senthil's blog

How to explain stuff

November 25, 2007


When it comes to explaining stuff, I used to wonder what is it that a person brings to the table that a book cannot, A book has so many advantages, you can make many copies of it, read it at your own convenience etc.

When I was a teenager, I used to get up at 4. I would bike a good distance in the cold morning wind to attend classes given by a certain famous local teacher. Thinking back I understand what an insane thing it was.

This particular teacher was very good at explaining and his passion was a presence in the room. It existed like a separate powerful, cheerful person, like a genie filling up the entire room. It is hard to enter into that room and get out without inhaling some of that spirit.

Frequently my teacher will digress about how his subject fits into life, he will talk about how to lead life right in the middle of talking about the structure of benzene and its chemical properties. It didn’t really feel out of place. In fact this kind of placing things in successively larger and larger contexts created a sense of meaning to what we were doing at that cold morning at 4 in the am. Why does it matter, why should I care.

Its easy to underestimate the importance of this. Or even if it does think that it has nothing to do with explaining stuff. A few weeks back I had to write a xyz-script to extract some thing out of proprietary format. I do not know xyz. But when I sat down to learn the few essentials that would help me get away with what I needed, I noticed that my mind went into a sort of paralysis. I felt stupid. I found myself re-reading things, not understanding what they mean.

My brain literally has a mind of its own. It wasn’t convinced that it had to focus getting this knowledge. It was a feeling quiet akin to having to pay a courtesy visit to a relative whom you passionately hate.

I could dive into python in a few days and start writing pythonic code. It might in fact be due to the essential simplicity of all things beautiful, but I want to stress that I found it easy right from the start. Not even after a few hours when I got a grip. From the start it was effortless, effortless since my mind was open to this learning. An openness that could only be inspired by a passionate leader.

So to be very clear, when you display passion for what you are explaining you engage the full attention of the listeners conscious and subconscious minds. You are already making what you are saying easier to understand by literally using more of the listeners mind.

Good explainers save your time. I wasn’t barely making it to the class. I was terrified of missing one. The cost of missing a class would be weeks spent trying the grasp the material, and even when I do its only in a shallow way. Few people bring this kind of value to their explanations. The crucial element that makes all the difference is sense-making.

Most books are just a pile of information. Information can be used to make sense of things. Sense making is not as straight forward as reading, sense making takes time to say the least. Information is a piece of wood. Sure sometimes you can hang on to a piece of wood and float on for a while, but only when you make a ship out of it can you cross the ocean. Information becomes formidable when you make sense out it. The promise of a good explainer is sense-making.

Certainly if they are saving time, good explainers must be worth a lot of money. More over its a knowledge, not an information economy. My friend explained to his boss about python, which convinced him to use it instead of their regular language, for a certain project. To go for a conservative estimate, he probably saved him a year. Explainers can save a life time when they bring clarity out of their experience between 2 alternative life paths.

At the very least, in scenarios of regular work places I witness these people regularly saving weeks to months of time by leaving better documents behind and by writing code that explains itself. According to Paul Graham, the difference between a good hacker and a great one is their ability to give good explanations. If we say a good hacker writes code that works and a great one writes code that is easy to understand, we get some clue as to how good explainers become great hackers. Code that explains itself is easy to understand. The skill to explain transforms ones intuitions for design.

If you imagine an explainer as a painter then his finished work will be called understanding. Explainers deliver understanding. For this reason, explainers have an intimate knowledge of what constitutes understanding. A side effect of being good at explaining could be being good at learning.

Good explanations need an understanding of how the mind works, the needs of your particular listener and even the kind of attitude you carry.

Simplest form

I do not recollect it, but my moms favorite story from my childhood is when she taught me subtraction. What is five minus two drew a glassy stare from me, and so she tried the slightly longer version. You had five laddus and you ate three. How many do you have. And the answer usually came back in what we now call hardware speeds. Sometimes with a defiant expression along the shades of “Are you kiddin me?“.

Even the loftiest scientific principles have a simple intuition behind them. Newton said that every theorem he worked out, he first started by dealing in simple examples. The bible would not have gone as far as it did if it had not used stories to convey morals. Examples help us to touch ideas, push them around, feel them. Only in this play can we learn to bend them. Gauss said one example is worth two books.

Finally an inability to explain in simple terms is the most direct evidence of an absence of understanding.

