The title is the thesis. Pre-Accident Investigations: an investigation before anything goes wrong. With it, Todd Conklin inverts the founding assumption of conventional safety work. Normally the incident is the triggering event. First the bang, then the review, then the corrective action. Conklin’s 2012 book asks the uncomfortable question of why an organisation should wait for harm to learn something it could have seen beforehand. In this logic, incidents are the second-best option: knowledge paid for late and in blood about a state that normal work had been signalling all along.
In the last essay, What Happens Between the Audits, I worked through two tools that come straight from this book: Learning Teams and Pre-Job Briefs. Here I am after the thinking behind them, less the tools themselves. The tools are the easy half. The hard half is the mindset that makes them mean anything in the first place.
The inversion
Conklin’s starting point is a simple observation about attention. An organisation that draws its safety knowledge from incidents has a data problem: serious events are rare, and the rarer they are, the less you can build a reliable picture from them. What does happen thousands of times a day is normal work that goes well. That, for Conklin, is where the real source of learning sits, in everyday normal operation and not in the rare exception. It is there that people make a continuous stream of small adjustments so that an imperfect system keeps working.
This is the operational sister of Hollnagel’s Safety-II: safety shows itself not in the accident avoided but in the capacity to handle variation. Conklin turns that into an instruction for the front line. Look at the work as it is actually done, and do it before the work fails. A “pre-accident investigation” is nothing other than disciplined curiosity about the normal case.
An organisation that only looks after the bang reliably learns one disaster too late.
Five principles, and the mindset behind them
The core of the book, later distilled in Conklin’s work into five HOP principles, can be put briefly. Error is normal: even competent, motivated people make mistakes, and a system built on freedom from error is built on an illusion. Blame fixes nothing: looking for the culprit ends the learning exactly where it would get interesting. Context drives behaviour: people act the way conditions invite them to, and whoever wants to change the behaviour has to reach the conditions. Learning is a deliberate activity, not a by-product: it happens only when an organisation actively makes room for it. And finally, how leadership responds to a mistake decides whether anyone will even mention the next one.
None of these principles is Conklin’s invention. Dekker, Hollnagel and Reason are visibly behind them. Conklin’s contribution is the translation: he takes an academically formulated New View position and makes it usable for a shift lead, in language that holds up on the shop floor.
Conklin alongside Dekker
In the book note on the Field Guide I wrote that Dekker supplies the lens and Conklin the tool. The image holds up on rereading. Dekker is an epistemologist: his subject is how we read incidents at all, how local rationality is reconstructed, why “the cause” is a construction. Conklin is less interested in that than in what a manager does differently on Monday morning. He is a practitioner, and the book reads accordingly: many anecdotes, memorable one-liners, little theoretical apparatus. Anyone coming from Dekker and looking for the next, operational step is in the right place.
This closeness to the shop floor is both the strength and the weakness. Conklin is narrative, anecdotal, occasionally repetitive, and empirically thin: much is asserted and propped up by plausibility, little by data. Anyone expecting academic rigour will find the book too loose. The greater danger, though, lies in the success of the idea itself. The five principles are perfect for a poster in the break room, and that is often where they end up: as HOP branding that adopts the language without touching the conditions. An organisation that puts “Blame fixes nothing” on the wall and still goes looking for the operator at the next deviation has read Conklin and not understood him. What he says little about, like much of the HOP literature, is power and incentives: what to do when leadership’s appetite for learning runs aground on a bonus system that rewards something else.
Who this book is for
Essential reading for line managers and safety leads in operational industries who want HOP as a practice, not as a slide deck. Anyone who has read the Field Guide and wonders what follows from it for their own operation will find the sequel here. Do not read the book for the tools, which you can get more compactly elsewhere. Read it for the inversion in the title: the insight that the most valuable investigation is the one you run while nothing has happened yet.
Sources
- Todd Conklin – Pre-Accident Investigations: An Introduction to Organizational Safety, Ashgate 2012 (main source)
- Todd Conklin – The 5 Principles of Human Performance, Pre-Accident Investigation Media 2019
- Sidney Dekker – The Field Guide to Understanding Human Error, 3rd ed., CRC Press 2014
- Erik Hollnagel – Safety-II in Practice: Developing the Resilience Potentials, Routledge 2018