How kids learn fascinates me. They don't do what you tell them to. Obeying orders is an abstract non-instinctual skill. Its simpler to just copy other people.

Plagiarism has advantages over obedience. First of all, it can be bootstrapped. Children are born without understanding language. Nobody can tell you how, in English, to speak English if you don't already understand English. More importantly, copying others is robust against deception. If everyone says you should obey the government while actually subverting it you should probably be a criminal too.

There is a teenage girl in my family who likes hanging out at my house while I work from home. She takes the initiative to do chores wherever she can be useful. This makes sense from an evolutionary perspective. She says she learns more per hour hanging out with me than going to school. (She complains high school doesn't teach how to do one's taxes. I should have her do my taxes.) It's more fun and useful too.

From what I know about primitive societies, they used little abstract verbal instruction. Most of the time spent learning to forage is spent foraging. Most of the time spent learning to cook is spent cooking. Elephants and chimpanzees spend no time on dialectic instructions at all.

What separates human pedagogy from other animals is our stories. Kids love stories, especially stories about people. Kids like stories about people so much the word "story" usually implies a story about people. I used to think the primary purpose of stories is to learn vicariously. But stories are often fantastical. I think the more important purpose is to provide examples of admirable behavior. It's no coincidence the most popular stories are all revolve around sympathetic capable hero protagonists.

Another fascinating attribute of kids is how immutable their preferences are. The kids in my family are all girls. They have little interest in how machines work[1]. They have no interest in weapons or dominance. They care about animals, medicine and relationships. The teenage girl wants to be a marketer so much she took the initiative to volunteer to market my business for free. (A job all three of our male co-founders loathed.) I learned marketing reluctantly after I accepted it was mandatory to achieving my other ambitions.

Your interests reflect your values. I frequently modify others' behavior by showing them how to manifest their desires better or easier. I have never successfully modified another person's values. Kids love me because I never try to shape them into something they don't want to be. My responsibility is to inspire, to lead and―occasionally―to explain.


  1. This isn't for lack of ability. One of them recovered a Linux system with a broken window manager on her own with no supervision…so she could log into Discord. ↩︎

New to LessWrong?

New Comment
12 comments, sorted by Click to highlight new comments since: Today at 6:01 AM

Something else I've noticed with very young children, usually starting between 18 months and 3 years old, is how much they love pretend play – and how this is often clearly pointed at practicing basic skills. Small children love to imitate what adults around them are doing: pretending to cook, to work on a laptop, to make phone calls, etc. (I've seen toddlers who can't actually talk in sentences yet make "phone calls" and babble into the phone.) At a slightly older age, kids start bringing stories from media they're exposed to into their pretend play. 

My impression is that many mammals do play, and that this is an important learning method, but that human children have much more sophisticated pretend play, which broadens what kinds of skills they can imitate and practice. 

This article has an interesting intersection with one of LW's other favorite topics: autism. Because autism (especially the high-functional, hyperverbal type common round these parts) is fairly described as an impairment in purely imitative learning, leading to an over-reliance on explicit verbal instruction...

which frequently leads to getting it wrong, especially in domains in which the explicit verbal instructions are lacking or in conflict with empirically observed behaviors. Your example of "everyone says you should obey the government while actually subverting it" is a good example of what seems to happen to high-functioning autists with respect to social rules, for example.

…autism (especially the high-functional, hyperverbal type common round these parts) is fairly described as an impairment in purely imitative learning…

I haven't heard this description before. I like it.

Most people talk a lot about how they hate hypocrites. Hypocrites say you're supposed to do one thing, and then they do another, and people don't like that. I can understand admitting that it is hard to live in accordance with your stated standards, but people shouldn't lie that they believe there is anything plausibly contextually good about a standard when they don't actually believe there is anything plausibly contextually good about the standard. Otherwise you can't hold people accountable to the standards that both of you say you think have something maybe-good about them.

Of course, there is an important distinction between lying and simulacrum level 3, wherein everyone understands the situation. People who aren't in on the simulacrum level 3 shouldn't be punished for wanting to understand the inconsistency. Once the inconsistency is explained, there is no problem. The explanation should be open for everyone to see, so as not to discriminate against those who still don't know. I don't think it's autistic to be unaware of the reasons for every weird inconsistency between word and action, and it's definitely not autistic to ask about them. 

No one is simulacrum-3-omniscient, and everyone is born with very little knowledge of simulacrum 3 situations. It would be poorly calibrated to expect a consistent flow of uninterrupted simulacrum 3 stability given how little most people know.

I don't think I disagree with this, except to note that it's rarely the case that social standards are explicitly, consciously hypocritical. More often, people simply don't notice the conflict between stated and actual standards.

Where I differ from many people is that, in case of a conflict between actual and stated standards of behavior, the correct thing to do is to endorse and formalize the actual standard, rather than trying to enforce the stated standard. This is because the stated standard, by virtue of never having actually been put into practice, is frequently insane if you try to actually practice it.

I don't take it for granted that saying something very beautiful but doing something contradictorily ugly and cynicism-inducing is less insane, nor, if it is necessarily sane, do I take it for granted that sanity is the thing we should be striving for in that case.

A person who is making a significant contribution to your business deserves to be paid appropriately for that contribution, even if she is too naive or unsure of herself to ask for it. ESPECIALLY if you owe her some familial duty.

Procedural knowledge is better copied than taught. But I talk to my kids a lot about causes and effects. Why something happens. What the deeper interrelations between things are. And my kids like it and try to anticipate consequences (how the story goes on). For example, I recently talked with my 10-year-old about matches and how they are ignited by friction with the phosphorus ignition strip. And we talked about what phosphor is and that it is part of the stuff that provides energy to the body (ATP). He already knew that there are little fires burning in the body i.e. that the body does burn oxygen internally - just not how, so he asked about it. And he could immediately anticipate that this would also stop working if the match sticks wouldn't work either - as told in Universal Fire.

Would it be better or worse if someone's takeaway from this post was that no one should reason about what makes a course of action or policy better or worse? That they should just copy other people?

What if copying other people meant burning suspected witches alive? What if some people who burn witches aren't really sure about the correctness of what they're doing, they care about that kind of thing, and yet they profess great certainty that their acts are in accordance with correct values? Should I not try to play to the part of them which is uncertain in order to prevent a cruel outcome, against their initial value statement?

Would you want me to try to give you values which aren't upstream from genocide, if you were born in a place which gave you values that were upstream from genocide?

The topics which are being invoked are way more fraught than is being implied. It's not obvious that you're not sneaking in a takeaway for a general topic by using this one question in a case where the takeaway doesn't generalize across the space of questions in that topic. On good faith, I don't assume you're trying to do that, but it's good to check; lampshade the possibility.

As a historical fact, did people actually burn witches because they were copying someone else's behavior, or because they had some verbal theory for why that is the right thing to do?

These two possibilities are not mutually exclusive; talking is a thing that people do. The correct answer is that it's the latter case (verbal theory) as an instance of the former category of cases (cases where people copy the behavior of others, such as fashions of thinking and talking).

I'm also not very sure that removing the ability to negotiate theories of objectivity or fairness, which are naturally controversial subjects, would make people more peaceful on average given it as a limiting condition on the deveopment of culture starting with the first appearance of any human communication; I expect it would make world histories more violent on average to remove such an ability.