Multiple readers have noted that this is very similar to a method called “Thinking At the Edge”, taught by Eugene Gendlin. I don’t know much about Thinking At the Edge, but what little I do know suggests that he and I have developed his focusing work in similar directions, perhaps for similar reasons. If you like this essay, you might want to check out Gendlin’s introduction to Thinking At the Edge.


There are a lot of different ways to dance. 

In one kind of dancing, you move your body around spontaneously in response to music. That’s a fun dance that even infants can do. But some dances have a more specific goal than “fun”. Partner dances enable conversation through touch and motion rather than symbolic language. Folk dances encourage participation in the soul of a community. Ballet pairs choreography with music to tell a story. If you’ve ever tried any of these specialized dances, you know that they take deliberate effort to learn.

Similarly, there are a lot of different ways to engage with the world in a spirit of curiosity. 

“Naturalism” (an allusion to 19th century naturalists) is the name I use for a specific way of engaging curiously with the world. It’s a method of inquiry that uses patient observation and original seeing to build models that were previously unthinkable. It takes longer to learn than the spontaneous curiosity of pure exploration, but this specialized method helps you to make deliberate progress in areas where your existing concepts are leading you astray.

It takes much more than one essay to talk about naturalism as a whole, so this essay is just about the very first piece: catching the spark.

Imagine two pieces of flint colliding with each other and sending sparks flying. By default, the sparks will go dark and fall to the ground. But if you catch one in a bundle of dried grass, and then blow on it gently, you can make a flame.

Curiosity is like that. It's being sparked all the time, but most of the sparks flare out and disappear. 

(If you watch your experience closely for just thirty seconds, with attention to teeny tiny searching sensations, you can probably feel it happening [1].)

This essay is about my strategy for turning some of those sparks into steady flames. When skillfully tended, a steady flame of curiosity can sustain a whole line of research, even when neither you nor anyone else knows how to think about the problems involved.

It took me a long time to come up with this method. I meandered a lot, and I am not done meandering. What I will share with you is where my meanderings have taken me so far. I hope you will make a better version that is closer to the place where you want to be.

So, how does a naturalist study begin?

I think there are several good answers to this. But like a hiker finding north, it usually starts with some kind of orientation.

What follows is an orientation procedure. It is the first lesson in my course on naturalism, adapted to text. The procedure is meant to prepare you to drive with curiosity toward a more intimate relationship with a region of territory where crucial data are likely to live. (More on that later.) 

I think my way of orienting is especially good for people who have some idea of what they're interested in studying, but not much of one (“Something about forecasting, maybe?”); and for people who know what bug they want to solve, but aren't relating to it with curiosity (“I just want to make the problem go away!”). 

This might not be quite the right way to help you catch the spark. When I guide someone through this one-on-one, we work together to design their orientation process. As you read, I recommend paying attention to your frustrations and desires, and holding a question like, “What might work better for me?”

This orientation procedure has three parts: articulating stories, squinting at stories, and choosing your quest. In each part, I’ll say a little about the point of it, show you one way it can go with a real-life example I work through on the spot, describe the instructions I followed to do it, then share another example before I move to the next part. There will be some discussion and philosophizing, and then, at the end, I’ll include an appendix with a couple more real-life examples.

If you want to follow along, this is the time to think of a spark you would like to catch [2].

The Orientation Procedure

Part 1: Articulating Stories

You’ll start by finding a handle for a potent felt sense that’s backed by curiosity. That handle will eventually become a torch as you carry your flame into deliberate exploration.

Example 1: Geometry

While designing a new cover for my greenhouse the other day, I used the Pythagorean theorem to figure out the dimensions of the roof. Something happened while I was doing those calculations. There was a moment of confusion, puzzling, and curiosity. A spark.

Let’s see if I can catch it.

(Context: I know almost zero geometry.)

Does it involve "calculation"? No, that doesn't seem right.

Something about "measure" or "physical dimensions"? Not quite, but "physical dimensions" is moving in the right direction.

The picture I drew on my whiteboard of the triangle representing the roof, that feels important. I remember tilting my head and puzzling at it. I thought things like, "Why does this work?" and "Why does it have to be this way?".

"Inevitability." That one clicks. 

"The inevitability of physical relationships." Yes. 

"Discovery." "Proof." "Comprehension." "The real world." All of these things.

Let’s try a sentence, now that the short phrases are clicking: I think I'm interested in geometry because… I want to discover the inevitability of physical relationships?

Almost. I would like… I want a different relationship with the physical world than the one I have. I want to have a kind of relationship where I believe deep down not just that it's governed by laws, but that I personally can discover those laws and prove the law-ness of them. And I think facility with (Euclidean?) geometry could shift me toward that relationship. 

Yeah, that feels right.

My story (a summary of all of that): I want mastery over the relative arrangements of physical properties.

So I started with a vague spark of curiosity, and now I have a clear articulation I can work with, like a flame on a single tiny twig.

Now that you've watched me demonstrate with a specific example, I’ll go back through and explain what I was doing in more general terms.

Instructions for Articulating Stories

  1. Find a felt sense of the general topic you're interested in. It doesn't matter if you don't know what it's called, or how to think about it. Just reach out with whatever part of you feels like reaching, until you recognize a sensation like, "Ah, yeah, something I care about is going on over there".
  2. Offer some words and images to your felt sense: words like "learning", or "overwhelm", or "that thing that happens to people when they spend too much time reading analytic philosophy". You're not necessarily looking for a single word or phrase that clicks. You're just locating relevant concepts and finding plausible handles for them.
  3. Weave some of those handles into a statement about your topic. I’ll call this your “story”. You’ll treat your story as (roughly) a hypothesis. 

    As an alternative: Your story does not actually have to be made of words. It could be a series of images, or maybe even a numerical model of some kind. 

    Or it could be highly poetic, an ungrammatical string of metaphors. 

    What matters is that you experience it as asserting something about how the world is, and that it be made of parts you can consider independently.
  4. Tinker with your story statement until it has the following properties:
    1. It is propositional (something that could be true or false).
    2. It is a short, big-picture summary of what you sort-of-probably believe about the thing you're interested in.
    3. Most importantly, it captures the core of the felt sense you've been talking with.

You’ve probably articulated your story correctly and completely if 

  • each part of it emotionally impacts you at least little as you read it, and
  • when you ask the felt sense, “Is anything missing?” you get a “no” or a sense of satisfaction.

(Throughout this essay, I’ll present instructions as well-ordered lists. In practice, the orientation process tends to be a lot more fluid than that. If you find yourself jumping back and forth between steps, taking things out of order, or inserting new steps I failed to mention, that’s totally fine. In fact it’s exactly what I hope you’ll do. These numbered lists are here for ease of reading.)

Example 2: Courage

(I worked through the geometry example in real time, but for this one I’m using a months-old memory. It’s a true story, but don’t trust the details.)

A while back, I was watching this video of ten-year-old Alex Shumaker playing the drums. And I was feeling some things about it.

The things I was feeling were happy and sad at the same time. My chest was warm and light, buzzing with electricity, but also constricted. My throat tightened, my eyes teared up. I felt admiration, love, and excitement. I also felt envy, regret, and longing. There was something important there. I wanted to know what, and why, and how it works. I wanted to catch the spark.

I was sad because I wanted so badly to be like that, and knew that I wasn’t. And I was happy because he was like that, and maybe I could be too.

Like what, exactly? What’s the “that” Alex is like?

