So I have a three-year old kid, and will usually read or tell him a bedtime story.

That is a nice opportunity to introduce new concepts, but my capacity for improvisation is limited, especially towards the end of the day. So I'm asking the good people on LessWrong for ideas. How would you wrap various lesswrongish ideas in a short story a little kid would pay attention to?

I'm mostly interested in the aspects of "practical rationality" that aren't going to be taught at school or in children's books or children's TV shows - so things like Sunk Costs, taking the outside view, wondering which side is true instead of arguing for a side, etc.

Pointers to outside sources of such stories are welcome too!

Edit: actually, if you want to share ideas of games or activities of the same kind, go ahead! :)

New Comment
19 comments, sorted by Click to highlight new comments since: Today at 3:42 AM

Sunk Costs - should be easy. Someone does something reasonable, then the situation changes, but the person refuses to change their path, explicitly speaking about the sunk costs. Because it's a bedtime story, the stupidity of the person could be astronomical. For example: there way a boy who wanted to ski in the winter, so he saved money and bought skis... but at the time the snow melted and the winter was over. But the boy insisted on his emotional and financial investment, so he kept wearing the skis during the whole year. Insert various unrealistic situations, resulting problems, and lost opportunities. The other kids were swimming in the lake, this one kid tried, but... wearing the skis... he almost drowned, and after he was saved he just kept sitting at the beach, envying the other kids. But he still refused to take the skis off, explaining how long he worked to get them.

Wondering which side is true instead of arguing for a side - make it so that no side is completely true; each of them is right about some aspect, but wrong about other. Two teams of children saw a mysterious animal; one team reported it as a pink elephant, other team as a brown bear; the Curious Kid refused to take sides and investigated, and it was actually a brown elephant.

Those are the kind of things I was looking for, thanks :)

I like your story ideas, but I wonder (seriously) about the need to crank everything up to the "astronomical" or obviously ridiculous. One of the things we are trying to do with our 2.5 year old daughter is keep our stories fairly realistic and avoid superstimuli. I'm amazed how hard it is to find books that don't involve talking animals with oversized eyes doing ridiculous things. Fortunately her favourites are the Charle and Lola books which involve two fairly normal kids doing everyday things in a fun way and using their imaginations. Not a lot of strict rationality techniques but plenty of good everyday problem solving. And she just loves them. Thanks to the op for starting this thread.. I'll give it some thought and try to come up with something..

I would say something absurd and when my son said I was wrong I would ask him for evidence (me: there is an invisible dinosaur in the room, son: it would crash through the floor, me: some dinosaurs are small.)

Playing the Devil's Advocate is a great way to teach wariness to just taking idea's at face value, and will also develop your child's ability to work out why the statements are wrong. Another practice that me and my mother would do is have a conversation, and let it flow to where it may. At some point, we would then stop the conversation and try to follow the flow backwards to the original point of the conversation. While it sounds mundane, I now look back on it as great practice in following my train of though, and seeing why I think what I think.

Mast tying: Frog and Toad Together, by Arnold Lobel, the "cookies" story. After eating too many cookies, Frog and Toad make them less and less accessible. Bonus because Frog and Toad are just generally awesome.

Not stories but...

Fermi estimates can be fun for kids (How many gumballs in this gumball machine is a classic) though for a three year old, it may be too advanced. How many action figures fit in this glass?

Three year olds can play "Guess the animal" which is 20 questions with some leeway on the yes/no part ("is it bigger or smaller than horse" is always our first question). Ingrain some binary search algorithms!


You seem to have the goal of exposing your child to ideas. Will you be expecting these ideas to take root? Your child might not get it, and your child might get it but not care as you do. If either of these outcomes happens, do not be discouraged. Memories that you spent time together matter most.

I encourage you to read Piaget theories of childhood development. Two sentences for you to consider: "At about two to four years of age, children cannot yet manipulate and transform information in a logical way. However, they now can think in images and symbols."

You seem to have the goal of exposing your child to ideas. Will you be expecting these ideas to take root?

Nah, it's more a strategy of "try a lot of things and see what sticks" - and also, I want to get better at teaching those kinds of ideas; if they don't stick now, they might later when the kid has grown and my teaching ability has as well.

