The Explanation LabAll 20 lessons

17 MINDS · How people actually learn

Cut the load you add.

You will learn why explanations overload people, which kind of load you can remove, and how to remove it. It takes 3 minutes.

Here's how a car engine works

EVERYTHING AT ONCE

The four-stroke cycle comprises intake, compression, combustion and exhaust phases, synchronized via camshaft-actuated poppet valves timed to crankshaft position at a 2:1 ratio, with spark timing advanced relative to top dead center according to load and engine speed, while the pressure-volume relationship over the cycle determines indicated work and thermal efficiency.

CylinderPistonRodCrankshaftIntake valveExhaust valveSpark plugCamshaftIntakeCompressionPowerExhaustP-V plotLegendRPMPoppet valves2:1 ratio

New things on screen: 17. Working memory holds about four.

Got all that? Almost nobody does. And the engine isn't the problem. The explanation is.

Watch the clip

The clip does the same thing: everything at once, then a reset. Watch the counter in the corner. It counts the new things on screen.

34 seconds · lesson 17 of 20 from the full video

Start again, one new thing at a time

  1. A piston slides up and down in a cylinder.

    2 new things so far

  2. A rod and crankshaft turn that into rotation.

    4 new things so far

  3. Two valves: one lets air in, one lets exhaust out.

    6 new things so far

  4. A spark plug lights the fuel.

    7 new things so far

Then the four strokes, one row each:

StrokeWhat happens
IntakeIntake valve opens. Piston moves down. Air and fuel come in.
CompressionBoth valves close. Piston moves up. The mix gets squeezed.
PowerSpark. The mix burns and pushes the piston down hard.
ExhaustExhaust valve opens. Piston moves up. Burned gas goes out.
7 new parts, plus 4 strokes in one table

Same engine. Same parts. It's the same amount of information as the paragraph above. This time it arrived at a pace your memory could hold.

The name for it

THE NAME FOR IT

Cognitive load theory

John Sweller · 1988

Working memory holds only a few new things at once. Intrinsic load comes from the subject. Extraneous load comes from the explanation.

There are two kinds of load in any explanation:

FIRST TRY

intrinsic loadextraneous load

SECOND TRY

intrinsic loadextraneous load

Same engine, so the same intrinsic load. Only the extraneous load changed.

TAKEAWAY

You can't cut the subject's complexity.
You can stop adding your own.

Cut the load you add

  1. Count the new things. Go through your explanation and count terms, symbols and parts the reader hasn't met. If one paragraph introduces more than three or four, split it.
  2. Name parts before you use them. Introduce the piston before the sentence about compression.
  3. Put things together that go together. Labels on the diagram, not in a legend. The definition next to the first use.
  4. Remove what doesn't serve the goal. The P-V diagram and RPM gauge were true and interesting. They weren't needed to understand four strokes.
  5. Build, then show the whole. End with the full picture, once the parts have homes.

Ask your agent

PROMPT
Explain the topic below without overloading a newcomer.
Before writing, list every term, symbol and part the explanation will need.
Then introduce them in an order where each one builds on what came before, with no more than three new items per paragraph.
Define each term the first time it appears, in the same sentence.
Leave out anything true but not needed for the goal I give you, and list what you left out at the end.
Finish with one short summary that uses all the terms together.

Goal (what the reader must understand):
Topic:
[paste here]

What good output looks like: an explanation that grows one piece at a time, a final summary that reads easily, and a short "left out" list you can choose to add back.

Sources