In the last three decades there have been a variety of studies of what is often referred to as "everyday" or "street" mathematics. These studies have documented a rich variety of arithmetic practices involved in activities such as tailoring, carpet laying, dieting, or grocery shopping. More importantly, these studies have helped to rectify outmoded models of rationality, cognition, and (school) instruction. Despite these important achievements, doubts can be raised about the ways in which theoretical conclusions have been drawn from empirical materials. Furthermore, while these studies rightly criticized prevalent theories of rationality and cognition as too simplistic to account for everyday activities, it seems that some of the proposed alternatives suffer from similar flaws (i.e., are straightforward inversions of the to-be-opposed theories, rather than more nuanced views on complicated issues). In this article we illustrate our skeptical view by discussing four case studies in Jean Lave's pioneering and influential "Cognition in Practice" (1988). By looking at the case studies in detail, we investigate how Lave's conclusions relate to the empirical materials and offer alternative characterizations. In particular, we question whether the empirical studies demonstrate the existence of two different "kinds" of mathematics ("everyday" and "school," or "formal" and "informal") and whether school instruction tries to replace the former with the latter.