In the video, he’s playing “Don’t Stop Believing” by Journey. It takes a while for the drums to come in, so at first he’s sort of just sitting there. If I were in that position, I’d be frozen, awkwardly gripping my drumsticks and sweating as everyone watched.

Only he’s not just sitting there. He’s moving around, singing along, and twirling his drumsticks. He hasn’t even started and he’s already so alive.

“Alive” fits, the click of a focusing handle. But that’s not all of it. As I watched, I had a battle going on in my head. 

(Before the second verse, he spins his drumstick six times, then hits the high hats [3] twice.) 

Part of me said, “This is the way to be. Alive and free and unafraid.” 

(Finally the drum part comes in, and Alex finds a rhythm on the snares.) 

Another part of me said, “He’s like that because he’s a child, and children don’t know about danger yet. Anyone can be fearless when they don’t see any reason to be afraid.”

(On the word “night”, he tilts his head back to sing with his mouth wide open, puts one hand over his heart, and continues drumming with the other hand.) 

“But the dangers that make me afraid are real, and I am not blind to them. I can’t be like Alex without somehow becoming ignorant again.” 

(Then he jumps to come down on two high hats at once.)

But the part of me that said “this is the way to be” would not give up. I was just so sure of it, so sure that there must be some way not to give up on ourselves just because we’ve grown up and learned more about the world. 

I kept thinking about the video for days after, and each time, it got a little louder: “Maybe it is possible.”

Maybe it is possible to be alive, even though.

Maybe it is possible to move without restraint, when you know that the danger is real.

Maybe when there is danger on all sides, and you have to act decisively while accepting risk, it is necessary to be like Alex, even while you are afraid.

For a week, these concepts swirled around in my head, chasing each other, weighing, asking questions: “Freedom.” “Maybe it is possible.” “Fear.” “Maybe it is necessary.” “Alex playing the drums.” “Freedom in the face of fear.” “The person I want to be.” “Courage.” “How?”

So after a few days, I found this story: “To be myself, in the world I find myself in, I will have to learn courage.”

Part 2: Squinting At Stories

Once you’ve articulated a story, it's time to transfer the flame from the tiny twig onto a torch that is more durable and mobile. You’ll do this by squinting at it. Uncovering its hidden assumptions. Inviting it to tell you way more about itself than you originally wanted to know.

This gentle interrogation has two purposes.

The first purpose, accomplished by listing assumptions implicit in your story, is to de-anchor yourself from your most familiar way of thinking about the topic.

Why do this? Eliezer wrote, “Curiosity requires both that you be ignorant, and that you desire to relinquish your ignorance.” But also, curiosity is most effective when you recognize your ignorance.

I believe that typos exist in this essay. I also want to correct them. But in an ordinary, habitual frame of mind, it’s difficult for me to recognize the typos when I see them.

Why are they difficult for me to recognize? Because I know what I meant to say, and I’ve rehearsed these sentences many times. The typos are invisible to me as long as I am hallucinating my expectations onto the page. If I really want to give myself an opportunity to notice where the symbols on the page differ from my expectations, I will have to do something a little unusual (such as reading the essay backward).

Concepts we have rehearsed many times tend to feel both unitary and inevitable. A year or two after learning the alphabet song and singing it over and over, I thought that one of the letters was called “elemeno”. When I got more serious about learning to read, I asked my teacher, “Which one is elemeno?” and I could not make any sense of her response. It took me a while to figure out that elemeno is in fact four completely different letters

A couch that gets stuck no matter how you try to shove it through your doorway may pass through easily once you’ve separated the backrest from the seat and arms.

The story you articulated in Part 1 is a product of your habitual patterns of thought and observation. It’s the way the world seems to you now, given all the grooves you’ve worn into your perceptual systems. But the real world is far more detailed.

By listing assumptions implicit in your story, you find the pieces your story is made of, and you begin to take it apart. The more pieces you break off, the more opportunities you create to consider the components independently, or in novel arrangements. You may begin to recognize your ignorance in places you never knew existed.

This approach (and the rest of naturalism) is especially useful when you’ll need a lot of original seeing to get anywhere worth going. But no matter what you’re hoping to learn, the efficacy of your curiosity will increase with even just a little bit of original seeing.

The second purpose of the gentle interrogation, accomplished by brainstorming questions, is to educate yourself about where you'll need to look if you want to shift your understanding toward a more accurate representation of the world. 

Unlike assumptions, questions point outside of themselves. They are like charters for nautical expeditions.

By asking questions inspired by assumptions, you can follow those questions out into the world, where you’ll encounter data related to your assumptions—and encounters with data related to your assumptions are often opportunities to change your mind.

Example 1: Geometry

My story: I want mastery over the relative arrangements of physical properties.

First I will list the assumptions implicit in this story, or in the parts of me that produced the story.


  • I mean something by "mastery".
  • Physical properties are arranged.
  • I care about relative arrangements, as opposed to absolute arrangements.
  • The thing I want is a kind of mastery.
  • There's a thing I want mastery over.
  • I'm interested in physical properties, as opposed to something else.
  • I'm interested in physical properties.
  • There are "physical properties".
  • Physical properties have arrangements.
  • Physical properties have relative arrangements.
  • What I am doing at mastery over the thing is "wanting".
  • The thing I am wanting mastery over is "the relative arrangements of physical properties".
  • Mastery is something you can have "over" things.
  • It's not just physical properties I want mastery over, but their relative arrangements in particular.
  • There are things with physical properties.
  • There is a space in which physical properties are arranged.
  • "Physical" is a thing.
  • "Relative" is a thing.
  • "Properties" is a thing.
  • There's a kind of relationship I can have with something that I call "mastery over".
  • The space in which physical properties are arranged is a kind of thing I can relate to in a mastery-over sort of way.
  • The difference between what I'm interested in and "physics" has to do with a focus on "relative arrangement".
  • "The relative arrangements of physical properties" is a thing.
  • "Geometry" has something to do with the relative arrangements of physical properties.

(This kind of list can go on indefinitely, but for me, five to ten minutes of listing is usually enough.)

Next I will tilt my head at these assumptions, wondering to myself "What is this? What's going on here? Is this right? Is it really so simple? Could I be confused somehow?".

I won't write about my experience of tilting my head at every single one of these. That would take too long, and I usually skip over some anyway (prioritizing by aliveness). I'll just pick a couple to show you.

"I mean something by 'mastery'."

  • Hm. (For the record, I literally tilted my head just now, not on purpose.) If this is true, I wonder how clear I am on what it is I mean. If it's false, then what am I doing with this "mastery" thing, other than meaning something by it? I do suspect it's possible to accidentally pepper your thoughts and sentences with "meaningless" words, words that serve some sort of auxiliary purpose. I don't think that's going on with "mastery"? But I am a little suspicious of this word. I feel as though I'm packing an awful lot into it, and it seems like the stuff is sort of swirling or tangled up. What is the stuff, and how easily does it untangle?

"Physical properties have arrangements."

