Note: There are two appendices at the end of this post. The first contains a summary of the steps I’ve described here. The second is a glossary of key terms I’ve introduced in this essay.


Once you’ve booted up some curiosity and original seeing—perhaps even identified a question that’s crucial to your story—it’s time to start making observations. At this stage, the purpose of these observations is not so much to find an answer to your question, as to get in closer contact with the world so you’re well positioned to ask better questions.

But how do you determine which parts of the world are relevant to your topic, especially when you know that your basic conceptualization of the issue may be flawed?

Standard approaches to this problem include using a working model to make testable predictions, and seeking expert advice. Here we'll take a more bottom-up approach; instead of investigating large models from the start, or relying on the models of others, naturalism focuses first and foremost on the immediate sense data you can personally gather. You'll direct your data-gathering (even in the absence of any coherent model!) by identifying something I call a fulcrum experience.


What is a fulcrum experience?

A fulcrum experience is a collection of sensations that would lead you to relate differently to your topic if you observed it closely.

For example, when I studied courage[1], “fear” turned out to be a fulcrum experience for me. After really paying attention to experiences of fear in detail, I automatically thought about courage, bravery, cowardice, and related topics quite differently from how I had before. In the past, I’d implicitly assumed that being courageous meant taking actions I am afraid of. By observing fear, I learned about a range of cognitive reactions to the experience of being afraid; I came to think instead that courage has much more to do with where my actions come from, and how that relates to my overall system of values, than with which actions I ultimately take.

Once I was really paying attention to my experiences surrounding fear, my previous conceptualization simply crumbled. After enough study, a new way of thinking gradually emerged, founded more on direct observation than on vague storytelling. Therefore, fear was a fulcrum experience

Locating fulcrum experiences is the first main phase of naturalist study. By “locating fulcrum experiences”, I mean 1) anticipating which experiences will have the power to shift your perspective in such a way that key features of your topic become clearly visible to you, and 2) anticipating where to find those experiences so that you can study them in depth. If you’re able to identify and explore fulcrum experiences, you’ll likely find yourself asking questions based on thoughts you were not previously capable of entertaining.

Suppose I’m concerned about something to do with where my beliefs come from. My current story is that, “Sometimes I believe things because I think they’re true, but sometimes I believe things for other reasons than that, and I’m worried about the beliefs that are indifferent to truth.” My current quest is, “What leads me to form beliefs other than attempts to guess the truth?”

It is of course possible to try answering a question like this by just consulting existing models of myself, or of people in general (“Well, probably motivated reasoning!”), but that’s not what comes next in the particular strategy I’m discussing. In this strategy, rather than doing my best to answer the question with whatever information I already have, the purpose of the question is to lead me toward fresh observations.

It’s often easy to make fresh observations: On the chair in front of me, the pattern of the weave makes many little squares. That’s an observation, and it’s one I’ve never made before.

The hard part is figuring out where to look. I’m relatively unlikely to learn much about “what leads me to form beliefs” by admiring the weave of my chair. “Identifying fulcrum experiences” means figuring out exactly when to pay attention so I can make observations that matter.


Guessing Which Observations Will Matter

To begin narrowing in on a fulcrum experience, ask yourself, “Where do the data live?”

If I’m interested in the linguistic features of Hoosier dialects, then the data live in Indiana, and specifically in the conversation of people born and raised there. If I’m interested in what causes some things to float while others sink, crucial data may live near the surface of a pond. Where do the crucial data live when I’m interested in “what leads me to form beliefs, besides attempts to guess the truth”?

Occasionally, a good answer to this kind of question will just spring to mind for me. More often, nothing is forthcoming. When that happens, I usually consult my memory in search of a reference experience: a situation I can walk through in my mind as reference material for generating guesses. I ask myself, “When has something relevant to this happened in the past?” 

When I look for a reference experience in my own memory related to “forming beliefs in ways other than guessing the truth,” I’m reminded of the time when I got a bad grade on my very first logic exam in college. I was pretty scared of math-like things due to awful experiences in highschool math classes, but this logic course was a requirement for the philosophy major, so I’d enrolled anyway. Since I was accustomed to getting straight A’s and I was also pretty set on majoring in philosophy, the poor test result hit me pretty hard.

After class, I remember going to the empty chapel across the street to be alone and cry. I had thoughts like, “I’ll just have to drop this class,” “I can’t be a philosopher,” and “If I just focus on religious studies [my other major], I’ll have a much more reasonable reading load anyway.” 

Beck Chapel at IU Bloomington

I think I was trying to find a shape for my mind to rest in that might comfortably accommodate this new information that I can’t automatically ace all my logic exams. I was forming beliefs about my abilities and best options, and whatever was driving that process had a lot to do with attempting to relieve psychological discomfort.

This reference experience suggests that if I want to learn about what leads me to form beliefs, besides attempts to guess the truth, crucial data might live in my attempts to relieve psychological discomfort[2].

Rather than stopping with your first guess at a reference experience, it’s usually worthwhile to search for several different reference experiences, then to make a list of possible fulcrum experiences suggested by them. There might be several places where crucial data live, and it’s often worthwhile to consider a range of options before narrowing in on just one. 

When I make a list of guesses about where I might find data on truth-indifferent belief formation, here’s what I get:

  • attempts to relieve psychological discomfort
  • decisions made while I’m feeling pushed or yanked around by something
  • attempts to reason while I’m aware that other people are watching
  • times when I’m considering multiple possibilities, some of which seem much better or worse than others
  • arguments on the internet
  • times when I’m receiving critical feedback

I often check how fulcrum-y a certain experience seems by using it to fill in the exclamation, "If only I really understood what was happening in the moments when __!" For example, “If only I really understood what was happening in the moments when I’m attempting to relieve psychological discomfort!” When I’ve chosen a really fulcrum-y sort of experience, completing that sentence tends to create a feeling of possibility; sometimes it's almost like a gigantic tome falling open. Not always, but often.


Preparing To Recognize Experiences As They Happen

Once you’ve got at least one guess at where the data live, the next question to ask is, “How will I know when I’m having the experience I’ve identified?” How will you know when you may have entered the natural habitat of crucial data?

