Much experimental research has been conducted in laboratory settings on human behavior related to public goods, common-pool resources, and other social dilemmas. These studies have shown that when subjects are anonymous and not allowed to communicate, they tend not to cooperate. However, to the surprise of game theorists, simply allowing subjects to communicate in a laboratory setting enables them to achieve far more cooperative outcomes. The replication of the experiment in laboratory settings in multiple countries as well as in some initial field experiments has only confirmed this important finding. However, while carefully conducted laboratory experiments do have strong internal validity, external validity requires further research beyond the initial field experiments that researchers have begun to conduct. In this paper, we report on a series of common-pool-resource field experiments conducted in eight indigenous communities in India that have very long traditions of shared norms and mutual trust. We used two experimental designs in all eight villages: a "no-communication" game where no one was allowed verbal or written communication and a "communication game" in which the same five participants were allowed to communicate with each other at the beginning of each round before making their decisions. The findings from these field experiments are substantially different from the findings of similar experiments conducted in experimental laboratories. Subjects tended to cooperate in the first design even in the absence of communication. Our findings suggest that the shared norms in these indigenous communities are so deeply embedded that communication is not essential to arrive at cooperative decisions. However, communication does homogenize group and individual outcomes so that communities that are overly cooperative tend to reduce cooperation slightly while those showing small deviations in the other direction move toward the optimal solution.