and Nazis had higher self-esteem than civil rights workers? It seemed impossible yet her view had empirical support. I could not refute her with anything other than my feeling and folk wisdom that prejudice can be used to boost self-esteem and that those with already high self-esteem would not need such a boost. My clinical experience backed this up. Patients who were prejudiced seemed to be low in self-esteem and needed to disparage others to feel worthwhile whereas those who were more egalitarian seemed more comfortable with who they were and so had no need to feel superior to some identified outgroup (cf. Kohut, 1971, 1977). But she seemed to have the data, and I had only intuitions, personal experience, and clinical lore. Kemis (this issue) has rescued me by differentiating between genuine and fragile self-esteem. I was thinking of what he terms genuine self-esteem whereas my student was referring to what the target article defines as fragile self-esteem. Although both have been referred to as high self-esteem in the literature, Kernis cogently shows that they are very different from one another. Genuine self-esteem is nondefensive and stable. A person high in this kind of self-esteem would be very unlikely to hold prejudiced views. Fragile self-esteem, on the other hand, is often defensive, contingent, and unstable. A prejudiced person could easily be high on this variable. In addition to helping me feel vindicated and providing an important reading for my student (I assigned it as soon as the editor of this journal told me it was officially in press), this differentiation is important theoretically. It means that simply knowing that someone has scored high on a measure of self-esteem does not tell us how such a person will behave or what motivated him or her to behave in that fashion. It means that we must investigate further to determine what a high score on a self-esteem measure indicates. It means that self-esteem is a more nuanced variable than has traditionally been supposed. Kemis is to be applauded for bringing this to the attention of the psychological community in such a convincing fashion. But he has done more; he and his colleagues have offered empirical ways to differentiate the two kinds of self-esteem and have brought disparate literatures, which seem to contradict one another, together to make a clear-cut theoretical case. This is a real achievement. Although I believe that the differentiation of genuine from fragile self-esteem represents a milestone in the study of self-esteem, that is not all there is to it. The specifics of how to differentiate the two is critical, and here everything is not so clear-cut. As is often the case, the devil is in the details. And it is to these details that I now turn. I would like to offer what I believe is some constructive criticism and ask some questions that I hope Kemis can answer in his response. I would also like to compare the views of self-esteem offered in the target article to some currently being offered in the areas of attitudes and motivation. I believe that these conceptions may be theoretically related and that connecting them could lead to a broader understanding of human functioning. Kernis differentiates between defensive and implicit self-esteem in his discussion of fragile self-esteem. If my reading is correct, he is saying that a conscious desire to present oneself in a positive light underlies defensive self-esteem (the subheading of this section is titled Is the Person Telling the Truth?). This is very different from the use of this term by most clinicians and even by those who conducted the empirical research using the Marlowe-Crowne social desirability scale Kemis invokes as critical to the assessment of defensive self-esteem. Thus in D. Weinberger's work on repression (D. Weinberger, 1995; D. Weinberger & Davidson, 1994; D. Weinberger, Schwartz, & Davidson, 1979), individuals scoring high on social desirability and low on a