To be clear, I’m talking about observing experiences in the wild, as they naturally occur. I’m not talking about designing experiments to elicit the experiences. By “How will I know when I’m having the experience I’ve identified?”, I mean, when you are going about your day and not thinking at all about this essay or your chosen quest, what could tip you off that it’s suddenly time to pay attention? 

In the case of the example, what could I notice in my immediate experience, while I’m doing the dishes or talking to someone at work, that would suggest I might be attempting to relieve psychological discomfort?

To answer this question, I typically use two different strategies, which approach the issue from opposite directions.

1. Consult the Reference Experience

The first strategy is to review the same reference experience(s) you used before. Play back through the experience in your mind, paying extra attention to what exactly the most important moments in the memory felt like from the inside. What was the character of your thoughts? What were you paying attention to? What sort of emotions were you having? Was anything notable going on with your body?

In my memory of the logic exam, the moments that seem most important happened in the chapel. The character of my thoughts and emotions was writhing, reaching, and contorting. So I might be able to recognize a similar experience by a feeling of writhing, reaching, or contorting inside my mind.

However, my reaction to the exam result was an especially intense version of the kind of experience I’m interested in. That is why it stood out to me as exemplary, even a decade later. I may be unlikely to encounter an experience with quite the same phenomenological character in the upcoming week.

2. Consult the Present Moment

To find the other end of the intensity spectrum for a given experience type, employ a different strategy: Identify the nearest analogue of the experience that you're able to perceive right now.

At the moment, I’m definitely not in crisis the way I was after the exam. Overall, I wouldn’t describe my mind as writhing, reaching, and contorting. But that doesn’t mean there are no similarities between my current experience and that one. 

For example, I’m certainly experiencing something I’d describe as reaching: I’m reaching for the next words to write, stretching forward toward my understanding of my topic, language, and the mind of the reader. Now that I’m tuned into that, I can even tell that this sort of “reaching” is a dominant part of my current experience; I just didn’t recognize that at first because it’s so much quieter than the version in my reference experience. 

I’m also aware of some discomfort: Physically, I’m a little restless. While I’m paying attention to it, I can even notice myself making tiny adjustments so that I can continue writing for a while longer despite this. I’m adjusting (a tiny version of contorting, perhaps) by taking sips of water, changing the way I lean in my chair, and exerting a sort of squeezing pressure on my thoughts to “stay focused” on my task. The adjustments feel a bit like running away, or like trying to ignore something.

Considering these subtle sensations beside the much louder ones from my reference experience, I can make a well informed initial guess at what “attempting to relieve psychological discomfort” might feel like in the most typical instance: I may notice that I’m trying to ignore something or that part of me is running away, and my mind might be reaching toward a certain direction at the same time. 

So in this case, my answer to the question, “How will I know when I may be in the presence of crucial data?” is “I may notice ignoring, running away, and reaching.”


Choosing Among Possible Fulcra

If I’ve used a few different reference experiences to identify possible fulcra, then I probably have a list of sensations I could watch for. 

When I extend my own list a little bit for “signs that I may be attempting to relieve psychological discomfort”, I get this:

  • ignoring something
  • running away from something
  • reaching for a particular outcome
  • a wrenching feeling accompanied by time pressure
  • puppeteering myself

I’d never watch for all of these at once; it’s usually better to choose one to three. There are a few good ways to choose. 


1. Choose an experience that’s common and obvious.

One way is to choose something that will definitely happen, and that you think will be pretty easy to notice. 

The benefit of using this criterion is that you’ll get in close contact with some bit of territory quickly and easily. Making some kind of contact, any contact at all, is the most important thing in naturalism, so picking the easy thing tends to be a solid strategy. The main downside of picking something certain and easy is that it’ll likely happen very often, in which case you’ll need to start narrowing your focus almost immediately, or else be overwhelmed.

In my list above, I’d choose “reaching for a particular outcome” if I wanted to go with something easy. I expect that some version of this is happening almost all of the time, because I think it’s probably central to the experience of decision making. If I’m right, I’ll have as many opportunities to notice it as decisions to make.


2. Choose an experience that excites you.

A second way to choose is to pick something that feels bright, juicy, or exciting. 

The benefits of this criterion are that 1) you’ll be more motivated to keep an eye out; and that 2) the thing seems juicy for a reason, and the reason might be a good one. The downside is that exciting things seem exciting from your current perspective, which might not be very trustworthy. Sometimes picking the exciting option biases you toward thinking about things in a way that’s already familiar. Still, I don’t think the danger is all that great most of the time, and going with whatever feels brightest is usually a solid option.

A juicy item on the list, for me, is “a wrenching feeling accompanied by time pressure”, but note that compound experiences like this present a special kind of challenge: To get good at noticing “a wrenching feeling accompanied by time pressure”, I would likely need to train attention to “wrenching” by itself, then “time pressure” by itself, before I began combining. This takes a bit longer, but it’s frequently worthwhile. Often I find that I’ve learned a lot from studying just one component that didn’t seem interesting at first on its own.


3. Choose an experience that confuses you.

A criterion I use a lot, but that I think of as more advanced, is to choose one of the poorly formed guesses. 

Take “puppeteering myself” from the list above, for example. This was my guess at what it would feel like to try to reason while I’m aware that other people are watching; but I’m not very clear on what I mean by it. It’s an extremely hand-wavy sort of guess. I wrote it on the list anyway because it seemed important, even though I couldn’t easily pin down what it might feel like in the moment. If I’d written this by hand to myself (instead of for an audience), it’s likely I would have written “puppeteering myself(??????)” instead of “puppeteering myself”, since I’m aware that I don’t really know what it is. This uncertainty indicates a known unknown, a clearly marked hole in my map where I’m pretty confident I need to fill things in.

The benefit of choosing a poorly formed guess is that there’s a very good chance I’ll make important progress right away, provided I manage to find any traction at all. The downside, of course, is that finding any traction at all with a poorly formed guess tends to require a lot of skill, and sometimes it just doesn’t work out. Basically, it requires that you practice every piece of the first phase of naturalism all at once and continuously. I use this criterion all the time myself, but I almost never recommend it when I’m guiding someone else through their first naturalist study.


