Drop Your Tools: What organisations can learn from studies on sense-making on the fireground.

“If Socrates had been foreman on the Mann Gulch fire, he and his crew would have been cremated while they were sitting there considering it.”

― Norman Maclean

On 5 August 1949, a highly select group of sixteen smokejumpers parachuted into Mann Gulch, a deep, narrow valley with steep slopes in the Gates of the Mountains Wilderness in Montana, to fight what looked like a routine forest fire burning quietly on the south side of the gulch. An hour later, thirteen of the smokejumpers were dead.

Let’s start with some of the events leading up to jump. Wind conditions were turbulent and unpredictable, it was so bad that one of the smokejumpers got so airsick he did not jump. Because of the wind conditions the smokejumpers were dropped from 2000ft rather than the usual 1200ft. The parachute carrying their radio did not deploy and the radio broke from the impact of landing.

But they landed safely, and they commenced their approach towards the fire from within the gulch. The team was led by Wagner Dodge, a veteran smokejumper, known to be exceptionally competent, physically tough and quiet. That said, for most of 1949, Dodge was involved with base maintenance, and he had not trained with the mostly rookie team he was leading on the day. They knew of him by name, and he did not have time to establish rapport with the team on route.

As they were approaching the fire, there was a sudden blowup with the wind in the river valley forming a funnel drawing in the fire. It was later estimated that the firestorm was burning at 2000 degrees, 300 feet deep, 200 feet tall, and sped across 3000 acres (about 17 rugby fields) in ten minutes.

Dodge witnessed the firestorm jump the gulch and moving rapidly towards them. With no time to talk through the options and choose the ‘best rational option’. He acted.

First he directed the team to run uphill. He very quickly realised that they could not outrun the fire. So he ordered them to drop their tools, and then he lit what is now known as a backfire (but a novel idea in 1949) in front of the firestorm, telling the crew to lie down in the ashes and let the fire wash over them. It would have worked, in fact it did work as he survived.

Yet, the crew didn’t listen, they kept running up the hill, tools in the hand and were overtaken and killed by the flames.

You might be wondering what does this tragedy that occurred so long ago have to do with organisations? How is it even relevant? Organisational theorist, Karl Weick spent years trying to understand why. Not why the fire behaved as it did — why trained, capable people, given a survivable instruction, could not bring themselves to follow it. Essentially in his own words and referring to this team as an organisation, he asked two pertinent questions: ‘why do organisations unravel?’ followed by: ‘and how can organisations be made more resilient?’.

His theories demonstrate some dangers that may be lurking in your organisation and it shows what happens when sense-making disintegrates. It further illustrates the importance of trust in a team, intergroup dynamics and role structure.

Picture of fires in Cape Town during fire season (2017)

Weick: when sense-making collapses

In Weick’s 1993 paper about the Mann Gulch disaster, he treats the event as what he calls a cosmology episode, pointing out that cosmology has to do with time, space, change and uncertainty. It refers to a moment when people suddenly feel that the universe is no longer a rational, orderly place, because something is happening that their assumptions cannot explain.

We can see this as it unfolds: the crew were dispatched to a small, manageable fire. They interpreted the information as coherent: the events that will unfold (change) in a rational, orderly way in time and space i.e. predictable, manageable, controlled. They did not account for the disruption of the fire suddenly changing direction, becoming unpredictable and uncontrolled. They also did not account for their lead telling them to drop their tools and thus abandon the reason they are there, never mind him lighting a second fire and telling them to lie in the ashes. The instructions became incomprehensible to the team, it could not be trusted, so it was ignored – with fatal consequences.

Both the sense-making and meaning of the sense collapsed together. As the event started unravelling, the team lost their structure, and they could not make sense, until they couldn’t even comprehend the one thing that would have saved their lives – following the order of their lead.

I think that Gary Klein’s data/frame model fits well here. And I used it with Weick’s process of sense-making in my PhD research design.

Klein: when expertise holds

Klein’s work initially drew from veteran fireground commanders. Observing them work, Klein found they almost never weighed several options against each other the way rational decision theory assumes. Instead, an experienced commander would size up a fire, recognise it as a version of something they’d handled before, and settle on a plausible course of action — usually the first one that came to mind — then briefly and rapidly imagine it playing out to check it wouldn’t fail. If the mental rehearsal held up, they acted. If not, they adjusted and rehearsed again. Klein called this the Recognition-Primed Decision (RPD) model, and it became one of the foundations of naturalistic decision-making research. The RPD makes a distinction between experts and novices, very clearly indicating that years of experience does not equate expertise.

