Episode 10

Rituals for moving on from disaster. A discussion about accident investigations.

March 3, 2021



In this podcast, Johan Bergström, Head of Department at the School of Aviation at Lund University, raises and addresses some fundamental questions about accident investigations:

1) What is the purpose of an accident investigation?
2) Why do we really do an accident investigation? What are the needs and who are the audiences?
3) How well do investigations fulfil the needs of those audiences?
4) What novel strategies can we think of in order to renew our approach to accident investigations? Who could we be inspired by?

[00:00:00] Nippin Anand: Welcome to another episode of Embracing Differences with me Nippin Anand. This podcast series is meant to bring you different perspectives and concepts in safety. The idea really is to create space for thinking and reflection, not to reinforce any grand theories or our prior knowledge on a subject. The aim is to learn and grow, not to remain stagnant. And of course, as I keep saying there is no reason for you to believe me or any so-called expert but keep an open mind and be prepared to challenge your beliefs if you truly want to learn more than what you knew yesterday.

[00:00:39] Nippin Anand: Today’s episode is very close to my heart it is often said that if you know your subject well you should have no difficulty in making clear and concise arguments and the proof is in the pudding. My guest tonight an immensely intelligent man gifted with excellent communication skills tackles one of the most difficult questions with such clarity and depth in just about 20 minutes. Johan Bergström from Lund University in Sweden asks a very obvious question that most safety professionals believe have an answer to, but delve deeper and you would find yourself in a difficult position to move beyond the obvious. Johan asks the questions – what is the purpose of an accident investigation? Why do we really do an accident investigation? What are the needs and who are the audiences? How well do investigations actually fulfill the needs of those audiences? What novel strategies could we think of in order to renew accident investigations? Who could we be inspired by? I think you will love this podcast!

[00:01:47]: Maybe we should start, Johan with yourself, a brief introduction about yourself.

[00:01:52] Johan Bergström: Thank you Nippin, for inviting me. My name is Johan Bergström. I work as associate professor or reader in risk and safety at Lund University in southern Sweden and since 2012 I’ve had the role of program director for our Master of Science program in human factors and systems safety. Also with several prominent alumni, I should say have already showed up as guest in this podcast. So that’s me.

[00:02:14] Nippin Anand: Excellent! And what would you like to talk about today, Johan?

[00:02:22] Johan Bergström: I will raise and try to address 5 questions about accident investigations and the way we do them, why we do them, the way we do them and how we could do them differently perhaps.

[00:02:35] Nippin Anand: Let’s get started then. So excited!

[00:02:41] Johan Bergström: That’s what I have prepared for this afternoon or morning or whatever wherever you are in the world. Five questions to give some structure to this talk:

So, my first question will be – Why do we really do Accident Investigation?
My second question will be – How are societal needs of accident investigations addressed by official Accident Investigation bodies?
The third question will be – How do these approaches differ from more conflict with contemporary safety science?
The fourth question to raise will be – What novel strategies could we use to renew and reinvent the ritual of Accident Investigation?
and then fifth and final question would be – What can you ask accident investigators do already now? What are some low hanging fruits for us?

[00:03:32] So, let’s address them one by one. The first question why do we really do Accident Investigation? What are the needs that we are trying to address? and my answer to this question will be rather short and not even my own to answer this question. I would go to what I consider to be one of my go-to pearl papers and that is Sidney Dekker’s, The psychology of Accident Investigation. It’s published in 2015 in the Journal theoretical issues in ergonomic science.

[00:04:02] In this paper, Dekker argues that there are full societal psychological needs which we try to address using Accident Investigation. First one being, an epistemological need of establishing a credible story of what happened.

[00:04:17] Second will be a preventive story of establishing what to do in order to avoid reoccurrence. The third social psychological need is the need for a moral story – a story of the boundaries around what we consider to be a profession, and then the fourth story is an existential story which deals with, explains and creates meaning and helps us to move on from suffering. Those are the four socio psychological needs.

[00:04:47] So not mine, but Dekker’s answer to my first question of why do we really do Accident Investigation? Well, it’s to address these four social psychological needs and I think this is a good answer to think with when it comes to this question.

