In June 2021, Zvi posted The Apprentice Thread, soliciting people to offer, or request, mentoring or apprenticeship in virtually any area. Gunnar_Zarncke offered advice on parenting, as the parent of four boys (incidentally, true of my grandmother as well) between the ages of 9 and 17, with the usual suite of observational skills and knowledge that comes with being a regular on this site. I responded with interest as my first child is due in November.

Gunnar and I are sharing our experience as an example of what a successful mentoring process looks like, and because his key points on parenting may be interesting to current and future parents in this community. I had several breakthrough-feeling insights which helped me to connect my LessWrong/rationalist schema to my parenting schema.

Gunnar and I began by exchanging messages about the parameters of what we were getting into. I was interested in his insight based on these messages and other comments and posts he had made on this site about parenting. We arranged a Google Meet video call, which confirmed that our personalities and philosophies were compatible for what we were undertaking.

We did not have a structured reading list, although I investigated resources as Gunnar suggested.  As we went along, Gunnar translated into English samples of notes taken by his children’s mother throughout their childhood and shared them with me. She had also systematically described the daily and weekly tasks a parent could expect in various development phases of the child’s life. I was an only child and have not parented before, so I found this extremely educational.

We had several video calls over the next few months and discussed a wide range of parenting-related topics. Gunnar also suggested this post, to report on our experience. I drafted the post, and Gunnar provided comments, which I merged, and after he reviewed the final version, we published it as a joint post.

By call number two, I was realizing that parenting was never going to be the sort of thing where I could read the “correct” book for the upcoming developmental stage, buy the “correct” tools, and thereby maximize outcomes. Instead, it would be a constant process of modeling the child’s mind, providing new inputs, observing behaviors, updating the model as needed, researching helpful tools, and iterating more or less until the kid is in its 20s. At first, this was intimidating, but I’ve come around to understanding that this just is the parenting process. This synthesis eventually gave me additional motivation and optimism. 

These calls gave me great comfort against anxiety about parenting, confidence, and a sense of human connection, all beyond what I expected. 

First call

Our first call was within a week of Zvi’s post. We described our backgrounds as people who were parented. Gunnar came from a large family; I came from a small one. We discussed how our parents nurtured positive traits in us and also touched on what our parents did that didn’t work. 

For example, my parents would frequently observe when other people were acting in ways consistent with the values they were trying to teach me, in addition to praising or otherwise rewarding me for acting that way myself. 

Gunnar's mother was mostly trusting of her children and "went with the flow," following her intuitions. His father was very fostering and offered a lot of practical education. He consciously created a safe environment. He said he learned this approach from his parents, who came from different backgrounds. Gunnar's grandmother came from a liberal Scandinavian family, and his grandfather came from disciplined Prussian family. His grandfather embraced his grandmother's liberal norms, which seems to have created a reliable high-trust environment for his father--despite difficult times during and after World War II. 

Gunnar segued into discussing general strategies for supporting children’s development. Highlights:

  • “Salami tactics”: Allow them to learn new behaviors and situations incrementally rather than all at once.
  • Developmental diary: Once a week, or more often, write down notes on what happened with each child during that period, what was effective parentingwise, what wasn’t. This was something that Gunnar came back to consistently. However, he is not confident that it is right for everyone, just that it was for him. I plan to do this as well.
    • Consistent reflection
    • Lessons to carry from one child to the next
    • Incorporate photographs
  • The saying goes, "Small kids, small problems; big kids, big problems." But the pattern goes like this:
    • With small kids you have a lot of very small tasks and problems: How to diaper. Why is the baby crying right now? Let's try this 5-minute game. Let's go to this 30-minute baby swimming class. We have to rock the crying baby for an hour until it finally sleeps. Oh, the baby is interested in this thing--oh, it's already gone. Why does X no longer work? Oh, Y works now.
    • As they grow older this switches to: Will they find friends at the new school? Taking the kid to soccer games every weekend--and staying there for cheering, photos, and small talk. Practicing math for hours before the exam. Working again and again on some fight between siblings. Helping to renovate the room.  Talking for hours about some conflict or problem.
  • Be alert to opportunities for teaching based on the child’s interests.
    • Model the behavior of conceiving and running experiments
    • Organize activities around projects (for example, in the garden)

Gunnar recommended several texts during this call and in a follow-up email, including:

We covered many topics in later calls, organized below by subject rather than chronology.

