Historical study has long used texts as a way to render the past thinkable. In the past half century, questions surrounding textual interpretation have focused on the role of historical analysis and its relationship with sociology, psychology, and semiotics -- particularly within the study of history and the philosophy of science. This is particularly true for studies involving curriculum history and analyses of the "work" of teachers where oral histories, narrative examinations, cultural linguistics and storytelling have all impacted the ways in which teachers are studied and analyzed. This paper argues that the decision of what types of text to "read," and accordingly what type of research to do when studying teacher education, is a type of intellectual self-discipline that obliterates the technologies of power which generate meaning in the discourse. Drawing from the work of Michel Foucault, the paper suggests that changing what "counts" as text does not actually change what can be written as history.