Project-simulator — The Flight Simulator for Project Managers

Double-loop learning: when skilled professionals stop learning

Published 2026-07-02 · The research behind Project-simulator

Here is an uncomfortable paradox: the most well-educated and successful people in an organization are often the ones who find it hardest to learn from their mistakes. Not because they are stupid – but because they fail so rarely that they have never had to practice it.

That was the core of Harvard professor Chris Argyris's classic article "Teaching Smart People How to Learn" in Harvard Business Review in 1991. His conclusion has become a foundational concept in workplace learning: the distinction between single-loop and double-loop learning.

What does the research say?

Argyris distinguished between two kinds of learning. Single-loop learning means correcting an error without questioning your underlying assumptions – like a thermostat that switches on the heat when it gets cold. Double-loop learning means also examining the assumptions behind your own actions: why did I set the thermostat to that particular temperature? In working life: why did I choose to solve the problem that way, and what does that say about how I see the situation?

His observation, built on many years of studying consultants and managers, was that skilled professionals almost always stay in the single loop. When something goes wrong, they look outward: the client was unclear, management set the wrong conditions, the team didn't deliver. Argyris called this pattern defensive reasoning – a way of protecting yourself from the feeling of having failed. And the more successful a person has been, the stronger the defense, because their identity is built on being the one who knows.

The insidious part is that defensive reasoning feels like rational analysis from the inside. The person feels they are getting to the bottom of the problem – but never examines their own part in it. So the same pattern repeats in the next project, and the organization's "lessons learned" become a list of what everyone else should change.

What does this mean for you as a project manager?

When a project starts to struggle, notice your first explanation. Is it only about others – the sponsor, the supplier, the team? That may be true, but train yourself to always ask the follow-up question: what in my own behavior made this more likely? Perhaps you signed off on vague requirements to avoid a difficult discussion. That question is the entrance to the double loop, and it costs only a moment of discomfort.

In retrospectives and lessons-learned meetings: lead with your own example. If the project manager says "one thing I should have done differently is...", it becomes safe for others to do the same. If the project manager only says what everyone around them should change, the group learns to protect itself.

Also pay attention to your most competent team members. Argyris's point applies to them most of all: the senior who never receives criticism may be the one least equipped to take it. Give feedback aimed at behaviors and assumptions, not at the person's competence.

How to practice this

Double-loop learning requires someone to hold up a mirror – and enough safety to look into it. In Project-simulator you practice difficult conversations against AI counterparts, and afterwards you get a debrief that points to exactly what you said and asks what you assumed in the moment. No face-saving needed, because nothing is at stake. Try it free for a week at project-simulator.com.

Want to practice this for real?

In Project-simulator you practice these conversations against an AI counterpart and get feedback grounded in research like this.

Try free for 7 days
Source: Argyris, C. (1991). Teaching Smart People How to Learn. Harvard Business Review, 69(3).