Words, deeds and the perception of consequences and responsibility in action situations.

Abstract

It is proposed that correspondence between people's norms and their behavior can be expected only when pertinent norms are activated in choice situation, and that activation depends upon how consequences and responsibility for social behavior are perceived. Using indexes of tendencies to become aware of consequences for welfare of others and to ascribe responsibility to self, 118 male college students were assigned to groups with different probabilities of experiencing activation of their norms. Within groups presumed to experience little activation of norms, no correspondenc e between norms and behavior was observed. Increasingly positive correlations were found within groups more likely to experience activation of norms. (For a summary index, r = .02 in lowest group, and r = .58 in highest.) The findings suggest an interaction between ascription of responsibilit y and awareness of consequences in activating norms. The question of how verbalizations concerning attitudes, opinions, and norms are related to overt behaviors is of continuing concern in social science. Deutscher (1966a) has compiled a bibliography of several hundred sources dealing with aspects of this question. Yet conclusions drawn by reviewers of this literature are widely divergent. DeFleur and Westie (1963), for example, labeled idea that there should be consistency between verbal behavior and actions the fallacy of expected correspondence, arguing that verbalizations and overt actions constitute separate universes of response. Yet Campbell (1963, pp. 157-162) asserted that true inconsistency is rarely found, and called what is reported in social attitude literature pseudoinconsistency. He cited failure to recognize that verbal behavior and overt behavior have different situational thresholds as basis for many inconsistency interpretations. The springboard for research reported here is often noted fact that verbalizations are commonly elicited or measured in one setting, while actions which they 1 This research was supported in part by Public Health Service fellowship MH-23, 855, and computer analysis was funded by Wisconsin Alumni Research Foundation and National Science Foundation. The author is indebted to Howard Schuman for his critical comments and suggestions. presumably might govern are observed in another (cf. Campbell, 1963; DeFleur &