Space, time and explanations

There is one aspect of explanation that make them clearly more song like(linear). The order in which you stack the ideas you are presenting will make make or break your explanation. And so comes about the question of what comes first and which ones after.

Ideally you make a graph of the idea space, with the connections representing the “depends on” relationship. And you can do a topological sort on this graph to guide your presentation. Unless you are writing a book you can do this by hand, using a pencil and paper. And before a short explanation you should at least give this stacking a passing thought.

In this context I should mention one draw back of using presentation software unwisely. Suppose you are explaining about an aspect of tectonic plate theory, and you are using a complicated diagram to explain some crucial land movement. In order to save time, you already made the diagram to show. This, I think is not always the best idea

The mind understands better when you slowly add layers of the diagram along with the appropriate explanations while doing so. When you present the diagram as a whole, you are buying too much distraction. Pointing to what you are talking about doesn’t totally offset this. The explanation should grow organically and so should the diagram you are using to explain it.

Help the listener to focus in all ways possible.


Empathy, in this context, is understanding what prevents someone from “getting” what you are saying. They might have some pre-conceived notions on their minds. The explainer must probe for this constantly and modify what he says subsequently even as they talk. And in order to get this information, your listeners should be comfortable with you. And this in turn depends on the attitude that you bring to the table. We all find it much easier to approach certain people with a question than others.[1]

One of the saddest things to say when explaining something to a kid or anyone for that matter is “You should have known this before”. This effectively puts the listener on the defensive. Exceptions aside, I have noticed that people who usually say this are only trying to distract you from their own insecurity. It might be that they are not sure about that part either, and they just panicked about having to explain it now. Knowing this has helped me quiet a bit. It gives away bad explainers instantly.

Another handy hack is to watch for these words: “Its really simple” versus “This thing is very complex”. Paradoxically people always seem to think that what they understood is very simple once they have understood it. Those patterns are so familiar to them that the thinking happens below their awareness. This is akin to seeing something and knowing that its blue. Ask a neurologist about what’s so simple about that one!

Holding things on paper

Try to make sense of both these sentences.

(1) The fly the spider the cat caught caught was swallowed by the old lady.

(2) The cat caught the spider that caught the fly the old lady swallowed.[2]

Both these sentences mean the same thing. But you probably understand the second one better. We have a working memory like a small scratch pad inside our brains. The process of making sense involves manipulating the words stored in this scratch pad. And as you can see, this working memory is quiet small. Which is why you couldn’t quiet catch the first sentence.

When you are explaining stuff you are putting a lot of stress on the listeners conscious mind. The conscious mind is the focus machine. And you want it to do just that. But then the listener should be able to hold everything in his head, manipulate them, finally understand them.

All I am saying is just take a paper and pencil and doodle as you go. By doing this you are effectively creating a memory aid for the listener. That’s all a paper and a pencil do: a memory aid. So the diagrams can be doodles, sketchy, imperfect and still have remarkable effect on the process.

Songs and maps

When you hand someone a large stone which is painted like a balloon, lets say you managed to pull it off very convincingly, what do you think will happen. The person will probably drop the stone. If they are lucky, on your toe.

Concepts which you are trying to explain are very similar. They may be heavy stones or just full of hot air. Your first job as an explainer is to give the listener an idea of what is coming. They say that the brain is a muscle, and just like a muscle it needs to be alerted of what is expected of it. Frequently we find it easier to understand things the second time around. We know when to focus, where to skim. We have a smooth function describing the difficulty/importance of concept landscape that we are trying to understand.

Voice modulation also falls under this category. An explanation is more like a song than a speech. When you modulate you voice you add another dimension of information far more important than words alone. You are adding how important each of the words, sentences are. You are adding profile information!.

And finally only by modulating your voice can you convey your passion. And if you are not passionate you shouldn’t be giving that explanation in the first place. Dogs can smell rats. That’s what they do. They don’t effort to do so. Its automatic. Peoples ability to sense a lack of passion can be compared to such instincts. The listening mind senses this lack of passion, and concludes that you don’t have anything of value to say. Remember the last time you tuned out of a lecture and try to recollect how they spoke. Passion is the sharp clear boundary separating good explainers from the bad.

[1] Thinking about the approachable people and the attitude they carry is something which will fill out to another essay. I shall leave it open for now except to note that its worth while to think about it.

[2] Mind Hacks: Tips & Tricks for Using Your Brain (Hacks) by Tom Stafford and Matt Webb.