The Nature of Cause, the Role of Antecedent Conditions in Children's Attribution, and Emotional Behavior

Abstract

The idea proposed by Dweck, Chiu, and Hong is both elegant and simple: They maintain that individual differences in implicit theories about the world affect people's judgments and reactions. It also should affect their emotional reactions (Lewis, 1992; Weiner, 1986). Dweck et al.'s claim fits well into Steven Pepper's (1942) World Hypotheses, the idea that certain overarching beliefs are used to explain more specific theories. In the developmental and personality literature, such notable theorists as Hayne Reese, Willis Overton, Richard Lerner, and Kenneth Gergen, among others, have maintained that in order to understand both experimental methodology and the interpretation of empirical findings, it is necessary to understand the world view from which they come. I can find no objection to that claim. The idea that world view may be equally applicable to how individuals operate in their worlds is parsimonious with this belief. In more specific terms, the problem that Dweck et al. set out to address is one that has influenced social psychology and personality theorists for some time. A review of the attributional theories is unnecessary because any interested reader can readily come upon them. The distinction that Dweck et al. draws-incremental versus entity world views account for a wide and diverse set of judgments and reactionsadds another dimension to attribution theory. They present convincing evidence for the association between two different types of attributions and between helpless versus mastery-oriented responses to personal setbacks as well as evidence on the retribution versus education perspective. I first briefly consider methodological issues. My concern is more on the issue of causality, and this receives most of my attention. Dweck et al. concede that individuals do not have one attribution or another, but their attributions are situationally determined. Thus, a subject might have incremental view in one context with entity view for another. Now this is interesting problem. Is it possible that incremental versus entity view for a particular domain is determined by the way attributions are measured? Is it possible, for example, that a subject might have both entity and incremental view in the same domain if we measured them in some other way? The authors' argument that there are, within domains, stable individual characteristics is not yet convincing. But of a bit more concern, especially from a logical point of view, is the fact that Dweck et al. appear to be arguing that there is a trait-like attribution in people, even in people who do not have a trait-like attribution. Is it not true that a person who has incremental view for a particular domain could be said to have a trait-like attribution? They attempt to get around this problem when they write that these findings suggest that it is more appropriate to view implicit theories and their allied judgment and reaction patterns as relatively stable but malleable personality qualities, rather than as fixed dispositions. This statement suggests-although under some protestation-that Dweck et al., in fact, are entity theorists even though the entity that they are talking about is the likelihood of being incremental theorist. This distinction between individual's attribution and the consistency of that attribution needs resolution, and I hope that this problem can be dealt with. My major concern, however, is with a more general problem: the issue of whether earlier behaviors affect later ones. Specifically, I wish to attend to the problem of causality. In several parts of the target article, Dweck et al. make clear that the attributional style of individuals determines their judgments and reactions. For example, Dweck et al. say that an entity theory of personality and moral characterfosters trait judgments and trait attributions, whereas incremental theory promotes a focus on more specific mediational processes. Moreover, they state that these studies suggest that not only does entity versus incremental theory of character predict a greater tendency to make trait judgments or attributions, it also predicts a preference for retribution versus education or remediation [italics added]. These statements reflect causal connections between the attribution and subsequent judgments and reactions. As far as I can determine, the attributional literature in general makes this same causal relationship. This being so, this criticism from Dweck et al. applies equally well to all of us interested in attributions. To what degree can we say that