Separate What Is Known, Wanted, and Assumed Before You Build a Long-Term Plan

Capture exact statements before assigning labels

Choose one plan and a short time horizon for the conversation. Instead of writing a broad topic such as “our future,” write a working question such as “What would need to be true for us to live in the same city next year?” Each person then contributes short statements in their original wording. Do not correct, merge, or answer them yet. “The program starts in September,” “I want a quiet room for weekday work,” and “We would probably visit family every month” look equally fluent, but they perform different jobs. The first may be checkable, the second expresses a preference, and the third predicts a pattern that has not happened. Keeping the original wording visible prevents the recorder from quietly turning another person’s wish into a joint promise. It also reveals compound sentences: split “The lease ends in June, so we will move in July” into a dated fact and a future assumption before discussing either part.

Require a source, scope, and date for confirmed facts

Place a statement in the fact lane only when both people can identify what confirms it. The source might be a current timetable, a written acceptance, an existing shared commitment, or a direct observation that both can inspect. Add the date checked and the narrow scope. “The course calendar lists a September start, checked on 11 August” is stronger than “The schedule is settled,” because it says exactly what was confirmed. A message remembered from last winter is not current merely because both people recall it. If the source cannot be found, move the sentence to assumption or unknown rather than debating whose memory is more accurate. The UK Government Analysis Function advises distinguishing evidence from opinion and rechecking information that may have aged. Applied here, that means facts can expire: write a review trigger beside details controlled by another person, institution, booking, or changing timetable.

Let preferences remain first-person answers

A preference answers “What would I choose or value under the current conditions?” It is neither a weak fact nor a claim that needs external proof. Write “I prefer to keep two evenings unscheduled” instead of “Two free evenings are necessary for everyone.” Ask the speaker to mark the preference as firm for this plan, flexible within stated conditions, or not yet clear. That extra field is descriptive, not a bargaining score. Two preferences can coexist without being averaged: one person may prefer a central location while the other prefers more space. The record should not invent “we prefer a spacious central home” unless both independently accept that sentence. The same rule protects changing preferences. A person may update their answer after seeing a real schedule without being accused of contradicting a fact. Record the new date and retire the old wording; do not rewrite the history to make the present answer look inevitable.

Turn every assumption into a checkable card

An assumption fills a gap between current information and a future outcome: “Remote days will stay the same,” “The commute will feel manageable,” or “Our visiting pattern will not change.” Do not call it false; mark it provisional. Policy Horizons Canada describes assumptions as statements accepted as true without proof and recommends surfacing them early for later testing. New Zealand’s Department of the Prime Minister and Cabinet also suggests reversing an assumption or testing it across more than one future scenario. For a personal plan, make that practical. Beside each assumption, write what would count as useful evidence, who will check, and by when. Reverse it once: if remote days decreased, what part of the plan would change? If the assumption survives both the ordinary case and a plausible variation, keep it as a working condition. If it fails, revise only the dependent part of the plan instead of declaring the entire conversation wasted.

Use status changes instead of winning arguments

End the sorting session by reviewing movements between lanes. A dated email may turn an assumption into a confirmed fact. A sentence first presented as a fact may become a preference when no external source exists. A broad prediction may split into a known boundary plus an unresolved assumption. Note the change without attaching blame: “Moved from fact to assumption on 11 August because the timetable is not published.” UNDP’s strategic foresight guidance says foresight is not prediction and emphasizes adjusting plans as conditions change. This makes revision part of the method, not evidence that someone failed. Close with only three outputs: facts ready to use, preferences accurately owned, and assumptions assigned for checking. Do not proceed to ranking options unless both people want that separate task. If an assumption cannot be checked now, leave it visible and decide what action remains reversible while it is unknown. A blank answer is more informative than a confident invention.

Keep exploring this topic