When major accidents or adverse events occur it is quite common for society to respond by setting up an incident investigation, through an ad hoc committee or permanent investigation boards or public supervisory bodies. In future, it is likely that such investigation processes will increasingly have to deal with incidents in systems that also have been an object of prior risk analyses.
The aim of this article is to discuss from a theoretical point of view how incident investigation processes can approach incidents preceded by relevant risk analyses and even criticize or challenge the established risk picture. The theoretical perspectives will be illustrated by material from investigation reports and other relevant comments on four specific incidents, cf. Table 1.