This article addresses the problem of treating generalizations of human activity as entities and structures that ultimately explain the activities from which they were initially drawn. This is problematic because it involves a circular reasoning leading to unwarranted claims explaining the originally studied activities of science teaching and learning. Unlike other fields within social science research, this problem has not been appreciated and discussed in the science education literature and the field thus needs to be reminded of it. A heuristic specifically developed for the purposes of this article is applied to two examples taken from a much-cited research in the field. Through the examples it is argued that the practice of creating entities out of generalizations of science classroom activities leads to a number of unintended consequences. It is further argued that the stated purposes in the two example articles would actually have been better served by investigating the entire processes through which the activities develop, as well as how the activities may change through teaching. The article concludes that through the search for explanations caused by underlying entities, science education research runs a risk of alienating its results from the activities from which it initially wanted to meliorate.