[sofayam]

This guy was not really trying to explain to hacker parents how they should teach their kids to ride a bike. As has has been adequately demonstrated in the comments they already know aaaaaaaall about that. His actual point, which seems to have whooshed past most people’s heads, is much more interesting: can you learn a thing more effectively by first simplifying that thing so radically that a seasoned user would find it useless? Also not exactly a totally new idea but, depending on context, just counterintuitive enough that you may miss it.

[adamcharnock]

I would love to see this approach in language learning, which I am fairly bad at. I'm very much driven by results rather than accuracy, and so often a teacher will correct me on the finer points of the language I'm using and I'll either no idea what they are talking about, or know that I am certainly not going to remember that detail. In either case I find it very demoralising.

If there was a "Learn to speak German like a 5 year old" course then I would love that. Give me something usable to motivate me further, then I can come back for more complexity when I actually want and need it.

But isn't this the case for all language courses? They start you slow and build up? I feel like it isn't, although perhaps it is just the courses I have seen. It seems to me that the people who teach languages generally really like languages, so they understandably revel in the details. I, however, do not (although I wish I did).

I realise now this is a bit of a rant. Apologies!

[cjohnson318]

I learned Spanish at a two year old level by focusing only on the top 24 verbs, in all (major) tenses first. So, ir, ser, estar, tener, haber, hacer, poner, poder, venir, ver, decir, dar, etc., in present, preterite, imperfect, future, conditional, subjunctive, progressives, and perfects.

Then I learned about 2000 nouns from several lists. Then I learned some common "weird" constructions like darse cuenta, and hace falta.

Now I can read paperbacks on the plane without a dictionary, and follow the plot. I can communicate simply, and send pretty good texts. I have a lot of trouble with TV and radio, but it's progress. This took 2 years with mainly self study and Duolingo for accountability. I don't think an app alone would do the trick.

[Alex-Programs]

> Learn to speak German like a 5 year old

This is pretty much the methodology behind "comprehensible input", where you consume lots of content that you can just about understand and "let your brain figure it out".

There's quite a lot along these lines. LingQ helps you learn as you read books, and I built https://nuenki.app, which gives you constant comprehensible input as you browse the web.

I also really like Language Transfer, which isn't really a comprehensible input course, but tries to draw parallels to English and talks through the etymology a little. The approach appeals to me.

[kjellsbells]

There are resources available for children, eg Muzzy is an animated immersion program for 5 year olds that (if you can bear the cheesy animation) might work. (In the US, available on Kanopy with a public library card.) And if you are good at searching you might also find resources that educators use for teaching recently immigrated children, although that implicitly requires a teacher/partner to be working with you.

Everyone learns languages in a different way. There are some people who like to be told what the basic rules of the language are and can use that to structure new sentences. Like giving someone K&R I suppose. Other people need to hear it. Personally, because I am only learning a language for practical purposes like travel, I'd love a course that dispensed with the grammar and taught contemporary phrases used in everyday life. For example, I am never going to ask and be told where the library is. But I'm very likely to hear, "cash or card?" or to ask "does this train go to Bologna?". So practicality for me wins early on, and then later I'd like to learn the top 500 words, and then the grammar structures.

[andrewla]

I have a theory about language acquisition that I've never had the time to fully explore, that developing an ear for a language is the critical first step.

To that end, my theory would be that a program of imitation & mimicry would be the most effective way to learn. That you would hear a native speaker say a phrase and attempt to fluently imitate it. Specifically, record your voice as you speak and listen to what you say and try to as perfectly as possible imitate the prototype phrase.

Learn vocabulary and grammar later; focus, like children do, on hearing the language and imitating its use. Learn reading and writing last of all; formal grammar and especially spelling are the pedals on the bike.

[DontchaKnowit]

Checkout Learncraft Spanish. The step they take to simplify that others do not is that they focus knly on grammer for a very lokg time, and just use english verbs and nouns.

For example, at a certain point the only words you will have learned are que and lo, so the quiz sentances will be like :

I want you to eat it -> I want que lo you eat

This prpgram has been extremely useful to me and helped me learn spanosh far quicker than other methods. They also use memory palace techniques and have an unusually effective way of organizing vocab learning

[larsiusprime]

Author of the piece here, I speak three languages (Norwegian, English, and Spanish) and am currently working on a fourth (Japanese).

