This article looks at how parts of a social stock of knowledge can become insulated from each other via their uneven distribution both "horizontally" across time and space, and "vertically" with respect to degrees of embodiment in unconscious habits and routines. It uses ideas from Alfred Schutz, Peter Berger, Thomas Luckmann, Michael Polanyi, and others to argue that this insulation can produce a highly dynamic structuring of knowledge, awareness of which has the potential to help explain the existence of ignorance, (...) misperception, and multiple interpretations in different social settings. This potential is illustrated with examples taken from a study of American economists, showing how an approach considering insulation can improve understanding of at least one currently influential branch of knowledge. The article also suggests briefly how such an approach might augment recent theories of habitual action by accounting for both stability and change, and even help with some longstanding epistemological problems in social theory. (shrink)