Child Cognition in General

We discussed more cognitive elements of parenting--the extent to which developing brains “need” new inputs and partially “know” what inputs they need but, if overloaded, will retreat to the familiar, and especially to you, and then consolidate.  Gunnar mentioned the Big Five as a good shorthand for observing kids’ personalities.  He shared the first of the translated parenting documents I mentioned above.

This discussion reminded me of Clark, Surfing Uncertainty, which I cannot recommend strongly enough.  After reading that book, I understood intellectually that brains seemed to be prediction/testing machines that thrive on stimulation, but I didn’t see that model as a frame to place over my parenting thoughts until Gunnar spoke about similar concepts in his own perception of parenting. This was a eureka moment for me.

Your kids spend even more cognitive energy on you than you do on them, because their survival depends on it (see also here). 

  • They will notice if you are stressed or worried.
  • They understand words you’re using before they can use those words themselves.


We discussed various ways to teach children before they are in school, and to augment what they learn in school.

  • Use homeschooling materials to assist them with their homework. (In Gunnar’s country, homeschooling is very rare; in mine, the USA, it’s a constitutionally protected right, consistent with Gunnar’s claim that the best homeschooling materials are in English.) I might never have considered these otherwise, because homeschooling in the USA is correlated with weird beliefs, and I was subconsciously assuming that homeschooling materials generated by weird-belief-holders would be somehow infected by the weird beliefs.  (Gunnar adds: They likely are infected by weird beliefs, but you can just keep the good parts.)
  • Avoid rote memorization, except where necessary--multiplication tables, for example.
  • Parents’ and teachers’ incentives are often misaligned (ideal methods for an entire room versus ideal methods for your own child).
  • Encourage kids to make testable predictions and bets.
  • At all verbal ages, you can talk to them in a more complex way than they are able to communicate, yet they will still understand some parts of it and absorb context and parts of meaning..


Conditioning works, but only on things you are consistent about. Corollary: if you’re not willing to be consistent on something, leave it out. (My parents used this on me when I learned how to whine. They agreed not to acknowledge anything I said in a whiny tone, and told me this would be their policy. According to them, it worked quickly.)

When the desired behavior is rare on its own, you can “cheat” by simulating the behavior (for example, in pretend play).

Rather than “No,” use “Yes, but” “yes, and” “yes, as soon as”. These are opportunities to show the child that you are also a person with needs, and to emphasize mutual responsibility.


  • Trust your instincts, yourself, your spouse, and offer the child a lot of trust.
  • The parent should behave such that the child unconditionally trusts that its needs will be met reliably.
  • Don’t lie to your kids.
  • Challenge them, but not so far that they feel physically unsafe.


Parenting is intense and challenging, not least when you are sleep deprived because of a baby’s sleep schedule. Observations:

  • Cultivate a support network of friends and other parents.
    • There is probably no substitute for in-person connection and support.
    • Talking helps.
  • Have a safe place to temporarily retreat to.
  • Consistently (perhaps a certain interval each day) set aside time for unstructured entire-family time.
  • The change in the marriage relationship requires focus and time to navigate.
  • Communicate feelings and stress with your spouse and provide physical support as needed.

Avenue for Neuroscience Research

Gunnar has an interesting, and possibly testable, hypothesis: One effect of puberty is to partially reset the values a child assigns to normative judgments, but not to procedural knowledge about reality. (Corollary: Whatever values you’ve taught your child will be more likely to survive if you’ve given them the information necessary to conclude that the value is correct.)  The cascade of puberty hormones could conceivably affect the chemicals in the brain responsible for adjusting weights of priors.  I don’t have enough neuroscience to develop this any further, but it’s “common knowledge” that many teenagers think their parents are idiots.  A biological explanation would explain how widespread the behaviors leading to this folk belief are.


