Types of knowing

There’s a really important difference in kind between knowing that the capitol city of France is Paris and knowing how to ride a bicycle.

The “knowing that” kind of knowledge is sometimes called declarative knowledge. It’s knowledge that you can call up consciously to mind and talk about—knowledge you can declare. It might not be easy to talk about; I would say that a physicist’s model of gravitation under general relativity is declarative, but it might take them a long time to explain it. But you can’t separate this kind of knowing about gravitation from the ability to consciously reflect on and talk about it: I think it’d be fair to say that a physicist with total amnesia doesn’t know general relativity anymore.

I’ll tag the “knowing how” kind of knowledge as “procedural“. What it means to have procedural knowledge is that you can engage in the procedure of using it. E.g., what it means to know how to ride a bike is that when you try to ride a bike, you succeed. I’ll use the word “skill” as a synonym for “procedural knowledge”.

I’m now going to make a string of assertions about these that are unproven but I think are obvious once said.

1. It’s pretty normal for declarative and procedural knowledge to be out of sync.

For instance, most people are actively wrong about how they turn while riding a bike. There’s a common misconception that if you want to turn left, you turn the handlebar to the left. This is exactly backwards: you actually turn the handlebar to the right first.

But you didn’t need to know that to be able to ride a bike. And knowing this probably doesn’t much affect how you’ll cycle in the future. And it probably wouldn’t have helped much for you to know this while learning to ride in the first place.

What mattered, probably, was that you kept getting on the bike until you stopped falling over.

2. Getting them in sync can be very powerful and doesn’t have any obvious downsides (other than opportunity cost).

Another example is the “unbendable arm” in martial arts. I learned this as a matter of “extending ki“: if you let magical life-energy blast out your fingertips, then your arm becomes hard to bend much like it’s hard to bend a hose with water blasting out of it. This is obviously not what’s really happening, but thinking this way often gets people to be able to do it after a few cumulative hours of practice.

But you know what helps better?

Knowing the physics.

Turns out that the unbendable arm is a leverage trick: if you treat the upward pressure on the wrist as a fulcrum and you push your hand down (or rather, raise your elbow a bit), you can redirect that force and the force that’s downward on your elbow into each other. Then you don’t need to be strong relative to how hard your partner is pushing on your elbow; you just need to be strong enough to redirect the forces into each other.

Knowing this, I can teach someone to pretty reliably do the unbendable arm in under ten minutes. No mystical philosophy needed.

This is a case where one can “get on the bike” enough to develop the skill, but knowing how it works can significantly speed up learning and can make the skill more reliable.

In the case of counter-steering a bike, it doesn’t really help in almost any case to know how it works as far as I know. But it certainly doesn’t hurt, and you never know when it might come in handy. For instance, I’m told that people who are learning to ride motorcycles are explicitly taught to counter-steer, which I imagine would be a lot more intuitive to someone who already knew and had experienced that counter-steering is normal.

3. Declarative knowledge is the kind we can easily talk about and share, and so sometimes gets transmitted in place of procedural knowledge.

If an American college were to create a one-credit class on bike-riding, I suspect it’d have a textbook, and lecture time.

Maybe that’s a little unkind. Although I have a lot of unkind things to say about the structure of American colleges….

Because we can talk about declarative knowledge, and can write about it, and can write standardized tests on some of it, there’s a common cultural assumption that this is what knowledge is.

So, sometimes you get these weird effects where, in order to transmit a skill, people will talk about the skill—akin to teaching the physics of riding a bike instead of putting people on bikes.

The same thing happens in a lot of martial arts classes. In most aikido dojos I’ve been to, the instructor will stand in front of everyone and show a move, and then talk for several minutes about nuances of the move. “Now, if he counters like this, then I’d do this other move here, which puts me in a good position because blah blah blah….”

(I’m quite guilty of this too.)

A huge amount of what gets called “education” consists of displaying a bunch of declarative knowledge and expecting students to store that info in a way that can be regurgitated at the right time—even though most of the time, what we’d prefer to transmit, if we knew how, is a set of skills like mathematical reasoning and artistic perception.

We basically know how to transmit skills: you find the analog of putting people on bikes. Unfortunately, it’s a lot easier to talk about doing that than it is to actually do.

4. Getting procedural knowledge from declarative knowledge is a skill set.

In my opinion, one of the most important skills of learning is being able to convert declarative knowledge into the act of getting on a bike, so to speak.

For instance, up above I gesture at how one might learn the “unbendable arm”. But right now that information is (probably) declarative for you: you know about it, but you (probably) don’t know how to do it. If you want the skill, then you will need to grab a partner and try it.

But notice that that, too, is declarative knowledge. At some point you have to actually go do it, rather than just know that doing it is how you’d convert the declarative knowledge into a skill.

In fact, this entire blog post is a display of declarative knowledge. Which is rather the point: if we want procedural knowledge but we have declarative knowledge, we need to recognize that the two are different and try to convert the latter into the former.

