Single And Double-Loop Learning

When a crisis hits, most teams default to Single-loop learning (SLL), fixing the immediate error without changing the underlying strategy. This is like a thermostat simply turning on or off. The more powerful approach is Double-loop learning (DLL). This means stepping back to challenge the very governing principles—the assumptions and goals—that led to the mistake. It’s the difference between adjusting the water temperature and questioning if you should be bathing in oil. This radical self-correction is crucial for mastering Mayday! scenarios, where entrenched biases can have lethal consequences.

Imagine a scenario where the consequences of just fixing the immediate problem—the symptom—were catastrophic.

Consider the Gimli Glider incident in 1983. A commercial aircraft, flying from Montréal to Edmonton, ran out of fuel mid-flight. Not over the ocean, but high in the sky. The immediate fix was an astounding piece of airmanship: the pilots managed to glide the plane to a safe landing on an abandoned racetrack in Gimli, Manitoba, saving everyone on board. However, the reason for the fuel starvation wasn’t just a simple miscalculation. The investigation uncovered a critical systemic failure: the ground crew had confused gallons and litres when calculating the required fuel, compounded by a recent switch from imperial to metric units (measurement standard) and procedural issues with the new fuel quantity indication system (FQIS). If the airline had only engaged in Single-loop learning, they might have simply disciplined the ground crew or fixed the faulty gauge.

But, as with almost all incidents across the aviation industry, Double-loop learning was the game-changer. The entire organization had to confront and change its deep-seated processes, cross-checking protocols and training on the new metric system. They had to ask:

Why did our system allow a single mistake to spiral into an emergency?

This shift in fundamental operational philosophy saved future flights.

When the pressure is on—the proverbial Mayday! moment—our brains naturally seek cognitive shortcuts. These mental shortcuts or cognitive biases, are efficiency tools, but under duress, they become Achilles’ heels. They push us toward Single-loop learning, which is the path of least resistance.

Think of Single-loop learning as taking an immediate, often mechanical, corrective action within the existing framework. If a pilot is flying too low, they pull back on the yoke. If a sales pitch fails, you tweak a few slides. It fixes the immediate symptom based on the established governing variable or rule. It’s safe, fast and comforting because it validates the existing worldview.

The danger is that in high-stakes situations, the existing rules are often the very source of the failure. An outdated protocol, an untested assumption about a competitor’s actions or a deeply held organizational belief (e.g., Our safety systems are foolproof) are all ripe for catastrophic failure.

Double-loop learning is where true resilience is built. This is the process of examining and challenging the governing variables themselves. Instead of just adjusting the aircraft’s altitude, you ask:

Why did my flight plan put me in this hazard zone in the first place? Do I need a fundamentally different route planning system?

Instead of merely tweaking the slides, you ask:

Is our product even solving the right problem for the customer? Are we targeting the wrong market entirely?

This process is profoundly uncomfortable. It requires acknowledging that our deeply held beliefs—our mental models—are flawed. It activates a significant level of psychological defensiveness because it questions expertise, seniority and established norms. This is why many organisations get stuck in the Single-loop cycle, endlessly perfecting a fundamentally broken process.

To truly master judgment and decision-making under duress, especially in crisis management (Enterprise Risk Management), you must deliberately build frameworks that enforce the stop and question moment. You need a structured approach that forces teams, when faced with an unexpected outcome, to not just ask:

How do we fix this?

but, more powerfully:

What must we believe to be doing this thing this way and are those beliefs still valid?

This shift from error correction to systemic re-evaluation is the ultimate form of organizational learning.