Invariably the first question we ask is the wrong one. Here's why
Brian Bugge died in May 2018 off Hawaii, on a training dive, with his oxygen cylinder turned off and his rebreather still in surface mode. Linnea Mills died in November 2020 in Lake McDonald, Montana. She was diving a drysuit for the first time, with the inflator hose unconnected and forty-four pounds of lead that she could not release. A twelve-year-old girl died in Texas in August 2025, on the first day of her open water course. She had almost no visibility, no depth gauge, and an instructor who appears to have been awake for more than twenty-four hours. Five divers died in the Maldives on a dive well beyond recreational depth limits, inside a cave system. As is the norm, within hours the community had rallied around the same question, whose fault was it?
Unfortunately, that question is an expected one because the legal system is framed around answering it, and predictably it is the one that fills the comment threads before the family has been formally told anything about the conditions and context.
We’d think such a question would help us make a difference going forward by holding someone ‘accountable’. Unfortunately, the research suggests something very different happens.
The line between acceptable and unacceptable is inconsistently drawn
Most Just Culture frameworks in current use ask where the line falls between acceptable and unacceptable behaviour and then tries to determine which side someone's actions landed on. Two researchers, Cromie and Bott, tested whether that line stays still based on how information is presented, who reads it, and how it perceived - it does not. Judgements of blame, and the appropriate level of perceived discipline, shifted according to the order in which the information was presented to the reviewers of the event. Those who understand how we make sense of the world will recognise that the same set of facts, read in a different sequence, produce a different conclusion. Many don’t recognise this bias and the impact is has on their decision-making.
The worrying thing is that if the order of the information changes the finding, then that means the finding describes the reader as much as it describes the person being judged. If you’ve sat or sit on a standards, quality management, or safety investigation panel, you should consider this as something worth paying attention to going forward.
Bitar and colleagues, working inside an international energy company, rebuilt their Just Culture process so that the questions about the system were asked first. The questions about individual choice came last, and only where the earlier ones had failed to explain the behaviour. Across 353 reviewed cases, around ninety per cent were resolved at the system level rather than the individual one. The organisation hadn’t lowered its standards; it had changed the order in which it asked its questions. This research paper is the basis of the Just Culture in Practice Guide that is available for free inside the ‘Curious Community’ resource here. https://www.thehumandiver.com/resources
Retributive and restorative responses are doing different jobs
There are many forms of justice, but the two we are most interested in when it comes to learning are retributive and restorative.
A retributive response asks which rule was broken, who broke it, how serious the breach was, and what consequence is proportionate.
A restorative response asks who was hurt, what they need in order to recover, and whose obligation it is to meet that need.
These are not different (strict and lenient) versions of the same process – instead they point to different outcomes and different learning opportunities. ‘Retributive’ is focused on ‘settling an account’ and ‘Restorative’ looks to repair harm so that learning becomes possible by ‘telling an account’.
Justice after harm can be thought of existing in three parts.
Substantive - question of what has to be put right.
Procedural - question of how decisions get made and by whom.
Restorative - question of how people are treated while that happens.
While an operation can get the first approximately right and still fail badly on the other two. As a consequence, those affected will experience the whole event as unjust.
I was privileged to have a private screening of “How to kill a mermaid”, the documentary about Linnea Mills, and it shows how difficult the restorative view is to hold. The investigators in the film are experienced and technically very good, and they see the system clearly. One of them, who is no friend to the instructor, says outright that “she was a product of the system that trained and certified her.” Another asks, “who taught the instructor, and who taught that person”, on up the chain.
Both then argue for a prosecution. Even though both can see the system perfectly well, the pull towards a single culpable individual overwhelms their system’s perspective.