I am grateful to Gunnar for his time, attention, and “gameness.” I am glad that this entire process happened, starting with Zvi’s initial post and ending with this post. I feel far more prepared than I did at the beginning, and I doubt that a person outside this community would have been able to get me there. I plan to implement is weekly development diary as a way to track trends, organize my own thoughts about parenting, and force myself to really think about what's going on.  Maybe most importantly, I have a role model for thinking hard about what's going on even with very young children. My only model for that before was cognitive scientists and their informative but ultimately clinical experiments.

I’ll give Gunnar the last word:

I enjoyed the mentoring tremendously. It is very rare to find someone so interested in parenting and taking the preparation so seriously. I felt myself and my advice highly valued. A good feeling that I hope many mentors share. Talking about my parenting experiences and insights also sharpened them and gave me more clarity about some of my thoughts on parenting. I highly appreciate all the note-taking that was done by Supposedlyfun.

One thing that I realized is how crowded the parent education market is and how difficult it is to find unbiased evidence-based material. I have been thinking quite a lot about this and hope to post about it sometime.

We have paused the mentoring for the time being and I am looking forward to how the advice works out in practice. We agreed on a call sometime after the family has adjusted to the new human being.

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

This is a Google Doc that I created based on the child development planning of my ex-wife. It contains an estimation of the effort (hours spent) for the first 8 years. 

And this Google Sheet contains activities and development program steps we developed. This contains what supposedlyfun called our Salami Tactics. 

Both are incomplete and just what I included during the preparation of our calls. I intend to add to it later. 

Thanks - I think the Sheet has restricted access.

Fixed, thanks.

Don't make the mistake that I made when I was pregnant. I spent too much time planning parenting strategies years down the line and not enough time researching breastfeeding, and I ended up struggling with a tongue tied baby who would not latch.

Before the baby arrives make sure that, as a minimum, you know the following:

  • The side-lying, laid-back, koala hold, cradle hold, cross cradle hold, dangle and clutch hold nursing positions.
  • the flipple technique
  • how to cup feed a baby
  • how to use nipple shields
  • how to use a breast pump
  • how to do breast compressions
  • how many wet nappies a baby should have per day
  • how wet a nappy has to be to qualify as wet
  • that you should never hold the back of a baby's head while nursing even if the midwife tells you to.

The best online resources are La Leche League, Kellymom and Analytical Armadillo.

Thank you for adding all these recommendations on breastfeeding. I can confirm that difficulties with breastfeeding cause royal pain. I don't see a way to pass on this knowledge via remote mentoring (plus, neither of us does breastfeeding). Maybe we should more clearly point out that this is a father's position, and of course, your mileage may vary. 

A note on methods.

Google Meet was adequate for our video calls.  I used my phone as a webcam, with a Bluetooth microphone and in-ear speaker. Gunnar had a similar setup, a headset with a boom mic. I had no trouble understanding him and vice versa so far as I know. The video and audio quality on Google Meet was more than enough for our purposes. We had very little in the way of lag, despite being eight time zones apart.

If I had it to do over, I would get a quality webcam to connect to my tower PC, as having Gunnar’s face on a full-size monitor instead of my phone screen would have been worth the money and effort.

I took extensive notes using Google Docs on the substance of what we discussed. (If I don’t do this during phone calls, I forget 80% of what was said, seemingly without regard for importance.) I shared my notes with Gunnar after the first call, and he commented, “wow, you took a lot of notes”.

Very interesting report. Some remarks:

