ABSTRACT
Since we have suggested that attitudes are important keys in understanding the long-range organization of behavior, any such lack of correspondence is a matter of concern to us. Although overt behavior is directly observable, attitudes are not, in the same sense. The notion of attitudes as stored dispositions is useful only in the degree that behavior can be seen as dependent in some way on such dispositions. If individuals are capable of behaving in ways that seem to correspond poorly with their attitudes, of what use is the notion of attitude at all?
