ABSTRACT
This chapter argues that it is worthwhile to investigate datasets to see if a particular set of cultural material can be adequately described as a culturally transmitted unit. It examines three problems or errors that can adversely affect our efforts to construct cultural phylogenies. One error that is relatively easy to make is to identify a basic element of behavior in one individual or group and associate it with a pseudo-ancestor that also shares the element. A second error is that the tales told about the history of cultural features do not reflect the true history of the material under investigation. A third problem facing those who wish to apply phylogenetic theory to anthropology is the mistaken identification of a polygenetic element as having its origin in one particular lineage that contributed to its history. Both methods depend on an appropriate choice of characters through which dissimilarity matrices representing the cultural objects can be constructed.