  • "Have"? Really, they "have" arrangements? Where do they keep them? I wonder what "properties" is doing that isn't covered by "physical". Arrangements like the things you do with funerals and wills and stuff after someone dies? Arrangements like the thing you're threatening to change when you tell someone you'll rearrange their face? Arrangement seems to have something to do with order. But is it about lower entropy, or is it more a description of the relationships among parts of a system, whatever its entropy? Is that why I talk about "relative" arrangements? But then why do I bother specifying "relative", if it's part of what I mean by "arrangement" already? What kind of "system"? Is a triangle a system? Multiple parts of me are trying to call bullshit on "physical". One part says that nothing isn't physical. Another part says that since mathematical objects like circles obviously aren't physical, it isn't physical properties I'm interested in at all. I also call bullshit on knowing what mathematical objects are. Maybe I should have said "properties of physical objects" instead of "physical properties". Anyway, whatever these things are, I seem to think that they're arranged, and it's interesting to think of non-physical things as having arrangements. I don't know what "physical" means. Maybe I should try to make one list of definitely physical things, and another list of definitely not physical things, and then see how confused I feel afterward.

And so forth. 

I don’t tend to think in words, and for me it doesn’t take nearly as long to tilt my head at one of these as it takes to read all of that (let alone to write it). But what I do with my mind at least has the flavor of those paragraphs.

When I look back over my list of assumptions, after tilting my head at them, I notice that my interest snags most on

  • "The thing I want is a kind of mastery."
  • "There is a space in which physical properties are arranged." (When I got to this one, I said "ohHO!" out loud.)
  • "'Geometry' has something to do with the relative arrangements of physical properties."

When I say that my interest snags on these assumptions, I don't necessarily mean that I'm more skeptical of these than of the others (though I may be). It's more like, "There is definitely something interesting going on here, and I want to know what it is."

Besides "snag", another way to describe it is "drawn into". I am drawn into these assumptions by my interest, my curiosity, the center of my quizzical squinting. I'm eager to engage with them.

Finally, I make a list of questions. I focus especially on those snaggiest of assumptions for input. 

Usually I don't write anything down during the head-tilting phase, but you can see in my accounts of head-tilting that this attitude generates many questions. The purpose of the list is to articulate and capture them. I'll just put a few here.

  • Are there other kinds of mastery?
  • What do I mean by mastery?
  • Do I want a thing?
  • What do I want?
  • How can I tell when I've mastered something?
  • Have I ever mastered something?
  • If I wanted to master something, how might I go about it?
  • Is "the relative arrangements of physical properties" a thing?
  • What kind of a thing is it?
  • Can physical properties be arranged non-relatively? Absolutely? Are there absolute arrangements?
  • What sort of properties am I talking about here?
  • What exactly does any of this have to do with triangles, quadrilaterals, conic sections, coordinate planes, and so forth?
  • And what does it have to do with Euclidean axioms and postulates?
  • What is the relationship between Euclidean axioms and the relative arrangements of physical properties?
  • What is "proof", and why does that feel so closely tied to "mastery"?
  • Why do I want to prove things?
  • What's so juicy about the Pythagorean theorem in particular?
  • What is this elegance business I like so much, and why does geometry seem like a good place to look for it?
  • What is geometry?
  • How is geometry different from topology?
  • I've read that topology is a generalization of geometry; if so, does that mean a geometry is a topology with certain properties? Which properties are they?
  • What would happen if I tried to understand Euclidean geometry very backward, like maybe from a perspective of topology and model theory?
  • What would happen if I tried to understand geometry very forward? What would that mean? Would it mean studying the tree outside my window? Or building a shed? Or starting with algebra?

As you can probably see, this list could get very long, and could roam around a lot. But it's also going in a direction, and it sort of has a destination that it's searching for. 

I don't know how apparent this is from the outside, but to me, the questions change as the list goes on. They become more spring-boardy. I started with questions like, “What do I mean by ‘mastery?’”, which had a little bit of a diligent feeling behind it. Gradually, I narrowed in on what I'm most curious about. As I did so, avenues of investigation presented themselves (“What’s so juicy about the Pythagorean theorem in particular?”. Subsequent questions tended to have a little bit more of whatever propels me to stick my hands in the dirt and figure out what is up with the world (“What would happen if I built a shed?”)—and specifically that part of the world picked out by the felt sense in Part 1.

So I ask a lot of questions, without much of a filter, but I also feel for what matters as I go.

Instructions for Squinting At Stories

  1. Make a list of assumptions you think are implicit (or explicit) in your story. 

    As you do, spend a little time on each substantive word (or symbol) in your story statement.

    Also, bother to write down the blindingly obvious. Include stuff that's so apparent, you tend to experience it as invisible. You might by default be inclined to restrict yourself to assumptions that seem shaky. Don’t. 

    (If you find that this, or any other part of Squinting At Stories, is taking too long, skip down to the appendix for troubleshooting.)
  2. Look over your list of assumptions, and tilt your head [4] at some of the items, like a puppy who's just heard a kazoo for the first time. For each one, try on an attitude of, "What is this? What's going on here? Is this right? Is it really so simple? Could I be confused somehow?" 

    Pay attention to which ones your interest snags on. Some of them will catch your attention more easily than others. Notice which assumptions stand out to you.
  3. Brainstorm questions around your topic.

    How many questions to brainstorm depends on a few factors. The main ones are how important the topic is to you, how much time you have to spare, and how easily the questions are coming. Usually I decide whether to set a five minute timer or a twenty-five minute timer to start, and commit to brainstorming for the duration. If I get to the end and want to keep going, I set another timer. Otherwise, I move on.

You’ve probably squinted at your story correctly if

  • you have a sense that, in several respects, your topic is rich in fascinating detail, and
  • you’re eager to uncover its secrets.

Example 2: Courage

(I don’t have notes from when I did this while first studying courage, and it happened in a slightly different form. I’ve done my best to remember my old perspective and recreate the thoughts, but I suspect there’s a lot more of my current perspective in here than my old one.)

My story: To be myself, in the world I find myself in, I will have to learn courage.


  • I want to be myself.
  • I am something.
  • I mean something by “be myself”.
  • I find myself in a particular kind of world.
  • The world seems at odds with “being myself”.
  • I mean something by “the world”.
  • I “find” myself in the world.
  • I am in the world.
  • Courage is learnable.
  • There’s no other way to be myself in the world than to learn courage.
  • Courage somehow makes being myself compatible with being in the world.
  • The world and I are incompatible while I am not courageous.
  • I do not currently know courage.
  • I mean something by “learn courage”.
  • Courage is a thing.
  • There is some kind of necessity in my future learning of courage.
  • It is possible to be other than myself.
  • If I’m living in the world and I am not courageous, then I am being something other than myself.
  • It is possible to not be courageous.
  • It is possible not to have learned courage.

