We Know, We Say Nothing

We Know, and We Say Nothing

THD-English| THD-Learning, Incidents & Just Culture
June 11, 2026•16 min read

(This first appeared in French and Italian. This is an AI-translated version, human checked.)

On 14 May 2026, six people died in an underwater cave in the Maldives, at sixty metres. Five Italians — a marine biology professor from the University of Genoa, her daughter, two young researchers, their instructor — and a Maldivian rescue diver.

What ought to give us pause, as divers, is not what is spectacular about this accident. It is what is ordinary about it. These six had every marker of the serious diver: training, certification, experience, years of practice. For most of us, they were more than that: a university professor, researchers, an instructor, an elite military unit.

The mechanisms by which serious divers die in perfectly predictable conditions are known, documented, and studied over fifty years. Our community's problem is not that it ignores them. It is that it prefers not to confront them. Out of respect for today's dead, and to prevent others tomorrow, we have to be willing to recall what these accidents have taught us at the cost of so many lives. That is what this article attempts.

In diving, the absence of an accident establishes nothing about the absence of risk. The diver who descends to fifty metres breathing compressed air and surfaces unharmed holds a precedent that seems to prove him right; at the next opportunity he will draw the intuitive inference: it was doable, therefore it was acceptable. The inference is false. Surviving a dive says nothing about its safety; it says only that one is alive to talk about it. This error of reasoning — confusing the quality of a decision with the quality of its outcome — has a name in decision science: it is called outcome bias.¹ It is documented across every high-risk field, from medicine to the trading floor.

Layered over this bias is another, more insidious, drawn out by the sociologist Diane Vaughan in 1996 from the decision that led to the Challenger disaster: the normalisation of deviance.² Every departure from a norm that carries no immediate consequence ends up shifting the norm itself. Margins are not crossed in a single step; they erode slowly, in practice invisible to those eroding them. Gareth Lock,³ who has carried this framework over to diving, establishes that the majority of serious accidents arise not from a single discrete error but from a slow accumulation spread over months or years.

The most commonplace example of drift is that of depth. A diver certified to thirty metres descends, dive after dive, a little deeper — thirty-three, then thirty-five, then forty. Nothing has ever happened; everything seems to confirm his intuition that those limits were needlessly cautious. The day the incident arrives — a buddy who strays, a regulator that jams, a little current — he finds himself in an environment he has never formally been trained to manage, and into which he had not realised he had ventured.

This double mechanism spares no one. Beginners are protected from it only for a while, for want of time to drift. Recreational divers imagine themselves sheltered by their comfortable margins — without seeing that it is precisely those margins that drift nibbles away. As for technical divers, they have more preparation, equipment designed to absorb failures, and training that has normally made them alert to these very mechanisms — outcome bias, normalisation, human factors. It is not enough. When they cross their own margins, they too fall victim to accidents — all the more tragic in that the very nature of their dives often makes those accidents fatal. When the community learns of these deaths, it does not judge. It recognises. And it carries on.

To the cognitive mechanism just described, depth adds another, purely physiological. Beyond a certain partial pressure of the narcotic gases breathed in — foremost among them nitrogen — the human brain stops functioning as it does at the surface. And it does not know it.

Nitrogen narcosis has been documented since the 1930s.⁴ Nitrogen, which makes up 78% of the compressed air a diver breathes from a cylinder, becomes narcotic as its partial pressure rises with depth. The medical literature⁵ converges on two points that appear to contradict one another: on the one hand, the first measurable effects appear as early as thirty metres; on the other, those effects go largely unnoticed by the diver himself. Cognitive studies — working-memory tasks, decision-making, reaction times — record systematic deviations that the subjects do not perceive in themselves. At thirty metres breathing air, everyone is impaired. What distinguishes divers from one another is not the impairment; it is the awareness one has of it.

