A distinction between the "hard" and "soft" scientific disciplines is a modern commonplace, widely invoked to contrast the natural and the social sciences and to distribute value accordingly, where it was generally agreed that it was good to be "hard," bad to be "soft." I trace the emergence of the distinction to institutional and political circumstances in the United States in the second part of the twentieth century; I describe varying academic efforts to give the contrast coherent meaning; I note the distinction's uses in disciplines' reflections on their own present and possible future status; and I document the consequential circulation of the antonym in settings where resources for science were distributed. To follow the history of the "hard-soft" distinction is to open a window on changing sensibilities about what science is, what values are attached to it, and what it is for. I conclude with speculations about more recent changes in the value-schemes implicated in the "hard" and the "soft" and about pertinent changes in the place of the "soft" human sciences in governance and production. I envisage a possible future in which the commonplace distinction might wither away.