Once I’ve chosen a possible fulcrum experience, what do I do with it?

At first, all you do with a possible fulcrum experience is watch for it. Try to notice when it happens.

I always make a plan to physically mark the event in some way, though this isn’t right for everyone. For example, I could tap my leg when I notice it, or wiggle my fingers. I call this action a marking gesture. Whatever my default response to the experience might be, taking some time to mark the experience with a neutral motion creates an extra little moment in which I can just observe. Many people also find that the right marking gesture makes it much easier to remember they’ve noticed, when looking back over their day.

But if you don’t like making a physical gesture, you can try using a phrase in your head, such as “I noticed.” What matters is that you notice that you’ve noticed.

Marking gestures (or marking thoughts) are especially valuable if you’re studying an experience that’s associated with unpleasant emotions, or with an urgency to fix or change something. For example, someone who’s studying experiences related to criticizing their partner might by default try to ignore the experience, perhaps because they’re ashamed of it. If you tap your leg when you notice “criticizing”, then there’s at least a moment where you get to remind yourself that an act of criticism and an act of observation are distinct, that the observation indicates success in one realm, even if the criticism indicates failure in some other realm.

In this earliest stage of observation, you might not stay with a particular plan for more than a day or two. Your guiding star during this phase should be a question like, “Is there something critical here?” or “Does this feel bright and shining to me?”

Suppose that on day one, I snap my fingers whenever I notice “ignoring something”. But by the end of the day, I suspect that I haven’t chosen quite the right thing to pay attention to. Perhaps what matters isn’t so much ignoring as something to do with “actively looking away” or maybe “struggling not to see”. So I make a new plan: I’ll tap my leg whenever I notice that I’m “actively looking away”.

There are a few ways you might end up suspecting that you haven’t been studying quite the right thing. Often, the way this happens is that you'll find yourself tapping your leg for some other kind of experience instead, something that you think probably “shouldn’t count” as whatever you're watching for. In the moment, you might have no idea why you've tapped my leg; but in my experience (both personally and with students), the experience frequently turns out to be more crucial than whatever you'd planned to study.

(I think this "marking things that shouldn't count" phenomenon is an especially excellent example of contact with the territory at work. I have a story that it happens when your perceptions are updating faster than, and largely independently of, of your conceptualizations, in response to relatively unfiltered observations.)

The other main way you might end up switching focus in this phase is by some kind of daily or weekly check-in, where you just ask yourself what has happened so far and how you think things are going. If your study feels dull and unfruitful, that’s probably a sign you should try something different. 

You could pick your next focus by asking yourself what key features were missing in the experiences you've studied, or what features they could conceivably have had that might have made them bright instead of dull. Make your next guess at a possible fulcrum experience by studying something that might have whatever the previous experience lacked.

This guess-and-check process is how you'll navigate toward a fulcrum experience. Even if your first guess isn’t quite right, over time you may be able feel your way toward a true source of crucial data.



Note: Many of these essays have a “troubleshooting” section. For the sake of brevity, I’ve chosen to shoot only three troubles per essay. If your main concern is not among them, I assure you that this probably does not mean you are uniquely confused or incompetent; it only means I ran out of space.


Problem 1: What if I’ve chosen a reference experience that I don’t want to pay attention to?

Although the ability to observe anything and everything with clarity is an extremely naturalism-flavored aspiration, getting there doesn’t necessarily mean staring directly at the most painful stuff from the outset. Gentleness and patience tend to prove more effective than ruthlessness with this kind of thing, especially in the long run.

When my topic is emotionally fraught and I don’t much feel like torturing myself with it, here are a few approaches I’ve found useful.

Approach 1: Study the opposite. Rather than observing “running away from something”, try observing “running toward something” instead. Rather than observing “the sinking feeling of realizing you’ve failed”, try observing “the uplifting feeling of realizing you’re succeeded”. Crucial data often live in inverse experiences as well.

Approach 2: Study something related but more pleasant. You might make a lot of progress by gradually circling around your topic, studying the systems surrounding it rather than the thing itself. If you’re interested in motivated reasoning, you could circle around that by studying reasoning more generally, motivation more generally, or goal-directed action. You’ll probably find components of motivated reasoning in these related topics, so that by the time you turn your attention to motivated reasoning in particular, much of what you find will be familiar, and perhaps not so hard to cope with.

Approach 3: Take a step back and study the resistance itself. If there’s some kind of powerful cognitive force compelling you not to observe something, it may be extremely valuable to get to know that force intimately. I’ve found that almost nothing is more likely to shift my relationship with a topic than closely observing the way my mind behaves around it. Try setting aside your curiosity about the topic itself for a while, and get curious about your own patterns of experience around the topic instead.

Approach 4: Build a toy. I’ll talk about toys in depth later in the sequence, but here’s a preview. A toy is a low-stakes scenario that affords opportunities to observe a chosen experience in a way that’s thoroughly under your control. 

While studying experiences of failure, for example, I played 2048 on my phone, and paid attention to all of my tiny mistakes. It didn’t really matter much that I was messing up, and it was easy for me to pause and resume at any time; yet there were many opportunities to observe a small, relatively innocuous version of the experience I wanted to study. 

Sometimes, starting with a toy is a good way to begin, because it lays a foundation for observation without throwing you straight into the deep-end of real-life experience.


Problem 2: What if I can’t find the right memory to use as a reference experience?

A reference experience doesn’t really have to be a memory. Memories are best when they’re available, but when they’re not, you can use anything that allows you to walk through an imagined scenario vividly, paying attention to the details of the experience. The core of this technique is about booting up System 1 and inviting those processes to steer, rather than leaving everything to explicit modeling, so the presence of phenomenological detail is more important than accuracy.

Option 1: Try making up a scenario that could happen, as though you’re telling a story, even if nothing like that has really happened to you before. Let it play through your head like you’re watching a movie, and let that movie be the reference experience that informs your guesses.

Option 2: See if you can remember a relevant scenario from a work of fiction, such as a scene from a book or movie. Imagine yourself inside the head of one of the characters, and ask yourself what exactly is going on for them.

