Structural Fixes Don't Reach Theory-in-Use

Definition

A recurring diagnosis (and complaint about the change-management literature Argyris reviews): leaders default to fixing Traps by changing the environment — new organizational structures, revised reward policies, flatter or more "organic" designs — on the assumption that different conditions will produce different behavior. Argyris argues this "will, at best, have a weak positive effect," and the effect won't persist, because "Traps cannot be reduced by focusing on environmental factors such as new structures and reward policies. Individuals operating in such structures or under such policies will use the theory-in-use that they already hold."

In the Book

Argyris cites Burns and Stalker's classic mechanistic-versus-organic organization distinction as a direct test of the claim: organic structures are designed to enable exactly the open, adaptive behavior Model II requires, yet a review of actual transcripts from people working inside organic structures showed they were still reasoning in Model I — because they had "never been educated in Model II theory-in-use and productive reasoning," the organic design gave them more freedom to act, but not a different theory-in-use to act with. "This inconsistency continues to flourish," he writes, in every case where structural redesign substitutes for addressing the underlying reasoning pattern.

He applies the same critique in Chapter 6 to the "ambidextrous organization" literature (Tushman and O'Reilly), which prescribes structural steps for balancing stability and innovation but never asks how long an organization's leaders privately knew about its problems before "discovering" them through the framework — leaving the underlying Traps untouched beneath a new structural overlay. Table 5.1 in the book lists eight factors executives themselves blame for stalled culture change (bureaucratic rigidity, fear, weak rewards, lack of executive follow-through); Argyris's rebuttal is that several of these factors — a victim mentality, blame-the-system reasoning — are themselves products of Traps, meaning the list mistakes symptoms of the underlying pattern for its causes.

Why It Matters

This is a caution against a very common intervention instinct — redesign the org chart, change the incentive, decentralize the decision rights — when the actual constraint is a reasoning pattern that will reproduce itself under any structure until it's addressed directly. It applies wherever "we changed the system and nothing changed" shows up: the new system still routes through the same people using the same theory-in-use, so structural change without a parallel change in how people reason under threat is treated as a "necessary but not sufficient" move, not a substitute for it.