Friday, September 7, 2018

On the magic of actually looking and intending to do stuff

I was talking with a friend a while back about how to get into a flow state while doing jiu jitsu and how to think things through when necessary without breaking the flow of things. It turned out to be helpful, which is always nice. When I mentioned that it still amazes me when things work out so perfectly like that, she mentioned that she has just come to expect it, ''and that's why she stopped asking how it works on a practical level''.

I had kinda thought that the reason she stopped saying "okay that's nice in theory but how do I actually do it" was that she understood how. Apparently it's just that she had come to expect that when talking to me that part will "magically" work out, and didn't want to mess with a good thing.

yeah the first few times the intention of trying something made it work kinda baffled me. I did actually spend some time trying to figure it out at some point, and then got worried that in doing that I'd fuck it up. I sort of tossed it in the bin of "meh sometimes the subconscious is just way smarter than the conscious" which is a handy wastebasket diagnosis, and just was happy its a thing. 
[...]
yeah, like seriously the discussion last night about signal to noise makes all the theoretical sense in the world, and all the sense practically when dealing with electronics, but trying to translate that into a practical difference today... like it doesn't make sense. Nothing about that conversation should have been easier to implement than [my jiu jitsu instructor] constantly telling us to stop thinking so damn much. But it was very different. And no, that doesn't make alot of sense to me how my brain translates that discussion into doing what I've been trying and failing at doing for months now. But I'm actually pretty ok with it not making sense. Because it works, and it works well and efficiently and seems to only be getting better. So given the trajectory, not messing with it seems the appropriate choice. I can always go overanalyze it to death later on if I hit a sticking point where "magic" doesn't work.

I told her then that I'd give her an answer, and kinda forgot to do it until recently. I figured it actually makes sense and is a decent explanation even without further context, so here it is:


It's tempting to think of the way we do things as similar to a computer program. Bake a cake by executing step one of the recipe, then step two then eventually you have a cake. No recipe? No cake.

That makes for a pretty lousy model for a lot of things. No one has a recipe for "standing upright" or "riding a bike". How do you initiate a right turn on a bicycle? Forget even trying to explain how to "contract your triceps" or whatever, let's pretend that that "doesn't count" and it's magically exempt from needing a recipe for some reason. How do you turn right? Most people would look at you like you're either retarded or asking a trick question -- and then give you the wrong answer. Put these people on a bike and they instinctively know to first turn the handlebars left so that you can lean right and begin turning right, but ask them explicitly and they'll miss this.

You learn things by doing them, and watching what happens. If you touch a hot stove, you probably won't do it again -- not because you consciously and verbally calculate "this would be a negative expected utility" but because it fucking hurts, and because you flinch away from doing things that hurt. When you wobble on a bike, you'll learn to associate leftward handlebars with rightward leans and stabilized rightward leans with right turns. Not in a "ask me what happens if the handlebars turn left" sense, just in a "close your eyes and picture riding a bike, and what happens when you jerk the handlebars left?" sense. It's "happens" because it really happens on its own. You supply the input imagery, and the corresponding output of "this is what would happen" comes out -- that shit that you have experienced over and over and learned to hard wire in response to the input. If I ask you to close your eyes and step outside on a bright sunny day, an image will come to mind. Notice how the sky already has a color, and you didn't get to "choose" it after the fact? If I asked you to believe that the sky was bright green, not blue, it would be hard to do. You could say you believe it, and answer questions like "what color is the sky?" with "green", but the moment you walk outside and see a green sky you're still going to be shocked. Plain old boring blue is what you actually anticipate, and you can't simply will that away by pretending otherwise. You'd actually have to see the sky be bright green and shit first.

Separately...

Separately from looking at the sky and taking in it's blueness, you also sometimes look at yourself looking at the blueness and learn to associate "myself looking at the sky" with "myself seeing blueness". This third person perspective of watching yourself see blue is entirely distinct from the seeing of blueness itself. The "quotation is distinct from the referent", and all that. This third person stuff is real learning still, but it is learning about yourself and your relation to the world around you, which is different than learning about the world around you itself.

These two things do go hand in hand to some extent. If you touch a stove, not only will you learn that gut level aversion and anticipation of pain and burns, but you will learn the idea that third person perspective of "if I touch that, I will burn myself and be hurt" in parallel. If I were to say "Bobby cranks on joint locks way too fast", you might cringe at the thought of rolling with him, if you internally translate this into a picture him popping your arm or something. That "translation" step is keeping your anticipations in sync with your verbal beliefs. If you skip that step, you still might decide "I'm not going to roll with Bobby", but if you ever forget and he pops your arm by throwing on an arm bar too aggressively, you'll think "what the fuck Bobby!?" because he violated your expectations by being bad. But once you've done that translation and internalized the anticipation that this is just what bobby does... then if you rolled with him and got your arm popped you would just be thinking "what the fuck Amanda? Why would I roll with someone who cranks on submissions???".

