Follow-Up to: On Juvenile Fiction

Related to: The Simple Truth

I quote again from JulianMorrison, who writes:

If you want people to repeat this back, write it in a test, maybe even apply it in an academic context, a four-credit undergrad course will work.

If you want them to have it as the ground state of their mind in everyday life, you probably need to have taught them songs about it in kindergarten.

Anonym adds:

Imagine a world in which 8-year olds grok things like confirmation bias and the base-rate fallacy on an intuitive level because they are reminded of their favorite childhood stories and the lessons they internalized after having the story read to them again and again. What a wonderful foundation to build upon.

With this in mind, here is my challenge:

Look through Eliezer's early standard bias posts.  Can you convey the essential content of one of these posts in a 16-page picture book, or in a nursery rhyme children could sing while they skip rope?

Write the story, and post it here.  Let's see what we can come up with.

This is not, by any means intended to be a simple challenge.  On the one hand, we are compressing a lot of information into a small space.  On the other, good fiction is not easy, and children's fiction is no exception.

We have two options.  We can humbly admit that we are not skilled writers of children's fiction and walk away, or we can determine that this is a task which needs to be completed, produce lots of really bad fiction, and begin the process of criticizing one another, learning from our mistakes, and growing stronger.

When I was a boy, I had a thick book of 365 short stories, some not even taking up a full page.  Each was self-contained, and I could flip open the book at random and find a story I hadn't read before.

How quickly would our community grow, both in strength and in numbers, if we could crowdsource a Rationalist's Book of Tales?

I know, I know.  It's optimistic. It's ambitious. Most of all, it seems really silly.

Let's do it anyway.

New Comment
38 comments, sorted by Click to highlight new comments since:

I wonder if poems designed to warn children against advertising could vaguely be included here?

Anyway, here's my attempt. I originally just planed it to be about confirmation bias, but it grew to encompass other things.

Little Johny thought he was very bright,
But the schoolkids did not -- they would laugh when he came in sight.
He could count, sing, and guess the weather.
Then one day, Big Bill said "Real bright boys will grow a feather."

"Ach!" he cried, "Could it be true?"
"Then I'm not bright, which makes me blue."
So he went home, and searched all over.
And then found growth on his head, clear as a clover.

"It is true, feathers are sprouting!"
"It's proof that I'm bright!" So he stopped pouting.
He ran to show his mom, nearly tripping over some eggs,
When he saw on TV "Bright boys will grow long legs."

So he waited for weeks and weeks for to find proof,
Worried over his brightness, and staying quite aloof,
Until one day, feeling in a pinch,
He grabbed a tape measure, and found his legs had grown a whole inch!

So he leaped off to school, but a scientist walked by,
And Johny overheard him say, that real bright boys could fly.
"The hair, the legs, from these I know
Of my brightness. The flying thus follows, so..."

Little Johny plotted of his grand display,
Standing high on a wall, he would proudly say
"Behold, I have proof that I'm bright!"
And he would deftly leap off, and soar into flight.

So he climbed up the wall, and made his speech,
But there his plan stopped with a screech,
For he hit the ground hard with a smack,
Leaving his leg all bloody and black.

As the other children laughed, he tried to explain,
Of the things that he heard, and why he had taken it to his brain,
"They came from on high, from people who knew
I looked at myself, and saw they were true."

They laughed, "You're too eager to believe, you fool.
Your feathers are just hair, all boys grow long legs as a rule.
Yes, if all you heard were true, you'd fly, but you'll find out,
That if you do logic with garbage, then you'll get garbage out."

So Johny thought wrongly, and got his leg in a cast,
He had sought fame in the schoolyard, but now that's all past.
He's taken the lesson to heart, no longer believes all he hears.
So he doesn't believe them when they say he's not bright -- brightness doesn't come from peers.

This deserves its own post.

You can fix the spacing on the above by adding a couple of spaces to the end of each line, instead of having an empty line in-between. see documentation

Thanks a lot, done!

I happen to think that Hans Christen Andersen did a pretty good job with "The Emperor's New Clothes". One of the best examinations of the various reasons why people groupthink and groupspeak, and it's not only directed at but is highly enjoyable by children!

