Skip to content
anymind

Method ·

How to Outline Your Thinking: A Field Guide

Outlining is not formatting. It is a small set of moves — split, nest, promote, prune — applied until the structure matches what you actually believe.

Most of us were taught outlining backwards. School said: plan the structure first — I, II, III, then A, B, C — and fill in content afterwards. Anyone who has done real work knows the lie in this. You rarely know the structure before you start. Structure is the output of thinking, not its prerequisite.

Practical outlining inverts the order. Write the fragments first, as they come, in whatever order they come. Then apply a small set of mechanical moves until the shape emerges. The moves are few enough to list on one hand, and each one is a kind of question you ask your own material.

Move 1: Split Until Atomic

A node that contains two ideas resists every other operation: you cannot move half a node, promote half a node, or schedule half a node for review. So the first pass is splitting — every sentence holding two thoughts becomes two nodes.

The question this move asks: is this one claim or several? You will be surprised how often a confident paragraph turns out to be three claims, one of which you do not actually believe.

Move 2: Nest by "Because" and "For Example"

Indentation is a semantic act. A child node should answer a question about its parent: why? (evidence, reasons), such as? (examples), how? (steps). If you cannot say which question a child answers, it is not a child — it is a sibling that drifted, or it belongs to a different branch entirely.

This is where outlining becomes thinking. Deciding that an idea sits under another is a claim about your argument's load-bearing structure. Get it wrong and you will feel it later, when the draft refuses to flow.

Move 3: Promote What Keeps Recurring

As material accumulates, watch for a telltale pattern: the same theme appearing as a grandchild in three different branches. That theme is not a detail. It is a top-level idea that has been living below its station.

Promote it. Make it a sibling of the headings it used to hide under, and move the scattered mentions beneath it. Demotion is the mirror image: a heading that never grew children was probably never a real section — push it down into the branch where it actually belongs.

Move 4: Zoom to One Branch and Finish It

A full outline of any serious project is too large to hold in attention, and trying to polish everything at once is how drafts stall. Zoom into a single branch — in AnyMind that is one keystroke — so the branch becomes your whole world. Finish it: split, nest, reorder, done. Zoom out. Pick the next.

Working memory is the scarcest resource you have. One writer here finished a doctoral thesis on exactly this move — not by being able to hold more, but by needing to hold less.

Move 5: Prune, Loudly

Every outline accumulates nodes that no longer pull their weight: duplicates, abandoned angles, notes-to-self that expired. Delete them. The tree is a working structure, not an archive of everything you once typed — mess you keep "just in case" is rent you pay on every future read of that branch.

(If something is worth keeping but not worth seeing, moving it to an archive branch at the bottom of the tree is an honest compromise. The point is that the active branches stay readable.)

A Worked Rhythm

In practice the moves run as a loop rather than a sequence:

  1. Dump fragments into an inbox as they arrive — no structure at capture time.
  2. At triage, move fragments into the tree (Move 2 decides where; Move 1 splits compound ones on the way).
  3. Weekly, scan the whole tree at one zoom level and apply Moves 3 and 5 — promote the recurring, prune the dead.
  4. When it is time to produce — a draft, a plan, a decision — zoom into one branch (Move 4) and walk it top to bottom. If the outline is right, the writing is mostly transcription.

None of these moves requires talent. They require cheap, instant tree operations — which is the real reason the tool matters. When restructuring costs one keystroke, you restructure constantly, and the outline converges on what you actually think. When it costs a dialog box, you stop restructuring, and the outline fossilizes at your first draft of the truth.