Children's Attributions 12
participated in several other studies in the region and is known to many of the participants’
families.
The experiment consisted in a version of the “surprising contents” task. Participants were
interviewed either in the hut of the experimenters or in their family’s hut. The agents used were a
doll named Soledad and the Catholic God (the Maya have adopted this religious entity into their
pantheon several centuries ago). We decided not to use the mother as stimulus in Yucatan as it
proved impossible to interview the children while their mother was away. In this situation, it
would not have been feasible to control for the possibility of the child thinking the mother had a
chance to see what was inside the container. The researchers used a container made out of a dried
squash, known in Yukatek as ho’ma, which keeps maize tortillas warm after cooking them. The
ho’ma has a small opening carved out on top, just large enough to put one’s hand through. Every
family visited by the experimenters owned at least one and usually several of these containers.
Although they may be occasionally used to store other objects, there was high consensus among
the participants that the normal, appropriate content was indeed tortillas, as measured by control
questions asked at the beginning of the experiment (“What is this container called?”; “what
would you usually find in it?”).
The ho’ma’s opening was closed with a piece of cardboard, so that children could not tell
what was inside. One of the experimenters opened the container to reveal a pair of shorts, a most
unusual content. The container was closed again and the experimenters then asked the set of
questions about the doll and God, in the following form: “What does X think is in the ho’ma?” In
this experiment, children were not asked questions about other agents’ behavior. However,
Barrett et al. (2001) obtained very similar results when a sample of US children were asked a