Some thoughts I have while tilting my head at a few of these assumptions:

  • “Courage is learnable.”: Learnable by whom? What makes something learnable or not learnable? Is there such a thing as “not learnable”? What exactly am I saying is learnable? What do I mean by “courage”? Are there some parts of it that can be learned, and some that can’t? Does it break down into parts? Is it a coherent thing that easily comes apart, or more of a swirling cloud that will dissipate on closer examination? If it’s learnable, what category of knowledge will be gained? Is it a skill? A capacity? A model? A perspective? Something else? Have people learned courage in the past? How have they done it? I have a hunch that this is a kind of thing people emphasize in childhood education, part of “being brought up right” or something; what does that mean, if anything, about what it takes to learn it?
  • “I ‘find’ myself in the world.” I couldn’t even get all the way through typing this one in the original list without tilting my head so much that I put “find” in quotation marks. What a strange sentence this is. I find myself in the world. When a person says that they find themselves in this or that situation, they’re emphasizing the accident of it, their having arrived in that situation without having chosen deliberately to get there. The “find” in my sentence here has that flavor. It’s as though I’m trying to deny responsibility preemptively. As though I’m saying, “Hey, it’s not my fault I’m here.” I wonder why that matters. Why is that an important part of how things are shaped in my head? And what do I mean by “the world”? Do I mean it in the modal semantics sense of “actual world”? I think not. But what do I mean? What aspects of the world do I care about here? What state of affairs do I have in mind when I say “the world”, while thinking about courage? And what is it I am finding, exactly? Is there something I’m picking out about me in particular, or am I talking about any given human? I think it’s a little of both. I wonder what about me in particular is relevant here.
  • “Courage is a thing.” Is it? Suppose it’s not a thing. Then what would be true? Some people, myself included, would be confused. Specifically, they’d be confused about something to do with virtue, fear, action, responsibility, conscience, danger, risk, safety, strength, or weakness. Probably they’d be wrong about how those things relate to each other, or to the world, or they’d be wrong about the internal structure of some of those things. Maybe they’d be conflating a lot of things that really need to be kept separate to understand the world accurately. Or maybe they’d be making some nonsensical distinctions. But even if everyone is confused enough to believe in courage, the way that some people claim others are confused enough to believe in free will, the concept is still doing something. What might it be doing? How does it influence behavior? How does it influence people’s conceptions of themselves and their communities? Anyway, if courage is a thing, then what kind of a thing is it? And how could I know?

Looking back through the list of assumptions, I notice that my attention snags on 

  • It is possible to be other than myself.
  • Courage is a thing.
  • I find myself in a particular kind of world.


  • What is courage?
  • Is courage a thing?
  • How can you tell when someone is being courageous?
  • Are there things that look similar to courage but are actually different?
  • What kind of a thing is courage?
  • What can be courageous?
  • Is there a difference between being courageous and behaving courageously?
  • Is courage a property of actions?
  • What does it mean to “learn” courage?
  • Can I “know” courage and choose not to be courageous?
  • Is every action either courageous or cowardly?
  • Is cowardice the opposite of courage?
  • What is cowardice?
  • What does courage have to do with fear?
  • What does it mean to be myself?
  • If I suppose that “myself” is a thing, and that it’s made of stuff I already think exists and am not even opposed to, what kind of a thing is it?
  • What do people mean when they say “just be yourself” or “I was not myself [when I yelled at you while drunk]”?
  • Where does courage come from?
  • Why might courage matter?
  • If a world is such that I can’t be myself without learning courage, what might be the properties of that world?
  • Why do I think that maybe I can’t be myself in this world without courage?
  • In what respect do I “find” myself in this world? What other ways can one be in a world?
  • Do I perhaps mean one very specific part of the world? Society, for example?
  • What am I scared of?
  • What is fear?
  • How can I tell when I’m afraid?
  • What happens when I’m afraid?

Part 3: Choosing Your Quest

After articulating your story and then squinting at it, your spark has (hopefully) become a steady flame atop a torch. 

Next, you will set out to illuminate some part of the world you think could teach you something. 

You’ll do this by finding a conceptual crux.

I don't think I mean "crux" in quite the usual CFAR sense, but I mean something very close. Rather than "a proposition that some other particular belief rests on", what I'm talking about is "something near the foundation of your whole way of thinking about things" (where "things", in this case, is everything to do with your topic).

Naturalism is not very interested in improving specific beliefs. If grokking whatever it is you want to grok were as simple as flipping the truth values you assign to particular propositions, or even adjusting the distribution of probability mass, naturalism probably wouldn’t be the most efficient method for you. That would mean you're merely incorrect, rather than deeply confused or lost. 

Naturalism is more interested in improving the concepts that your beliefs are made of. It helps you recognize when the world is shaped in a fundamentally different way than you originally thought—like when elemeno turns out to be L, M, N, and O.

Your quest will be a question with the potential to do that.

And once you have such a conceptually crucial question, you’ll set out to answer it, packing your other questions like provisions for your journey.

Example 1: Geometry

My story: I want mastery over the relative arrangements of physical properties.

I’ve taken a moment to get back in touch with the felt sense behind my story. I can tell I’m in touch with it because several of the words are hitting me in the chest: "mastery", "relative arrangements", and "physical properties" all go "Pow pow pow! Yeah, those things!". 

So now I will look over my list of questions.

The first one that jumps out to me a little is, "What do I mean by mastery?". It jumps out at me because when I imagine seeking the answer to that question, the person I am afterward is in a better position to understand the thing I care about. They're closer than me to the region of territory picked out by the felt sense I am holding. "Yes, crucial data over there!" is the name of the feeling, for me.

But it's not a very strong feeling, for this particular question.

"What kind of thing is [the relative arrangements of physical properties]?" That one jumps out a lot.

When I offer this question to the felt sense behind my story, the connection is like a gigantic tome falling open. I have a hunch that whatever is over there, it will not only result in me gaining crucial information, but will also lead me to ask many new questions that I can't even conceive of yet, much better ones than I've asked so far.

I get about the same thing with, "How is geometry different from topology?"

The second and third of these both seem like fine quest objects, but I suspect that I'll get further faster with the second. “What kind of thing is [the relative arrangements of physical properties]?” is the sort of question I expect any high school geometry textbook could help me with, for example. It also seems like looking at a tree all by myself could be productive.

By contrast, I expect the third one (“How is geometry different from topology?”) will require reading stuff written for people with very different math backgrounds than mine.

So I have chosen the first quest of my geometry adventure: 

My quest: What kind of thing is the 'relative arrangement of physical properties'?

Instructions for Choosing Your Quest

  1. Set aside all the words from part 2.
  2. Look back at your story from part 1, and re-connect with the felt sense that generated it.
  3. Hold onto that sense, and look over your list of questions.
  4. Imagine trying to investigate each question. Notice when you expect that pursuing an answer to one of them could change the way you think about or relate to your topic—when it could change your story about what’s going on. (Those questions are conceptually crucial.)
  5. Consider each crucial question, noticing what you would most enjoy attempting to answer, and what could get you traction quickly.
  6. Choose your favorite question as your quest.

    As an alternative: If you have multiple excellent questions and don’t know which path to take, you could plan to spend one short block of time investigating the first, and another short block of time investigating the second. This might give you more clarity on how to proceed.

You’ve probably chosen a good quest if,

  • when you hold the question up to the felt sense of your story, it’s clear that the question matters,
  • you’re at least a little excited, and
  • part of you is already making plans to investigate.

Example 2: Courage

My story: If I want to be myself, in the world I find myself in, I’ll have to learn courage.

It’s remarkable how much fear is in the felt sense of this. And resolve. “If I want to be myself” is somewhere between an invitation and a challenge. “In the world I find myself in” comes with this image of swirling white electric chaos, with my body in the middle. “I’ll have to learn courage” is like pressure and heat in my chest, and an image of a rhinoceros lowering its head toward the ground and stamping the dust as it prepares to charge. Overall, the story comes from a feeling of opposition to limpness, plus concern for not just myself but also the society around me and where we’re headed.

As I look back over my list of questions, the ones that feel most likely to quickly change how I relate to the topic are

  • What kind of a thing is courage?
  • What happens when I’m afraid?
  • What does it mean to be myself?

Of these, I think I’d like to start with, “What happens when I’m afraid?” I like the groundedness of it. It’s just so obvious where it’s suggesting I look. 

Plus, I’d be shocked if I didn’t learn something crucial about courage while improving my understanding of fear.