For the expert in the RPD model, action and understanding become one phase that occurs at the same time as the expert recognises a situation as a familiar pattern. The key in the RPD is pattern recognition built from experience: a novice, facing the same fire, has no repository of prior situations to match it against and has to reason it out from first principles, slowly, with far less confidence.

Klein, G. (1997). The recognition-primed decision (RPD) model: Looking back, looking forward. In C. E. Zsambok & G. Klein (Eds.), Naturalistic decision making (pp. 285–292). Lawrence Erlbaum Associates

A few years later, he expanded on the RPD model with the Data/frame model (and recently with another model called the plausibility transition model, which is a blog for another day). These are all sense-making frameworks.

In the data/frame model for sense-making the premise is that when faced with a situation, people ‘frame’ the incoming ‘data’ in their minds. The frame is a theory-in-use or an active perspective describing and perceptually changing the situation. When facing a situation with new data coming in, and especially if the new data is not matching the current theory we are using, is to justify the data to fit our current theory (elaborate), ignore the data to preserve our current theory, question the data, seek more information or reframe. Each of these will lead to different actions and outcomes.

In situations like Mann Gulch or a busy emergency department with high stakes and pressures, it is incredibly important to note small changes or shifts, and accepting a mismatch between information, explanation and mental frame could be highly detrimental.

Klein argued that perceptions regarding consequence will influence what happens next, for example if people feel that their opinion on voicing an anomaly will not matter, then they might ignore incoming data, or justify it – i.e. “the team lead knows best, and would’ve seen what I am seeing”. We can see it in the Mann Gulch situation where the smokejumpers could not comprehend what was happening, as they could not let go of their story that this is a quiet, routine fire, until it was too late.

Klein’s data/frame theory holds that a frame organise the data people see. As he puts it, the more central a frame is to someone’s thinking, the more likely they are to explain away an anomaly rather than revise the frame itself. The Mann Gulch crew were running on the frame that this is a small, routine fire and every cue that should have broken it, the sudden blow-up, Dodge lighting a second fire ahead of them, arrived too fast to force a reframe. Klein’s fireground commanders avoid that collapse partly through speed of recognition, but just as much through social capital: years on the same crew, trust already banked in the one giving orders. Dodge’s crew had neither the time to reframe nor the trust in hand to follow a reframe that radical.

Klein G, Moon B, Hoffman RR. Making Sense of Sensemaking 2: Macrocognitive model. IEEE Intelligent systems 2006 https://api.semanticscholar.org/CorpusID:12538674}

Weick studied many other disasters and he later extended this into a wider allegory he called “drop your tools”: across firefighting, aviation and business, he found the same pattern of skilled people who could not let go of the equipment, procedures or identities that had made them experts, even once those tools had become a liability rather than a help. The tools were who these people understood themselves to be, firmly entrenched in their identity — which is exactly why dropping them, in the instant it mattered, was nearly impossible. Without their tools, they lost their identity.

Sense-making in the ED: what the data shows

Our SenseMaker® study across five Cape Town emergency departments (ED) asked the ED team, for a specific remembered moment. Then we asked the following: In situations like your story, when information and explanation do not match, it is best to: Check and reinterpret the information/live with the difference/look for additional information.

We found that when people said that they are feeling swamped with information being too much or too early, they were more likely to say that they would ‘live with the difference’ of information and explanation not matching. Further, ‘filtering’ the response with another question regarding their role in the situation, the decision-makers were also more likely to accept the difference, and not question it.

This is where Weick’s Mann Gulch scenario becomes valuable. A decision-maker has more than information riding on their frame – they also have their authority riding on it. Revising the frame means admitting the original call was wrong, which is a costly act. That’s precisely Weick’s “drop your tools” argument: the tools people can’t let go of are often decisions and roles as much as equipment, because the tool and the identity built on it are the same thing.

But Mann Gulch also shows this isn’t only an internal, cognitive cost — it’s also a social one. Reframing in front of others means asking them to follow you into a new, unproven interpretation, and that only works if there’s enough standing trust to spend. Dodge was the one person at the Mann Gulch incident who did reframe, radically and correctly, and it still failed: he couldn’t get his crew to follow him into the escape fire, because he hadn’t banked the social capital with them that a reframe that extreme requires. Where that trust hasn’t been built, or where spending it visibly risks the decision-maker’s standing, living with the difference is simply cheaper than asking. It’s the same failure Mann Gulch shows in extremis: not that reframing is cognitively impossible, but that getting a frame to disseminate through other people costs something most decision-makers under pressure don’t have.

