Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio.
This is: parenting rules, published by Dave Orr on the AI Alignment Forum.
(crossposted from my nascent substack)
Way back in 2012 I wrote up on livejournal (I told you this was a long time ago) a few parenting rules we lived by. This is the one livejournal post I regularly reshare, so here it is on a more modern platform. Our kids are older now (11 and 13) but with one exception I think these really hold up.
That one exception is praise, where the research on praise seemed clear in 2012 and has since largely failed to replicate and certainly doesn’t have the effect size that everyone thought, so that one I no longer stand by.
Here they are:
Try never to lie. If kids ask a question and they aren't ready to hear the answer, just tell them that. This doesn't mean you have to go into every gruesome detail, it's fine to couch your answer at the level you think they'll understand and that you have time for, but they're smarter than you probably think.
This does extend to things like Santa Claus and the Easter Bunny. We've told them those stories with the attempt to treat them just like any other fictional story. When Jackson point blank asked if Santa was real, I told him, "No, but it's a fun story and fun to pretend."
There's a common pattern with kids to tell them things that are untrue but scary as a joke, like "Be careful not to slip down the drain!" Don't do that. Kids have trouble distinguishing fake warnings from real ones.
However, saying untrue things as a joke is fine in the right context. "Elephant toes" is a fine answer to a question about what's for dinner. (As long as it's not true.) People say untrue things all the time, and taking the time to evaluate whether an adult is telling the truth is a useful skill. But until the kids are good at it, the untruths should be completely implausible, then can get more plausible as they get more on to you. Fun game, actually.
The most difficult time for this one is when they want something that you don't want to give them. Like if mommy is downstairs and I'm doing bedtime, it's very tempting to claim that Katy is busy doing sometime important that can't be interrupted rather than just admitting she needs a break, or it's my turn to answer the late night call.
Remember that every interaction is a repeated game, and your goal is not to win this one iteration, but to win the series. So if a child is crying because she wants something, even though it feels like a win to give in now (she stops crying which is better for everyone, you haven't really given up much), it's disastrous in the repeated game because she learns that she can get what she wants by crying.
The flipside of that is that you have to let them get what they want in other ways. If you say no and they have good reasons why you should give in, or even an attempt at good reasons, sometimes you have to give in. You want them to be thinking critically and trying to persuade you.
Here's an example. Katy put down a couple of dollars on the counter, which Jackson took, leading to the following conversation:
Katy: Jackson, please leave those there.
Jackson: But this one is mine.
Katy: No it's not, I just put it there.
Jackson: It looks just like the one I got last week!
Katy: It's not the same one, I just put it there like 30 seconds ago!
Jackson: But money is spongeable.
Katy: ...
Katy: Ok, you can have it.
Because money being fungible is a great reason, even if it's not completely persuasive in this particular instance, and "spongeable" is awesome. If he'd started crying, the answer would have been a much more solid, no-more-negotiation "no."
Almost never bluff. This is related to the first two points, but is really more like the second. If you threaten a consequence and don't follow through, they'll figure that out really quickly. Which leads to the following rule: be very ...
view more