The goal of this paper is to articulate a set of design principles that support transfer. In doing so, I will first describe a perspective on transfer in which an idea is not so much abstracted from its original context and applied to a new context (as transfer is often conceptualized), but one in which the learning context is invoked in a new context — for example, when opportunities to extend a classroom investigation on light are recognized and pursued outside of class — in what I will refer to as intercontextuality (cf., Wagner, 2006; Engle, 2012). After characterizing a range of context domains that may be positioned intercontextually, I will then argue that such intercontextuality is fostered in classrooms that are themselves intercontextual: where out-of-class contexts (e.g., physical, social, functional, and temporal contexts) are invoked by students in scientifically consequential ways as they develop and vet ideas. I describe episodes of such intercontextuality in detail, with transcripts, field notes and artifacts from a course, Scientific Inquiry (Atkins Elliott, Jaxon, Salter, 2016), that shows evidence of high transfer (Atkins and Frank, 2015). Finally, I will argue that intercontextuality in class is supported by disruptions (Ma, 2016) to traditional instructional practices that confer power on students, i.e., a devolution (Brousseau, 1997). In this final section, I contrast structures and expectations in a traditional physics course with those from Scientific Inquiry, describing how those structures represent a disruption and devolution that sponsor intercontextuality and, therefore, transfer.