Building teams that don’t drop into Mann Gulch

The key takeaway from Mann Gulch, Weick, Klein and the ED study is that you need to build resilience and ways for collective sense-making prior to needing it. I find that there is often an assumption that teams will simply work together, because they are in the same environment, or doing the same job or formally part of the same team. We should be challenging the assumption and spend more time deliberately building teams, and nurturing intergroup dynamics.

Here are a few ways that this could look like/prepared for:

  • Build resilient structure before you need it. Weick speaks about a “virtual role system”: even in a crisis, people can mentally hold the structure of the group — who does what, who to trust — if that structure was clear and rehearsed beforehand. In a team, that means naming roles and decision rights before the pressured moment, not improvising them inside it. Simulations can be very helpful here, not just how to follow the guideline, but also what to do if roles, structures and sense-making disintegrates.
  • Build recognisable patterns deliberately. Fireground commanders didn’t arrive with good instincts; years of fighting fires built the library they were drawing on. After-action reviews, deliberate debriefs and naming the specific cues experienced commanders used turn one-off experience into recognisable pattern for the whole team, not just the individual who lived it.
  • Notice when you’re being asked to drop your tools. The instruction that saves you is often the one that looks, in the moment, like it contradicts everything you know. Teams that have discussed this possibility in advance — that expertise sometimes means abandoning the very things that make you feel like an expert — are more able to act on it when it counts.
  • Build social capital before it is needed: Dodge’s order would have worked, but the tragedy clearly shows that it is not enough to shout out orders in a crisis. Trust needs to be built prior to a crisis, and this speaks to how we build and nurture teams. This has been demonstrated in several well-documented disasters. It also speaks to paying attention to the everyday language, the habits and rituals in a team. Build the team intentionally – and cross those boundaries/disciplines and siloes before the crisis.
  • A plan is a frame, and it is essential, but preserving regardless of change is a problem. Teams need to know when to abandon the plan. Watch for expectancy violations. Commit to action, but hold the interpretation loosely enough to change it.

Two halves of the same question

Within particularly the Western Cape Province of South Africa, the word ecosystem is used often to describe how the health system behaves. There is a parallel worth flagging between Weick’s reference to a cosmology episode and resilience theory in ecology. Holling’s adaptive cycle describes ecosystems moving through phases of exploitation and conservation, where accumulated structure gradually becomes rigid and over-connected, until a disturbance the system would once have absorbed instead triggers a rapid, discontinuous collapse — a release phase, sometimes called “creative destruction.”

What follows is an uncertain reorganisation, in which surviving structure and memory get recombined into something new through what ecologists, (and Weick call bricolage). In Mann Gulch the fire’s blow-up was the disturbance that crossed the threshold: the cosmology episode is Weick’s name for that same kind of release, only of interpretive structure rather than biophysical structure. I’ll be writing a follow-up post exploring this comparison properly — how Holling’s release-and-reorganisation cycle maps onto sense-making collapse and recovery, and what that means for building teams and organisations that can reorganise rather than simply fall apart.

Summary

This post particularly demonstrates how teams actively construct the situations as they are responding to it, and what happens when the situation being enacted turns hostile faster than the group’s structure or experience can keep up. If you’ve watched a team hold together under real pressure, or watched one come apart, you could probably relate.

I would like to end as I began, with a sentence from Norman MacLean. This is something to think about: “if the major purpose of your team is to put out fires so fast they don’t have time to become big ones, then you won’t learn about putting out big fires.”

Post story notes:

As the Mann Gulch fire continued to burn, it took 450 men five more days to bring the fire under control. Weick’s main source the Mann Gulch fire, was from a Norman Maclean’s book written by 28 years after the incident, using mostly historical and case reports as. Maclean was a firefighter, and was at the Mann Gulch fire (after the smokejumper incident). His knowledge of Mann Gulch was used to build a model of how wild fires spread using mathematical predictive models. Completely unrelated to Mann Gulch, Maclean also wrote A River Runs Through It that became a block buster film.

Sources


Discover more from Dr Charmaine Cunningham

Subscribe to get the latest posts sent to your email.

Scroll to Top

Discover more from Dr Charmaine Cunningham

Subscribe now to keep reading and get access to the full archive.

Continue reading