Parents actually read it to their kids, even theists. I always found that amazing. It's like they're not actually aware of the point of the story on some level.

"Cinderella, dressed in yella / Had a theory she would tellya / How much evidence could she ignore / Before her listeners start to bore? / One, two, three--"

James Thurber wrote an amusing one-page parody of The Tortoise and the Hare that serves as a nice explanation of publication bias.

I think it's great that people are identifying existing children's tales that serve this purpose. Since they are already written, and in many cases out of copyright, they are the low-hanging fruit we can use to jump-start a Rationalist's Book of Tales.

Dr. Seuss wrote some great rationality stories for children. I Had Trouble In Getting To Solla Sollew, The Sneeches, and The Butter Battle Book, just off the top of my head.

I can offer a rationalist jump-rope couplet: A long walk starts with a single step, Solve the first problem first and the next one next. One, two, three, four...

The beats seem wrong for jump-rope - or have I got that wrong?

This bit works:

a LONG walk STARTS with a SING-le STEP

But then it gets messy:

SOLVE the first PROB-lem first AND the NEXT one NEXT

It's the "problem first" section that has to get squished into a single beat that'd be tough... Doable - but doesn't "feel natural", so it probably wouldn't get used.



Yeah - that scans much better :)

Could even be sung:

take the FIRST step FIRST and the NEXT one NEXT



13 years later: did anyone end up actually making such a book?

Here's my first attempt. It was meant to be about confirmation bias. On rereading it, it seems more like it's about finding the positives in bleak situations. I guess that's important, too.

Nothing much happened this week, you tell your mum at week's end. Because 'nothing' never gets mentioned, when talking to your friend.

Because inside all that nothing, there are things both here and there. If a room is mostly empty, You'll find a table or a chair.

On Monday you did nothing, except you stubbed your toe.

On Tuesday you did nothing, but it begain to snow.

On Wednesday you did nothing, except you found that money.

On Thursday you did nothing, except you spilled that hunny.

On Friday you did nothing, except you cleaned that fluff.

But now remembering your empty week, do you think of all that stuff?

So telling your friend about that room, Or your mum about your week. Now that room is full of stuff. Your week is not so bleak.

Teach Timeless Decision Theory via a Choose-Your-Own Adventure Book?


I tried to do some of them, but I found that I often felt that I needed more prerequisites even when Eliezer's posts did not themselves have any prerequisites. I guess EY can assume that we know certain things which I cannot assume 5 year olds know (for example, that your parents don't have infinite money).

I'll keep hacking away at it, I guess, but I'm having a hard time finding a starting point.

Hmm. Try copying Aesop's approach. Bad thinking should have bad consequences that can be played out (unexplained) in the plot, after which time the explanation of the moral will be easier to follow.

Keep in mind that Aesop's Fables didn't originally come with morals! That's a strictly Victorian perversion.

The Fables were originally meant to be understood and interpreted by everyone who heard them, themselves.

(edit) I will further note that many of the "traditional" morals pasted onto some of the Fables don't actually make much sense when looked at critically. Does "slow and steady wins the race" make sense in itself, much less as an explanation for "The Tortoise and the Hare"?

A much more plausible moral would be "talent is good, but hard work is better".

Actually I think it is an important lesson.

A good, modern, quick example might be:

Amy and Betty are both sick and they have medicine to make them better - they have to take one tablet a day. Amy gets lazy and misses a few days, then tries to make up for it by taking a whole bunch at once. Instead of getting better - she gets really sick from overdose. Whereas Amy's friend Betty who continued to take the medicine one day at a time, every day - got better.

The moral here is that for some activities (eg taking medicine, exercise, maintaining a relationship or learning a new skill..). you have to put in small amount of regular, continuous effort - rather than thinking you can make up for it by overdosing on a large amount at one time.

Alternative moral: "No matter how great you are, there exists a level of pride sufficient to bring you down."

Note that one probably doesn't want to aim the program at children who are too young. For one, magical thinking is a normal part of child behavior until a certain age, so trying to impose a rationalist mindset on those kids probably won't do much good before they're ready for it. Also, (this isn't my field of expertise, but was pointed out to me by a friend who's studied developmental psychology in more detail) young children do need a feeling of safety about the world (visible, for instance, in the way that daily routines are important for them) - imposing a rationalist "question everything" mindset from too early an age probably isn't the best way to go.

