Radiocarbon dated features at the newly investigated Dawson Creek site in southern Ontario, Canada, document recurrent Amerindian occupations between the ninth and fifth centuries B.C. Artifactual, botanical, and faunal remains from six hearths evidence consistent fall season visits by task groups using Vinette 1 ceramics. Indicated reliance on nut-collecting and deer-hunting is duplicated at small Early Woodland sites elsewhere in Ontario and the Great Lakes region. Early thick ceramic types are clearly associated with scheduled seasonal activities. Possible differentiation of artifact assemblages by site function and season underscores the need for more discriminating definition of Early Woodland culture in the Northeast.

In a recent article in American Antiquity, Muller (1984) challenged the claim that Mississippian salt production was a specialized activity. He made a useful distinction between special activity sites and specialized production in his study, but by limiting his investigation of Mississippian specialization to a single task at a special purpose site (the Great Salt Spring in Galatin County, Illinois), he fails to prove or disprove his hypothesis that no non-leadership specialists existed in Mississippian society. Muller's article also raises some issues about the relation of economic specialization to levels of social complexity in prehistoric cultures, and about measuring the degree of economic specialization in a society from archaeological remains. These issues are addressed in this comment.

Yerkes's comments fail to place my article in its context of other research on the lower Ohio Valley. He reacts more to what he thinks I believe than to what I actually wrote. Yerkes seems to want to identify Mississippian specialization on the basis of artifact traits rather than on the basis of actual organization of production. Moreover, he does not present any additional significant evidence to support the idea that true specialization was present in Mississippian society, in the American Bottom or elsewhere. The need to reject lower-level explanations of data before proposing higher-level explanations is not sufficiently appreciated. Finally, there is not sufficient appreciation of the degree of importance of the concept of specialization to the investigation of the social evolution of hierarchical systems.

In a recent article based on ethnoarchaeological research in Mesoamerica, Cannon has suggested that presence/absence statistics are generally preferable to percentage statistics. The case presented is weak, and seems only to reflect a specific instance rather than a general rule.

Comments are offered on the inappropriate use of Texas-derived projectile point typology at certain Mesoamerican sites. Moholy-Nagy et al. (1984) recently applied labels to point groups from the Maya site of Tikal that are misleading in terms of the original type descriptions. Similar misuse of the Texas typology in central Mexico and in the Tehuacan Valley is noted.