Option 3: Talk to other people about their experiences. Try to get them to tell you relevant stories from their own lives.

I sometimes find that supplementing my work with a combination of these options is valuable even if an appropriate memory sprang to mind for me immediately.


Problem 3: What if I think my fulcrum experience will never happen, or that it will happen too infrequently to be useful?

For example, what if I’m watching for experiences of “running away from something”, but I’m skeptical that this will happen to me much in the upcoming week?

Option 1: My first recommendation in this case is usually to wait three days, just in case you turn out to be wrong by at least an order of magnitude (which people sometimes are). If you’ve already spent your entire past failing to observe a certain experience, you probably won’t lose much on the margin by continuing to fail for three more days. You can always make a new plan after that.

Option 2: Plan to look for smaller versions. Rare experiences are usually composed of common components. Their novelty almost always comes from the particular arrangement or intensity of their component phenomena. By studying component phenomena piecemeal, or by studying similar experiences that are lower on the intensity spectrum, it is frequently possible to find crucial data on rare experiences while observing the phenomena afforded by just a few days of ordinary life.

To increase your sensitivity and observe lower-intensity versions, use the exercise I describe above: Identify the nearest analogue of the experience that you’re able to perceive right now. Or, use an evening check-in to expand this exercise: Think back through your day and identify the nearest analogue you’re able to perceive in the memories you’ve formed since waking up.

To study smaller components of a large experience, take a guess at which parts the large experience is made of, then pick just one little piece to watch for. Rather than the compound experience “writhing, reaching, and contorting”, try to watch only for “writhing”.

Option 3: Study something related but more common. Rather than studying a rare experience like “doubting your sanity”, try studying related every-day experiences like “feeling confused” or “being uncertain”. Even if you start out with something quite far from the experience you’re interested in, you might be able to move toward more closely related experiences once you’re in touch with better data.


What does this part of the process feel like?

“Locating fulcrum experiences” is characterized by meandering and honing. 

When studying something that involves a lot of unpleasant experiences, this phase can feel like confronting a hopelessly tangled mass. One person I was working with described this approach as “massaging conditioner into matted hair and carefully teasing out the knots, when part of me just wants to get a razor and cut it all off”.

At other times, it feels to me a lot like walking a labyrinth (the spiritual kind, not the maze kind), moving toward the center and then looping back out again, working my way inward in a series of patient spirals. You're getting in contact with pieces of the territory, one after another, feeling around in a sensitive and direct way for what matters. Sometimes the meandering goes on for just a few days, but it’s common for this phase to continue for two to four weeks before you find something worth studying in depth.

The whole time, it's good to check back in with earlier parts of the process: What is my story? What does it rest on? What is my quest? Where do the data live? What is it like there? Is there something to learn here? Could this impact my story?


When should I move on?

I said at the beginning that in phase one, you’ll begin observing, not so much to find an answer to your question as to position yourself to ask better questions. That position is one of closer contact with the territory, and you’re ready to move on when you have made some contact.

There’s rarely a sharp distinction between phase one and phase two of naturalism. It all blends together. But insofar as there’s a distinction, you are probably ready to move all the way into phase two when you seem to have latched onto a certain kind of experience. One way to tell you’ve latched on is that you have a lot of questions, and the questions are all about what’s going on with [whatever]. When you ask yourself whether you’d like to continue studying that experience, or to keep poking around elsewhere some more, you find that you want to stay with it.

When that happens, it’s time for phase two: Getting Your Eyes On.

Appendix 1: Process Summary

  1. Guess which observations will matter.
    1. Ask, “Where do the data live?”
      1. Ask, “When has something relevant to this happened in the past?”
    2. Test your guess by filling in, “If only I really understood what was happening in the moments when __!"
  2. Prepare to recognize your fulcrum experience in real time.
    1. Ask, “How will I know when I’m having the experience I’ve identified?”
      1. Review your reference experience, and pay attention to what the most important moments felt like from the inside.
      2. Identify the nearest analogue of the experience that you’re able to perceive right now.
  3. If you have a list of guesses for how to know you’re having a fulcrum experience, choose one that
    1. is common and obvious,
    2. excites you, or
    3. confuses you.

Appendix 2: Glossary

Fulcrum experience: A collection of sensations that would lead you to relate differently to your topic if you observed it closely; the location of crucial information.

Reference experience: A situation you can walk through in your mind, and use as reference material for making guesses.

Marking gesture: A neutral action you take when you notice the experience you've been watching for, to help you notice that you've noticed.

Locating Fulcrum Experiences: The first phase of naturalism, in which you make informed guesses at which subjective experiences intersect with crucial information.



  1. ^

    I think that courage is among the most foundational skills of epistemic rationality; but I suspect this is not immediately obvious to many readers, and I didn't want to make this sequence mainly an attempt to communicate about courage. I decided to mostly use examples that are more obviously rationality-flavored. I left in this one last courage-related example because it feels important to me to point out the following: Inasmuch as rationality involves thinking clearly while uncertain or otherwise threatened, the ability to act with integrity while afraid is indispensable to a rationalist.

  2. ^

    You may recognize that “attempts to avoid psychological discomfort through directed belief formation” is basically another way of saying “motivated reasoning”, the first hypothesis my existing model supplied in the previous section. But can you see how this new way of expressing it comes at the phenomenon from a different direction? From the inside, perhaps, instead of from the outside. Or from the bottom up, instead of from the top down. For this reason, it will prove far more valuable in a naturalist-style study.

New to LessWrong?

New Comment
31 comments, sorted by Click to highlight new comments since: Today at 6:18 PM

Am I the only one who earnestly tried to make sense of this and previous posts and couldn't, beyond the obvious "I pay attention and notice things"?

Ok, I'm going to put some effort into building a bridge here, even without further detail about where things are going wrong for you. It'll mostly be shots in the dark, so I will probably miss. But I expect this is a pretty common response, so I'll give it a try anyway.

I'll start by focusing on, "What does this particular essay contain besides 'I pay attention and notice things'?"

