How can societies learn for or from catastrophes?

The maxim “learning from damage” refers to the general assumption that we, as individuals, are capable of learning from experienced calamities. We take it for granted that this also holds true for groups, enterprises, or the respective society as a whole, as well as the (disaster-) management organisations and institutions for prevention and mitigation involved. The idea of learning from disasters is expected to be better prepared or to be more efficient the next time. Thus, we assume that the results of these learning processes are preserved as "knowledge" in the collective memory of a society, and that patterns of practices of learning have been adopted on this basis. However, looking closer at post-disaster learning, there is some evidence for the opposite: Analysing past calamities fairly often reveals hardly any societal learning (e.g. Fukushima) and, if so, “learning” often turns out to consist of quick fix solutions with unintended side effects (e.g. the suicidal Germanwings pilot), or the disaster memory rarely lasts more than two generations (e.g. South Pacific-Tsunami 2004).