This article draws on empirical material to reflect on what drives rapid
change in flood risk management practice, reflecting wider interest in
the way that scientific practices make risk landscapes and a specific
focus on extreme events as drivers of rapid change. Such events are
commonly referred to as a form of creative destruction, ones that reveal
both the composition of socioenvironmental assemblages and provide a
creative opportunity to remake those assemblages in alternate ways,
therefore rapidly changing policy and practice. Drawing on wider
thinking in complexity theory, we argue that what happens between events
might be as, if not more, important than the events themselves. We use
two empirical examples concerned with flood risk management practice: a
rapid shift in the dominant technologies used to map flood risk in the
United Kingdom and an experimental approach to public participation
tested in two different locations, with dramatically different
consequences. Both show that the state of the socioenvironmental
assemblage in which the events take place matters as much as the
magnitude of the events themselves. The periods between rapid changes
are not simply periods of discursive consolidation but involve the
ongoing mutation of such assemblages, which could either sensitize or
desensitize them to rapid change. Understanding these intervening
periods matters as much as the events themselves. If events matter, it
is because of the ways in which they might bring into sharp focus the
coding or framing of a socioenvironmental assemblage in policy or
scientific practice irrespective of whether or not those events evolve
the assemblage in subtle or more radical ways.