Imagine putting together a jigsaw puzzle that works like the board game in the movie “Jumanji”: When you finish, whatever the puzzle portrays becomes real. The children playing “Jumanji” learn to prepare for the reality that emerges from the next throw of the dice. But how would this work for the puzzle of scientific research? How do you prepare for unlocking the secrets of the atom, or assembling from the bottom-up nanotechnologies with unforeseen properties – especially when completion of such (...) puzzles lies decades after the first scattered pieces are tentatively assembled? In the inaugural issue of this journal, Michael Polanyi argued that because the progress of science is unpredictable, society must only move forward with solving the puzzle until the picture completes itself. Decades earlier, Frederick Soddy argued that once the potential for danger reveals itself, one must reorient the whole of one’s work to avoid it. While both scientists stake out extreme positions, Soddy’s approach – together with the action taken by the like-minded Leo Szilard – provides a foundation for the anticipatory governance of emerging technologies. This paper narrates the intertwining stories of Polanyi, Soddy and Szilard, revealing how anticipation influenced governance in the case of atomic weapons and how Polanyi’s claim in “The Republic of Science” of an unpredictable and hence ungovernable science is faulty on multiple levels. (shrink)