[00:05:02] So, my second question then, having dealt quickly with that first question, my second question is how are these needs typically addressed in Accident Investigation as conducted today? The epistemological need of establishing what happened here or official accident investigations are typically dedicated to finding one final and not too controversial story. Typically, a chain of events. It could consist of linear at least, deterministic cause and effect relationships which only to a limited extent. If at all, accounts for and embraces the complexity of failure in sociotechnical systems. This story is often dedicated to great forensic detail essentially an exercise of building credibility for the epistemological story of what’s happened. So that’s the first social psychological need addressed by Accident Investigation.

[00:05:58] Next, the preventive story of how to avoid reoccurrence? That’s typically addressed as one of removing or improving system components or actors or adding bureaucratic layers of routine and structure. This tent, if anything to lead to a drift into bureaucratization which comes with unintended consequences which are not the topic of this session but it indeed comes with unintended consequences of its own.

[00:06:22] Third question or third socio psychological need, a moral need, and how do accident investigations then typically address the need for a moral story? They typically do by focusing on professional moral such as airmanship, seamanship, moral judgment based on outcome rather than local rationality. So, people are then judged in relation to often arbitrary regulatory norms. For you, maritime people, an example would be ColRegs – Collision avoidance regulations at sea and we hold you as morally accountable or we make moral judgments based on your adherence to these norms which again are often arbitrary at best.

[00:07:04] The fourth psychological need, the existential need. It’s typically addressed by an attempt by accident investigations to regain the trust in the risky business which we have developed and conducted in the first place.

[00:07:18] So, we know what happened, we promised to have fixed the problem. It is still safe to fly or buy a ticket for a cruise! There is also here, the dimension of we know what happened to your loved ones which is healing in itself and the clear example of how just telling the epistemological story is a way to address the existential need is when we are not able to even come up with a narrative of what happened in the first place.

[00:07:43] So, the loss of Malaysian the airplane 370 is still a mystery and that leads to great existential crisis and suffering and the spending of huge societal resources into finding that plane.

[00:08:06] My third question is – How do these approaches to Accident Investigation which I’ve just described differ from or even conflict with contemporary safety science?

Let’s go through them again one by one. You seem it’s sort of analytical framework here, how I’m using Dekker’s paper. First of all, the epistemological need, again contemporary safety science as we see it is rooted in complexity theory. In complexity theory, one event can embody multiple realities or to paraphrase Healy, an epistemological pluralism to contemporary safety science. It’s not controversial that different actors in an event will experience different local realities and the best we can do then is to describe those without judging which one is more true than the other but rather in terms of what emerges from the interactions and relations between these multiple and sometimes conflicting realities, embracing differences, if you like.

[00:09:05] The preventive need – avoiding re occurrence. Contemporary safety science is keen on understanding and modifying the conditions under which people work and the resources and adaptive capacities that enable and constrain work. So, in order to come up with a preventive story, rather than focusing on the removal or behavioral change at the level of individual components or actors we would rather look at sometimes really simple changes to the work system. Indeed, you could argue that additional procedures in the constitute changes of work and they do but typically by constraining rather than enabling the adaptive capacities of such work and contemporary safety science tries to be more balanced in how we describe the macro and meso level conditions that enable and constrain work and how to manipulate those rather than the individual component or actor.

[00:10:05] Third, so how does the contemporary safety Sciences address the need for moral boundaries? By trying to move from the ‘who’ is responsible towards the ‘what’ is responsible. For contemporary safety science the moral story of accidents should focus on us as the societal actors who want to buy cheap flight tickets or view Penguins up close at a cruise close to the South Pole or us as a political system in which the capitalistic and nationalistic system produces certain macro level behaviors as deregulation, fragmentation of various sectors and actors in a large scale social technical system. System such as railway or energy would be good examples so contemporary safety science would dare to tell the macro level moral stories of us that’s what I’m trying to say there.

[00:10:59] Then for the existential story I would say that contemporary safety science is not particularly good at dealing with the question of meaning-making and moving on from suffering. Some work has been done on rituals for restoration for instance, of rebuilding trust and relationships between involved actors such as first victims, second victims, that’s sharpened stuff and organizations. But this is still a rather unexplored field within our science so a window of opportunity for good research to come perhaps. That concludes my attempt to address my third question of how safety science differs or contrast with how these social psychological needs are addressed in classic Accident Investigation.