"Taking the pedals off the bike" advice for language learning:

- Learn pronunciation first

A lot of people never master a native (or semi-native) accent, but if you sit down and figure it out, it's easier than it seems at first. Getting the native pronunciation down matters because otherwise you literally can't distinguish between certain words, and it will be a habit you'll never unlearn. It makes everything else easier. It also gives you massive cred with native speakers who will overlook your atrocious grammar and paltry vocab because "wow, you sound good! I'm impressed." It's taken both as a sign of respect and that you're putting in the effort, and makes you punch above your actual weight. This does wonders for confidence and makes you less shy about trying to learn by speaking & listening, which is crucial.

Gabriel Wyner does a good job explaining it in his book "Fluent Forever" (his method is pretty cool too but I have some critiques of it overall, "learn pronunciation first" is the best single lesson to transfer to other language learning methodologies).

[ericrallen]

If you haven’t checked out the Pimsleur[0] app, you might find it useful for learning a language.

It’s the only language learning system that has ever worked for me. It focuses on speaking and every day language rather than reading, writing, and memorizing vocabulary, conjugations, etc.

It isn’t cheap, but it’s designed for you to learn enough to not need it anymore.

It probably doesn’t work for everyone, but it did feel like a different approach than many of the other language apps I tried in the past.

[0]: https://www.pimsleur.com/

[anarchonurzox]

That sounds rather like the way the Pimsleur approach teaches. It drills fundamentals of grammar through fairly basic "travel vocabulary," but once you have that foundation you can go pretty far.

[bntyhntr]

If you read reviews for kids foreign-language language books on Amazon, you'll see a fair amount of adults reading it for themselves mixed in. That's a little more self-directed but the vocab and sentence structure is organically restrained and the books are fun! caveat: I've only read two books in this manner incidentally, but I knew some people who did this kind of thing on the side during our college language classes.

[ted_bunny]

I would think such a course might ingrain bad habits. The case system sounds strange when used incorrectly, and it'd be much harder to re-learn it if the wrong version started to become familiar to you.

[globular-toast]

It's funny really. I assumed by the context of being posted here on HN that this wasn't literally about teaching children to ride a bike.

I'm reminded of the approach taken by the book Structure and Interpretation of Computer Programs (SICP): start with this magical thing called Scheme and learn simple programming techniques and general principles like abstraction, then gradually add the "pedals" back until you've basically learnt to program in assembly and write a compiler for your high-level language.

[nagonago]

I agree, but to be fair I think that point could have been made clearer in the post.

A similar but related lesson: the best way to teach something is to design a task that is just difficult enough that the learner can figure it out on their own.

When I was reading parenting books in preparation for my own kids, this is one consistent theme that kept coming up, sometimes called "scaffolding." The idea is that you provide a safe environment, design a task that is just the right level of difficulty, then let the child figure it out themselves. (For example, rather than directly holding a kid climbing up a ladder, let them climb it by themselves while you stand by to catch them just in case.) As a result, they develop more independence, self-confidence, and the lessons stick.

"Every time we teach a child something, we keep him from inventing it himself. On the other hand that which we allow him to discover by himself will remain with him visibly for the rest of his life." -- Jean Piaget

[d1sxeyes]

Piaget is well known in teaching circles as a philosophical father of pedagogy. A (slightly) less know pedagogist is Vygotsky, who invented the term “Zone of Proximal Development”. The idea is that kids can learn from others and from experimentation if you can design activities where you take the skills they currently have, consider the skill you want them to acquire, and build steps between them that a child can succeed in. To develop this example: once a child can walk, they can learn to balance by being given a task which allows them to safely experiment with falling over and staying upright. Once they can balance, you can experiment with moving while balancing. Once they can move forward with balancing, they can learn stopping safely. Finally, they should be ready to learn how to pedal.

If you don’t allow them to complete all the previous steps, they may just keep failing at the next task, because they’re not yet in the “zone” to be able to acquire the next skill.

If a child can’t balance annd move forwards unaided, they won’t be able do the next thing (pedalling) even with help.

Children have different skills and capabilities and Vygotsky is not prescriptive about who needs to help, and the ZPD theory often encourages learning from peers rather than adults (parents/teachers).

[nmeofthestate]

Folk sometimes can't see the Linkedin post for the trees.

[Maxion]

Don’t Miss the Post for the Trees – Here’s Why Most People Do

You see it all the time. A post comes through your feed. It's insightful. It’s bold. It’s… mostly ignored.

Why? Because people don’t actually engage with the core idea — they react to what they think it says.

The same thing happens in business: - Founders get stuck in the weeds of their product without seeing the bigger market opportunity. - Teams hyper-focus on the tech, missing the customer pain point. - Investors hear the pitch, but miss the deeper vision driving it.

