The Logic of Failure is a popular translation of what appears to be some pretty hefty scholarly literature (I think-didn't bother to actually check 30 years of literature in German), that is hindered by becoming largely accepted wisdom. Dorner is a cognitive scientist who based this book on a series of studies of how people interacted with computer models: desertification in the Sahel, the economy and politics of a small town, predator and prey interactions. These studies, along with some examples drawn from recent events like Chernobyl and military history, are used to explain failure a consequence of a lack of understanding of complex systems.

Complex systems, interconnected networks with time-delays, buffering units, hidden keystone variables, and unclear indicators, are everywhere in the real world. Unfortunately, human minds tend to think linearly and concretely. Dorner documents several pathological thinking styles he encounters in his experiments. Some people over-correct, making dramatic changes while chasing a pointer that drowned out any data in induced oscillations. Some people get lost chasing irrelevant details, asking for more information rather than acting. And some people get trapped in methodism, following a predetermined course of action in complete disregard of the information coming in.

Against this, Dorner advocates for having a clear mental model of a system, discrete objectives, and a holistic sense of possible higher-order effects. Make small changes, seek steady states, and do not try and race a chaotic system. He points towards 'wisdom' with maddening vagueness. If there's a major problem with this book, it's that it's been overtaken by the zeitgeist. Dorner's methods are now children's toys rather than cutting edge science. We all 'get' networks and complexity, but we still lack the language to truly understand them.