This distinction, seldom taught in training, is in fact decisive. It has one precise practical consequence: narcosis does not betray itself in what is felt. It betrays itself in what is done — in the decisions, the omissions, the approximations one would never allow oneself at the surface. Far from announcing itself as a hallucinatory drunkenness, it degrades reflection first. The actions learned through repetition still hold — partially, enough to keep up appearances — and the diver leans on them to compensate, without perceiving that his capacity to reason about a novel situation has already deteriorated.

A stubborn belief, especially alive among experienced divers, would have it that one grows accustomed to narcosis with practice. Controlled studies refute it. The most cited⁶ subjected eleven experienced divers to repeated compressions equivalent to 55 metres over five consecutive days. The performance measured objectively — reaction times on cognitive tasks — showed no improvement: narcosis went on degrading their capacities to the same degree. Their subjective assessment of the narcosis, by contrast, fell away from the third day. In other words: with repetition, divers do not adapt to narcosis. They adapt to their perception of it. What experience teaches them is not to function better under narcosis; it is to feel it less. The danger increases — it becomes greatest, in fact, exactly where the diver believes he has it best in hand.

Among the documented symptoms of narcosis is the narrowing of the field of attention — tunnel vision. Attention closes on a single object or idea and loses access to the rest. The same narrowing occurs under panic too, independently of any narcosis.

At sixty metres, when an incident occurs, a diver under narcosis does not undergo one tunnel vision — he undergoes two that reinforce each other. Narcosis has already narrowed his attention; panic narrows it further. Cognition has no exits left. It is through this double narrowing that deep diving on air changes in nature. It is no longer a dive riskier than another — it is qualitatively other. At sixty metres breathing compressed air, the gap between an unremarkable dive and a fatal one no longer rests on the diver's competence; it rests on the absence of the unexpected.

To these individual mechanisms one last is added, which is no longer individual at all: it operates in the relationship between divers. Harder to say — and that is precisely why it must be said. In any dive where some lead and others follow, responsibility shifts imperceptibly. Those who follow delegate their judgement, sometimes without realising it. Those who lead carry — without always realising it — the safety of the group. The effect is not symmetrical: the followers' margin of error is not added to the leader's, it is exported onto him. If the leader drifts, everyone drifts with him. And no one knows it.

This is what diving jargon calls a trust-me dive. The term is misleading, because it suggests a conscious ceremony in which one would formally hand oneself over to someone. The practice is the opposite: the trust-me dive is tacit. No one asks, no one authorises, no one signs anything. Someone is there, more qualified, more experienced, better known. And he does not say no. The others take that silence for a green light and enter the water.

This dynamic has a particularly cruel consequence: those who lead die too. It is a constant feature of group accidents in diving. The leader who drifts survives no better than those he draws along — he dies with them, sometimes first. This is not a detail. It is the proof that he had not seen coming the danger he was making others run. No one, in these accidents, acts with bad intent. But the accumulation of mutual trust carries them all away.

One might think this accident exceptional. It is not. It belongs to a known statistical family. The Divers Alert Network has been compiling diving deaths since the 1980s. The annual reports are public, accessible, and instructive to read — not for raw figures, which warrant suspicion (the databases are heterogeneous, the denominators uncertain), but for their regularities. Year after year, the same factors recur. On the recreational side, deaths are rarely caused by a single spectacular event. They almost always result from a cascade: a predisposing factor — often an underlying health condition — a trigger — gas management, current, equipment, panic — and an emergency ascent that ends in drowning. Drowning appears on more than 80% of death certificates, but it is rarely the cause. It is the end point.

On the cave side, the pattern is steadier still. Sheck Exley, one of the pioneers of modern cave diving, published in 1979 a text that remains foundational: Basic Cave Diving: A Blueprint for Survival. From a systematic analysis of cave deaths in the United States, he identified a handful of recurring factors — no training, no continuous guideline, failing gas management, excessive depth, insufficient light. Peer-reviewed statistical analyses published thirty years later validated his framework against data accumulated between 1969 and 2015. Cave deaths almost always combine several of these factors. Nearly all of them would have been avoided had even one of them been corrected.

