Model I Defensive Reasoning vs. Model II Productive Reasoning

Definition

After analyzing thousands of cases, Argyris identifies four values that govern most people's actions in threatening or embarrassing situations — Model I: (1) be in unilateral control, (2) win and do not lose, (3) suppress negative feelings, (4) behave rationally. Model I's purpose is to protect the self against disruptive change, and skilled use of it produces a "defensive reasoning mind-set" for explaining past actions and designing future ones. Model II is the alternative, productive-reasoning theory people widely espouse but rarely enact: (1) seek valid, testable information, (2) create informed choice, (3) monitor vigilantly to detect and correct error. Model II is, for almost everyone, still only an espoused theory — "very few individuals can routinely act on their espoused values, and they are often unaware of this limitation."

In the Book

Argyris re-reads the State Department and Intel cases through this lens: Rusk advocated his position, evaluated the ambassadors' silence privately without testing that evaluation publicly, and let the meeting end rather than risk losing control — textbook Model I, even though his espoused goal was to build a Model II culture of openness. Grove's version looked different on the surface (hard-driving and confrontational rather than soft and conflict-avoidant) but obeyed the identical governing values: control, winning, suppressed feelings, dressed up as rationality. Argyris uses this to make a pointed claim about leadership style — that soft power and hard power "strengthened the Traps in their respective organizations" identically, because style is surface and Model I is the underlying operating system in both cases.

The book also traces how Model I becomes self-sustaining: to use Model I effectively, the actor must not reveal that the values in play are self-protective, so the reasoning has to disguise itself as impartial and rational even to the person using it. This is what produces "skilled unawareness" — competence at defensiveness that feels, from the inside, like objectivity.

Why It Matters

This gives a content-neutral test for whether an interaction — a negotiation, a performance review, a policy debate — is actually oriented toward learning or is running on self-protection dressed as reasonableness: is testable information being surfaced and genuinely open to disconfirmation, or is control being maintained under cover of "just being rational"? Because Model I is largely invisible to the person enacting it, the concept also explains why telling people to "be more Model II" rarely works — the gap has to be surfaced through their own behavior, not their self-report.