According to me, this essay is not about "I pay attention and notice things". This essay is about "I pay attention to *particular* things, I employ *particular strategies* for deciding what to pay attention to, and I make very careful guesses about *what* I will notice so that I am able to pay attention to the right things at the right times."

I rather doubt that was anywhere near sufficient to build a bridge between this essay and wherever you are, let alone between you and all the rest of my naturalism writings, so in my next (probably much more sprawling) comment I will try to dig into some of the implicit stuff underneath my summary. (However I would love to hear whether this comment on its own was at all helpful to you, in case I'm wrong.)

I will now try to communicate an implicit conjecture in this essay that I think of as "Conservation of Attention". This may possibly speak more directly to @spxtr, who suggested that our internal models of cognition may be dramatically different. I apologize that I will probably do this in a rather round-about way; "If I had more time, I would have written a shorter letter."

Conservation of Attention states (roughly) that attention can be redistributed, but not increased.

I think that this very short, rough statement of the conjecture is technically false, in the following way: Brains and bodies share resources, and it is indeed possible to increase or decrease the overall resources available to the entire system, or to shift resources between cognitive and physiological processes. For example, I expect that attention really does decrease overall during starvation, and increase overall during recovery from starvation (or perhaps even during recover from moderately low blood sugar). According to my understanding of what's up with caffeine, as adenosine builds up over the course of the day, it decreases the possible *expenditure* of resources by the body and mind, regardless of what resources are physically available; so a shot of caffeine (which binds to and thereby incapacitates the adenosine receptors) may increase attention overall by widening the valve of ATP expenditure (or something along those lines). I also expect that sufficiently intense exercise tends to reduce the availability of cognitive resources, as those resources are redistributed to the muscular and cardiovascular systems. So in fact, attention can be increased or decreased globally by the activities of the rest of the body.

(There is also something going on with certain drugs, such as mescaline especially but IME also high doses of THC, that looks on the surface a lot like "increasing overall attention", perhaps by a caffeine-like reduction-of-restrictive-mechanisms that is far more precisely targeted to whatever cognitive processes I'm pointing toward when I say "attention". But my model of psychoactive drugs feels to me even more shaky than my model of metabolism, and anyway I'm pretty sure my observations are also consistent with a model where attention is merely redistributed in unusual ways during drug trips.)

The truer version of my conjecture runs more like this:

1) Every human contains a control system that is constantly governing the expenditure of resources on attention;

2) the reference point of the attention expenditure control system is an output of larger control systems that ensure overall economy of resource expenditure in a mind and body; and

3) the economics of cognitive resources are determined by the structure of the mind/body system, and perhaps also by the patterns of environmental stimuli a human encounters.

4) Therefore, it is at least extremely difficult and perhaps impossible to globally increase a human's attention in the long term without dramatically changing their overall life circumstances.

5) However, temporary reallocation of attention is clearly possible, and can be accomplished deliberately and therefore strategically.

Why do I lay out this model here?

Because I do not think that any part of what I do in naturalism involves "paying more attention" than anyone else, including my past self. I think it is generally quite ridiculous to "try to pay more attention, in general". I won't go so far as to say that I'm certain it can't be done, but I think that it at minimum requires an enormous lifestyle change that is probably not a good idea for most people most of the time.

(Indeed, I think this is a common failure mode among people who "get into mindfulness". They mistake the reallocation of attention for an increase in attention, and thereby fail to take deliberate, strategic control of their attention. They reallocate in ways that may tend to reduce stress, but there is so much more to be gained from the deliberate development of novel patterns of attention allocation over time.)

Similarly, I do not "notice more things [in general]" (although it may be the case that I more frequently devote more attention to the things that I notice). According to my model, "noticing" is the rapid reallocation of attention. A person "notices" something when they suddenly start paying much more attention to it than they were paying moments before.

According to my tentative working model of ADHD, ADHDers really do "notice more things": their attention reallocates more frequently and more rapidly. Autistics, by contrast, tend to reallocate more of our attention, on average, when reallocation occurs. (I have sometimes been accused of "trying to make everyone autistic", and that would perhaps be an accurate assessment if I were suggesting that people "pay more attention" or "notice things more". But I'm not suggesting that! I am suggesting much more strategic reallocation of attention.)

So in these terms, "Locating Fulcrum Experiences" is a method of preparing yourself to rapidly reallocate a large amount of attention away from its default state and toward a carefully chosen set of stimuli in precisely the moments when you are a tiny little bit aware that the set of stimuli is present.

Which (further implicit conjecture) is not necessarily easy or straightforward! (Perhaps this is where we disagree?) And thus I have written an essay explaining how to do it.

This essay is about "I pay attention to *particular* things, I employ *particular strategies* for deciding what to pay attention to, and I make very careful guesses about *what* I will notice so that I am able to pay attention to the right things at the right times."

Right, I got that. Hence the fulcrum metaphor. I would really like to be able to figure out these particular things and learn these particular strategies, though mostly in a research application, where there seem to be unnoticed gems hiding in plain sight.

Almost certainly not.

Could you perhaps highlight a particular phrase that baffles you?

I suspect my experience is somewhat similar to shminux's.

I simply can't follow these posts, and the experience of reading them feels odd, and even off-putting at times (in an uncanny valley sort of way). At the same time, I can see that a number of people in the comments are saying that they find great value in them.

My first guess as to why I had trouble with them was that there are basically no concrete examples given, but now I don't think that's the reason. Personally, I get a strong sense of "I must be making some sort of typical mind fallacy" here. Something about our world views, or internal models of cognition, or something along those lines must be dramatically different. It's hard to communicate just how difficult it is for me to make sense of these posts without being rude.


Thanks for the interesting read, and I love the watercolors.

I... wish I could find a particular phrase. It's like I am trying to find that fulcrum (lever?) that would shift my perspective in a way you describe so that I finally get what you are talking about, but it keeps escaping me the moment I zoom in on anything in particular.

Just to check, when you say "this and previous posts", do you mean all the stuff in Intro To Naturalism as well, or are you trying to make sense of just the posts so far published in Nuts and Bolts on their own?

