There has been a strong tradition of assuming that concepts, and their patterns of formation might be best understood in terms of how they are embedded in theory-like sets of beliefs. Although such views of concepts as embedded in theories have been criticized on five distinct grounds, there are reasonable responses to each of these usual objections. There is, however, a newly emerging concern that is much more challenging to address - people's intuitive theories seem to be remarkably impoverished. In fact, they are so impoverished it is difficult to see how they could provide the necessary structure to explain differences between concepts and how they might form in development. One response to this recent challenge is to abandon all views of concept structure as being related to people's intuitive theories and see concepts as essentially structure-free atoms. The alternative proposed here argues that our very weak theories might in fact do a great deal of work in explaining how we form concepts and are able to use them to successfully refer.