THE IDEA
Two kinds of learning
Single-loop learning asks: are we doing things right? It adjusts actions within existing assumptions. The thermostat notices the room is cold, turns the heating on, notices the room is warm, turns it off. The goal (target temperature) is never questioned. Only the action (heating on or off) changes.
Double-loop learning asks a different question: are we doing the right things? It doesn’t just adjust actions - it examines and changes the assumptions, goals, and rules that govern those actions. It’s the thermostat that asks: is this the right temperature? Should we even be heating this room? Is the goal comfort, energy efficiency, or something else entirely?
Chris Argyris observed that most organisations - and most people - are stuck in single-loop learning. They get better at executing their current strategy without ever questioning whether it’s the right strategy. They refine their processes without asking whether the processes serve the right goals. They solve problems faster without asking whether they’re solving the right problems. Double-loop learning is harder, more uncomfortable, and vastly more powerful - because it changes the frame, not just the picture within it.
IN PRACTICE
Questioning the question
A sales team misses its quarterly target. Single-loop learning: what went wrong with our execution? Did we make enough calls? Was the pitch effective? Did we follow up quickly enough? The team adjusts its tactics and tries again. Double-loop learning: is the target realistic? Is our market definition right? Are we selling the right product to the right people? Do our incentives reward the behaviour we actually want? The first loop improves the execution. The second questions whether the execution is aimed at the right thing.
A school’s test scores drop. Single-loop: how do we improve test preparation? More practice exams, better revision materials, targeted tutoring. Double-loop: are these tests measuring what we actually care about? Is “improving test scores” the same as “improving education”? Are we optimising for the metric at the expense of the mission? The first loop makes the school better at producing scores. The second asks whether scores are the thing worth producing.
A person keeps ending relationships at the same stage - around the two-year mark, when the initial excitement fades. Single-loop learning: how do I keep the excitement going? Plan more dates, try new things, make more effort. Double-loop learning: is “sustained excitement” the right goal for a long-term relationship? What assumption am I carrying about what relationships should feel like? Is the pattern telling me something about my expectations rather than about my partners? The first loop tries harder at the same game. The second questions the game.
WORKING WITH THIS
Getting to the second loop
The barrier to double-loop learning is emotional, not intellectual. Questioning assumptions means admitting that what you’ve been doing might be based on flawed premises. It means accepting that working harder at the current approach might not help. It means confronting the possibility that the strategy, not the execution, is the problem.
The practical trigger is persistent failure despite good execution. If you’re doing everything right and it’s still not working, the first loop has been exhausted. The problem isn’t how you’re doing it. It’s what you’re doing, or why.
To practise double-loop learning, ask: what are we taking for granted? What assumptions would need to be true for our current approach to succeed? Are those assumptions still valid? If we started from scratch, knowing what we know now, would we make the same choices? These questions feel dangerous because they can destabilise everything the current approach is built on. That’s exactly why they’re valuable.
THE INSIGHT
The line to remember
Single-loop learning makes you better at the game you’re playing. Double-loop learning asks whether you’re playing the right game.
RECOGNITION
When this is in play
You need double-loop learning when the same problem recurs despite repeated improvements in execution. When working harder produces diminishing returns. When the team is excellent at delivering something that doesn’t seem to matter. When the response to every setback is to refine the approach rather than question the direction. When someone asks “why are we doing this?” and the only answer is “because we’ve always done it.”