I suspect that, for both of these points, you'll want to be aiming at a somewhat older age than five.

Kids under five anecdotally have a harder time learning to count, learning that there's still the same number of balls after you re-arrange them than before (many pre-schoolers will count them all over again), etc. So a math program for pre-schoolers needs to be built from smaller building-blocks than a program for adults. I imagine it's similar with rationality.

That needn't mean teaching math or rationality to pre-schoolers is silly. My mom played a lot of math games with me in the zero to five years, and while it's hard to tell effects without larger sample-sizes, and despite the Judith Rich Harris results, it seems plausible that this gave me math farther into my bones than I would otherwise have gotten it. (E.g., a large fraction of my pre-age-eight memories involve mathematical concepts, usually ones I was just sponaneously thinking about.)

As to the feeling of safety, there're a lot of early rationalist thinking skills that can be taught without eroding the kid's trust in their parents and stable home. Math, noticing how you're feeling and what effect that has on your actions, asking why gadgets work and enjoying the exploration, turning questions into testable hypotheses...

My mom played a lot of math games with me in the zero to five years

What sort of games? I should be running my daughter through these!

A couple of people asked about the games your mum put you through as a kid. I'm really interested too.

Care to share? :)

My mom played a lot of math games with me in the zero to five years

What sort of math games does a mother teach a zero year old baby? I can't imagine it would have much interest beyond "how many breasts do I have to choose from for my next meal?"

I seem to recall Richard Dawkins expressing a desire to write something like this in the future.

Anyone know him professionally? Get him in on the thread.

I've heard this story that we have to teach things to children at a young age in order for them to fully embrace it before, but is there any evidence of this actually happening? Moreover, what's wrong with people opting into being rational?

You probably learned one language (probably English) when you were 1. Maybe you've learned one or more since. If so, how effortful is it to speak one of these later languages compared to your native language? How does the speaking ability of others, who have learned your language later in life, compare to your own?

Now imagine that the habits of subjecting our beliefs to criticism, of examining them from every angle, of seeking the truth with an open mind, were as effortless as speaking your native language. This is what we're trying to do.


I agree with your underlying point, but the way you framed your argument is less convincing because it doesn't coincide with the findings of linguists.

It's not the number of languages you learn, but the age at which you learn them. So a child might learn 4 languages as easily as learn 1. But once childhood is over, no matter how many languages you already know (or don't know), learning more is always difficult.

I was actually speaking of age, rather than number -- if I gave the opposite impression then I must have written my comment poorly.

This is a version of "what's the time, Mr Wolf".

Participants: one "it" and several other players.


(IT stands against a wall, eyes closed or blindfold. PLAYERS stand some distance away facing IT.)

IT: "I'm the man who's found a truth"

PLAYERS: "we don't believe you"

Repeating round:

PLAYERS "we have bits of evidence" (they take a step forward)

IT (tries to guess where they are by sound, if he does he lunges and touches a PLAYER, and says) "I refute you" (he only gets one lunge per round)

Repeat the round until either

  1. the PLAYERS sneak up behind IT, "we have N bits of evidence and you're wrong" (they lunge, touch IT and he loses)

  2. IT touches the last PLAYER with a lunge "I refute you all", and he wins

I'm sorry, but I think this is exactly wrong. Framing the exchange of evidence as a competition is part of what makes people so irrational in the first place.

Much real scientific evidence exchange is a good-hearted competition between academics. You have to want to refute mistakes, but you also have to tolerate losing if you can't refute the opposition. This was intended to teach both sides.

In my opinion, this is a very Traditional Rationalist way of thinking.

I can think of two ways of dealing with people taking sides in arguments. The first is to encourage each participant to state their opinions in a humble way. This will reduce the loss in status participants experience when they change their mind. The second is to condition participants to believe that correctness should be totally unrelated to status (the final strategy mentioned here).

Your game associates the search for truth with something that is much more clearly a competition. Competitions generally have a large status component associated with them. Thus the game works against successful use of both strategies.



nice work. but i don't know if the word bayesian should be in the title.

also, why have a scientist say something so stupid?