This talk will highlight the tension between standardization and improvisation that shaped, and continues to shape, the nuclear industry – in the former Soviet Union and elsewhere. This tension was significant not only for personnel recruitment, training, and management style, but informed technical aspects such as reactor design choices, and the organizational set-up of the entire nuclear sector.

I focus in particular on improvisation as an often neglected aspect of organizational structure and behavior, and its significance for nuclear safety. Drawing on work that has pitted organizations prone to “normal accidents” (Perrow 1984/1999) against “high reliability organizations” (Roberts 1990, Weick 1987), as well as on Diane Vaughan’s work on the “normalization of deviance” (1996, 1999, 2005) and Constance Perin’s “culture of control” (2005), I argue that the idea of control (of both technological artifacts and practices) is of limited value in a branch of industry  that involves complex, high-risk technologies. The safety of such technologies lies precisely in the ability of operators to improvise, to make decisions on the fly, occasionally in violation with written rules, and to correct course independently and competently in a volatile environment.

Based on primary sources from Russian archives and interviews with nuclear industry veterans, I show how the technologies and practices that both formed, and grew out of, the development of the Soviet nuclear industry shaped the ways in which the Soviet state reacted to the Chernobyl catastrophe. With the ongoing nuclear disaster at Fukushima in mind, I will raise a few questions that relate these fundamental tensions to nuclear emergency response in general, and in the future.