Brian Bugge's death shows the other weakness of the retributive route. He had entered the water with his rebreather in surface mode two weeks earlier. His oxygen was on that day, the fail-safe held (the unit turned itself on), and the dive was uneventful, so nothing was picked up and no conversation followed. The same slip in similar conditions produced a routine dive on one occasion and a fatality on the other. The configuration that killed Linnea Mills would likewise have been identical in a version of that dive where she was recovered in time. I described the Maldives tragedy as ‘normal’ in structure but abnormal in scale. Those same behaviours are present on many dives, but they don’t end up with a mass casualty event.
In each case, only the outcome differed, and it is the outcome that triggers blame. The research shows that a response fixed on the outcome cannot see the conditions or context that produced it.
The harm reaches further than the person closest to the event
This isn’t just a diving ‘thing’. Researchers surveyed 147 practitioners across the outdoor and adventure sector about 171 serious incidents, 73 of which were fatal. More than half reported hypervigilance about risk on returning to work. Anger, guilt and frustration came next, followed by depressed mood, and around one in five had seriously considered leaving the sector.
Where the harm turned up is the more significant finding for diving. It was not confined to those at the scene. Coordinators, managers and directors who were nowhere near the incident reported their own patterns of distress. Nearly two thirds of all respondents coped with it on their own. More than nine in ten identified counselling, peer support and a prompt debrief as things that would have helped, and most of them received none of those.
Diving carries the points made by these researchers and has very little infrastructure to catch it. Dr Laura Walton provides some resources, and DAN has recently released a Psychological First Aid course for those involved in adverse events.
Christopher Richardson recently wrote a long piece on moral injury, and I added something about ‘second victims’ (albeit that term causes some angst as the word ‘victim’ can remove agency from those involved). He describes something many in the industry have struggled with.
The people harmed around Linnea Mills included the instructor, who did not intend any of it and will carry it for the rest of her life. They also included the diver who reached her, did everything a rescuer could, and was then telephoned by the dive centre and told the death was his doing. A fourteen-year-old was asked to help recover a body, and the assistant was injured during the recovery and blamed as well as hurt. The same spread is visible in the faces of the dive team throughout the Bugge documentary. A retributive process cannot hold these people as harmed, because its logic requires them to be sorted into the culpable and the aggrieved.
The response to a death can add its own harm
Researchers in the healthcare space interviewed forty-two people with experience of what happens after a serious incident in their domain. They identified six ways in which the organisational response created a second layer of injury on top of the first. People were left feeling powerless, inconsequential, manipulated, abandoned, de-humanised and disoriented. They call this compounded harm, and it is produced by the process rather than by the event.
The narrative told in the documentary to celebrate Linnea’s life maps onto that list closely. There were eight days of silence from the dive centre after the death. The coroner was told that Linnea had panicked, which the physical evidence does not support, and for a period the death was recorded as accidental. The dive computer holding the best account of her descent was removed and not available when needed. Years of litigation then ended in a settlement that produced no adjudicated account of what happened.
What the family were looking for was an understanding of what happened to their daughter and an acknowledgement that this understanding was true. The film closes with her brother learning to dive and descending into Lake McDonald three years after Linnea could not complete her dive - surrounded by the people who loved her. That repairing process sits outside every formal channel available to them. In recreational diving there is no organisation whose job it is to offer one. It exists in healthcare and US Wildland Fire Fighting Service, and other domains too.
Why families reach for punishment
A colleague of mine from Lund University asked what relatives are actually looking for after a violent loss as his MSc thesis. Mads began from a hypothesis that meaning comes through three things working together: justice, learning and punishment. He then interviewed people who had lost someone in an accident and ran focus groups with people who had not. His hypothesis held up in part, though how much of each theme any individual needs turned out to be highly personal.
Why I bring it up is because the relationship he found is the part I believe that applies directly to diving.
Learning has to happen if anything is to be prevented.
Justice can be reached through either learning or punishment.
Punishment becomes more relevant the less the intention there is to learn.