There is one fact to add, which should give us still greater pause. The dead are not mostly beginners. Many are experienced, certified divers with hundreds of dives behind them. It is, in fact, typical. Beginners have not had time to accumulate the false confidence that would lead them there. The deaths befall divers who looked competent — who believed themselves competent, and whom the community believed competent. Which is exactly what every reader of this article should hear as a personal warning.

If drift is inescapable and silent, what can be set against it? To each of the three mechanisms named above — narcosis, drift, uncritical following — there answers a discipline of its own.

For the first — nitrogen narcosis — the answer is to recognise that air has a depth beyond which it is no longer a defensible gas. This limit is not an absolute number; it depends on what the dive demands of the brain. The more demanding the context — cave, wreck, current, low visibility — the sooner it is reached. To believe one is holding up at sixty metres because one held up at fifty is to fall back into exactly the outcome bias described above. Air has a zone of relevance; beyond it, another mix is needed.

For the second — silent drift — the remedy is to make that silence audible. A dive is planned point by point before entering the water; it is gone back over phase by phase after coming out. The logbook serves not to record what happened, but what held only by luck. An outside eye — a partner who dives in the same register, but who has not lived the dive from the inside — sees what drift blinds. Without these fixed points, drift operates as predicted: invisible until the incident.

For the third — tacit following — the silent delegation must be broken. This runs through two technical gestures. The first is to build the team around the precise competences this dive demands, and not around a partner's general standing: an instructor is not a universal guarantor, and experience of one diving environment does not transfer to another, however similar. The second is the old rule of cave divers: a raised thumb ends the dive, called by any member of the team, owing no justification to anyone. And that gesture should call forth from the others not disappointment, but gratitude: the gratitude of having, on one's team, someone who knew how to see and to stop before all of them came to be in danger.

But these rules live only if the culture around them upholds them. One must dare to speak — to voice the doubt, the discomfort, the fear when it comes. One must dare to say no — without apologising for it, without having to account for some failure of nerve. One must, above all, listen to those who speak — without judging them, without dismissing them as merely overcautious, without reducing their words to passing fatigue or a bad day. Silence does not count as consent; a word counts as a veto; but a word carries weight only if someone is listening.

None of these disciplines makes diving safe. All of them restore what the mechanisms had slowly erased: the diver's lucidity about what he is actually doing.

Let us return to the Maldives. Of the six dead of 14 May 2026, five made that dive by choice. The sixth died during the dive to recover the bodies, and his case is of another order. Let us stay with the five.

And let us acknowledge this first: nothing extraordinary happened that day. No unusual equipment failure. No gross blunder that a reasonably attentive eye could have seen coming. A dive like the ones practised every day, everywhere in the world, in similar conditions and with the same choices. That day, it fell upon them. Tomorrow, it will fall upon others.

It is precisely in this that the accident is not one. The word suggests the unforeseen, the exception, the external event. The reality is altogether different: what happened is the normal output of a practice that, statistically, must produce this result. The surprise is not that it took place; the surprise is that it does not take place more often. International accident research has been measuring this for fifty years.

I expect that writing these lines will earn me the charge of indecency. To criticise the dead, who cannot defend themselves, would be easy, petty, foreign to the diving community. I believe the opposite. The indecency is to look away. The indecency is to let these five people die a second time in our collective silence — the silence that says, 'let us not speak of mistakes, it would be disrespectful,' and that thereby guarantees the statistic its stability. So long as we go on recounting limit-pushing dives as feats and passing over deaths as bad luck, the next deaths will resemble the last. And we know it.

These five people did not die of bad luck. They died because an entire community — ours — tolerates, normalises, and sometimes admires the practices that led them there. The only tribute that counts is not the respectful silence of the weeks after the accident. It is to look squarely at the machinery that carried them off, and to name it to those preparing to repeat the same actions.

Nothing in the foregoing has been said against diving. Quite the opposite. Diving well practised — the diving that can see its own mechanisms, that can say no when it must, that debriefs with honesty what it has just lived through — remains an encounter with a world and with sensations that nothing else affords.

