Ongoing efforts to include social contextual dynamics in research on mathematics education require theoretical frameworks that allow researchers to zoom their focus out to include the social sphere (Lerman, 1998). This is especially important when studying teacher development, as teachers learn in a wide variety of social contexts. Here we illuminate two key concepts that are helping us better understand the affects of various communities on the development of our cooperating teachers. These concepts are Wenger's (1998) modes of participation and regimes of accountability. By looking specifically for variation among different communities' regimes of accountability, and the modes of participation they engender, we are able to better understand the development (or lack thereof) that we are seeing take place in our participants. [For complete proceedings, see ED500860.]