The demand for a prosecution is therefore not simply an appetite for revenge, and the industry (and community) keeps making this mistake. Prosecution is what remains when the other routes to justice have been closed off. Linnea's family were offered no inquiry, no account and no credible sign that the system intended to change. The dive centre went silent, the record was wrong, the best evidence went missing, and the settlement produced nothing anyone was obliged to read. Punishment was the only remaining path to justice, so that is the path they took. While some changes did happen, they fundamentally didn’t change the systemic conditions that led to this event. The same process applied to the Texas case, where nothing moved until public pressure forced it to.
If you do not want families and communities reaching for punishment, the answer is not to ask them to be more reasonable. The answer is to make learning visible enough that it becomes the more attractive route. This can be scary for an organisation, because it could infer that issues were known about in the past. Margaret Heffernan described this as wilful blindness.
Who is going to do this work in diving? Be part of the change.
Healthcare has a duty of candour and a regulator. Aviation has occurrence reporting protected in law in some countries. Recreational diving has neither, plus there is no mandatory reporting and no confidential route through which the industry might learn from a death before the next one happens.
In the Texas case the sheriff's office took brief statements and closed its investigation in under ninety minutes, and only public pressure weeks later reopened it. What is left is civil litigation and, occasionally, criminal investigation. Both are retributive by design and both arrive years too late.
None of that prevents an instructor, a centre owner or an instructor trainer from working differently. Or a diver having a different perspective on social media:
Ask the system questions first, and in that order, before anyone reaches a conclusion about a person.
Start from how it made sense to those involved to do what they did and what conditions they were operating under.
Recognise that the line you are about to draw will move depending on what you read first and how severe the outcome was.
Treat everyone touched by the event as harmed, including the person closest to it.
Ask them what they need rather than informing them what will happen to them.
Accountability that looks forward, at what has to change, tends to produce learning. Accountability that looks backwards tends to produce silence and a thinner account next time.
The next fatality or serious injury will produce the same question within hours of becoming public. Which question are you personally going to ask?
References
Bitar, F.K., Chadwick-Jones, D., Nazaruk, M. and Boodhai, C. (2018). From individual behaviour to system weaknesses: the re-design of the Just Culture process in an international energy company. Journal of Loss Prevention in the Process Industries, 55, 267–282.
Cromie, S. and Bott, F. (2016). Just culture's "line in the sand" is a shifting one; an empirical investigation of culpability determination. Safety Science, 86, 258–272. doi:10.1016/j.ssci.2016.03.012
Dallat, C., Mitten, D., Slay, S., Mitchell, V. and Ajango, D. (2025). The silence is deafening: exploring the impacts of serious incidents on practitioners across the outdoor and adventure programs work system. Human Factors and Ergonomics in Manufacturing & Service Industries, 35(4), e70017. doi:10.1002/hfm.70017
Dekker, S. (2025). Restorative Just Culture: From Disciplinary Action to Meaningful Accountability (4th ed.). CRC Press.
Dekker, S.W.A. and Breakey, H. (2016). 'Just culture': improving safety by achieving substantive, procedural and restorative justice. Safety Science, 85, 187–193.
Møller, M.M. (2023). Why Do We Ask Why? Finding Meaning After a Violent Loss. MSc thesis, Division of Risk Management and Societal Safety, Lund University. https://lup.lub.lu.se/student-papers/record/9106422
Ramsey, L., Hughes, J., Hazeldine, D., Seddon, S., Gould, M., Wailling, J., Murray, J., McHugh, S., Simms-Ellis, R., Halligan, D., Ludwin, K. and O'Hara, J.K. (2024). Humanising processes after harm part 2: compounded harm experienced by patients and their families after safety incidents. Frontiers in Health Services, 4, 1473296. doi:10.3389/frhs.2024.1473296
Related
If Only… documentary and field guide: https://www.thehumandiver.com/ifonly
Linnea Mills through a human factors and systems lens: https://www.thehumandiver.com/post/linnea-mills-death-hf-systems-lens
The death of a child in diver training. There are no 'silver bullet' solutions: https://www.thehumandiver.com/post/learning-from-tragedy-dh
