The Field Guide › Paper
Helmreich, Merritt and Wilhelm trace Crew Resource Management through five generations, and the arc they describe is really the story of a good idea repeatedly failing to stick until its rationale was made explicit. The first generation, launched by United Airlines in 1981, was borrowed wholesale from corporate managerial-style training, heavy on psychological testing and games unrelated to flying, and pilots dismissed it as charm school. The second, prompted by a NASA workshop in 1986, renamed Cockpit to Crew Resource Management and grew more aviation-specific, but still drew charges of psycho-babble (the word synergy came in for particular scorn). The third broadened CRM to flight attendants, dispatchers and maintenance staff, which was necessary but, the authors suspect, diluted the original focus on reducing error. The fourth folded CRM into every part of technical training and checklists under the FAA's Advanced Qualification Program, proceduralising it, sometimes to the point that pilots could describe CRM only as training to make us work together better, having lost sight of why that mattered. Attitudes measured years after initial training routinely decayed even with recurrent refreshers, and a persistent minority of pilots, known to their colleagues as Cowboys, Boomerangs or Drongos, rejected the concepts outright regardless of how they were taught. The fifth generation, which the authors propose, replaces technique with a single explicit rationale: human error is inevitable, so CRM is error management, a set of countermeasures to avoid error, trap it before it is committed, and mitigate its consequences when it is not trapped, adopted under an explicitly non-punitive organisational stance influenced directly by James Reason. The paper's other major finding is that CRM did not export well: courses built around American assumptions of low power distance, individualism and low uncertainty avoidance met resistance in cultures where hierarchy is taken for granted and challenging a captain is itself the violation, using Hofstede's dimensions to explain why identical training produces different uptake in different national cultures. For a corpus about psychological safety this is essential grounding, not because it introduces a new idea but because it documents, with unusual candour, what happens when a psychological-safety-adjacent programme is taught as compliance rather than as reasoned practice: acceptance decays, a purpose-shaped hole opens up, and some proportion of any population will not be reached by training alone, whatever the generation. Its limits are that the evidence for each generation's effectiveness is largely attitudinal and anecdotal rather than outcome-based, a limitation the authors themselves state directly, since the safety metric that would settle the question, the accident rate, is too rare an event to yield a clean answer. (Text drawn from the 1999 International Journal of Aviation Psychology paper, 9(1), pp. 19-32.)