“The only true voyage of discovery would be not to visit strange lands, but to see with other eyes.”

— Marcel Proust, La Prisonnière

Walk into a recurring strategy meeting and you can watch a room decide nothing, fluently. The agenda is familiar. The issues are known. The dialogue is smooth. The head of sales asks about capacity; the head of product gives the standard answer about roadmap velocity. Everyone speaks. Everyone nods. The meeting ends on time. And nothing has happened.

You already know how it runs. The room knows which questions are safe to ask, which answers are acceptable to give, and which silence reads as agreement. The question no one asks is the one that might disturb the arrangement. Call it frame stagnation: the set of assumptions driving the business survives every meeting untouched, protected by a layer of busy conversation.

The questions are socially negotiated in advance. Ask a predictable question and you get an answer retrieved from memory, not built by thought. The room is not reasoning. It is reciting.

Now, be careful with that claim, because a fast answer is not always a lazy one. An expert can answer in a heartbeat and be exactly right; speed can be deep pattern, not autopilot. So fluency by itself proves nothing. It becomes a warning sign in one place only: when a room meant to be examining its own assumptions glides straight over them, never touching the frame underneath.

Do not confuse this with a lack of rigour, because the two look alike. A rigorous meeting and a stagnant one can produce the same fluent exchange. The difference is whether anyone is thinking, or merely matching a known stimulus to a stored answer. The danger is not that the room is lazy or stupid. It is that fluency feels like progress, so a room can recite its way through a decision and mistake the smoothness for alignment.

Underneath sits a switch between two modes of thinking, and a familiar question keeps it stuck in the wrong one. In a high-stakes room, you learn the rules fast: challenging a timeline is safe, challenging a mandate is not. That learning hardens into a loop, a predictable question triggering a rehearsed answer, and the assumption underneath is never examined.

What breaks it? A question the room has never heard. Your mind reaches for the script, finds nothing that fits, and has to switch from retrieval to live analysis. This is why such a question wins not by being clever but by being unfamiliar. Cleverness can be parried with another rehearsed line. Real unfamiliarity cannot, because there is no rehearsed line to reach for.

So how do you know the switch has flipped? Silence. In a scripted exchange, the gap between question and answer is almost nothing. In a reset, there is a distinct pause. Read that pause carefully, though, because it does not always mean the same thing. It can be insight forming. It can also be confusion, or offence, or someone quietly building a way around you. The pause only tells you the script has broken; what comes after it tells you whether the break did any good. So do not fall in love with the silence. Watch what fills it, because that is where the fears, incentives, and hidden assumptions that were steering the room become briefly visible.

The cost of never reaching that pause is an institution that sleepwalks through its own stewardship. It recites safe questions and safe answers while the frame that needed examining sails through untouched, and the failure, when it comes, is a surprise everyone half suspected.

The method has a real danger, and it has to be named. Surprise is a power, and like any power it can be abused. A fresh question must never be an ambush dressed as inquiry, a tool to embarrass a colleague, trap a subordinate, or perform dominance. Used that way it triggers defence, which is worse than autopilot, because a defended mind is even less able to think than a scripted one. So the governing ethic is curiosity, not judgment. The aim is to disarm the room’s guard, not to attack the person behind it. The same question, asked to open a mind or to corner one, produces opposite results, and only the first is worth asking.

You cannot memorise these questions, because the moment one becomes a script it loses its power. What you can learn are the angles that reliably break autopilot, and notice what they share: none can be answered by pulling a position off the shelf. Each forces the room to build one.

The most useful is the pre-mortem. Ask: it is three years from now and this initiative has failed, what was the one reason? The point is not pessimism. It is permission. The question lifts the social pressure to perform certainty and gives the room a safe way to say what it already suspects but cannot yet admit, because the hypothetical carries none of the blame the plain statement would. The thing everyone sensed and no one would voice arrives as an answer to a hypothetical, which is the only form in which it could be spoken.

Other angles work the same way. Constraint removal, if the budget were infinite, would you still run this exact plan, separates the merit of an idea from the scarcity that was hiding its weakness. The agency audit, who are we really trying to impress here, the customer or the board, exposes whose approval the work is really serving. Each angle does one job: it asks something the room cannot answer from memory, and so forces it to think. A scripted question maps to a cached answer in an instant. A built question has no cache to draw on, so the mind must leave autopilot to make one, and the assumption it was quietly protecting surfaces in the building.

Watch a room recite the same reassurances for a year, then go silent for five seconds when someone finally asks what would have to be true for this to be a mistake. That silence is the first real thinking the project has seen. So when the meeting is fluent and nothing is moving, stop hunting for a better answer and ask a question the room cannot already answer. The enemy of good governance is not error. It is sleepwalking. Break the predictability, and you restore the inquiry.

Decision Rule — The Probe Protocol

When the room is looping, ask one question that cannot be answered from memory, and read the response.

• The probe. Pose a question that does not fit the meeting’s usual pattern, asked from curiosity, never to corner anyone.
• The read. Immediate fluency means the question was safe and the script held. A noticeable pause means you have touched a real assumption.
• The follow. Stay in the pause. Let the room build a new thought rather than rescuing it with a rehearsed one, and watch what fills the silence.

If every question you ask is answered instantly, the room is reciting, not deciding.

If the answer comes without a pause, you asked a question the room had already scripted.