Arvind Murwaha’s recent essay1 opens with a scene:
An exhausted CEO at a strategy session asks her team whether they’ve agreed. Heads nod. The meeting ends. Six months later, the initiative fails. The team had apparently reached consensus, but the decision was weak. It was neither sufficiently rigorous nor deserving of the team’s commitment.
His diagnosis is one with which I agree wholeheartedly. Consensus is comforting, high-quality decisions require productive friction, and the facilitator’s duty is to help engineer honest disagreement rather than manufacture harmony.
His prescriptions, too, are sound. In certain circumstances, though, they may lean unnecessarily heavily on the facilitator’s skill and individual team members’ courage.
Murwaha notes, “As facilitators, we must help groups navigate the ‘Groan Zone’ — that uncomfortable, messy divergence of opinion — without rushing them to a premature convergence just to relieve the tension.” He proposes four in-meeting interventions: make the decision rule explicit, separate cognitive from affective conflict, run a pre-mortem, and use gradients of agreement.
Let’s look at what an explicit decision rule does. When the leader says, in effect, “I’m the decider, but I need your candid input,” the dissenter is no longer staging an act of bravery. They’re discharging an assignment.
Hidden-profile research tells us why this works. The person who uniquely holds a piece of information faces an implicit question: “Why am I the one bringing this up?” Absent a warrant to speak up, they’ll usually default to the perspectives shared by the rest of the group.2 The decision rule not only makes valid dissent permissible but also makes it obligatory.
The pre-mortem conscripts the whole group into imagining failure at once, so no one is the solitary naysayer. Gradients of agreement remove the agree-or-disagree binary that faces the eighty-percent-convinced person. These are all very solid techniques.
I do need to push back on one piece of advice. Murwaha says, “Facilitators must act as the filter. When the debate turns personal…we must intervene and reframe.” If applied rigidly, categorizing cognitive conflict as “good” and affective conflict as universally “bad” might sometimes suppress critical information.
To be clear: I'm not defending affective conflict in Murwaha's sense — the personal attacks, the grudges, and the status games. Those deserve filtering. What I'm pointing to is different: cognitive content that carries an affective cost. The information itself is task-relevant, but saying it out loud creates relational discomfort. It's not that the speaker is angry; it's that the sentence, once spoken, might make someone else feel corrected, bypassed, or exposed. This isn't affective conflict. It's decision-relevant information wrapped in social risk. And a filter calibrated for the former can inadvertently catch the latter.
For example, "I think the timeline is aggressive" makes it to the whiteboard during a pre-mortem. "The last time we tried geographic expansion, we pulled back after eighteen months. I was here for that, unlike most of the people in the room," might not. Not in open discussion, and not on a sticky note that can be attributed to a person in the room.
This isn't an accusation of dishonesty or even a criticism. It's just uncomfortable because saying it implies that the people driving the initiative don't know what the person speaking knows. That implication, however mild, may be enough to keep the sentence unspoken. The concern can't be constructively reframed and redirected unless it's given a voice. That may not happen in a meeting.
If the meeting can’t tolerate the sentence, the sentence has to be spoken somewhere else. The obvious tactic is to elicit it before the room convenes and in a way that severs what was said from who said it. The reason the historical parallel goes unspoken is that attribution carries an uncomfortable cost. Remove the attribution, and you remove most of the cost. The concern that could never survive eye contact across a conference table can survive an anonymous prompt answered alone, before the meeting.
This is the logic behind the kind of pre-meeting elicitation I’ve been working toward.3 It doesn’t ask anyone to be braver than they are. A facilitator working from a synthesis of that input walks into the room already knowing a tension exists, and can open the question as a planned feature of the agenda rather than waiting for someone to risk their standing to raise it.4
That much I’m confident about.
Surfacing a buried sentence isn't the same as resolving it. A synthesis can show a facilitator that someone holds a dissenting view—but not whether that view is right. You learn a pattern exists; you don't learn enough to evaluate it. The anonymity that makes the sentence sayable may strip away the accountability needed to act on it.
So the honest conclusion isn't that good pre-meeting design rescues these in-meeting techniques. It's that pre-meeting processes and in-meeting methods aim at different things, and neither reaches all the way. The meeting can't surface the uncomfortable sentence. The upstream process can give it voice, but can't resolve what that voice reveals.
That's not a failure of method. It's a boundary condition. The facilitator who knows where their tools stop is more useful than one who believes good process fixes everything. Sometimes the most valuable thing you can do is surface a problem clearly enough that those with authority to act can no longer pretend it isn't there.
Stasser, Stewart, and Wittenbaum (1995) found that telling a group in advance who held expertise on which topic substantially increased the surfacing of unshared information. The assignment didn’t make people more confident; it made their contributions less presumptuous. I interpret their findings to mean that suppression doesn’t reflect timidity; it results from the absence of a warrant for one-person assertion.
This is the design stance behind Reflect & Engage, the pre-meeting orientation process I’ve been writing about here. I explored the same logic through a different analogy in “Sweeping the Ground Before You Walk“ — detection without detonation — including the caveat that revealing a buried disagreement doesn’t, on its own, tell a group what to do about it.
Three objections to anonymous elicitation deserve more than a footnote allows. Anonymity shelters the sniper as readily as the truth-teller. Asynchronous reflection tends to improve what gets said, but a synthesis can't reliably separate a substantive worry from a self-serving one. It can also set a room guessing who said what, and a concern specific enough to be decision-relevant is often specific enough to identify its author. The synthesis is itself an act of interpretation: whoever performs it decides what to elevate, relocating the problem of bias rather than removing it. And the convergence-seeking habits of most methods quietly sand off the outliers that the process exists to preserve. I have no definitive answers. These are the questions that decide whether anonymous elicitation is a useful instrument or facilitation theater.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.