[00:11:50] Moving on to my 4th question which is really the main topic of this brief talk – what novel strategies could we think of in order to renew and reinvent the ritual of Accident Investigation? Who could we be inspired by? and I think that we should be inspired by the fine and performing arts, theater, screenplay, or prep literature and I will give you just some examples and 1st an example of epistemological pluralism as embedded into HBO’s miniseries on Chernobyl. It’s really interesting that you note how the director Johan Renck, works the camera in episode one versus episode 5 if you’ve seen it. In episode one he embraces local rationality and very much the feeling of what the hell is going on here. We see soon in face shots of actors no big picture, uncertainty, doubt conflicting views and then in episode five we get zoomed out views, much more certainty and clarity in terms of sequence of events and causal relations. We get the same scene but different perspective and narrating techniques within the same piece of art. It’s really interesting way of incorporating multiple epistemologies and realities into the story and I hope that you will be encouraged to try ways of doing that yourselves.

[00:13:12] The next would be examples from the fine and performing arts or the moral stories and especially for seafarers. I would encourage you to look at how morale and seamanship or perhaps even though not explicitly told like that is portrayed in the operas Peter Grimes by Britten and Wozzeck both produced in the first half of the 20th century. These operas are both experimenting with notions of causality, heroes, villains, outliers, madness, power, relationships, class and they very much involved us, the society, surrounding context and events and perhaps us, who construct others as alien or hostile in Wozzeck, this is the by his captain and eventually the entire community badly treated private soldier in Grimes, misunderstood fisherman. Both portrayed very much in relation to us. I see the need for an opera on the Costa Concordia and it would very much be written with reference to both, Peter Grimes and Wozzeck. Sorry if I get carried away but that’s because I find these approaches to be much more interesting in terms of how they introduce a moral account and in how they allow for complexity, conflict, arbitrariness and shift of perspective in contrast to normal truth tending confidence restoring nature of Accident Investigation as currently and typically done.

[00:14:42] I also think if sort of moving away from fine and performing arts, I think we should get inspired by investigative journalism. While news can indeed be sensational and demanding justice by retribution, investigative journalism can also seek answers in organizational and ideological and political history. One example which we also use in teaching is the short little book, the train crash which stopped Britain written by iron Jack and published in 2001 I think, in which is seemingly simple derailment accident is portrayed in the lights of the entire British political history or modern British political history, essentially. Also, in the wake of the 737 Max disasters I think there are currently several really interesting stories about the history of Boeing and its emergent relations to the US state and the FAA in the harsh competitive environment in which they were falling behind Airbus essentially and these stories can really serve as a good source of inspiration on how to tell stories of disaster, in fact better stories of disaster.

[00:15:51] There are also examples from the academic world of interesting accounts of accidents, Snook’s Friendly Fire, Diane Vaughan’s Challenger launch decision are examples of how accidents not only can be portrayed using academic discourse but also contribute to the development of our theories and we use both these pieces in in our teaching as well.

[00:16:15] I am now dealt with four of my 5 questions for this talk and we’re moving on to the 5th. The 5th one is what can you as accident investigators do already now? What are some low hanging fruits? Without engaging opera librettists and composers, which you should do, by the way, but perhaps in the next step. First, I would say in your analysis, open up the dimension of time, space and locality. With time – why start with the accident and go backwards in time? My colleagues in the Netherlands are currently doing the other way around. They are arguing that much more information is gained by starting from the beginning somewhere where it makes sense and then going forward in the narrative rather than the other way around.

[00:17:03] Don’t work with minutes work with years. Again, investigative journalisms are good at this typically better I would argue, that accident investigators who not always tend to isolate events into their operational environments but open the dimensions of time and space, embrace several realities within the same report or event without judging which one is right and which one is wrong. Embrace those different realities and hire scholars of humanities to join your accident Investigation boards, hire scholars of history, language, arts etc. and revolt!

[00:17:46]: Break formats, refuse to use words such as errors or course. Experiment with other language, see what happens. Some almost get fired. You might not. Experiment and revolt and get inspiration from others. Get inspired by the Danish maritime Accident Investigation board. Oessur has already been on the podcast. The reports are published online in English. Get inspired by them. They really play with some of these notions that I’ve just introduced be inspired by Diane Vaughan’s contribution to the Columbia Accident Investigation report. Read Snoop and Vaughan. Visit the opera, get inspired by art and literature.