(I've read about Piaget - though not from Piaget himself; more in Psychology textbooks and Wikipedia - and have been reading up more generally on psychology of development and learning; I know that at that age plenty of parts of the brain aren't fully developed yet, so I don't have high hopes for teaching everything at a young age)

You've already seen other responses from people who tell what they do OTHER than stories. I would go with that. For my kids, I ALWAYS tried to answer any queries in a very informative way, and give both sides when it was an opinion question. I talked about origins of racism, why we like to buy things we get (the people selling/making it deserve money so they can buy things, an econ lesson). I would constantly tell them relatively ridiculous things and get them used to challenging me. One of my proudest moments was when my 4 year old accused me of clapping twice after I tried to explain an echo as sound bouncing off the houses across the street.

For stories at that young age, I just admired their concentration, their commitment to the stories. I would tell some stories where the characters were essentially version of their aunt their mom their cousins. There would usually be hidden stuff underground, but I wasn't too concerned about loading "real" teaching in to these, I just loved watching their concentration. The nice thing about telling kids stories is they LIKE it if you repeat yourself, if you like the story. SO I could tell very similar stories each night and only vary them slightly to keep myself a little amused and to explore what might work with the kids. I found making up stories to be more taxing than reading stories, but it was fun some of the time.

consider: "At about two to four years of age, children cannot yet manipulate and transform information in a logical way. However, they now can think in images and symbols."

Tangent: I remember, and remember remembering, and have evidence of remembering at least one situation to which I applied what passed for rational thought at the time, which I later found out matched with an event when I had just turned two (I had been assuming throughout elementary school that I had been around four at the time, but a serendipidously found video that corrected me; I definitely didn't start ordering memories chronologically until at least age 5). It stuck with me because I made a couple predictions based on experience, decided to act on one (my parents' fears about me falling in the lake overestimated the risk), and acknowledged and ignored the other (if I ignored them and ran toward the water I'd inevitably get caught and called back, but it didn't matter because I was right). (It turned out that I was right about the second prediction, and subsequently failed to test the first one.)

This always gave me issues with developmental psychology literature, until I got to LW, admitted to myself that the literature probably knows what it's talking about, and I was probably just weird.

When pressed for this I tend to take a particular story, and try to strip it down into it's abstract form - something like - protagonist and companion undergo journey, face 3 opposing characters, arrive at destination and have life lessons confirmed. Then redecorate it with different particulars from other stories - switch males for females, dinosaurs for insects etc. Often this process stimulates my own creativity and i end up with something fairly new, but it guarantees an acceptable bedtime story. Know your audience though. A three year old child mainly wants amusing images. Mild scatology tends to go down a treat as well.

Sounds like something I occasionally do, or try to do (mostly because my kid asks for stories with dinosaurs); do you have a handy list of abstract stripped-down stories by any chance ? (or do you use an existing one like the Seven Basic Plots?) Do you have a usable list of life lessons?

(I'm slowly collecting and organizing material like this, it's something I find pretty interesting, not only for telling stories to my kid ...)

Nope, I generally just plagiarize the last film or piece of fiction I've read. Sooner or later I'll probably end up traumatizing a niece because the last film I saw was by Lars Von Trier. As for life lessons, the typical stuff you teach children of that age group - younger kids need to learn empathy, self control etc. Older ones can be introduced to basic social commentary - sometimes people have to be mean to be nice, sometimes people get sick and go away, but this is nobodies fault, sometimes your imagination is unreliable etc.

At this point, it might be more beneficial to teach values than anything abstract. For example, a lot of people here are crippled by akrasia. You could hedge against that by telling your kid stories where hard work pays off. That might make teaching other things later a lot easier. Too many stories have a protagonist who did nothing to become awesome.

A fun project! And one I'm trying to do for my kids.

One thing that worries me a little about trying to tell parables about these sorts of concepts is that, outside mathematical formalism, most real-world examples are not clear cut. Most fallacies, for instance, have versions that are useful real-world heuristics. Take post hoc ergo propter hoc. It is indeed strictly a fallacy to deduce that and event was caused by the event that immediately preceded it. But "What did you do differently just before it broke?" can be a really useful diagnostic question. And most of the time when you're a small kid, an adult pulling an appeal to authority on you really does know better than you do.

I'd worry less about trying to introduce abstract concepts to small kids and do more modelling/engaging/reinforcing of general curiosity, questioning, reasoning, and trying to figure things out for yourself. If they get that, they'll be able to pick up the abstract concepts for themselves, whether you are an effective teacher of them or not.

Kids seem remarkably immune - or even resistant - to adopting explicit 'morals' from stories (I know I was, and my own kids seem similar). But they do soak up general approaches and underlying values.

The best moments are when the kids ask about something. But for me it's often a fine balance between giving them the immediate answer (satisfying their curiosity and rewarding asking), and using it as an opportunity to build their ability to work things out for themselves.

You might find this useful -- may help with improvisations :-)