THE DEMARCATION OF SCIENCE: A PROBLEM WHOSE DEMISE HAS BEEN GREATLY EXAGGERATED BY STEVE FULLER JL/ arry Laudan has recently argued that the demarcation of science from non-science is a pseudo-problem which should be replaced by the more modest task of determining whether and why particular beliefs are epistemically warranted or heuristically fertile. In response, Thomas Gieryn has rejected this exercise in philosophical self- restraint, arguing that the scientific community itself takes measures to distinguish itself from others who compete for cognitive authority and its attendant political and economic benefits. As a result, rhetorical strategies develop— “boundary work” Gieryn calls them— which are proper objects of sociological study. In this paper, I shall argue that Laudan is wrong to think that there is nothing at stake in demarcating science from non-science, but I shall also argue that the importance of the demarcation problem has not been fully appreciated by Gieryn. My critique will reveal a basic con- fusion that needs to be dispelled before the full significance of demarcating science from non-science can be seen: namely, the failure to distinguish the relatively constant social role played by what has been called “science” from the historically variable social practices that have played the role of science. In addition, I shall claim that this confusion is itself one of the key ways in which a social practice retains its status as science. Laudan’s^ history of abortive efforts at solving the demarcation problem may be reconstructed as two general arguments for rejecting the project Pacific Philosophical Quarterly 66 (1985) 329-341 0031-5621/85/0300-0329-$01.30 Copyright © 1985 by University of Southern California 329