[00:18:27] And that’s it, that’s my five questions about the status and potential development of the craft or perhaps now art of Accident Investigation. We’ve gone through why we do accident investigations in the first place, how the societal needs of accident investigations are typically addressed by official investigating bodies, how those approaches differ from or conflict with contemporary safety science? what novel strategies we could use to renew and reinvent the ritual? and what you ask investigators could try to do already now?

[00:19:03] I will wrap this up and end with a quote which I always read to our graduating students that at graduating ceremonies at Lund. It’s by the late complexity scholar Paul Cilliars and he wrote this,

“The claim that our understanding of complex systems cannot be reduced to calculation, means that there will always be some form of creativity involved when dealing with complexity. Creativity should not be understood in terms of flights of fancy or wild abandoned but also in terms of a careful and responsible development of the imagination. Imagining the future will involve risk but the nature of this risk will be a function of the quality of our imagination. It’s important that we start imagining a better future and for that we need better imaginations – reading books, listening to music, appreciating art and film is not a form of entertainment to be indulged in after we’ve done our serious work. These creative activities stimulate the imagination and thereby transform the frameworks we apply when apprehending the world.”

Thank you for listening everyone!

[00:20:08] Nippin Anand: Brilliant! Thank you! This is very informative. You kept repeating the term ‘contemporary safety sciences’ and also in the title, the term you use, ‘ritual’. I mean as somebody with a deep interest in anthropology, I was just itching to ask you what do you mean by ritual? Why do you use the term ritual in this in the title of this podcast?

[00:20:28] Johan Bergström: I mean ritual as a societal practice with a particular aim in this case to help society move on from disaster. To me accident investigations form a ritual and that’s not to be derogatory in any way towards the ritual. But that’s what I see it as. An important ritual, an important ritual that we should take seriously. But it’s social practice that we engage in together or that we give someone the power to conduct, in order to help all of us move on from harm and suffering. So, it’s about explaining, creating meaning out of suffering and to help us deal with suffering, to help us deal with the fact that the systems that we have produced and developed create suffering. In order to cope with that sort of the existential need again is them to introduce a ritual and that ritual we call, ‘Accident Investigation’ and that we could fill with many different activities which is what I’m hinting at.

[00:21:33] Nippin Anand: Yes, I mean growing up in India one of the ways we used to cope with uncertainty was witchcraft. That’s the reason I found it resonates with me because it’s a very classic example of a ritual.

[00:21:44] Johan Bergström: It is and it serves its purpose it dusts things too as it helps us with certain things and without calling Accident Investigation witchcraft ritual and a meaningful one. Absolutely we need good witches to conduct our accident investigations.

[00:22:03] Nippin Anand: Thank you Johan for giving us the opportunity and time. Really appreciate it and learn so much from you.

[00:22:09] Johan Bergström: Very inspiring to talk to you guys.

[00:22:11] Nippin Anand Thanks everyone for joining and we will see you again in the next episode.

What did you think? I thought the conflict between fact finding, preventing, finding who or what to blame and bringing an end to sufferings in the wake of an accident was beautifully articulated by Johan. He does not only problematize it so well but he also provides some low hanging fruits in terms of what we can do as action investigation bodies and safety critical organizations to move forward.

[00:22:44]: Thank you for taking the time to listen to this podcast. I hope the time you spent was worthwhile. If you think the podcast has made you think, slow down and reflect, I have achieved my purpose. Please share it with others in your community so that messages reaches far and wide.

I spend a lot of time thinking, researching and producing meaningful content. If there is a specific top you’d wish to know more about, please let me know. If I can, I will make every attempt to create something meaningful and valuable to you. If you have a topic that you would like to discuss with me, please feel free to be in touch. Particularly if there is something you don’t agree with. Disagreements are always fun.

I wish to also remind you that all my podcasts, related reference material and transcript for each podcast is available on my website, www.novellus.solutions. You can also get in touch with me on the same website or through LinkedIn, Twitter or my personal website, www.nippinanand.com

Plays mentioned in this podcast to gather some inspiration from 🙂

Berg: Wozzeck (HQ) – 1970 film version [English Subtitles] – YouTube

Peter Grimes 1969 – YouTube

Reference paper from Sidney Dekker (a must read)