Double-loop learning — Organizational Development

Chris Argyris and Donald Schön distinguished two kinds of learning:

Single-loop learning: detect an error and correct the action, leaving the governing variables — goals, values, assumptions, rules — unchanged. "We missed the target; work harder."
Double-loop learning: detect an error and question the governing variables themselves. "Why is that the target? Is it the right target?"

Most organizations are competent at single-loop and structurally incapable of double-loop, because double-loop learning requires questioning things that senior people have committed to publicly.

Evidence: Moderate. The espoused theory versus theory-in-use gap is well documented and easily demonstrated, and it remains one of the most useful observations in the field. Model I and Model II are clinically derived from extensive observation rather than experimentally validated, and Argyris's own finding — that people trained in Model II revert under stress — has not been systematically tested. The ladder of inference is a facilitation heuristic, not a cognitive model.

Use when: the same class of problem recurs despite repeated corrective action; when an organization is efficiently pursuing the wrong objective.

Do not use when: the goal is genuinely right and the execution is genuinely poor. Not everything is a governing-variable problem.

Espoused theory vs. theory-in-use

Argyris's most useful distinction:

Espoused theory: what people say they believe and do.
Theory-in-use: what their behavior actually implies they believe.

The gap is usually invisible to the person, and visible to everyone else. The diagnostic move is not to ask people about their values; it is to observe behavior and infer the theory-in-use, then feed it back non-punitively.

Espoused: "we want challenge and debate." Theory-in-use, inferred from behavior: "challenge is welcome from peers, not from juniors, and never in front of the board."

Model I and Model II

Argyris found nearly everyone's theory-in-use follows Model I, whose governing values are:

Achieve your intended purpose.
Maximize winning, minimize losing.
Suppress negative feelings.
Behave rationally.

Model I produces defensiveness, face-saving, undiscussable topics, and "undiscussability of the undiscussable" — the fact that the topic cannot be raised is itself unraisable.

Model II governing values:

Valid information.
Free and informed choice.
Internal commitment to the choice, and constant monitoring of its implementation.

Model II behavior means advocating your view and inviting inquiry into it, stating the data behind your conclusion, and testing rather than asserting.

The ladder of inference

A tool for making reasoning visible, from bottom to top:

Observable data.
Data I select from it.
Meanings I add, from culture and personal history.
Assumptions I make.
Conclusions I draw.
Beliefs I adopt about the world.
Actions I take — which then determine what data I select next.

The reflexive loop at the top is why two competent people can observe the same meeting and disagree completely about what happened.

Practical use in facilitation:

"What did you see or hear that led you to that?" — walk down the ladder.
"Here is my data, and here is the meaning I put on it. What do you see differently?" — make your own rungs visible.
Balance advocacy with inquiry, deliberately, in a ratio you can count.

Organizational defensive routines

Actions or policies that prevent embarrassment or threat while preventing the organization from identifying and reducing their causes. Their signature: they are anti-learning, and they are protected. Examples:

The reorganization that avoids naming an underperforming executive.
The "communication problem" diagnosis that avoids naming the disagreement.
The consultant hired to say what employees have said for two years.
The meeting after the meeting, where the real discussion happens.

Examples by audience

Health care: single-loop is "the checklist was not completed; retrain the team." Double-loop asks why a checklist designed elsewhere does not fit this workflow, and whether the measured compliance rate is measuring the right thing at all.
Software: single-loop is "we missed the sprint; improve estimation." Double-loop asks whether fixed-scope sprints are the right control mechanism for this class of work, and who chose them and why.
Executive: the undiscussable in most executive teams is the performance of a peer, or the viability of a strategy the CEO announced publicly. Both are known by everyone and raised by no one. The practitioner's job is to make the undiscussability itself discussable — which is why this work requires a contract that anticipates it. See entry-and-contracting.

Limitations

Argyris's own finding: people trained in Model II revert to Model I under stress, which is exactly when it matters.
Requires skilled facilitation and real safety.
Confrontational if handled badly; the ladder can be weaponized as "you're making assumptions".