My quest: “What happens when I’m afraid?”

Reflection and Conclusion

Why this particular way?

The orientation procedure is meant to prepare you to drive with curiosity toward a more intimate relationship with a region of territory where crucial data are likely to live. 

Which is a bit of a mouthful. Let's break it down.

Why use "curiosity" to drive?

One way to categorize people is to ask what usually drives them to learn. 

Many things can be learned out of duty, out of ambition, or out of a desire for the problem to be solved. But naturalism is geared for situations so novel, fraught, or hopeless that your whole way of thinking about them is totally inadequate. In those situations, there's no way around it: You need rationality.

And there's a reason curiosity is the first virtue or rationality.

With duty, there's a particular thing you have a duty to do, and your thoughts can hang themselves on that frame. If you can't conceptualize your duty, though, you won't know what is right or wrong, so there's no way to steer.

With ambition, your frame is made of the abilities, skills, or resources you want to command. But if you're too lost to know which resource you're after, ambition won't help much either.

With goal-oriented problem solving, your frame is a pre-defined problem. But if you used broken concepts to define that problem, you won't find traction, because no ground exists where you'll try to put your wheels.

(I conjecture that this is usually what's happening when people try and fail to solve the same problem over and over again.)

It seems to me that all three of those engines (duty, ambition, and problem solving) share a sort of inward gaze.

Duty is about you being a good person. When you use it to drive, there's a constant dialog between your behaviors and what you think is right.

Ambition is about you becoming powerful. When you use it to drive, there's a constant dialog between the resources you have and the resources you want.

Goal-oriented problem solving is about what shape you want the world to take. When you use it to drive, there's a constant dialog between the state of the world and the one you envision for it.

You might argue that curiosity is about what you want to know, but I think I would disagree. That perspective on curiosity is problem solving, ambition, or duty in disguise.

Pure curiosity is about your reflection of the world. Although it tends to take the world piece by piece, it does not fundamentally rest on a desire to know a particular thing for a particular practical purpose. It does not feel like, “If only I knew this, then I could accomplish that.” Instead, it feels like, “I must fill this hole in my knowledge, or I will be forever incomplete.”

A person driven primarily by curiosity is so in love with the world that they want to become a perfect reflection of it. And to do that, you put your perceptions of the world in dialog, not with yourself, but with your other perceptions of the world.

So when your concepts are broken, inadequate, or nonexistent, curiosity has a special advantage. It doesn't rely so much on your concepts to judge.

Why drive toward "a more intimate relationship with a region of territory"? What do I even mean by that?

When I was six or seven years old, my dad wouldn’t let me open my Christmas presents early. But I wanted desperately to know what was inside the opaque boxes. So I got Dad’s blowgun, snuck the presents into the basement, and shot the darts through the cardboard. 

In some places, the darts penetrated almost all the way through. In other places, they got stuck on the surface of whatever was in there. The idea was to infer the object from its outline.

This method… needed some improvement, and did not actually cause me to know what I was getting for Christmas. Still, I was literally piercing through a region of territory, over and over again, to learn whatever I could about it through direct observation. 

That is the kind of intimacy I’m talking about.

I read a story once about a boy who wanted to master the wind. He went to an old wizard whom he’d heard could direct the wind, and asked to be his student. The boy was ready to do anything: he'd read every book in the library; he'd study late, night after night; he'd listen and research and write essays until he knew every rune, magic word, and alchemical recipe in existence, if that was what it took.

But the wizard refused to lecture. He offered no textbook.

Instead, he made the boy stand naked on the roof of a tall building during a storm. That is the kind of intimacy I'm talking about.

When Darwin went to the Galapagos Islands, he collected many finches, and he measured each of their beaks. That is the kind of intimacy I'm talking about.

It's easier to find an answer when you know what the question is. But what about when you don't? What if you know that you're as yet unable to formulate the question whose answer you need? How can you make progress anyway?

What you can do is cut out as many intermediaries as possible. 

Do not ask an encyclopedia to tell you about the world. Do not think harder and longer, waiting for something to emerge from the implications of your existing beliefs. Instead, seek out opportunities to encounter crucial data. Directly expose your own sensorium to reality. Become intimate with the territory.

When successful, this orientation procedure draws out your own felt senses and hooks them up to your own curiosity. By this method, it leaves you undisguised and naked, with a longing for the world to impact you. It prepares you to learn without mediation, which is essential when your intermediary concepts will lead you astray.

Why go "where crucial data are likely to live"?

If you want to learn about birds, digging in the dirt won’t [5] help.

To get anything in particular out of directly exposing your own sensorium to reality, you have to be strategic about which bits of reality it's exposed to. You're bound to learn something while in contact with reality, but what you learn will depend on where you are. You can see and process thousands of things without ever getting close to anything that might answer your questions. 

If you want to draw a map of the coastline, it's necessary but not sufficient to have your eyes open. You also have to go to the sea.

What comes next?

Once you've stopped superimposing so many preconceptions onto your observations—once you've lowered your broken map—you can turn your curiosity toward the mountains and shadows that currently surround you. You can look at them through a longing to understand something, which is how you will deliberately move toward a destination.

That's what it's like in my mind, anyway, at the end of this orientation procedure.

But you're not me. You'll probably have to do something at least a little different to get all the way there. The goal is to look directly at the world with genuine curiosity and awareness of what you care about. How you do it is up to you, but that's what's needed to move on.

The rest of naturalism is about figuring out how to go where crucial data live when the map you're starting with is all wrong.

In the next essay (unless my plans change before then), I'll show how I begin to seek out experiences that are rich in crucial data.


A Note On Duration

For both of the below examples, I followed my instructions from this essay. I also time-boxed each step to five minutes (five minutes for story articulation, five minutes for assumptions, five minutes for head-tilting, five minutes for questions, five minutes for quest choosing.). Only "assumptions" and "questions" took the full five minutes (and I could easily have kept going).

If it's taking you a lot longer than this, then either you feel good about how long it's taking and you should carry on, or you feel bad about how long it's taking and you should probably change something.

If you want to change something, maybe try one of these:

  • Do a time-boxed version and run with whatever results from that, even if it's not perfect.
  • Warm up with a brainstorming exercise: Write down as many animals as you can in one minute. Make a few notes about what strategies you used for coming up with animals, and what other strategies you could have used but didn't. Then spend another minute naming animals. When you're done, come back to the "assumptions" and "questions" sections with the same mental posture you used in your second minute of animal naming.
  • Skip "Squinting At Stories" entirely, and try the whole thing again after you've been questing for a week.

Example 3

Story: When I'm trying really hard, I'm afraid that if I let go, things will fall apart.


  • Sometimes I try really hard.
  • Sometimes I try.
  • There's a difference between trying a little and trying "really hard".
  • Sometimes I try only a little.
  • At some point I've experienced a thing in my mind and/or body that I'm now calling "trying really hard".
  • When that "trying really hard" thing happens, there's something I'm afraid of.
  • The thing I'm afraid of while trying really hard is what will happen if I "let go".
  • If I let go, things will fall apart.
  • There's a thing I can do called "letting go".
  • Things can "fall apart".
  • There's a particular thing I mean by "fall apart".
  • Sometimes I'm afraid.
  • Sometimes I'm afraid of something that will happen if I act in a certain way.
  • I don't want things to fall apart.
  • I want something besides "things falling apart".
  • I think something bad will happen if I let go.
  • I think something bad will happen if things fall apart.
  • I think something will happen if I let go.
  • It's important to me that I prevent whatever it is I think will happen if I let go.
  • It's important to me that I prevent things from falling apart.
  • The fear of letting go happens specifically while I am trying really hard.



