Akiva M. Liberman
National Institute of Justice
Exploring the Boundaries of Rationality: A Functional Perspective on Dual-Process Models in Social Psychology
This chapter reconstructs “dual-process” model in social psychology in functional terms. I argue that the basic idea behind early dual-process model concerned whether certain biases in judgment were the result of nonrational intrusions on judgment or whether they were instead the result of bounded rationality. This idea was implicit in early conceptions, but has not been fully explicated. After describing the theoretical idea and its genesis, I illustrate how it was productive in generating empirical research. In the final section, I contrast this functional reconstruction with more conventional descriptions of dual-process model.
Much of the chapter can be thought of as a discussion of the meaning of the word heuristic. The notion of heuristics as it emerged in cognitive psychology was central to the development of dual-process model in social psychology and to the functional argument implicit in them. Chaiken (1980) made this connection clear in her work on the heuristic view of persuasion and later the heuristic-systematic model (HSM; Chaiken, Liberman, & Eagly, 1989). Therefore, although my functional reconstruction is relevant to a variety of dual-process model, I focus on the HSM. (I briefly discuss models of stereotyping as well.) The HSM is also pitched quite broadly. Although initially developed for persuasion, it was intended for broader application, as indicated in “The HSM Within and Beyond the Persuasion Context” (Chaiken et al., 1989). This breadth of conception is important for my critical discussion of processing distinctions in the chapter's last section.
Social psychologists have long been concerned with errors and biases in judgment. One recurring question has been whether these errors reflect suboptimal attempts to acquire accurate knowledge about the world or whether these errors reflect quite different underlying motives, such as self-esteem maintenance and group identification. As Markus and Zajonc (1985) put it in their Handbook chapter on the Cognitive Perspective in Social Psychology:
Was faulty information processing to be explained by assuming an interference from “hot” factors, such as motives, moods, and emotions, or was it simply a matter of various inadequacies in the programs that operated on information? (p. 177)
In the 1960s, “hot” mechanisms (i.e., motivational) were emphasized in social psychology. Cognitive dissonance theory suggested that motives for psychological (vs. logical) consistency led people to distort their representations of the world, and even of their own experience (Festinger, 1957; Festinger & Carlsmith, 1959). Around 1970, self-perception theory offered the alternative “cold” (i.e., cognitive) theoretical explanation for these distortions (Bem, 1967, 1972). At the same time, attribution theory's description of people as lay scientists (Kelley, 1967, 1973) also helped cold cognitive processing become the dominant perspective. But even in the context of attribution theory, the hot side reemerged in concern with self-serving attribution biases, in which people systematically take more credit for success than for failure. Whether such biases were driven by cold information-processing limitations or by hot self-serving goals was vigorously debated (e.g., Miller & Ross, 1975; Nisbett & Ross, 1980; Tetlock & Levi, 1982).
Such debates ultimately turn on the goals (not necessarily explicit) or functions of people's judgments. As Smith, Bruner, and White (1956) asked, “Of what use to a man are his opinions?” 1 Various functional taxonomies were proposed, such as Katz's (1960) four func-