  1. About literature: The most interesting book to me was "Babyjahre" by Remo Largo, a pediatrist who led the Zurich longitudinal child study and basically just explains what happens empirically in different areas of development in different ages. The book is about the ages 0-4. I assume there must be something similar in English. There's a kind of book that's a real nuisance, namely those that explain developmental stages and behavior like a horoscope: It's very vague but seemingly concrete and people tend to interpret things until it fits.
  2. Polgar: You can find Scott Alexander's articles about the Polgar family approach here and here.
  3. If your child is sick, Googling may have a different effect than it does in case sick adults: In the latter case, people tend to believe they have something deadly afterwards, while in the former you often just learn that certain "symptoms" are just normal for the age.
  4. Stated like state, "avoid rote memorization" is not really just parenting advicing, but a statement about learning. It depends on method, context and detail whether that is sensible - for example, flashcard learning seems to work quite well.
  5. "At all verbal ages, you can talk to them in a more complex way than they are able to communicate, yet they will still understand some parts of it and absorb context and parts of meaning.." Sure, but the art is to calibrate how complex your communication should be. If I read a book to learn physics and I take one aimed at PhD students, I may also be able to understand part of it, but a lot of it is just wasted time.
  6. "I don’t have enough neuroscience to develop this any further, but it’s “common knowledge” that many teenagers think their parents are idiots."  It's surprising that the English wikipedia article on puberty does not comment on behavioral development, or maybe I overlook it, and the German one only has a short paragraph lacking sources, translated by google as "During puberty, both parents and teenagers report that they no longer feel so close to each other. One reason for the problems that arise between child and parents can be the child's improved judgment, which tends to question and criticize the behavior of the parents. In addition, the roles of young people in their lives change as they mature physically and they want to be treated as adults accordingly. The adolescents also want to take on more responsibility for their leisure time. However, the parents want to protect their children from harm and thus adopt a counter-position. Most of these arguments are only superficial and do not endanger protective family ties. The young people also quarrel with their siblings more often. Puberty is often referred to as the “second phase of defiance” because of the quarrels. "

Sorry to be a buzzkill, but what are you trying to achieve here?

It is my impression from the literature that once controlled for genetic confounders the long-term effect of parents on cognition [and a host of other factors] is 0.  Why spent so much effort if the net effect is nil?

I have spent a lot of thought on this, and I am well aware of all the studies posted by Gwern over time that show the effect of parenting on cognitive ability, educational attainment, etc. to be small or zero. I buy that your influence on intelligence and character is neglible. Still, it wouldn't be adaptive for the child's brain to throw away useful knowledge (though it might question its valence). There are also a lot of things that can make the parent-child relationship easier. Scott Alexander wrote that there seem to be shared-environment effects influenced by parents that are beneficial in other ways (I can't find the exact quote, but this is close). A more fruitful parent-child relationship - esp. with multiple children - seems to fall into this category. You can also see it as being about trust. And it is also possible that there are effects at the tails of the distribution that are hard to find with usual study designs. 

Consider all those teenaged Olympic medallists. It wasn't just their genes that got them there. It takes opportunity as well. This is what parents can provide. And yes, the parents have the genes as well, but all that means is that they are in a position to provide those opportunities. Whether they do or not is their choice. These are the choices that distinguish good parenting from bad.

Think of parenting as being the primary stage of a rocket.

It is easier to break a system than to improve it. With parenting, this may mean that if you listen to advice like "parenting has no effect, you can skip it," then you may stop doing what you as a parent usually would (i.e., foster and take care of the kid), and end up with worse results.

I think it's often worth making a comparison to "what am I trying to achieve in my relationship with my partner? With another adult, you're not trying to mold them into a better human.  You're trying to enjoy the time you have together now, and next month, and over the decades.

There's a lot of parenting that makes a difference to how much you both enjoy the next 18 years. Like the "teaching them not to whine" thing - you will both be happier if they have other ways of getting their needs met.

This is only tangential, but "does being friend with them pushes me into becoming a better human" is a pretty good question to ask about your friends. And if the answer is a resounding"no", considering not being friend with them is a good next step.

I'm deeply confused by this question. I think we are not coming from the same factual background. Do you have some literature reviews or meta-analyses showing that parental input (other than genes) has no effect on cognition, and what the "host of other factors" are? 

Not to hide the ball--if the list of other factors isn't "literally everything that matters," I'm going to claim that the other factors about which we have no evidence are worth the effort, and if it is "literally everything that matters," I'm going to look at the tails of the distributions and, given that I chose to create this eventual conscious entity, conclude that I have a duty to chase tails above a certain low probability threshold.

I haven't gone through the literature you mention, so I may be off, but I am puzzled by your comment. It seems obvious to me that the effect of parents on very important factors is clear.

