The Field Guide › Paper
The paper that reframed high reliability organisations around cognition rather than structure, and introduced 'collective mindfulness' as the mechanism by which reliable performance is continuously re-accomplished. Where Perrow's normal accident theory (in this corpus) treats interactive complexity and tight coupling as macro-structural givens that dominate outcomes, and where the earlier Berkeley HRO work (La Porte & Consolini, also here) catalogued the conditions of reliable organisations, Weick, Sutcliffe and Obstfeld supply the missing micro-level process account: what people in effective HROs actually do, moment to moment, that keeps small errors from cumulating into catastrophe. They identify five processes that together produce a state of mindfulness, understood as a rich awareness of discriminatory detail coupled to a capacity for action: a preoccupation with failure (treating any lapse, and any near miss, as a window on the health of the whole system, and treating the liabilities of success (complacency, inattention, habituation) as themselves failures); a reluctance to simplify interpretations (cultivating requisite variety and 'conceptual slack', divergent analytical perspectives, and a scepticism that double-checks rather than defers); a sensitivity to operations (the shared, effortful 'bubble' of situational awareness held collectively in the moment); a commitment to resilience (capacity to cope with, contain, and bounce back from surprises that anticipation failed to prevent, often through ad hoc epistemic networks that self-organise around a problem and dissolve when it passes); and underspecification of structures, the process later reframed as 'deference to expertise' in Weick and Sutcliffe's Managing the Unexpected, in which hierarchical rank is deliberately subordinated to expertise so that decisions migrate to whoever has the relevant knowledge, wherever they sit in the formal order. That fifth process is the direct conceptual sibling of the tempo-migrating authority La Porte and Consolini observed on carrier decks, and its logic is squarely a psychological-safety logic: it works only where lower-status members can act on, and speak to, what they notice without waiting for permission. The paper's account of preoccupation with failure is built explicitly on error-reporting climate: it cites Edmondson's (1996) finding that better-led nursing units reported more errors because openness, not infallibility, was what distinguished them, alongside the organisational habit of rewarding rather than punishing those who report their own mistakes (the engineer sent champagne for owning a costly error; the sailor commended for reporting a lost tool that grounded all aircraft). Mindfulness, the authors argue, does not merely coexist with the structural dangers Perrow describes but actively counters them: it increases comprehension of complexity and loosens tight coupling, treating technology as an equivoque to be interrupted and redirected rather than an imperative to be suffered. The five processes have become one of the most widely used vocabularies in practitioner safety and psychological-safety writing.