I went through both, and tried to see what I am missing here, since I remember then your earlier writeups on noticing resonated with me back then. Sadly... I have made no progress.

I realize you totally did ask me to review a draft copy of this and I didn't get around to it, but a pedagogical thing I notice here is that it'd probably have been useful to me (and I anticipate to others who follow your work less closely) to have two examples you were working through, which were as different as possible. 

The example of "why do I form beliefs?" here is an introspective one. While it's entangled with the outside world, the active territory you're studying mostly lives inside you. If the domain you're studying is more of a concrete thing that lives "out there in the world", how do you engage with it? (This is similar to a question I previously asked about "okay, but, like, how would I Do Naturalism to learning about neural networks?". You did have an answer which made sense)

A lot of the examples and mechanisms here feel fairly Logan-flavored, and one side effect of that is I'm not sure which of that flavor is more intrinsic to the technique, and which is just an artifact of "Logan wrote this as opposed to some other naturalist".

(I think by this point I've got enough of a handle on things to answer this sort of question for myself, but it's something I'd anticipate someone getting confused about, maybe without noticing they were confused).

This passage touched a bit on the thing I'm interested in here:

If I’m interested in what causes some things to float while others sink, crucial data may live near the surface of a pond. Where do the crucial data live when I’m interested in “what leads me to form beliefs, besides attempts to guess the truth”?

and an example of a "parallel example" here might have been "what's up with buoyancy and/or gravity?", and following that through the different subsections might have been helpful.

Yes, that seems right to me.

However, I've deliberately focused on "training rationality" in this sequence, so I think it makes sense for pretty much all of the examples to be "internal" in this way? This isn't the only way to use naturalism, but it's the use case that I'm most personally interested in focusing on, and the one where I have the most PCK.

Gotcha. I'm not sure if you already did and I missed it, but it might be helpful to spell that out somewhat in the intro. (I think it might even make sense for the sequence to have a somewhat more specific name if that's the goal)

Hm.  Well, the very first sentence of the sequence is, "Naturalism is a general-purpose procedure for advancing one’s art of rationality." Which is true but incomplete, like most statements I know how to make, but it does seem to me to frame this sequence appropriately. But apparently I was wrong in your case, at minimum!

I think I can make some small edits to that essay under "What is this sequence?" that will help.

I think one thing is that I'm currently thinking a lot in terms of "a good art of rationality should involve contact with the territory [of things other than your mind and the rationality-or-lack-thereof-that-lives-inside it]" (which I think is a thing you also believe?)

I was expecting the sequence to primarily (or at least have a significant focus) on studying something-other-than-your-mind, while also having good self-awareness in the process. 

Hm, perhaps you imagine that I'm about to spend a whole sequence advocating for studying your mind without engaging with the world?

I suspect you're employing a distinction that I do not honor.

I mean, to use the example in the essay, when are these fulcrum experiences of eg "puppeteering myself" going to show up? Not when I'm sitting in a lounge chair introspecting. Most of what I am actually doing with myself lately is 1) writing, editing, and publishing essays (and making paintings for them), 2) navigating the third trimester of pregnancy (dealing with a ton of stuff about how my body works, learning about birth, learning about tiny children, working with midwives and doulas and so forth), and 3) setting up my environment so that it will support me and my family when we have very little attention to spare (building garden boxes, getting the wild-bird-feeding arrangement just right so I don't have to fix it later, setting up my recovery space, improving systems around food and other maintenance activities, other stuff). So when I notice "puppeteering myself" (if I choose to study that), it will happen when I realize that someone more skilled at construction than me is watching me secure hardware cloth with fence staples. It will happen when a commenter points out a possible flaw in my attempt to navigate the tricky balance between fidelity and brevity in written examples. It will happen when my bush strokes do not accomplish the visual effect I believe I have imagined, because I was in fact imagining a social role rather than any particular physical shape or hue. It will happen when I have a disagreement with my primary care doctor about the risks and benefits of taking a certain medication at this point in pregnancy. It will happen when I try to figure out why the Stellar's Jay near my feeder keeps making sounds that imitate a red-tailed hawk.

The entire time, no matter whether you topic seems more stereotypically "internal" or "external", naturalism it is really about your internal experience in contact with the external world. There's not really such a thing as an internal topic or an external topic. The methodology focuses on the interface.

hmm, this way this convo is going feels... maybe more fighty than I intended (or at least wanted in retrospect)

I do think "I'm holding a distinction that you don't hold" is probably relevant, and I am interested in talking about it more. But maybe what I want to do next is go reread the previous post and actually think about the sort of things I might want to study.

>I think one thing is that I'm currently thinking a lot in terms of "a good art of rationality should involve contact with the territory [of things other than your mind and the rationality-or-lack-thereof-that-lives-inside it]" (which I think is a thing you also believe?)

Yes, I do indeed think that a good art of rationality should involve contact with mind-independent territory. Constantly. Relentlessly. I... I think that's a thesis of the naturalist program? It is why I'm all "and you will observe these things in daily life: as you engage with your projects at work (which may involve coding, or math, or cooking, or whatever it is you do), as you read bedtime stories to your children, as you learn underwater basket weaving, whatever." Like, do not just sit here and read and think until you think you have things figured out inside of your head. Go do stuff, go try to understand how the world works in practice. Not in the context of this essay, not in the context of a one hour class or a four day workshop, but in the context of how you actually navigate the world on a daily basis.

I'm not sure how to respond to this, and for some reason I want to know if @Duncan_Sabien has anything to say.

Claim: this sequence is almost one hundred percent about studying something other than your mind, and what's happening is a confusion between tools and purposes.

At a very coarse/gross level of understanding, the way that we gather information about objects is by hurling other objects at them, and watching the interaction. This is one way to think about light—we throw trillions of tiny photons at an object, and the way they bounce off gives us information about the object (its location, shape, surface properties, etc).

Ditto sound waves, now that I think of it (the photon analogy is from Brian Greene's The Elegant Universe).

The key point is that we never quite interact directly with the object. We do on human scales; there's a thing we call "direct interaction" that makes sense to talk about. But actually what's going on is that we're perceiving photons that are out there hurtling through the void, and constructing understanding via extrapolation about what those photons interacted with a fraction of a second earlier.

