Reader situation
In a disagreement, both sides keep restating their positions and interpreting new information in their favor. Every example seems to confirm what each person already believed. The discussion goes in circles, and neither side learns anything.
Central idea
A discussion becomes more useful when you can describe what would alter your conclusion. Define relevant evidence before interpreting every new detail in favor of your current view.
Practical steps
- State belief. Say clearly what you believe about the question. Vague positions are hard to test. “Remote work reduces collaboration” is clearer than “Remote work has downsides.” Stating the belief precisely helps you and others see what’s actually being claimed. Include the scope: “Remote work reduces collaboration in teams that are new to each other” is narrower, and often closer to what people actually mean.
- Explain confidence. Say how confident you are and why. Are you very sure, based on years of direct experience? Moderately sure, based on a few examples? Uncertain, but leaning one way? Confidence levels help others understand how firmly you hold the view, and they make it easier for you to revise it. Be honest about where your confidence comes from. Strong feelings aren’t the same as strong evidence.
- Specify evidence. Before looking at more examples, describe what evidence would change your mind. For example: “If teams that moved to remote work produced as many cross-team projects as before, I’d reconsider.” Specifying this in advance prevents you from dismissing inconvenient evidence later. Ask the other person to do the same: “What would change your view?” If neither of you can name anything, the disagreement may be about values rather than evidence, which is useful to know.
- Seek counterexamples. Actively look for examples that don’t fit your belief. Ask others in the discussion for situations where the opposite happened. Counterexamples don’t necessarily prove you wrong, but they show the limits of your view and the conditions under which it applies. One well-documented counterexample can be more informative than several examples that agree with you. When someone offers a counterexample, ask about its context rather than dismissing it: “What was different about that team?”
- Reconsider. Review the evidence against the threshold you set. If it’s met, revise your view. If it’s partly met, narrow your view to the situations it fits. If it isn’t met, you can hold your view with more confidence, while staying open to new evidence. Tell others what you’ve concluded and why. Explaining your reconsideration helps others think through their own positions. It also makes it easier for others to change their minds, because you’ve shown that revising a view is a normal part of discussion rather than a loss.
Worked example
Fictional example
The people and details are invented to illustrate an approach. It reports no outcomes.
In a Circle for operations leaders, Sven and Aiko disagree about whether four-day weeks reduce output. Sven states his belief: “In customer-facing operations, a four-day week reduces output.” He says he’s fairly confident, based on his own team’s trial. Aiko says she believes the opposite, based on two organizations she’s worked with.
Instead of continuing to argue, Sven asks: “What would change either of our minds?” Aiko says she’d reconsider if several organizations with similar customer service needs reported lower output. Sven says he’d reconsider if teams similar to his kept their response times steady over six months.
They ask the Circle for examples. Three members share their experiences, including one team very similar to Sven’s that kept response times steady after adjusting schedules. Sven narrows his view: “It seems the outcome depends on how schedules are designed. Our trial didn’t change rotas, which might explain our results.” Aiko acknowledges that her examples were from organizations with lower customer demand. The discussion ends with a clearer and more useful conclusion than either started with. Other members say the thread was one of the most practical the Circle has had.
Companion tool: Evidence threshold prompts
Use these prompts before or during a disagreement.
- My belief, stated precisely: ________
- Scope: In what situations do I think this applies? ________
- Confidence: ☐ Very confident ☐ Fairly confident ☐ Uncertain
- What my confidence is based on: ________
- Evidence that would change my mind: ________
- Question for the other person: “What would change your view?”
- Counterexamples I’ve found or been offered: ________
- My conclusion now: ☐ Revise ☐ Narrow ☐ Hold, with more confidence
Boundaries
- Some disagreements are about values or priorities, not evidence. Defining evidence won’t resolve those, but it can show that this is the real difference.
- Not every discussion needs this level of rigor. Use it when the question matters.
- Be fair to others’ evidence. Hold their examples to the same standard as yours.
Member action
Ask a discussion partner what evidence would change either of your positions. Answer the question yourself first, so the request feels fair.
Discussion prompt
What evidence would change your mind about a view you hold strongly in your field?
