The Primacy of Affect
in Political Evaluations
DAN CASSINO AND MILTON LODGE
In large part, the cognitive revolution in psychology has been fueled by the metaphor of the mind as a computer. The human mind is seen as a remarkably sophisticated Turing machine, a device capable of perceiving and manipulating symbols in accordance with set rules. For some purposes, the metaphor is apt; for others, it can be terribly misleading.
Rather than acting as a symbol-processing device, recent arguments have held, the brain is primarily for feeling, not thinking (Le Doux 1996; Damasio 1994; 2003). To the extent that this is the case, we should question any process in which emotions are thought to be based on rational evaluation of evidence. This includes standard models of opinion formation, according to which we like or dislike something because of what we know about it: one party in an election is favored instead of another because that party's agenda will be better for the voter.
Experimental evidence, however, casts serious doubt on the viability of this reasoning. In studying the ways in which individuals evaluate issues (in this case, affirmative action and gun control), Taber and Lodge (2006) find a strong influence of prior attitudes on the evaluation of arguments. Individuals with strong prior attitudes tend to perceive affectively congruent arguments as being stronger than affectively incongruent arguments. Furthermore, when the same subjects are allowed to seek out information, they tend to search for information that would bolster their own opinions. The judgment process, in which information is integrated into an individual's existing knowledge about an object, then, seems to come after the affective process, in which the individual forms his or her likes or dislikes about an object.
In our view, this is a necessary consequence of the organization of social objects in the mind. While objects in memory can be organized by