We are in full agreement with Leonard, and many others, that the sampl
e size-richness relation is an important concern for archaeologists, a
nd we appreciate his evenhanded discussion of our paper However; we di
sagree with Leonard regarding the point in the research process when v
ariable prehistoric behavior can be used to explain differences in ric
hness. Leonard and others argue that we must always begin studies of r
ichness by controlling any observed relation to sample size; we can on
ly consider prehistoric behavior as an explanation of richness after w
e account for sample size effects. In contrast, toe suggest that follo
wing his recommendations will obscure aspects of behavioral variation
that are important to understanding prehistoric societies. Considerati
ons of prehistoric behavior must therefore be central in studies of ri
chness.