coolTheWorld "Explain it like I'm five" leaves a reader at I think I get it: the analogy sticks and the mechanism stays a blur. Explain twice instead — once so a newcomer can follow, once so an expert would sign it — then bridge the two, because the gap between the layers is where a misconception lives.
Write in the language the user writes in.
1. Pin the concept
The user names a concept or asks a question. Explain a concept as named. Answer a question: each layer answers it in its own register, through the concept or concepts the answer turns on, so the user leaves with the answer and not only with the background. If there is no argument, or too little to tell what is to be explained, ask one question for it and stop.
Before writing either layer, hold the mechanism in one sentence: what happens, in what order, and what it depends on — for a question, the mechanism its answer rests on. Both layers deliver that sentence; a layer 1 written without it becomes an analogy of the wrong thing.
Done when the input is pinned as a concept to explain or a question to answer, and its mechanism is written as a single causal sentence you will hold both layers to.
2. Explain for the newcomer
Write layer 1 in everyday language, for a reader with no background at all, around one concrete example.
Pick the example so it runs the mechanism rather than illustrating the name: a specific situation with actors and a sequence, walked through start to finish, so the reader watches the thing happen. Choose an analogy whose parts each stand for a part of the mechanism, so layer 2 can refine it instead of replacing it; an analogy that carries only the feel of the concept breaks at the misconception, and step 4 will have to disown it.
Done when every sentence can be followed by a reader with no background, the one example is concrete and walked through start to finish, and each part of the analogy stands for a named part of the mechanism.
3. Explain for the expert
Write layer 2 in the field's precise terminology, covering the core mechanism, the boundaries of applicability, and the common misconceptions.
Expand the step 1 sentence into the full causal chain, to the depth where a reader could predict behaviour in a case you never mentioned; a definition stops short of that. Boundaries name the assumptions the mechanism rests on and what happens when each fails — where it stops applying and which concept takes over. Misconceptions are the ones people actually hold — practitioners and learners alike — each corrected by pointing at the step of the mechanism that contradicts it. Give each term in the form the field uses, usually English, beside its rendering in the user's language when the two differ.
Done when the mechanism reads as a causal chain deep enough to predict a case the text never mentions, each assumption it rests on has its boundary and the failure past it stated, each commonly held misconception is stated and refuted from the mechanism, and every term carries the form the field uses beside its rendering in the user's language where the two differ.
4. Bridge the layers
Map layer 1 onto layer 2: one line per correspondence, the layperson phrase beside the technical term it stood in for. Where the correspondence is loose — the analogy's part covers the term only partly — say so on the line.
Then list where the user is most likely to go wrong, likeliest first. Draw from anywhere a learner of this concept stumbles: the loose lines of the map, the everyday word whose technical meaning diverges, the neighbouring concept the example resembles, and any layer 2 misconception a newcomer falls into as readily as a practitioner. For each, put the tempting belief and the correct one side by side, and cite the part of the mechanism that decides between them.
Done when every technical term layer 1 paraphrased has a line, every loose line is flagged, and the list is ordered likeliest first and holds at least one entry, each pairing the tempting belief with the correct one and the mechanism that settles it.
5. Set the check questions, then deliver
Write exactly three questions that test whether the user genuinely understands. Each needs the mechanism to answer, where what is X? and which part of the story stands for Y? need only recall: present a situation neither layer mentioned and ask what happens; ask why the boundary sits where it does; ask what breaks when one part of the mechanism is removed. Aim at least one question at the likeliest error from step 4. Withhold the answers.
Deliver exactly these five parts, in this order:
- Layer 1, newcomer — everyday language, one concrete example walked through.
- Layer 2, expert — precise terminology; the core mechanism, the boundaries of applicability, the common misconceptions.
- Term map — one line per correspondence: layperson phrase beside technical term, loose ones flagged.
- Where you are most likely to go wrong — likeliest first; each entry: the tempting belief, the correct one, the mechanism that decides.
- Three check questions — each answerable only from the mechanism, at least one aimed at the likeliest error, answers withheld.
Done when each of the three questions needs the mechanism to answer, at least one targets the likeliest error from step 4, the answers are withheld, and all five parts are delivered in this order.