Map the Disagreement

A disagreement map is a short description of what the parties differ about, what each must establish, and what an acceptable outcome would look like. It matters because people often trade answers to different questions and mistake activity for progress.

Suppose a team says, “We should return to the office.” One person may dispute the facts: remote work reduced delivery speed. Another may accept the facts but dispute their meaning: speed is not the only value. A third may agree about both yet reject the proposed remedy: mandatory attendance is too broad. Until those questions are separated, every new statistic is liable to miss its target.

Write the contested proposal at the top of a page. Under it, record each party’s strongest version, not the easiest version to defeat. Then ask what decision must actually be made, by whom, and by when. This converts a cloud of irritation into a set of answerable questions.

First exercise: take a current disagreement and complete: “We agree that ___; we disagree about ___; the decision required is ___.” If you cannot fill the final blank, you may be having a conversation rather than deciding a case.

A dispute type names the kind of question being contested. Four practical types are fact, definition, value, and policy; each needs different support, so identifying the type prevents an irrelevant response.

Type	Core question	Useful support	Common mistake
Fact	What happened or will happen?	Records, observations, forecasts	Arguing preferences
Definition	What should this case be called?	Criteria, boundaries, examples	Treating labels as self-evident
Value	What is better, fairer, or more important?	Standards and tradeoffs	Hiding the chosen standard
Policy	What should we do?	Benefits, costs, feasibility, alternatives	Proving a problem but not a remedy

Real disagreements often contain several types. “The rollout was reckless, so pause it” includes a factual question about the rollout, a definitional question about “reckless,” a value judgment about acceptable risk, and a policy proposal to pause. Deal with them in a sensible order: establish the relevant facts, state the evaluative standard, then test the action.

Sorting drill: rewrite each sentence as a question. “This price is unfair” becomes “Against which fairness standard should this price be judged?” A question exposes the missing work more clearly than an accusation.

The burden of proof is the responsibility to provide enough support for a claim to deserve acceptance. It is not a permanent punishment for one side; it follows the claim, the requested change, and the practical risks of being wrong.

The person proposing a change normally carries an initial burden: explain the problem, show why the proposal addresses it, and compare it with reasonable alternatives. A person making a serious factual allegation carries the burden for that allegation. Meanwhile, a defender who claims the status quo is harmless also assumes a burden; “do nothing” has consequences and is not automatically neutral.

Set the standard in proportion to consequence. Choosing lunch needs less evidence than accusing a colleague of misconduct. A reversible pilot can proceed under uncertainty that would be unacceptable for an irreversible migration.

Burden statement: “Because I am asking for ___, I need to show ___ to the standard of ___.” Then write the other side’s corresponding burden without making it impossible.

Scope is the boundary around the people, period, cases, and consequences covered by a claim. A claim with uncontrolled scope becomes vague enough to evade testing or broad enough to collapse under one exception.

Mark four boundaries: population, time, domain, and strength. “Meetings reduce productivity” could mean all meetings, in every team, forever, always reduce every form of productivity. A defensible version might be: “For this engineering team during the next six-week delivery window, recurring status meetings longer than thirty minutes are likely to reduce uninterrupted build time.”

Boundary	Prompt	Example
Population	Who or what is included?	This product team
Time	During which period?	Six-week delivery window
Domain	Under which conditions?	Recurring status meetings
Strength	How certain or universal?	Likely, not always

Narrowing is not cowardice. A precise claim tells the listener what evidence could confirm or weaken it. It also permits partial agreement: perhaps the argument holds for status meetings but not design reviews.

Scope edit: circle every universal word—all, never, everyone, proves. Replace it only if you can name a justified boundary. Keep a strong term when the evidence truly supports it.

A position is the stated demand; an interest is the need or concern the demand is meant to protect; a constraint is a limit the person may be unable to change. Separating them reveals room for agreement without pretending incompatible demands are identical.

“No launch before Monday” is a position. The interest may be avoiding weekend incident duty. The constraint may be that the support vendor is unavailable on Sunday. Another party’s “launch Friday” may protect a contractual milestone, with a constraint imposed by billing. The positions clash, but the map may reveal alternatives: limited release, delayed announcement, or staffed support.

Ask “What does that protect?” to uncover interests and “What cannot change?” to uncover constraints. Do not assume every claimed constraint is fixed; record who controls it and what evidence would show flexibility.

Map template: Position / Underlying interests / Real constraints / Suspected preferences / Questions to verify. Keep “suspected” separate from “confirmed” so empathy does not become mind-reading.

Success criteria are observable conditions for judging whether a conversation or decision worked. They keep “winning” from shrinking into making the other person surrender in the moment.

Choose criteria at three levels. A reasoning criterion asks whether the conclusion is adequately supported. A relationship criterion asks whether people can continue working and speaking honestly. An action criterion names the next step, owner, and review point. A good outcome may include a narrower decision, a pilot, or a documented unresolved point.

Examples include: both parties can state the other’s case accurately; the decisive factual uncertainty has an owner; the proposal has a reversible test; and nobody is pressured to affirm something they do not believe. Notice that agreement is not mandatory. Sometimes the successful result is a clean, respectful description of a genuine conflict.

Success card:

• Decision: What must be decided?
• Evidence: What would be sufficient today?
• Relationship: What conduct must remain possible afterward?
• Action: Who does what next, and when is it reviewed?
• Stop rule: What would make continuation unproductive or unsafe?

Read the card aloud at the start of a consequential meeting. It turns process expectations into shared property.

A derailment is a move that shifts attention away from the question currently at issue. It may be accidental—people associate freely—or strategic, so the remedy is to label the shift without speculating about motive.

Common derailments include changing the subject, demanding impossible certainty, attacking a person, multiplying minor objections, and using a vivid anecdote to erase broader evidence. Respond with a parking lot: record the new issue, say whether it matters, and return to the active question.

Derailment	Neutral response
New topic	“That may matter. Can we park it and finish the pricing question?”
Impossible standard	“What level of confidence is proportionate to this reversible step?”
Personal attack	“Let’s test the claim independently of who raised it.”
Objection cascade	“Which objection would change the decision most?”
Anecdote replaces pattern	“What does this case show, and what cannot it establish alone?”

Do not call every uncomfortable objection a derailment. A challenge is relevant if answering it could alter the claim, evidence, standard, or remedy. The map should be revisable when a real missing issue appears.

A mapping workshop is a timed exercise that produces a shared picture before advocacy begins. It is useful for workplace, civic, or personal disagreements where conversation has become repetitive.

Minute 0–2: each party writes the decision and the other side’s position. Minute 2–4: compare wording and correct only for accuracy. Minute 4–6: sort issues into fact, definition, value, and policy. Minute 6–8: assign burdens and list evidence. Minute 8–10: agree on scope, success criteria, and the first answerable question.

Use this checklist:

• [ ] One decision is written as a neutral question.
• [ ] Each position is stated in terms its holder accepts.
• [ ] Fact, definition, value, and policy issues are separated.
• [ ] Every material claim has a proportionate burden.
• [ ] Population, time, domain, and strength are bounded.
• [ ] Interests and constraints are marked as confirmed or suspected.
• [ ] Success includes reasoning, relationship, and action.
• [ ] Parked issues have an owner or a reason for exclusion.

Finish with a one-page map, not a speech. The map is successful when a newcomer could explain what must be decided, where the disagreement lives, and what evidence or judgment would move it.