The answer to your question has to do with that translation from words to experiences. You had been getting caught up in the third person explicit beliefs that "seeing myself do these things would result in seeing myself do better and seeing myself do those things would result in seeing myself do worse", and never making it to the stage of looking in the first person and seeing things change as your idea of what an appropriate response to the reality coming at you changes.

I wasn't doing anything especially right so much as seeing and not getting dragged down the path where things go especially wrong. Let's compare your interactions with your instructor to your interactions with me.

You started the discussion by saying "how do I jiu jitsu better and more smoothly?". He answered "think less", and you said "I don't know how!". But you doomed yourself to failure at step one by asking the wrong question. You didn't really care so much about "knowing how to do it" so much as actually doing it successfully. It's like wanting to have an accurate mental representation of what color the sky is, so you ask him "what color should I believe the sky is?" (emphasis mine), he says "red", and you say "I don't know how to make myself believe that!".

You were caught up in this anxiety that "If I don't know how to do things, I can't possibly do them!", and then chasing after a third person perspective that is so clear that somehow you'd know how to do it in the first person too. "Do these steps, and you'll believe the right thing".

When you'd ask me questions like "what color should I believe the sky is?" and I'd say things like "red", I could see that you were reading my response as saying "[you should believe that the sky is] red", rather than just "the sky is red". So when you'd say shit like "But I don't know how to believe that!", and sometimes adding "but I really think it's blue!". I'd just say "Shh. Let me show you something", and then I'd just take you outside and point. You'd see that the sky actually is red despite how crazy that seemed, and it would all start to make sense. "Oh. It's sunset. How did I not think of that before". From that moment on, your anticipations would be calibrated to the actual reality, and you would find yourself believing what you set out to believe, just not knowing how you did it (or in your mind, how "I" did it, maybe).

And it's kinda funny looking back at it now, since I totally should have just answered your question. "You believe the sky is red by looking at the sky when it's red. Look, I'll show you", and then taking you outside to show you both that the sky is red AND that the act of Amanda looking at the red sky changes her beliefs and anticipations to the sky being red.

Instead though, we just iterated this dynamic where you'd say "but I don't know how to force myself to believe that!", and instead of letting you block your own understanding by preempting it with the "but I don't know how to believe it!" nonsense, I'd say "Doesn't matter, just look". To me, it was an invalid objection because you don't have to understand how to believe it or that looking is how to create beliefs, you just have to look. And I just figured that you'd also notice the fact that when you actually looked at things your beliefs would change, so I never pointed at it explicitly and never really picked up on the fact that you were taking "doesn't matter, just look" as "be less curious" rather than as "try it first, because that's where the answer to your curiosity lies [and if it's still not clear by that point by all means ask!]". So what ended up happening was that you'd say "[even though my priority is for this to work, I'm not going to listen to your solution yet because] I don't know how to believe it!", and I'd say "Shh, bad Amanda. Look". After enough iterations you were eventually conditioned to shut up and look, because when you looked you anticipated getting your solution, and when you tried to object that you "didn't know how", you anticipated hearing "shh, bad Amanda", until you ended up looking anyway.

So by the time you started a conversation with me about getting into flow like states when rolling, and I said "I'd love to do a mind dump on the topic right now", you just said "okay, sounds great" and expected that "somehow" afterwards you would find yourself putting it into practice. That "somehow" is simply that I explained to you how to think about it, and you understood. That's really all it is, and it happens naturally when there is nothing coming up to stop it.

It's only weird when it violates your completely mistaken expectations about what you "can" and "cannot" do. If I were to say "here's how you bake a cake:", you would think it only makes sense that you'd be able to bake a cake after listening to me. When I say "here's how to handle when someone is leaking sadness that they don't want to look at", it only makes sense to you that you'd understand and come out of it with an ability to put that understanding to some use.

When I say "here's how to determine how much to think", it's only weird to you because you were starting from the frame of "trying to change" how you think, and you've gotten used to this idea that you "can't" change things like that. It's understandable, of course. You've tried many times and only failed failed failed. Sure seems like one of those things where it's just "how your brain is". And you really couldn't intentionally change your mind. You didn't know how to. If I were to tell you to believe the sky is red, and even if I were to "shh" your "but I don't know how" for long enough for you to shut up and actually try it, you would have just stubbornly tried to demand yourself believe something that didn't feel right and it wouldn't have worked. However, that doesn't mean that changing how your mind works is "impossible" or even all that difficult once you know what to do. It just means you didn't know.