Just, for example, in what kind of job or studies the children choose. On average, children are much more likely to have jobs similar to those of the people they know well, just because they have much more inside knowledge. Children of parents with PhD are much more likely to get a PhD, etc. If parents influence this, why would they not have influence in other factors?

Maybe the studies only account for cognitive factors such as IQ and not in shaping the children decisions?

Still, it seems also clear that parents can actively worsen those (induce anxiety, fear, traumatise...), which would already be a strong incentive to learn and try to educate children as well as possible. Are such effects not long-term? Anecdotal evidence suggests me the contrary: one has to actively work to heal from it, meaning that the effect is long-lasting. And if having a bad influence on one's children is possible, why should it not be possible to have a good one? Again, anecdotal evidence seems to show that some families are actually better than others helping their children develop (of course, this could be entirely a genetic thing).

Other ideas that come to mind are income and orphanages. Income seems to have a large influence in IQ, why should then parenting not have it? And people who used to live in orphanages are said to often have cognitive problems (I have no idea, just repeating "commou knowledge" for this one).

Am I writing about effects not accounted for in those studies?

well-done rigorous studies (i.e. those done with twins and/or adoption studies) mostly find that almost every aspect of offspring traits has a large genetic component and almost zero shared environment component. 

These traits range from cognitive (like IQ) to degree of spirituality to traits like anxiety (which are harder to measure, hence the correlation coefficients can occasionally be lower due to measurement error). Indeed, it is hard to find traits that do not have a large genetic component! (interestingly homosexuality often regarded as the genetic trait par excellence has one of the lowest genetic components of common traits)

Yes children whose biological parents have a PhD are much more likely to have a PhD. How does this prove that parenting style is the mediating force as opposed to genes?

Severe abuse can certainly hurt children - it is easier to fuck something up than to significantly improve on a natural process. But the abuse has to be quite severe to have a measurable impact on long-term life outcomes - although of course it will still be an awful experience and memory for the child (and obviously I'm not saying you should abuse your child!). 

There is indeed a fad to ascribe all kinds of maladies to this or that happening in childhood. Actually, this idea is quite age old (Freud?). Doesn't make it true though. Well-done studies that control for genetic components as opposed to shoddy studies or anecdotes mostly don't pick up on any of the effects that these theories claim (outside of severe abuse). Certainly, it seems plausible on the face of it and it is undeniably popular - but the lack of strong scientific evidence for these theories should make on skeptical. 

Differences between the First and Third world are so large that these have very large effect on life trajectories. 

If one believes the actual research done on this question then one can only conclude that the differences between middle-class parenting styles in the First World outside of severe abuse don't make much difference in long-term outcomes for the child.

Thanks you very much! That's a great answer!

To be fair, it had never occurred to me that completing a PhD could be genetically driven... It seems quite plausible, actually (however, not completing one seems much less correlated to me).


Yes, I'm more skeptical now! Still, I'm also skeptical of these studies, though (maybe because I have not reviewed them myself --nothing persona!--, which I actually doubt I will do...). 

AFAIK, there were very well know studies strongly relating IQ to genes (which implied that e.g. black people have lower IQ) that are now being refuted and the new ones link it much more to the socio-economic status of the family. I guess the difference is only the degree of correlation (you say it yourself that socio-economic status has an impact), but what I heard/read is that difference is quite large.

In addition, I can not square these studies you mention with other good research. E.g. gratitude journaling seem to have a huge effect improving mental health, and this is something that can be easily taught at home.

My guess to put everything together is now that the world is very messy, such effects very difficult to measure, and that most education is good enough to avoid trauma but still very sub-optimal (difficult to find the signal within lots of noise). In addition, probably, educating techniques work differently for different children (which would not be surprising and would largely explain that education is sub-optimal).


I'm interested in your thoughts :-)

Your mental flexibility and willingness to chance your mind is commendable. 

I recommend taking a look at Plomin's "Blueprint".

even on IQ short-term gains can be made by intensive training. However, it effects seem to always wash out in the long-term. 

Thanks, I'll take a look.

Possibly they're trying to beat the odds?

To some degree, sure.