That’s how we develop the skill of extracting skills from declarative knowledge: we get on the bike.

5. Getting accurate declarative knowledge of our skills is also a skill set.

I think it’s a good idea to be able to get declarative and procedural knowledge in sync, and sometimes this means changing your declarative knowledge, as in turning a bike and also with the unbendable arm.

This point is less central to the rest of Turbocharging Training, so I’ll leave the question of how to develop this skill as an exercise to the reader.


Most of what I’m going to be talking about in the rest of the Turbocharging Training model is skill-acquisition. The main take-away from this post, in my mind, is that skills are different from facts about skills—but we often tend to miss that fact, so there’s a skill to noticing the difference.


Turbocharging Training

I have a theory of learning. If it’s basically right, then there are ways to significantly accelerate how quickly and well we learn things in general. And anecdotally this seems to work.

Originally my plan in graduate school had been to finish my Ph.D., get a position as a professor, write a bunch of grants for studies to test this learning theory and figure out how to apply it, and then a decade or so later open an alternative high school that would teach based on these ideas. The hope was that we could focus on making students skilled at learning in general, develop their emotional and social intelligence including compassion, and give them the knowledge and skill base they’d need for whatever market could plausibly arise as technology and worldwide culture changed the demands of people entering the work force.

Instead, shortly before finishing my Ph.D. in 2012, CFAR offered me funding and a team right away. I just had to focus more on “rationality” and on training adults.

To which I said “Sure!”

Until quite recently, CFAR didn’t have the resources to test this theory very well, so I’ve been just collecting anecdotes and trying to articulate the theory. Along the way, courtesy of Leah Libresco‘s love of alliteration, the theory got deemed “Turbocharging Training” in 2013.

I’d like to try articulating that theory in more detail now. But caveat emptor: this theory is often intuitively compelling, but we haven’t yet asked nature its opinion on the theory’s truth.

I intend this post to be sort of a table of contents. The theory has a lot of components, so this will take the form of a series of posts. Below is a list of the posts as I currently intend them, roughly in order, though both the content and order might change as I write the series. I’ll try to come back and hyperlink the posts here as I write them.

  1. Types of knowing. It turns out that it’s really hard to talk usefully about skills, and this creates a funny kind of illusion about what learning is.
  2. Familiarization and play. A lot of American stereotypes of what learning looks like is an attempt to skip the majority of the work. Properly understood, the bulk of learning is fun, enriching, and a little bit surprising.
  3. You learn what you practice. People tend to assume that the intent of practice and the result of practice are the same. They usually aren’t, and this matters a lot.
  4. The rule of intensity. In the same way that you can learn to notice what “getting defensive” feels like, you can learn to notice what “learning faster” feels like. This helps you change what you’re doing to increase your learning rate in real time.
  5. Learning by resting. Across all skilled disciplines I know of, people often report coming back from a hiatus more skilled than when they left. I think this is unsurprising and is something we can use intentionally.
  6. The meta-skill. The components of Turbocharging are skills that improve skillfulness. This means they can be used to improve Turbocharging. The self-reference here has really nice practical structure and helps to unify the theory into a discipline.


Why “yin”?

People keep asking me what “yin” is, as I use it. And why I think it’s important.

The really short and maybe unsatisfying answer is “Read this.” I think of yin as more like bicycle-riding than like understanding facts or theories. If I were to describe good bike-riding, I might do best by pointing at people who are trying to ride bikes and saying “That person is doing well because of X, and that other person is doing poorly because of not-X, I think.” That’s what I tried to do with my post on grieving well.

But people who have read that post still ask me about this, which makes sense. A lot of the main ideas of yin are buried or implied in that post rather than stated directly.

So, I’ll try to be pretty direct here, to the extent that I can.

(Trigger warning: references to death, child mortality, and losing faith in religion. Nothing gruesome — but if it doesn’t pull your heartstrings then I haven’t done my job. And some of the things I link to might be harsher.)

Continue reading

Red Queen Races

I want to summarize an idea here that others have already described. It comes up a lot in my thinking about mental structures, so I want a place I can link to that briefly captures the idea. I also think it’s a wonderful example of a well-specified mental mechanism.

In the way I think of it, a Red Queen Race is a situation with these ingredients:

  1. There is an optimizer A that wants to increase some variable X.
  2. Optimizer A can trade some kind of resource R to increase X.
  3. There is an optimizer B that wants to decrease X.
  4. Optimizer B can trade some kind of resource Q to decrease X.
  5. A and B have comparable effects on X. That is, if A increases X by some amount, then it’s within B’s capabilities to decrease X by that same amount, and vice versa.
  6. A and B don’t coordinate.

Once you have this situation, then A will trade all the R that it can, and B will trade all the Q that it can, but the net effect on X will be roughly zero.