  • What does it feel like to "try really hard"?
  • What does it feel like to "try"?
  • Are there multiple kinds of trying?
  • Are there kinds of trying that do not fall on a linear scale from "only trying a little" to "trying really hard"?
  • Do the different kinds of trying feel different?
  • Do the different kinds of trying have different purposes?
  • When I try only a little, is there fear?
  • When I try a lot, is there fear?
  • Is there a difference between trying "a lot" and trying "really hard"?
  • What is "trying"?
  • When I'm trying really hard, what is the fear like?
  • What is fear usually like?
  • What is fear trying to do, in general?
  • What is fear trying to do while I'm trying really hard?
  • When I'm afraid of something while trying really hard, am I imagining a particular state of the world, or is something else going on?
  • What would it mean for "things" to "fall apart"?
  • Which "things" do I not want to fall apart?
  • Have I ever experienced "things falling apart"?
  • If things fell apart, how would I know?
  • Can one individual thing fall apart, or is it some kind of systemic property?
  • What is the bad thing I think will happen if things fall apart?
  • Can I prevent things from falling apart?
  • Is it sometimes true that if I let go, things will fall apart?
  • What might I do *after* things fall apart, if that happens? How bad is it exactly?
  • Does trying really hard prevent things from falling apart?
  • What does "trying really hard" actually cause?
  • Can I choose whether to try really hard or to let go?
  • Is there anything it's almost impossible to do at the same time as "trying really hard"?

Quest: What happens when I'm trying really hard?

Example 4

Story: Seedlings are tricky.


  • Seedlings are trickier than adult plants.
  • There's something I mean here by "tricky".
  • The trickiness is a property of the seedlings.
  • All seedlings are tricky.
  • There's such a thing as a "seedling".
  • There aren't any seedlings that aren't tricky.
  • There are several different ways in which seedlings are tricky.
  • Lots of different things can go wrong with seedlings.
  • When I want one plant, I should sew ten seeds.
  • I don't have much control over how many of the seeds I sew grow up into adult plants.
  • It's very hard to help new gardeners grow plants from seed without them becoming discouraged.
  • Mold is an especially tricky thing about seedlings.
  • It's not that I'm dumb about seedlings, it's that the seedling part of the universe is mysterious.


  • What causes damping-off mold?
  • Are there kinds of mold that grow on the surface but don't affect the roots?
  • Is there some clever way to keep seedlings moist enough to grow, but not so soggy that mold grows, without manually checking the moisture level every few hours?
  • What causes seeds not to germinate?
  • How many of the seedlings that never break the surface of the soil do in fact germinate?
  • What happens when you spray a lot of copper fungicide on a seedling?
  • Are some plants more resistant to damping off mold than others?
  • What causes those plants to be more resistant, if so?
  • What role does nutrient balance play in the germination of seedlings?
  • Is it perhaps better to let seeds germinate in a damp paper towel before transplanting them to soil?
  • Is there something about growing in tiny containers that makes things harder, such that more seedlings would prosper if sewn directly into large containers or into the ground?
  • How much does the actual germination rate of my seeds tend to differ from the germination rates claimed by professional horticulturists?
  • What factors affect seedling health that I've completely failed to consider?
  • Could I grow damping-off mold on purpose without even planting any seeds?
  • How does industrial agriculture treat seedling care?
  • How do horticulturists working with very rare and expensive plants handle seedling care?

Quest: Can I find a plant that's very difficult to cultivate from seed and raise it to maturity?

Coda: Empiricism, Objectivity, and the Soul of Science

(You're still here? It's over! Go home.)

(Ok, well, if you insist…)

In earlier drafts of this essay, readers asked me, “How do I know what kind of spark I’m supposed to catch?”

This is a straightforward question to which I unfortunately do not have a straightforward answer. The closest I can come to a short answer is, "Try catching any spark whatsoever, and see how it goes," which I recognize might not be satisfying.

The realer answer is, "Wrong question." I think I can explain why it's a wrong question, and I think that in the process of explaining, I can probably give you the thing you were groping toward by asking it. You’ll likely get a much better sense of where all this “naturalism” stuff is coming from along the way, too. 

It's going to be a long and convoluted process, but I’ll give it a shot.

I said that “Catching the Spark” is a text adaptation of the first class in my naturalism program. But that’s not quite true. 

In fact, there’s a zeroth-class beforehand that’s a combination of “interview” and “going over the syllabus”. For people who either decide to continue with the program or are on the fence, I give an assignment to complete before the first propper lesson: “Spend two weeks paying attention to patterns in what you're actually drawn to in the moment. As you do, hold the question, ‘What feels important or exciting but also daunting/confusing/impossible?’” 

In the first lesson, sparks are drawn from memories of those experiences. 

Ultimately, I’m working toward a general approach to applied rationality research and collaboration, so I’ve tended to encourage people to pursue rationality-related topics. Plus, naturalism is designed to be a method for researching such topics, which may succeed where other approaches are likely to fail.

But I expect this particular procedure will work well for any object of curiosity. When I imagine taking a shortcut like, "Look, if it would show up on your bugs list at a CFAR workshop, it's fine,” I feel... disappointed. That scope is so narrow. I really want to find out what happens when lots of people try this in lots of different domains with lots of different goals.

I am interested in combating a kind of listlessness or impotence I perceive among the modern intelligencia, rationalists included. 

I have a perception that many of the people around me think that they are not allowed to come to conclusions based on their own observations. And as a result, they do not bother to observe carefully or deliberately. Since they have so often rehearsed the assertion that subjective experience is untrustworthy and anecdata doesn’t count, when they want to know something that isn’t readily googleable, it does not occur to them to personally check. (That’s my story so far, anyway.)

I have a guess about how this has happened. 

In the early 20th century, there was a lot of excitement about the scientific study of phenomenology. But the methodology turned out to be really tricky to reconcile with what the larger global scientific project was trying to do. The pendulum swung very hard away from phenomenology, reaching its anti-subjectivity peak in the 60s with behaviorism, to the extent that the very existence of subjective experience was frequently denied, even by psychologists and philosophers.

Today, it's common for people to reject out of hand any empirical observation that is not obtained through double blind randomized controlled trials, peer reviewed, and published in a reputable scientific journal. Many people think that's what science is. After all, you can't just believe everything written by some crackpot blogger tinkering in his basement lab. 

And, in one sense, they're right. There are good reasons that what we tend to consider "scientific discoveries" are usually the result of a whole bunch of coordination, orchestration, and bureaucracy. Establishing anything with certainty is very hard, and as readers of a rationality blog know better than most, humans are incredibly proficient at duping ourselves into confirming whatever it is we wanted to believe all along.

But in the early history of modern science, circa the founding of the Royal Society of London and a couple hundred years afterward, people didn’t really know that yet. And science looked very different.

John Aubrey spent a bunch of time walking around the Avebury henge monument recording details of the structures and their surroundings. In 1663, he presented his findings to the Royal Society, thereby creating the modern field of archeology.

In 1665, Antoni van Leeuwenhoek and Robert Hooke shared their drawings and descriptions of things they observed while looking at a whole bunch of stuff under microscopes, thereby creating the modern field of microbiology.