What these lines have tried to name are the silent habits that, little by little, turn it into something else. To see them and to say them is what makes it possible to go on practising it.

Notes

  1. Concept formalised by Baron, J. & Hershey, J. C., "Outcome bias in decision evaluation," Journal of Personality and Social Psychology 54 (1988), p. 569-579. PDF available: https://www.sas.upenn.edu/~baron/papers/outcomebias.pdf

  2. Vaughan, D., The Challenger Launch Decision: Risky Technology, Culture, and Deviance at NASA, University of Chicago Press, 1996. Publisher page: https://press.uchicago.edu/ucp/books/book/chicago/C/bo22781921.html

  3. Lock, G., Under Pressure: Diving Deeper with Human Factors (2019); see also The Human Diver (https://www.thehumandiver.com) and the documentary If Only… (https://www.thehumandiver.com/ifonly).

  4. First systematic experimental demonstration: Behnke, A. R., Thomson, R. M. & Motley, E. P., "The psychologic effects from breathing air at 4 atmospheres pressure," American Journal of Physiology 112(3), 1935, p. 554-558. DOI: https://doi.org/10.1152/ajplegacy.1935.112.3.554

  5. For an overview, see Nitrogen Narcosis In Diving on StatPearls (NCBI/NIH): https://www.ncbi.nlm.nih.gov/books/NBK470304/; and Clark, J. E., Moving in extreme environments: inert gas narcosis and underwater activities, Extreme Physiology & Medicine 3 (2014), which establishes that narcosis directly contributes to up to 6% of diving deaths — free full text: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4337274/

  6. Hamilton, K., Laliberté, M. F. & Fowler, B., "Dissociation of the behavioral and subjective components of nitrogen narcosis and diver adaptation," Undersea & Hyperbaric Medicine 22(1), March 1995, p. 41-49 (PMID 7742709): https://pubmed.ncbi.nlm.nih.gov/7742709/

Bibliography

Decision biases, normalisation of deviance, human factors

Physiology of narcosis

  • Behnke, A. R., Thomson, R. M. & Motley, E. P., "The psychologic effects from breathing air at 4 atmospheres pressure," American Journal of Physiology 112(3), 1935, p. 554-558. First systematic experimental demonstration of the narcotic effects of compressed air at depth. DOI: https://doi.org/10.1152/ajplegacy.1935.112.3.554

  • Kirkland, P. J., Mathew, D., Modi, P. & Cooper, J. S., Nitrogen Narcosis In Diving, StatPearls (NCBI/NIH), last updated 2023: https://www.ncbi.nlm.nih.gov/books/NBK470304/

  • Clark, J. E., "Moving in extreme environments: inert gas narcosis and underwater activities," Extreme Physiology & Medicine 3 (2014). Free full text via PMC: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4337274/

  • Hamilton, K., Laliberté, M. F. & Fowler, B., "Dissociation of the behavioral and subjective components of nitrogen narcosis and diver adaptation," Undersea & Hyperbaric Medicine 22(1), March 1995, p. 41-49. PMID 7742709: https://pubmed.ncbi.nlm.nih.gov/7742709/

Diving accident research

David Charbonnier is a French cave and technical diver based in Rome. Full Cave certified and a CCR diver, he learned to dive in the 1990s and is still learning. For him, diving is one of the last forms of true human exploration — an encounter with a world and with sensations that nothing else affords. "We Know, and We Say Nothing", his first essay on diving, was prompted by the Maldives cave fatalities of May 2026; five of the six victims were Italian.

David Charbonnier

David Charbonnier is a French cave and technical diver based in Rome. Full Cave certified and a CCR diver, he learned to dive in the 1990s and is still learning. For him, diving is one of the last forms of true human exploration — an encounter with a world and with sensations that nothing else affords. "We Know, and We Say Nothing", his first essay on diving, was prompted by the Maldives cave fatalities of May 2026; five of the six victims were Italian.

Back to Blog

Contact Menu

© 2026 The Human Diver