Research Report The Curse of Knowledge in Reasoning About False Beliefs Susan A.J. Birch 1 and Paul Bloom 2 1 University of British Columbia, Vancouver, British Columbia, Canada, and 2 Yale University ABSTRACT—Assessing what other people know and believe is critical for accurately understanding human action. Young children find it difficult to reason about false beliefs (i.e., beliefs that conflict with reality). The source of this difficulty is a matter of considerable debate. Here we show that if sensitive-enough measures are used, adults show deficits in a false-belief task similar to one used with young children. In particular, we show a curse-of-knowledge bias in false-belief reasoning. That is, adults’ own knowledge of an event’s outcome can compromise their ability to reason about another person’s beliefs about that event. We also found that adults’ perception of the plausibility of an event mediates the extent of this bias. These findings shed light on the factors involved in false-belief reasoning and are dis- cussed in light of their implications for both adults’ and children’s social cognition. Reasoning about what other people believe is often essential for predicting and interpreting human action. A wealth of research has investigated children’s appreciation that the mind can misrepresent reality—that is, their appreciation that people can hold false beliefs. Most research on children’s false-belief rea- soning has utilized some variant of the displacement task (e.g., Baron-Cohen, Leslie, & Frith, 1985; Wimmer & Perner, 1983). For example, subjects are told a story about Sally, who puts her candy in a box and leaves the room. In her absence, another character moves the candy to a basket. When Sally returns, where will she look for her candy? The right answer, that she will look in the box, requires attributing a false belief to Sally. Four- year-olds tend to do fairly well at such tasks, but younger children tend to fail (see Wellman, Cross, & Watson’s, 2001, meta-analysis). Younger children tend to answer in accord with their own knowledge, saying that Sally will think the candy is in the basket. The source of children’s difficulty is a matter of considerable debate. Some researchers interpret children’s difficulties on these tasks as reflecting a conceptual deficit: Perhaps young children lack a concept of belief or a concept of mental repre- sentation more generally (e.g., Gopnik, 1993; Perner, Leekam, & Wimmer, 1987; Wellman, 1990; Wellman et al., 2001). An alternative view is that young children’s problems are due to more general cognitive factors such as memory and processing limitations, and thus not necessarily indicative of a conceptual limitation (e.g., Fodor, 1992; German & Leslie, 2000; Leslie, 1987; Onishi & Baillargeon, 2005; Roth & Leslie, 1998; Zait- chik, 1990; for a discussion, see Bloom & German, 2000). According to one version of this alternative, children have the same bias in perspective taking as adults, only to a greater extent. In earlier work (Birch & Bloom, 2003), we showed that 3- and 4-year-olds are more susceptible than 5-year-olds to a cognitive bias found in adults, the curse of knowledge (see also Bernstein, Atance, Loftus, & Meltzoff, 2004; Pohl & Haracic, 2005). We adopted this term from Camerer, Loewenstein, and Weber (1989) and use it to refer to the tendency to be biased by one’s own current knowledge state when trying to appreciate a more naive perspective, whether that more naive perspective is one’s own earlier perspective (as in the hindsight bias) or someone else’s perspective (see Birch & Bernstein, 2007, for a discussion). We proposed (Birch & Bloom, 2003) that the exaggeration of the bias in younger children may partially account for why younger children experience greater difficulty than older chil- dren on false-belief tasks (see also Birch, 2005; Birch & Bloom, 2004). If this is correct, then one might expect adults to also experience problems in false-belief reasoning if sensitive- enough measures were used. The logic mimics that of Diamond and Kirkham (2005), who showed that when sensitive measures are used, even adults have difficulty on a problem (the dimen- sional-change card-sort task) that is particularly difficult for children. We should emphasize, however, that such a finding of adult difficulty in the domain of false-belief reasoning would not prove the conceptual-deficit account false. It is possible, for instance, that children’s problems are due to both an exagger- ated curse-of-knowledge bias and conceptual limitations. Address correspondence to Susan Birch, Department of Psychology, University of British Columbia, 2136 West Mall, Vancouver, BC, Canada V6T 1Z4, e-mail: sbirch@psych.ubc.ca. PSYCHOLOGICAL SCIENCE 382 Volume 18—Number 5 Copyright r 2007 Association for Psychological Science