In 1752, Benjamin Franklin described how he captured electrical charge by attaching a kite string to a Leyden jar and flying the kite near storm clouds, which strongly suggested the electrical nature of lightning.

In the late 1700s and early 1800s, John Audubon attempted to make a complete pictorial record of all the bird species of North America, discovered twenty five new species in the process, and dramatically advanced the field of ornithology.

In 1830, Charles Darwin set out to study geology by observing and documenting volcanic activity in the Galapagos Islands, became very curious about differences among local finches along the way, and ended up completely revolutionizing natural history itself.

I think that as a species, we sacrificed something vital in our wholesale rejection of scientific phenomenology. We lost the feeling that when you aren't sure what to believe, and there's no reputable publication around to tell you, sometimes you can, 


go check.

The world discussed in all those Google Scholar search results is in fact the very same one you yourself are sitting in right now. When you wonder how the world is, it is often possible to just... look.

But what does that have to do with phenomenology? Isn't phenomenology about what happens inside your own mind? What does it have to do with understanding the mind-independent natural world?

This is a lot like asking what eyes have to do with literature.

In a sense, eyes have nothing to do with literature. Literature is stories and essays and poetry. It involves things like plot tension, and narrative structure, and linguistic rhythm. It exists in the realm of language, not the realm of optics or physiology.

But if you're severely far-sighted and you lose your glasses, it will suddenly become immediately obvious that eyes have everything to do with literature. Without eyes (or ears, or fingertips, depending on the format), you have no access to literature. Written language is universally mediated by vision.

Similarly, everything you can possibly interact with—not just cognitive phenomena like confusion and defensiveness, but also the fraying threads of a kite string as they stand on end with static charge, and the liquid crystal display screen on your calculator as you double check the statistical claims of an epidemiological study—presents itself to you as immediate subjective experience.

So even if you can coordinate all the lab subjects, and technicians, and grant committees, and publications, to minimize observer bias as you try to learn whatever it is you want to know,

it still seems

rather irresponsible, to me,

to completely neglect the practice of phenomenology, all together. 

It also seems irresponsible to routinely learn through direct observation without getting very good at taking your subjective experiences as object, examining them in detail, and taking as little as possible for granted. 

One of my favorite things about the rationality community is its recognition of Bayesian evidence.

Unlike most other modern cultures, it understands that probabilities exist independently of authoritarian decree, that it's possible to update incrementally on many pieces of relatively weak evidence. I like this for several reasons, but one is that it means a rationalist (theoretically, at least) has a lot more personal power to learn things on their own, without the support or hindrance of large institutions. It makes them more like Ben Franklin and Robert Hooke, and less like Lab Technician Number 73.

Don’t get me wrong, I think it's good to be Lab Technician Number 73. It took many such people to build the vaccines for covid-19. We all understand the value of that work.

But I happen to really like Ben Franklins, and I think we need more of them than we have—especially in the field of applied rationality, which is as much the Wild West as volcanic geology was in Darwin’s day.

I think that more rationalists could be more like Ben Franklin. But as a culture, we're so stuck between our fetish for empiricism and our fetish for objectivity that we don't even notice the way in which empiricism is opposed to objectivity. 

So when we feel a spark of curiosity, and consider setting out to personally investigate, we find that our shoelaces are hopelessly knotted together.

I would like to un-knot our shoelaces.

You ask me what kind of spark you’re supposed to catch? I ask you who is doing the supposing, because it sure isn’t me.

The appropriate topic for your study as a naturalist is between you and God. Hook wondered what fleas look like up close. I’ve wondered what courage is for and how it works. My dad raises salamanders and studies their genetics. As long as you're willing to look at the real world with your own eyes, curiosity can fuel your project, provided you’re bold enough to catch the spark and take it somewhere worth going.

I now do this kind of work independently. If you like it and want there to be more, you can support me through Patreon.

New Comment
21 comments, sorted by Click to highlight new comments since:
  • Oh man, what an interesting time to be writing this review!
  • I've now written second drafts of an entire sequence that more or less begins with an abridged (or re-written?) version of "Catching the Spark". The provisional title of the sequence is "Nuts and Bolts Of Naturalism".  (I'm still at least a month and probably more from beginning to publish the sequence, though.) This is the post in the sequence that's given me the most trouble; I've spent a lot of the past week trying to figure out where I stand with it.
  • I think if I just had to answer "yes" or "no" to "do I endorse the post at this point", I'd say "yes". I continue to think it lays out a valuable process that can result in a person being much more in tune with what they actually care about, and able to see much more clearly how they're relating to a topic that they might want to investigate.
  • As I re-write the post for my new sequence, though, I have two main categories of objections to it, both of which seem to be results of my having rushed to publish it as a somewhat stand-alone piece so I could get funding for the rest of my work.
  • One category of objection I have is that it tries to do too much at once. It tries to give instructions for the procedure itself, demonstrate the procedure, and provide a grounding in the underlying philosophy/worldview. It's perhaps a noble goal to do all of that in one post, but I don't think I personally am actually capable of that, and I think I ended up falling short of my standards on all three points. If you've read my sequence Intro To Naturalism, you might possibly share my feeling that the philosophy parts of Catching the Spark are some kind of desperate and muddled. Additionally, I think the demonstration parts are insufficiently real and insufficiently diverse. When I wrote the post, I mostly looked back at my memories to find illustrative examples, rather than catching my examples in real time. A version of this with demonstrations that meet my standards would have real-time notes, both from myself and from other people who have used the procedure.
  • The second category of objection I have is that the post relies heavily on a framing that I think is about 50% inaccurate. I think the posts makes it *sound* like this is a procedure that I designed for the purpose of curiosity cultivation. But it really isn't. The curiosity stuff is more like a side effect, or perhaps an undercurrent of the way I do things-in-general. I think the way this procedure *really* came about is that I guided some people through the larger naturalism process, and kept being annoyed by how many weeds I kept having to pull well over a month in. The procedure is largely meant to pull those weeds at the very beginning. It's really more about creating space for original seeing than it is about catching a spark of curiosity, but I'm also not sure there's any one coherent thing that it's about. It's just "some stuff it turns out is a really good idea to do before you go investigating something in the particular way I'm calling 'naturalism'".
  • But I do still think the procedure itself is pretty good.
  • I'm skipping over the "claims" questions, 'cause I think the post is pretty light on claims. It's mainly an instructional post, not an argumentative one. But feel free to follow up if you want to poke me about particular claims.
  • Follow-up work: All the naturalism stuff. I did the worldview sequence already (that was Intro To Naturalism). Now I'm doing the Nuts and Bolts sequence, which is more or less the "how-to". Most of my hopes and dreams for this and related posts, though, come in the sequence after that: I want a collection of case studies and demonstrations. Basically there will be a whole series of instructional posts similar to this one, and I want several people (not just myself) to follow those instructions while focusing on stuff in the original Sequences; so like people demonstrating what it is like, in detail, to learn to actually change your mind, or to make your beliefs pay rent in anticipated experiences, and I want this kind of resource to accompany many key posts on Lesswrong along with a handful of chapters from books like Scout Mindset and How To Measure Anything.
  • This was all very rambly and chaotic because that is in fact the state of my thoughts right now with respect to this post. Apologies. I welcome attempts to direct me with further questions.


I am quite excited for LessWrong to have more exercises in general, and I'm excited for people pushing forward the art of rationality by developing exercises to teach core skills.

I always quite appreciate Logan's approach to giving very phenomenological descriptions of how to go about skills, breaking them down into small pieces that show all the teeny nuances of what might be going on in your mind.

