Argument Recruitment

When faced with the task of judging which of two possibilities is true, an “argument strategy” with the following elements seems eminently reasonable:

• Element 1: Accumulate as many arguments as possible that bear on the issue.

• Element 2: Examine each argument carefully and determine which of the two alternatives that argument supports and how strongly the argument supports it.

• Element 3: Weigh the cumulative import of all the arguments pointing toward each option and then pick the alternative that “wins” this competition.

• Element 4: If a degree of certainty that this winner really is correct must be rendered, make that judgment according to how convincing was the victory.

Several proposed accounts for overconfidence start with an assumption that people spontaneously (even if nonconsciously) do something like apply this strategy when confronted with general-knowledge questions of the usual form (e.g., “Potatoes grow better in: (a) warm climates or (b) cool climates”). Overconfidence is attributed to peculiarities in what happens in specific elements of the process. Variants within this class of explanation differ according to exactly what those peculiarities might be. The argument models of Koriat, Lichtenstein, and Fischhoff (1980) and of Griffin and Tversky (1992) have been especially prominent and have enjoyed some measure of success in their ability to account for actual judgment data. This then raises the question of whether there are differences in how people from various cultures tend to carry out the activities in the argument process described previously that, in turn, might contribute to the cultural variations in probability judgments at issue here.

There is reason to believe that this speculation could have some validity. A review of psychological, historical, and philosophical literature conducted by Yates, Lee, Shinotsuka, and Sieck (2000) indicated that traditional Chinese learning has emphasized a kind of collective empiricism. This is embodied in special reverence for (and emulation of) ways of doing things that have proved their mettle over long periods of time. (Aphorisms such as “Why reinvent the wheel?” should seem especially compelling given this point of view.) Furthermore, it appears that Chinese tradition has had little respect and use for deliberation techniques such as debate, which pit competing ideas head to head in open forum (see also Nisbett et al., 2001). This would suggest that, when bringing to mind arguments pertinent to some issue, as in Element 1 of the strategy sketched previously, Chinese argument recruitment customs should be distinctive. Yates et al. (2000) reported data consistent with this possibility. In one study, participants were asked to list arguments for and against each of the alternative answers to general-knowledge questions posed to them. American participants were almost twice as likely as Chinese participants to bring to mind arguments that disagreed with the alternatives they ultimately indicated to be the ones they thought were the correct ones, and the incidence of such choiceinconsistent arguments cited by the Japanese participants was more than twice that for the Chinese (and significantly higher than that for the Americans). The idea certainly needs to be examined more carefully and extensively, but results like these suggest that one plausible contributor to Chinese participants being so distinctively highly confident in the correctness of their chosen answers to general-knowledge questions is that they see little reason to doubt that those answers are indeed correct.