John Dewey and the National Research Council NRC both discuss the problem of translating scientific research into contexts of schooling, but differ about the proper solution to the problem. The NRC would solve it by empirical investigation. Dewey finds value in that approach, but also wants educational theorists to construct general heuristics to guide scientists in creating research agendas of intrinsic interest to education practitioners. Dewey's plan faces an apparently insurmountable difficulty. General heuristics of the sort Dewey needs are not widely recognized to exist. In light of the presumed fact of education's inability to articulate a general framework to guide teaching practice, the NRC plan becomes preferred by default. I propose that Experience and Education provides a frame of reference from which pedagogical practice appears as a field of scientific endeavor in its own right. Conceptualizing teaching as a science suggests ways to rework the NRC plan.