An interesting study by J. Kiley Hamlin, Karen Wynn and Paul Bloom, in the psychology department at Yale, showed that infants as young as 6 or 10 months tended to want to play with *good guys* rather than *bad guys*. It is hard to believe that such behavior would be taught by parents at such a young age.
In the Hamlin et al. experiments, triangles, squares and circles (with eyes) were manipulated by researchers so that one of the shapes helped a second shape up the hill, only to be pushed down the hill by the third shape. Infants given a choice of which shape to play with after this skit inherently shied away from playing with the “bad shape” that had done the pushing and almost universally wanted to play with the “good shape” that had helped another get up the hill.
This research is consistent with a belief that we’ve long asserted that we have been bred for evolutionary reasons to be able to assess trustworthiness, because people in earlier societies who were not good at judging trustworthiness were more likely to die (e.g., knifed in the back; abandoned by a partner) and thus less likely to be able to pass on their genes. Thus over time, we’ve become especially good at judging trustworthiness. And experiments show that if you assess what percentage of residents in a community believe that others can be trusted, it turns out to be significantly predictive of what percent of wallets dropped in these communities by researchers will be returned to the address listed in the wallet with the money still in the wallet. This suggests that our assessments of trustworthiness (at the community level, with some noise for individual optimism or paranoia) are fairly accurate.
The study appears in the November 22 issue of Nature in “Social Evaluation by Preverbal Infants”. Abstract here.