We don't talk about "studying the photons" when we describe looking at an object, though. We gloss over that step, handwave it away.

This sequence is, I think, about looking at things other than your brain.

But it focuses on the analogue of the photons themselves. It's looking through your own phenomenology, to understand what's really going on out there. It's saying (roughly) "notice how these photons bounce off this way, and these other photons bounce off that way, and these other photons get absorbed, and see what you can reasonably conclude about the object, given those facts."

So there's a heavy focus on your own perceptions, and your emotional reactions, and so forth, but it's in service of understanding the object that is upstream of [your brain reacting in such a way].

Honestly, I was somewhat surprised to hear @Raemon 's complaint, and at first bewildered/taken aback, because it hadn't even occurred to me that this sequence might be mistaken for being about minds/brains. But of course it does talk a lot about the internals of one's experience, so I understand the confusion! Ray's complaint isn't coming out of nowhere!

But to tack on yet another analogy, I feel sort of like just ... reassuring the complaint away? In the same way that, if I were teaching a parkour class and a student was like, wait, why are we doing pushups and stretches, I thought we were here to do vaults, I would be like yes, yes, don't worry, we are absolutely getting to the vaults, but this is important preparation for the vaults, and will help you build up the strength and physical vocabulary necessary to be non-lost once we start working on the vaults, which is coming right up, actually.

@Raemon FYI there isn't internet at our place since ~26h ago so Logan probably hasn't looked at this or any other responses yet.

I don't know that a response here is specifically needed. I thought about adding a line "to be clear, I think there's clearly enough content to warrant a course/practice/sequence focused on the thing I now-understand-this-sequence-to-be-about, so, like, I'm not saying you should have written a different thing or whatever."

I meant my second paragraph to just be "here's what I was expecting, fwiw" (without claiming that was, like, an appropriate thing to expect) and the first paragraph to be "here's some model I have on my mind that I'd interested in talking about, as part of a longer term conversation about how to push the art of rationality forward", but, not necessarily super relevant to this sequence.

Something got weird in the way the other thread went, and I'd like to try rebooting from this save-game-point and trying again?

In Catching the Spark, you have two examples – geometry, and courage. I think it was useful to have a) two different examples (generally having multiple examples is useful for triangulating. Three examples is better for the actual "triangulation" part, since readers have fewer degrees of freedom to accidentally misinterpret things in, although each example is kinda costly both from pure writing time and the final post being longer, so, dunno if that's worth it)

Something that feels particularly helpful about geometry and courage is that one is starting from more of an outward facing direction, and the other a more inward facing direction. In both cases, learning about the thing will include both external stimuli and internal thought processes. But, I expect for me, and I predict for some others, seeing how noticing external-facing and internal-facing stimuli might differ (and how it might be the same) is probably useful for actually learning the lessons you intended to convey.

As previously stated, I think by this point I have some sense of what things you tend-to-mean to be common to both or distinct, so I don't feel that confused for my own sake, but I think past-me would have been confused.

More generally, whether or not my ontology of inward facing or outward facing is even relevant here, the general principle of "pick two examples that are as different as possible while still being fairly central examples of the concept you're trying to convey" seems useful, so that people learn the thing at whatever-level-of-generality you were aiming at, rather than either overfitting or assuming it applies to literally everything.

I continue to have a feeling like "I don't know how to interact with this without either falling into Ray's ill-fitting framework and thus further confusing both of us, or completely failing to engage with his framework".  On the other hand I'm not sure I have tried, and perhaps I ought to try. Experience suggests that I will only succeed if I first ensure I don't lose sight of whatever it is I can see from my current perspective, which is probably much of what I was up to in the previous thread, and that is probably why it seemed to you to "get weird"; I was halfway trying to say the relevant truths as I see them, and halfway trying to communicate with you, and I was not doing either of those things well as a result.

I do completely agree with everything @Duncan said in the other thread, and have said very similar things myself in the past, which perhaps I will post below as like supplementary reading or something. But I think it's missing you because when I imagine you reading it I also imagine you responding with "But there really is a difference between trying to learn about things like 'updating incrementally' and trying to learn about things like 'electromagnetism'." 

And I think I agree with (my imagined version of) you. Even if learning about things like "updating incrementally" requires learning about things like "electromagnetism" (which it does, at least in my own rationality framework and I suspect also in Eliezer's, for whatever that's worth), there is something different about setting out to study "updating incrementally" vs setting out to study "electromagnetism".

What is different, according to me, is that studying electromagnetism by roughly naturalist methodology is way easier. It's easier because the thing you're trying to study is far less entangled with the tools you are using to study it. Things outside of the mind stay put in a way that things inside the mind do not. Which is why I made How To Observe Abstract Objects.

And it's also most of why I made the course I talked about in the previous essay, and why I said that I wish everyone would start there. It's almost the same set of tools, but it's aimed at nature instead of at minds. This sequence is that course, but beefed up a bunch to contend with the unique difficulties of turning the cognitive tools on themselves, or something along those lines.

But I feel that I'm wandering away from communicating with you again. I think you approximately think that I should interweave my nature study course with this rationality sequence, because it would help people better understand what "naturalism" is, as a whole, and how to apply it in full generality. (Do you in fact think something like that?)

I think that if I were attempting to meet the goal of teaching people how to apply naturalism in full generality, I would probably agree with that. But in this particular sequence, I am not; I have learned over the years that I am capable of accomplishing at most one or two things at a time. Something about my cognitive and perceptual style, perhaps. And so even my "naturalist rationality sequence" is broken into three separate sequences: Intro to Naturalism, The Nuts and Bolts of Naturalism, and whatever I'll end up calling the demonstration sequence. I do very much like the idea of eventually synthesizing all three sequences into a single concise guidebook, but I think that I simply am not smart enough to write that guidebook without first writing each piece of it. And then, even beyond that not-yet-existant "Naturalist Rationality" guidebook, there are other guidebooks I would love to create, such as something like "Naturalist approaches to finding traction in pre-theoretic fields", which I expect will require just as much groundwork as I have so far put into "Naturalist approaches to mastering the basics of rationality".