Here’s a quick handful of possible examples:

  • Cheetahs & gazelles. Cheetahs eat the slowest gazelles, and the slowest cheetahs starve. So each gazelle, and each cheetah, has incentive to run faster. Because the two groups can’t coordinate, they both end up running as fast as they can, but are in basically the same position they would have been in had they been running half as fast as they can: the same gazelles and cheetahs die. The only difference is that both groups are now expending enormous amounts of energy. (In this case R = Q.)
  • Writhing in agony. As I described recently, it looks like movement and touch can distract the nervous system from the signals that the brain interprets as pain. So if you have a process that wants to decrease the amount of pain reaching the conscious mind, that process would encourage more movement and touch (like swinging a fist or rubbing the hurt area). But pain is a warning system, and it works in part by grabbing the attention of the conscious mind. So, the pain system might “turn up the volume” in order to get the signal through. If this process runs away in a Red Queen Race, you might end up with someone thrashing around as much as they can in order to block the pain signal, and with the pain being unendurable.
  • Waking up to an alarm clock. A friend of mine tells a story of him going to war with his subconscious mind about waking up when his alarm clock goes off. After a few days of setting an alarm, he would wake up way too late and find that his alarm clock had been turned off, with no memory of having done so. He then set it across the room, which worked for about two days until he woke up late again with the alarm clock across the room having been turned off. He even tried locking it in a briefcase that requires a combination — and after a few days he found he had woken up late with the briefcase popped open on the opposite side of the room.

There are lots and lots of examples of this phenomenon. It’s also one of the most prominent examples of not having efficiency in a system that I’m aware of. (Here I’m thinking of efficiency as achieving a goal with minimal input, in analogy to a machine being efficient by doing work (in the physics sense) with minimal entropy increase.)

Sometimes this can be good: technically, an arm-wrestling match has the ingredients for a Red Queen Race, but one of the benefits you can get out of doing this is that people who try their hand at it get stronger over time, and they can use that strength elsewhere.

But most of the times I think to point out a Red Queen Race, it’s because the race is devouring resources for no good net effect.

And I think that by default it happens a lot.

Looking into the Abyss

Sometimes I’ll hear a conversation flirt with acknowledging really terrible things in the world. Someone will bring up how people are starving to death in parts of the world, or how children are sold into slavery and often horribly violated and abused, or some other wicked thing that should not be. And at that point, sometimes, I’ll hear someone say something like this:

“I try not to think about that too much. I’m afraid that if I do, it’ll be too much for me to handle.”

I want to point at this question of what we can or cannot handle. I think this is a really important question. I suspect that what we believe we can endure guides what we can and cannot clearly see; after all, if looking too closely at (say) our own mortality would destroy us, maybe it makes sense not to look. If we want to see everything in our lives clearly, on the other hand, maybe we first need to know we’ll be okay after we see whatever we find.

So, how might we come to endure whatever we find?

Well, in short, I think there’s something wrong with the question. I think we already can endure whatever we find. What I think we need instead is a sort of skill with trusting this fact about ourselves.

I know this is a bold claim. Some horrors do seem to shatter us. People do, in fact, seem to experience psychological trauma and get triggered. If we can already weather any hellish storm, why do we sometimes break?

I think it’s because we misperceive our relationship to horror.

I think it’s because “endurance” might be the wrong word.

I suspect it’s because we try to look away.

Continue reading

The art of grieving well

In this post, I’m going to talk about grief. And sorrow. And the pain of loss.

I imagine this won’t be easy for you, my dear reader. And I wish I could say that I’m sorry for that.

…but I’m not.

I think there’s a skill to seeing horror clearly. And I think we need to learn how to see horror clearly if we want to end it.

This means that in order to point at the skill, I need to also point at real horror, to show how it works.

So, I’m not sorry that I will make you uncomfortable if I succeed at conveying my thoughts here. I imagine I have to.

Instead, I’m sorry that we live in a universe where this is necessary.

Continue reading

Proper posture for mental arts

I sometimes teach the martial art of aikido. The way I was originally taught, you had to learn how to “feel the flow of ki” (basically life energy) through you and from your opponent, and you had to make sure that your movements — both physical and mental — were such that your “ki” would blend with and guide the “ki” of your opponent. Even after I stopped believing in ki, though, there were some core elements of the art that I just couldn’t do, let alone teach, without thinking and talking in terms of ki flow.

A great example of this is the “unbendable arm“. This is a pretty critical thing to get right for most aikido techniques. And it feels really weird. Most people when they first get it think that the person trying to fold their arm isn’t actually pushing because it doesn’t feel like effort to keep their arm straight. Many students (including me once upon a time) end up taking this basic practice as compelling proof that ki is real. Even after I realized that ki wasn’t real, I still had to teach unbendable arm this way because nothing else seemed to work.

…and then I found anatomical resources like Becoming a Supple Leopard.

Continue reading