ABSTRACT
In a seminal study, Bartlett (1932) asked British university students to read a culturally unfamiliar Native North American folk tale called “the war of the ghosts” and retell it to others in writing who in turn passed it on to others. Over several generations of retellings culturally unfamiliar concepts were distorted and replaced by more familiar concepts; for instance, a canoe was replaced by a rowboat. Bartlett reasoned that unfamiliar concepts such as canoe are more difficult to represent and thus less likely to be represented and transmitted than familiar concepts. However, recent studies following up on Bartlett’s work (Boyer and Ramble 2001, Barrett and Nyhof 2001, Atran 2003) have drawn the seemingly opposite conclusion, namely that under some conditions, counterintuitive concepts (i.e., concepts that violate intuitive expectations) are actually better recalled than relatively more intuitive concepts. Barrett and Nyhof (2001) asked their subjects to remember and retell three of six different Native American folktales containing an equal number of intuitive and counterintuitive concepts. A content analysis of what they remembered showed that subjects remembered significantly larger number of counterintuitive concepts than intuitive concepts. In subsequent experiments, Barrett and Nyhof (2001) and Boyer and Ramble (2001) constructed a number of stories containing an equal number of three different types of concepts: the minimally counterintuitive concepts (i.e., concepts that violate intuitive expectations regarding one or two feature values) such as “furniture that flies in the air”, ordinary concepts such as “furniture that stays where you put it”, and maximally counterintuitive concepts (i.e., concepts that violate intuitive expectations regarding multiple feature values) such as “furniture that flies in the air, melts when try to grab it, and is made of uranium.” When subjects from a variety of cultural backgrounds were asked to recall the stories containing an equal number of three types of concepts, they recalled a significantly larger number of minimally counterintuitive concepts than both ordinary and maximally counterintuitive concepts. The results regarding better recall for minimally counterintuitive concepts visa-vi ordinary concepts are in accordance with a host of studies showing that unusual or distinctive stimuli are generally better remembered than stimuli that are not (see Waddill and McDaniel 1998 for a review). A number of factors have been proposed to account for these effects including:
more attention being paid to the unusual stimuli resulting in richer spontaneous elaboration and encoding for them and less attention and time being spent on processing the usual stimuli,
unusual stimuli being encoded in their own category, different from the one into which the common stimuli are clustered,
unusual stimuli being distinctive in the retrieval set that is formed.
Boyer and Barrett have argued that it is the economy of representation as simple negation of a feature value and the ability to invoke the counterintuitive concept to make predictions because of its set of non-violated feature values that makes minimally counterintuitive concepts easier to recall than ordinary concepts. They argue that maximally counterintuitive concepts are harder to recall because they have little “inferential potential” as they violate too many intuitive expectations.