I have not yet tried to learn the skill in this post – I haven;'t yet had time to set aside for reading it properly and applying it to things relevant to me (ideally I'd like curated posts to pass a bar like that. But in practice, the curation team is often busy and has to curate things we haven't fully vetted). But, I am fairly confident based on past experience with Logan's writing that even if this particular technique doesn't work for me, if I were to practice it I would learn things about how to think about gaining new cognitive skills.

Thanks Raemon! Those are some very nice things you said about me blushes. I hope this post doesn't disappoint.

Most of my previous writing has been pretty "here's a bunch of thoughts and experiences I had in no particular order, which I'm telling you about so I can digest them better myself". The series I'm working on right now is a lot more "I've put a lot of actual work into building a coherent curriculum and I'm writing this for other people rather than just for myself". (Most of it is a more mature version of the noticing stuff you've been trying to get me to post here for ages.) So if you (or anyone else on the LW team) ever do get around to engaging with the strategy directly, I'd love to hear how it goes, and I'd especially love to hear what you wish had been different about my presentation of it. I know how to write for people who specifically subscribe to my personal blogs, but I... really do not know how to write for LW.

A lot of people helped me write this post! They gave me excellent feedback, and it would have been way worse without their help. Thank you so much timepoof, erratio, Phoenix Eliot, Nora Amman, Robin Goins, Matt Goldenberg, and Duncan Sabien. Y'all are the best.

This is a incredibly detailed examination of how to take curiosity further. I haven't try to apply it yet, but it fits broadly with the way I make sense of intuitions about what would be relevant for specific research directions.

I'm also curious about how this relates to my proposal of different types of curiosity stoppers.

Make a list of assumptions

I spent some time on this a few years ago and I'd say it's akin to backing selection effects out of your data, i.e. hard. I have a strong wish for there to be more material in existence outlining how people who are good at this use various heuristics for it.

One other thing that occurred to me reading the post is 'desires aren't curious' and it's interesting to think of all desires as a form of writing the bottom line first (which is fine sometimes, but good to keep track of the side effects).

To your points about phenomenology, I think it would be great if we could better teach the method of phenomenological reduction as Husserl described it. There's a crappy thing people do these days they call "phenomenological reduction" in academic literature that's more like some narrative-conforming version of "examine your assumptions".

Unfortunately I don't know of any great resource to walk people through the method, and Husserl seemed to just intuitively know how to do it such that he couldn't or at least didn't communicate clear, detailed instructions. In particular people tend to get stuck on the second motion of reduction, the reduction proper, or perform the first motion, epoche, in a linguistic-only way that doesn't get to the heart of the experience being bracketed.

This gives me a hypothesis that phenomenological methods aren't more popular among rationalists because there not currently something like "legible" and that makes it hard to evaluate if they do something or not. Sure, we can point to people who used phenomenological or proto-phenomenological methods and tell a story that it was causally important that they used those methods, but I think we also lack stronger legible evidence that those methods mattered and were not superfluous. Cf. the academic mathematician cultural norm of publishing just the proofs with little to not explanation of intuitions.

This is a good point.

I think one of my central original seeing exercises does walk people through (a version of?) phenomenological reduction, and I wonder if you'd agree. It goes like this.

  1. Find an object. Doesn't matter what it is, but this will probably be easier if it's a concrete, physical object you can hold, such as a pen or a box of tissues.

  2. Write a sentence about what the object is and how you're relating it.

  3. Set a timer for two minutes. During that minute, snap your fingers whenever you notice something new about the object, something you haven't already snapped your fingers for. If you're not sure whether something "counts" as worthy of a finger snap, err on the side of inclusion.

  4. After the two minutes are up, list the strategies and tactics you used to get finger snaps.

  5. Then extend the list, including strategies and tactics you didn't use, but perhaps could have, to get even more finger snaps.

  6. Set the timer for another minute, and continue snapping your fingers every time you notice something new about the object.

  7. Notice if any of your snaps felt like they were borderline, especially the ones that felt that way because it wasn't clear that your object is quite what you were snapping about. In the next round, plan to count those sorts of borderline cases as straightforwardly worthy of finger snaps.

  8. Set one more one minute timer, and keep snapping when you notice something new.

  9. When the time is up, write a new sentence about what the object is and how you're relating to it.

(Once you've done this a few times, you can do the whole thing very quickly in one smooth motion, no finger snaps required.)

Edit: A cautionary note may be appropriate here. A couple people who have done this exercise reported "feeling like they're going crazy", or variations on that. I don't think there's actually much danger, but if things start to feel frightening or terribly uncomfortable in the middle of this, I recommend that you stop, do something normal like making dinner, and then evaluate what happened and whether you want to try again.

‘What feels important or exciting but also daunting/confusing/impossible?’

Merely reading this question, inline with everything else, no special pauses or anything, caused me to feel like I do when I recall some of my favorite music swells. Chest grows suddenly warm, and feels heavy-with-significance (not heavy with dread, the good kind of heavy).

Instructions like these may well teach Transformers to generate useful insights.

I'm pretty sure you don't mean these things, but I don't know what you do mean.

I'm just talking about the recent, surprisingly coherent text autocomplete engines like GPT-3. (They wouldn't need help staying motivated/curious, but they could use instruction in concept manipulation.)

They are probably talking about the machine learning model, like GPT-3.

Footnotes (each footnote is a reply to this comment)

Footnote #2: 

If that line made perfect sense to you, go straight to “The Orientation Procedure”. Otherwise, here are a few options.

1. If you’re sort of panicking and just need something object level to grab onto immediately, try to catch a spark of curiosity about “confusion”.

2. If you’re overwhelmed by all the sparks you see and are anxious because you don’t know what kind you’re supposed to catch, take either of the other options, or read the sub-essay in the coda at the very end. (It might help, it might not. If you’re not stuck at this point, I recommend saving the coda for later.)

3. If you’re stuck on, “What even is a spark of curiosity and how do I know one when I see it?”, read the main essay without trying the procedure. Then, watch for things that might be sparks of curiosity in your daily life over the next week or two, and come back to these instructions once you think you may have found one.

Footnote #1:

If this isn’t readily apparent, notice when the direction of your gaze shifts, and ask yourself why it happened. When the answer is of the form, "I wanted to know X" or "I wondered Y", the shift in your gaze was preceded by curiosity.

Footnote #5: 

Unless you happen to be studying burrowing owls, I suppose.

Footnote #4: 

This might be exactly what CFAR calls "boggling". I'm calling it something else, because no matter how many times people try to explain boggling to me, 1) I can't quite understand what it is, and 2) I feel very sure that they are somehow doing it wrong.

Footnote #3: 

I do not know what drums are called. I'm just guessing. Sorry, percussionists.

By the way, if you want to really understand Pythagoras's theorem, Cut The Knot has a collection of 118 proofs.

As for the 'tricky seedlings' example, another question would be:

I would expect plants whose seeds are dispersed by animals to be more k-selected and have less 'tricky' seedlings compared with plants whose seeds are dispersed by the wind. Can I think of any counterexamples?

For the question about mould and seedlings, there are some interesting spin-off questions that you could come up with e.g.

I heard speculation somewhere that endothermy may have evolved to protect against fungal infections. If heat is good at killing off fungi, why do compost heaps work better when they are insulated so that they get hot on the inside?

(And then fungi evolved to cope with endothermy.)