I Am Talkative
Better at being heard

Storytelling

Explaining complicated things simply

A good explanation has a shape, and finding it is a matter of technique rather than talent.

Female narrator in glasses reading out loud from book while sitting at desk with microphone and recording audiobook in cozy study
Female narrator in glasses reading out loud from book while sitting at desk with microphone and recording audiobook in cozy study · Photo via Pexels

The people who explain well are not simply clearer thinkers — they follow a sequence, and the sequence can be learned.

Starting with why

Which is skipped constantly.

Before explaining what something is, establish why the listener should care.

Which means naming the problem it solves, the question it answers, or the thing it makes possible.

Without this, the listener has no framework to attach the information to and no motivation to build one.

And it should be genuine rather than manufactured, since a forced relevance claim is worse than none.

The concrete before the abstract

The most reliable ordering principle.

Give a specific example first, then the general principle.

Not: here is the definition, and here is an instance of it.

But: here is a situation, here is what happens, and this pattern has a name.

Which matches how people actually build understanding, and which is the reverse of how most textbooks and most experts present things.

One thing at a time

The discipline.

Each step should depend only on what has already been established.

Which requires listing the steps and checking each one for hidden prerequisites.

The common failure is a step that quietly requires something not yet explained, at which point the listener is lost and everything after it is wasted.

And the listener rarely says so, which is why explanations fail silently.

Building a mental model

Rather than transmitting facts.

The objective is for the listener to be able to reason about the thing, not to recall what you said.

Which means giving them the structure — how the parts relate, what causes what — rather than the details.

A person with a correct simple model can work out the details; a person with the details and no model cannot.

And it means being willing to give a simplified model first and complicate it later, which some experts resist and which is how learning actually proceeds.

Useful simplification

The judgement involved.

A simplification that is wrong in a way that will need unlearning is worse than a slower correct explanation.

A simplification that is incomplete but not misleading is fine and is necessary.

Which is the test: does this leave something out, or does it teach something false?

And flagging the simplification — "this is roughly right; there is more to it" — preserves the listener's ability to go further.

Analogy, carefully

Powerful and hazardous.

A good analogy is one the listener already understands deeply, that shares the relevant structure, and whose limits you state.

State where it breaks down, since listeners extend analogies past the point of usefulness.

Avoid stacking analogies, which compounds the errors.

And check that the source domain is actually familiar to this listener, which is where analogies most often fail.

Checking understanding

Rather than assuming it.

"Does that make sense?" produces yes regardless of the truth.

What works: asking them to explain it back; asking them to apply it to a new case; asking what they would predict in a specific situation.

Each of these reveals whether a model was built or words were received.

And the answers tell you exactly where to go back to.

Handling the gap in knowledge

When you realise something is missing.

Go back and fill it rather than pressing on, since everything after depends on it.

Do it without making the listener feel deficient, which is what stops people asking.

"I skipped a step" is better than "you should know this".

And establishing early that questions are welcome changes how much you find out.

Written explanation

Where the differences lie.

No feedback, which means anticipating the confusions rather than detecting them.

Which makes testing on a reader more important, not less.

Structure must be visible: headings, numbering, summaries.

Diagrams do work that prose cannot, particularly for anything spatial or relational.

And a summary at the top serves readers who need only the conclusion.

The Feynman test

A practical routine.

Write the explanation as if for someone with no background.

Where you find yourself using jargon or hand-waving, that is where your own understanding is thin.

Go and fix that, then rewrite.

Which uses explanation as a diagnostic of comprehension, and which is why teaching something is the fastest route to understanding it.

Mei-Lin Chow
Writing & Languages, I Am Talkative

Mei-Lin is a translator and editor who works in four languages, and is patient about the difference between fluency and confidence.

More from Mei-Lin →

Also by Mei-Lin Chow

Storytelling

Structuring a longer story

Anything beyond a few minutes needs a shape the listener can feel, and there are only a handful that work.

Esme Carrington··3 min read

Storytelling

Telling a story to a hostile room

Narrative works differently when the audience is sceptical, and the usual techniques can make things worse.

Esme Carrington··3 min read

Writing

Writing that persuades

Argument on the page follows different rules from argument in conversation, and the main one is about what you concede.

Mei-Lin Chow··4 min read

Public Speaking

The first minute

Openings are the most disproportionately valuable part of a talk and the part most often wasted on housekeeping.

Esme Carrington··3 min read

Public Speaking

Teaching and training a room

Training is not presenting, and the most common mistake is treating a workshop as a talk with breaks.

Esme Carrington··3 min read