Learning from accidents: the power of narratives

A narrative structure

Andrew Hopkins’ accounts of disasters have proved invaluable to explore the multifaceted aspect of such events from an organisational perspective (Hopkins, 1999, 2000, 2005, 2008, 2012). His writing style is often portrayed as one of great storytelling. James Reason wrote “Andrew Hopkins is a consummate storyteller as well as being an internationally known expert on the breakdown of hazardous socio-technical systems. I believe that only stories such as those told here can capture the subtle influences of organisational culture and embrace the complex interactions between causes and conditions” (Reason, 2005, vi).

Hopkins writes, commenting on his first analysis of a major event “the book was an examination of these failures coupled with suggestions for correcting them. Most of my subsequent books have followed this model.” (Hopkins, 2016, 34). When looking closer into his stories, one finds a narrative structure based on sequences associating:

o  description (1),

o  assumption (2),

o  explanation (3),

o  comparison (4),

o  recommendation (5) and,

o   counterfactual reasoning (6) 

Let’s illustrate it with an important theme in safety, auditing. The example used here is based on the Longford book (Hopkins, 2000).

Auditing and narrative structure

First, the description. Hopkins writes, “it is clear that there was plenty of bad news around, plenty of news which, had it reached the highest levels of the company and had been acted on, would have averted the incident. But the Exxon audit missed it all” (Hopkins, 2000, 93).

Second, Hopkins identifies an assumption of the organisation which consists in believing in the value of a ranking of sites in terms of safety performance through score cards. “Six months prior to the explosions, Esso’s health and safety management system (called OIMS – Operational Integrity Management System) was audited by a team from Esso’s corporate owner, Exxon (…) Esso’s managing director reported to the inquiry that the audit had shown that most elements of the safety management system were functioning at level three or better” (Hopkins, 2000, 81).

Third, an explanation is provided as to why this assumption is not valid. “It is worth pointing out that an audit whose purpose is to identify hazards which have been missed does not lend itself to this score card approach” (Hopkins, 2000, 86), and as a consequence, “one of the central conclusions of most disaster inquiries is that the auditing of safety management systems was defective (…) ” (Hopkins, 2000, 54).

Fourth, a series of comparisons, with other disasters’ accounts, good practices or theoretical properties follows.

·        From disasters’ accounts, Hopkins indicates that “following the fire on the Piper Alpha oil platform in the North Sea in 1987 in which 167 men died, the official inquiry found numerous defects in the safety management system which had not been picked up in company auditing” (1999, 29).

·        On the side of good practices, he mentions that “BHP coal had clearly learnt the lesson of its previous auditing failure. It had understood that the hallmark of a good audit is that it must be thorough enough to uncover the bad news about safety and convey it upwards to the top of the corporation.” (1999, 35).

·        As an insight from theory, he refers to high-reliability organisations. “The strategy which HROs adopt is collective mindfulness. The essence of this idea is that no system can guarantee safety once and for all. Rather, it is necessary for the organisation to cultivate a state of continuous mindfulness of the possibility of disaster (Hopkins, 2000, 140) but in the case of Longford “safety auditing, an ideal opportunity to focus on the possibility of failure, was turned into an opportunity to celebrate success” (Hopkins, 2000, 142).

Fifth, a recommendation is developed, here, the following one. “A rigorous audit needs to examine the hazard identification strategy and make some effort to seek out hazards which may have been missed, so as to be able to make a judgement about how effectively hazard identification and control is being carried out” (Hopkins, 2000, 86)

Finally, a counterfactual reasoning concludes with a plausible alternative to the disaster. “Had it reached the highest levels of the company and had been acted on, it could have averted the incident” (Hopkins, 2000, 93)

The power of the narrative structure

Why is this narrative structure successful? One answer is that it engages readers into a learning process, from problems to solutions. 

First, one finds a point of departure, a description (1) of what happened. In the example, it is a problem of auditing. It contrasts on the one hand the presence of many bad news available when interviewing people retrospectively, with, on the other hand, positive auditing results.

Second, an assumption (2) of the organisation and associated with auditing is then introduced. This assumption considers that a good principle of auditing is to rank sites with score card (from 1 to 5) based on predefined level of compliance to a defined standard. This is a narrative step increasing the likelihood of echoing practitioners’ experience, beyond the specific organisation involved in a particular case.

Consulting companies and corporations in many industries have indeed devised such practices to ensure the monitoring of safety critical operations. Different levels of achievement can be expected, and are audited by auditors, often through minimum time spent in the field, checking paperwork instead. This is a warning to any readers who feel the same, or to other who are content with their current auditing strategies, despite their flaws.   

Third, description (1) and assumption (2) are followed by an explanation (3). Auditing without a field strategy combined with an understanding of the technological risks involved in a particular process plant becomes a shaky exercise as no link is established between real practices, preventive measures and hazardous scenarios. First, auditors relying on paperwork miss the operational realities of daily context, second, without sufficient knowledge about the hazardous processes and their designed preventive barriers, they cannot assess a situation. 

Fourth, comparing (4) this situation with events in previous investigations of disasters (Piper Alpha), known good safety practices in other companies (in this case, BHP) and research findings on safety (collective mindfulness) show how the organisation can be seen as having failed to learn, first, from the findings of previous accidents, and failed second, to achieve a degree of success which is achieved elsewhere, in other companies. It emphasises the idea that disasters do not happen very often precisely because other companies do far better than this. It also is reassuring to know that a problem can be fixed.  

Fifth, recommendation (5) are then built up on the previous sequence of description (1), assumption (2), explanation (3) and comparison (4). In the case of auditing and current limitations, the proposition is to increase time spent in the field by empowered auditors in relation to targeted observations connected to major hazard prevention.

Sixth, through a counterfactual reasoning (6), the reader is now ready to consider highly plausible another story, a story in which the prevention of such an event was possible. This counterfactual reasoning feeds back to the description of the case, a description which could have been different…“had things been done differently”.

Visualising the narrative structure

Hopkins explains that his mathematical background inspired him to build some form of logical arguments (Hopkins, 2016). Yet, more than a deductive kind of logic, his books are plausible accounts of disasters based on a persuasive narrative structure. rather than logic. The following figure tries to visually capture the quality of this structure by connecting the sequences together through what could also be described as a learning loop…from description to  counterfactual reasoning through assumption, explanation, comparison and recommendation (figure 1).

Figure 1. Learning narrative structure

Hopkins’ example shows the importance of the way narratives are built for learning. There is more to explore from Hopkins’ accounts of disasters, but this way of building arguments retrospectively and generating rich narratives is a core aspect of his invaluable contribution to the field.  

 Full article (with more topics covered) available here:

https://www.academia.edu/40327081/Storytelling_or_theory_building_Hopkins_sociology_of_safety               

References

Hopkins, A. 1999. Managing Major Hazards: the Lessons of the Moura Mine Disaster, Sydney: Allen & Unwin

Hopkins, A. 2000. Lessons from Longford: The ESSO Gas Plant Explosion, CCH, Sydney.

Hopkins, A. 2008. Failure to learn: the BP Texas City refinery disaster. Sydney, NSW: CCH.

Hopkins, A. 2012. Disastrous Decisions: The Human and Organisational Causes of the Gulf of Mexico Blowout. CCH.

Hopkins, A. 2016. Quiet Outrage. The way of a sociologist. Sidney, CCH Press.

Reason, J. 2005. Preface, in Hopkins, A. Safety, culture and risk. The organizational causes of accidents. CCH. Sydney: CCH Australia.