[Evaluating Greenberg's hypotheses regarding American Indian language families]