I did, though. And it turns out that the answer is as simple as "look". Coming at it from the frame of "what is the right way to think when doing jiu jitsu?", it's suddenly no longer mysterious that you'd learn something that you could put to use. And it's no wonder that you weren't able to do it yourself, since you didn't actually have a good understanding of the considerations involved, just that the right answer looked different than the only thing you knew how to do. And you struggled to get a sufficient answer from anyone else, because they didn't really get what you didn't get and why you were overthinking too much or how to guide you into thinking less because they didn't have an answer to the "how do I believe it though?" question either.

And when you were willing to put all that aside, forget about what you "can" or "cannot" do, what you "know" or "don't know" how to do, and just listened to my mind dump on the topic... and when I took the time to walk you through all the considerations and all the ins and outs that I use when determining how to think, and checking in to make sure you can picture it and make sure you understand... How is that going to lead to anything but you walking away with an understanding that you can begin to put into practice as you see fit?

2 comments:

  1. I think that, as you say, "notice what you're actually doing" (or "look") is generally good advice, but "Stop thinking so much" is often a kind of typical bad advice given by skilled practitioners who've forgotten what it's like to be a novice.

    They see the novice struggling. They ask themselves what *they* do in that situation. The answer they from their own perceptions is "I employ an automatic response". Their perception is one of just immediately doing the thing without thinking about it. So they tell you that's what you need to do.

    What they've forgotten is the part where they taught themselves do this automatically, and to not think about it, by doing it 1000 times effortfully and with careful intention. They had to struggle through it, to calculate slowly and with great uncertainty what the needed their body to do, and then to force their body to move they way they knew it needed to move. They had to do this so many times that now their body just automatically does the thing.

    Once you've made a guitar chord shape 1000 times, you no longer think about what strings and frets your fingers are on. But the first 100 times, you have to think about exactly what your fingers are doing every single time. If you're playing the chord badly, either you haven't practiced it enough, or you've practiced it *wrong*, which is actually much worse. In either case, "stop thinking about it so much" is not going to improve the sound of the chord. Thinking less is an *outcome* that will occur when the player is (i) executing a skill that has been learned deeply and (ii) not detecting any subsequent problems with the way the chord sounds.

    That said, I put a nontrivial probability on not having understood what you're saying in this post.

    ReplyDelete
  2. Yeah, you definitely can't skip right to "have well calibrated automatic responses" without doing the hard work of learning and calibrating them first, and sometimes skilled practitioners can forget what it's like to be a novice.

    In this case though, it's a bit different. There are better and worse ways of learning, and the "think less" advice was intending to point her to a better way of learning rather than just an instruction to "already know". If you do nothing until you have a solution, then if your standards for "acceptable solution" are too high to satisfy, then you end up doing nothing. Not only is this a horrifically ineffective method of grappling, it's a terrible method of learning as well. If all you do is nothing, then you learn nothing too, since you already know how "do nothing" will fail. If you come up with shitty solutions quickly, at least you get to test in which way they're shitty and learn from that.

    There are lots of ways to match your rate of analysis with your rate of tests. You can roll fast and think little, perhaps doing more thinking and reflection after the fact. You can sit there and work every possibility out in your mind before stepping on the mat, or you can roll slowly with someone who is willing to wait for you to solve the problems before taking the next step. These are all valid approaches that have their place and time. His point, as I see it, isn't that she shouldn't be doing thinking *ever*, but that when you're training your ability to roll at full speed, now is not the time.

    It's not that he forgot what it's like to be a novice who doesn't know where to move their body, it's that he never really understood what it's like to be a control freak who doesn't fully understand the value of just wingin' it sometimes. He understands jiu jitsu and can teach it well, but mindsets aren't his area of expertise, so helping her get out of an ineffective one wasn't something he knew how to do.

    It's true that "think less" is very unclear and oversimplified, but I don't see that as the important problem with it. If that was the only problem, she could have asked for clarification and presented arguments to be addressed (e.g. "I don't already know what to do, so I *have* to think!"). The big problem, which is the point I wrote this letter to her to explain, was that she was responding with "I don't know how!" when she should have been saying "Why?". She had gotten so caught up in "I don't know how to change how I think" that she never got to study the question of "How *should* I be thinking, anyway?". All I did to solve the "problem" of how to change her mindset is ignore the problem altogether and address the question of what her mindset *should* be, exactly. It worked not because I'm special or "magical" or anything, but because the answer to "How do I change my mindset?" is simply to answer the question "what should my mindset be". The way to see the sky as red is to look up and see a red sky, and the way to think less is to look at the idea of thinking more and find it to be out of place.

    ReplyDelete