One of the most important themes in social psychology for the past three decades has been the conditions under which actions produce changes in the actor's dispositions. A major and distinctive contribution of our field has been to take the conventional view of attitudes affecting behavior and to stand it on its ear-or at least to recognize the companion truth that one's actions can bring about changes in one's attitudes. Most people in my generation grew up thinking that attitudes, values, and other dispositions grew out of experiences, out of things that happened to us and things that we were told. These attitudes were interesting because, we thought, they were good (though not infallible) predictors of behavior. Attitudes can be cheaply measured and such measurements can form bases of important predictions concerning consumption, voting behavior, ethnocentric discrimination and other consequential social decisions. Then along came Leon Festinger (1957) whose most important message (at least for me) was that people can be constrained to act in various "uncharacteristic" ways and, under some circumstances, the need to accommodate or to come to terms with those uncharacteristic actions results in various kinds of dispositional changes. This profound message was consolidated by Daryl Bem (1972) and is now part of the enduring legacy of experimental social psychology, so much of a part in fact that we forget that the message was indeed strange and controversial not very long ago.