People miss the post for the trees.

Here’s the thing: breakthroughs happen when you push past surface-level reactions. The best founders? They’re not just building products — they’re connecting dots others miss. The best marketers? They’re not just optimizing campaigns — they’re shifting narratives. The best investors? They’re not chasing trends — they’re seeing past the noise.

If you want to stand out in a noisy world, here’s the question to ask yourself: Am I reacting to the surface? Or am I leaning in to understand what’s really being said?

The magic is always in the nuance. The signal is often buried in the noise. The big ideas? They’re the ones that most people scroll past.

So… don’t.

(I'm sorry I had to)

[throwway120385]

I've been told that it's easier to learn to ski without poles by an experienced instructor. Apparently his pupils would all spend their time trying to figure out these new "pole" things and very little time actually learning how to balance if he started there. I also think it's probably better to hike without trekking poles even on difficult terrain until you master balance. They're more for supporting your upper body if you have a 75+ liter pack full of gear than for keeping you from eating shit if you step on an unsupported rock.

[marklubi]

I've taught several snowboarders how to ride and the teaching process is about putting them in a position where they forget about the distracting things.

Always start them on a blue (or blue-black) slope, because it forces them to learn to use the edge of the board. Lot of easy drills like side slipping and simply turning your head to control direction.

If you start a snow boarder on a green run, it always results in them catching an edge and eventually face planting. Not a fun experience.

[keeganpoppen]

that might be a good analogy because, generally speaking, skiing is much, much harder without poles (hence an expert would never do it (unless they are hitting an olympic vert ramp or something)), but it is exactly for this reason that the poles distract you from learning the "right" lesson-- people use their poles pretty much randomly at the start, and the poles help them... to do the wrong thing. but once you have the real crux of skiing down-- body position and balance + using your edges and weight shifts to turn-- poles are completely trivial to add.

great analogy!

[analog31]

When my kids learned to ski, it was without poles. The poles are an encumbrance, not only on the slope, but getting onto the tow rope.

Likewise, music. The most popular method -- Suzuki -- starts kids out without sheet music. Reading comes later.

In both cases, it's also just less gear to manage. The benefit becomes obvious in group lessons and recitals.

[tarentel]

I took skiing lessons a couple years ago. This is how it was done. I can't say it was any easier as I had never skied before and I never will again.

[netghost]

Another take I've had is that we were wrong in the past. Pedaling isn't the tricky part of riding a bike. Balancing is. Training wheels let you learn to pedal.

[conor-]

Maybe that's part of the reason why "strider" bikes are becoming a lot more common for toddlers to learn how to ride a bike. They're effectively what the OP describes, a bike with no pedals that you run and then balance on.

[dr_dshiv]

I love riding bikes with no pedals. You do feel like a dandy.

https://en.wikipedia.org/wiki/Dandy_horse

Also, is it shocking to anyone that pedal+chain bicycles were invented in 1885, less than 20 years before powered flight?

https://en.wikipedia.org/wiki/Bicycle

[01HNNWZ0MV43FF]

Sort of a sit-down, off-road push scooter!

[wrsh07]

There's a great (underrated) video on YouTube about training a neural network to manage a double pendulum

It changes gravity! This is sort of like managing a double pendulum under water^ and then trying it in the air

The video: https://youtu.be/9gQQAO4I1Ck?si=Hs_3GxrZgmhr2xSn

^ haha not sure this would be easier

[jodrellblank]

That is an interesting video! Your writing made me expect the neural network had changed gravity as hacky part of its solution, but it's the creator trying to train a neural network to balance a simulated double pendulum and failed, and then wondered if he could train a simpler version in a simulation with no gravity and high friction to make it easy, then see if the solution can gradually be retrained to cope as he adjusts those to full gravity and normal friction. The visualisations in it are very good.

[photochemsyn]

Made perfect sense when I read it, because this is also how to teach snowboarding. A reasonably proficient snowboarder will think 'it all about going back and forth from the front edge to the back edge smoothly' but a beginner needs to learn how to first stand in balance over the board without falling on the face or their back.

[HenryBemis]

Balance bikes. The dude discovered balanced bikes.

[loglog]

Which were also the historical precursor of bicycles: https://en.wikipedia.org/wiki/History_of_the_bicycle#19th_ce...

[smohare]

I’d like to say, “Is this not obvious?” but I’m not sure I fully appreciated what follows until I had a sufficient level of exposure to abstract mathematics. A core element of problem solving is reduction, be it to logically equivalent formulations or to more atomic components that can be reasoned about together.