But I am again only half-talking to you, I see. Perhaps I should shut up for now, see if you have anything for me to listen to, and try again later.

That all honestly seems pretty reasonable.

A thing I found a bit confusing was, like, you list ‘what’s up with how things float in water’ as a thing you might study and get in contact with. Which contributed to a sense that, like, central examples of the things this course was meant to be about studying would include both things like ‘courage’ and things like ‘stuff floating on water’. You also listed ‘how to apply comparative advantage’. So I think I’m still confused about what you see the focus area of this sequence as.

I'm still feeling fairly surprised/confused that this conversation has felt so frame-clashy. I keep not being very surprised by most of the words you're saying, but then somehow it still feels like there's a stronger disconnect than I expect.

"But there really is a difference between trying to learn about things like 'updating incrementally' and trying to learn about things like 'electromagnetism'." 

I maybe want to clarify, I don't think there's (necessarily) a difference between learning about things like 'updating incrementally' and learning about things like 'electromagnetism'. My claim is "the average reader will think there is a difference (whether there is or not), and, if it's an assumption running through your work that they are the same, or close-to-the-same in many ways, readers may be confused because they aren't tracking that this is an assumption of yours.

Does that still feel ill-fitting-frame-y?

(I think there were other frame-disagreements at play here, but this was the part where I was feeling most confused about having apparently talked past each other or whatnot)

I just noticed that you have a post called "Noticing Frame Differences", and I'm gonna go read it (in the next few days) in case that turns out to help.

Honestly I don't expect it to much, it's mostly just covering the basics of "frames exist". (It does cover how, like, even pretty similar frames can be subtly different in ways that are difficult to track, but, like, that's just enough to make me sadly aware of what's happening but not necessarily be able to do anything about it)

Here is the "supplementary reading" I mentioned, (which perhaps belongs instead under Duncan's comment in the other thread).

In an earlier draft, my opening sentence for this essay was, "Once I’ve identified a question that’s crucial to my story, it's time to start observing the world—not so much to find an answer, as to position myself to ask better questions." A beta reader highlighted "it's time to start observing the world", and said this about it (paraphrased): "I found this to be a confusing first sentence, because it makes it sound like this part of naturalism is about 'observing the world'; but as I understand it, this part of naturalism is about 'observing my reactions to the world'."

To which I replied:

>I think part of what's going on here is that we have... not quite different ontologies, I think, in the sense that we think different entities exist... but at least different conceptualizations of the world and minds at a pretty fundamental level.

>I don't know how to talk about this well yet, despite an awful lot of attempts (like the "Intro To Naturalism" sequence), but I'll take a stab at talking about it anyway.

>There is no such thing as "observing the world as distinct from observing my /reactions/ to the world." Instead, the real distinction that actually exists is between "learning about the world by observing my reactions to it and knowing that what I'm observing are my reactions" and "learning about the world by observing my reactions to it and not knowing that what I'm observing are my reactions". There's no such thing as looking directly at a cup, in the way it's most natural to imagine. All you can do is point your eyes toward a cup, and be aware of whatever you experience as a result.

>When I take a naturalist approach to cup observation, I "try to look at my experience as I direct my eyes toward cups". This is not because it is actually possible to observe cups any more directly than that, but because framing it as a study of my own experience is a really powerful strategy mitigating the damage that map/territory conflations ordinarily cause to attempts to learn about things and solve problems.

>So when I'm "observing my experience of a cup", I am actually doing my very best to observe the outside world. In fact I'm attempting to observe the outside world more accurately and precisely than I ordinarily could while merely "trying to observe cups", because when I am aware that I'm observing my own experience, I am also aware that what exists in the world is distinct from my experience in various ways, even if I don't understand what those ways are. My awareness of that distinction grants me much more freedom to hypothesize about other ways the actual world could be, and so whatever conclusions I draw about cups as a result contain less interference from my own reactions to cups. By observing my reactions to cups, I end up forming cup models that have less to do with me, and more to do with cups.

>Like imagine that you see a straw in a glass of water (for the first time). This seems weird to you, since straws are supposed to be straight, so you deicide to study it and try to figure out what's going on.

>Imagine two different ways of approaching that study.

>The first way, you start with the question, "Why do straws bend when you stick them in water?" 

>The second way, you start with the question, "Why do straws look to me as though they bend when you stick them in water?"

>In the second case you're studying your reactions to straws in water, but you're more likely to end up more quickly with models that involve how light works, because you recognized from the outset that "the straw itself bends" is not actually what you observed; what you actually observed is "the image of the straw that appears in my mind under these conditions bends".

>A major thesis of mine is that "everything we cast our attention on is like the staw, to varying degrees", and I tend to go "oh naturalism is an especially good idea for this person in particular" when I hear about their problem/curiosity/interest and it makes me think "ah yes, that right there sounds especially much like a bent straw".

(Rationality, I claim, is absolutely chock full of bent straws, much more so than the vast majority of other fields of study.)

My beta reader found this response helpful, but they also were starting with a different set of thoughts than you are, I think. Still, I do wonder whether it does anything to bridge the gap between us.

Behind the painting

This is my first time attempting a painting in this particular style. (Well, it's actually draft 3; but it's part of my first series of attempts.) I'm not sure what the style is called, but it's the one where most of the subject as it exists in the world (or the artist's visual field) is deliberately indistinct, or not depicted at all. My automatic inclination (at least up to this point in my study of visual art) is to attempt to convey every part of my subject in full detail. Instead, I tried to choose a tiny number of details to convey more precisely, just the ones most evocative of the heart of my visual experience of Beck Chapel's exterior in Fall. Not only do I think this is a far better way to capitalize on the properties of watercolor as a medium, but it's meant to echo the central strategy of this post, which is to aim limited attentional resources precisely at the experiences that matter.

Much gratitude to @Duncan_Sabien for digitally editing this and all my other LW illustrations. It's a pain the butt to get this kind of indistinct frameless thing to blend nicely into the particular white of the LW background. He's also tweaked a couple other elements that didn't turn out quite the way I'd hoped.