Apply the problem-vs-polarity test — if new information…

Why This Is a Rule

Barry Johnson identified a fundamental distinction that most leaders miss: some tensions are problems to solve, and some are polarities to manage. Problems have correct answers — you can resolve them permanently with enough information. Polarities are interdependent pairs where both poles are necessary and neither can permanently win: centralization/decentralization, individual/team, stability/change, short-term/long-term.

Applying problem-solving to a polarity produces organizational thrashing. The leadership team decides "we need more decentralization!" (correctly diagnosing over-centralization) but treats it as a problem to solve rather than a pole to balance. They decentralize aggressively. Two years later, the downsides of over-decentralization (silos, inconsistency) trigger the reverse swing: "we need more centralization!" This oscillation repeats indefinitely because the underlying tension isn't resolvable — both poles are necessary.

The diagnostic test is binary: "Can new information or analysis make one side permanently win?" If yes → it's a problem; solve it. If no → it's a polarity; design management systems (Design early warning indicators for polarity drift…) rather than searching for resolution. Misdiagnosis in either direction wastes enormous effort.

When This Fires

• When a tension keeps recurring despite multiple "resolutions" — strong signal it's a polarity, not a problem
• Before investing significant effort in resolving any persistent organizational tension
• When two factions in an organization each advocate for opposite approaches and both have valid arguments
• During strategic planning when allocating effort between resolution and management of known tensions

Common Failure Mode

Treating polarities as problems: "If we just get the right balance of centralization and decentralization, we'll be done." You'll never be done. The right balance shifts continuously as context changes. If you've "resolved" the centralization/decentralization question three times in five years and it keeps coming back, it's a polarity. Stop searching for the answer and start designing the management system.

The Protocol

(1) When facing a persistent organizational tension, apply the test: "Can new information, better analysis, or a creative solution make one side permanently win while the other is fully abandoned?" (2) If yes → it's a problem. Solve it with the standard resolution toolkit (root cause analysis, experimentation, decision-making). (3) If no — if both sides are genuinely necessary and neither can be permanently abandoned → it's a polarity. (4) For polarities: map both poles' positive and negative results. Design early warning indicators for drift (Design early warning indicators for polarity drift…). Establish regular review cadence to check balance. (5) Communicate the polarity diagnosis to stakeholders: "This is not a problem to solve but a tension to manage. Both sides are right."