Stage 1: Start from a live question and surface your stake
Do not begin with a topic. Begin with a live question.
Weak:
"I want to write about delegation."
Stronger:
"Why did delegation make this process slower even though responsibility was clearer?"
Then surface your stake:
- Why has this stayed unresolved?
- What decision could change?
- What answer do you want to be true?
- What would be inconvenient if true?
A writer with no stake may produce detached information. A writer who hides their stake may produce advocacy disguised as inquiry.
Stage 2: Reconstruct the ground truth before interpreting
Reconstruct before interpreting. Ask for these details:
- exact sequence
- dates
- people or roles
- numbers
- constraints
- what was tried
- what changed
- what did not change
- records or artefacts that support the account
Memory is evidence of experience, not a perfect recording. Mark estimates and missing details rather than smoothing them away.
Stage 3: Make the interpretation inspectable
State the explanation in plain language. Then identify these parts of it:
- mechanism
- assumptions
- borrowed ideas
- causal leaps
- simpler explanations
- local conditions
- omitted facts that weaken the story
The goal is not to make the explanation impressive. It is to make it inspectable.
Stage 4: Find the strongest opposition
Find the strongest alternative, not a ceremonial objection. Ask:
- What evidence contradicts the position?
- What does the explanation fail to explain?
- What would a sceptic say is cherry-picked?
- Under what conditions would the opposite be true?
- What would change my mind?
If the answer to the last question is "nothing", you do not have a conclusion. You have an identity claim.
Stage 5: Commit to a position and mark its boundary
Force a provisional commitment:
- What do I now believe?
- What did I stop believing?
- What is the smallest defensible claim?
- What is the strongest defensible claim?
- What stronger claim can I not defend?
- What remains uncertain?
- What adjacent questions should stay separate?
This is where an idea earns enough shape to become writing material.
The output is a Thinking Document, not an outline
A useful Thinking Document is an epistemic record. It preserves these parts:
- the inquiry
- trigger and stake
- ground facts
- observations
- interpretations and mechanisms
- assumptions
- speculation
- present position
- countercase
- boundaries
- unresolved questions
- raw material
- claims not ready for use
- connections to other thinking
- readiness verdict
Do not polish it into an essay. Its messiness is useful because it preserves the provenance of the idea.