p.640 The organic, loosely coupled organization, flexible and responsive, dedicated to learning new solutions to new problems, has become the accepted wisdom (e.g., Drucker, 1992; Nonaka, 1991).

p.640 tightly coupled, meaning that perturbations are transmitted rapidly between subsystems with little attenuation

p.642 it becomes increasingly more difficult to measure and evaluate indiviual performance as the organization becomes more complex.

p.643 Weick... argued (1987,p. 112) that accidents occur because the human beings who manage and are integrated into these complex systems are insufficiently complex to sense and therefore anticipate the system's problems. [JLJ - no, I would say that accidents happen because the cues and symbols that we methodically use everyday to determine how to 'go on' - that usually we can read with clearness and detail - are occasionally ambiguous, or we get lazy, and briefly mis-attend, or we fail to cross-check our results, where complexity or uncertainty would have us at attention. Whatever a human does, no matter how complicated the task, there is always a choice to attend or not, to be mindful or not. We can perhaps get by with brief periods of inattention. But there is no excuse for lengthy periods where we ignore the signs and symbols of critical performance - our performance dashboard. There is no excuse for ignoring well-crafted and -intentioned alarms to awaken us to an alert state. There is no excuse for not being mindful. Accidents do not just happen. Accidents are symptomatic of ignoring the cues and important signs freely available (yet perhaps buried in noise and requiring effort to interpret) to our perceptions.]

p.643 Most practical learning is by trial and error.

p.644 March, Sproull and Tanner (1991) suggested an alternative to the scientific strategy for organizational learning. They argued that collectives which are internally diverse are able to apply that diversity to small samples of critical events and so experience them 'richly', probing their varied manifestations and interpretations more completely. Thus the paucity of data is balanced by the richness of the analysis.

p.645-654 [Description of Admiral Rickover's HRO]