The First 20 Recursive Bees

June 27, 2014

{% previous not-for-its-own-sake power-word-case-study %}

Quoth the Feedback Loops

When discussing what truth is, many wise philosophers (I am not being facetious here) in an attempt to deek the seeming-necessity of abstracting away all of their definitions define truth as such:

The sentence {snow is white} is true if and only if snow is white.

Stop and appreciate the beauty of that reasoning. Something fascinating just happening there. Somehow the same phrase was used to describe itself, and in doing so, lead us to an interesting conclusion about what truth really is.

Truth, however, is boring1. What I want to focus on today for a little while is how this sentence is constructed, why it works. And the answer to that question, like so many things, is itself in a strange superposition of being blatantly obvious and remarkably profound.

Quotation marks. Quotation marks are why the sentence works. However, for clarity’s sake, I’m going to put {curly brackets} around the constructs whenever I’m talking about them. I am going to make my own quotation marks because I am the king of this blog. {X} will talk about the fact X, while {“X”} will talk about the sentence “X”. This convention frees me to save “real” “quotation marks” for, like quoting things and expressing doubt and referring to sentences and the million other little ways we use quotation marks without ever really thinking about them.

But I digress. Onwards, with gusto.

You see, when discussing {“snow is white”} (the sentence) we are talking about a wrapper around {snow being white} (the fact). Usually quotes are used to denote the object from the referent – i.e. {cheese} is something you eat while {“cheese”} is a word. It’s a really clever system to differentiate between talking about something and talking about talking about something.

Meta-levels. Get used to traveling between them. We’re going to do a lot more of it in a second here.

To get back to my point, there is clearly a relationship between the objects {snow is white} and {“snow is white”}. If {snow is white} were false, we probably wouldn’t ever say the sentence {“snow is white”}, just like we never say the sentence {“snow is orange”}. However, the converse isn’t true. {Snow is white} isn’t true just because we say {“snow is white”}.

For the most part, saying things are true doesn’t actually make them true. But let’s not get ahead of ourselves just yet.

You have probably heard about the paradox between the sentences {“the next statement is true”} and {“the previous statement is false”}. When ordered beside one another, a curious thing happens. If you accept the first statement as true, then you must accept the second true as well – but then you can’t believe the first, so you don’t believe the second, and, well, your brain explodes if you don’t give up soon enough. In fancy math words, we say that this feedback diverges – it’s impossible to find a consistent way of believing it.

This raises the question, however, of whether all mutually-referential statements necessarily diverge.

It’s pretty obvious that they don’t, but that didn’t mean it’s super easy to come up with striking examples. So I asked my super-smart friend Malcolm Ocean to help, and he came up with these three constructs:

  1. {“I hate myself”}
  2. {“Only a loser would say that”}
  3. {“I hate myself for being such a loser”}

They’re not a fantastically joyous set of thoughts, but I’m sure you or someone you know has at one point or another known the inside of a depression loop not unlike this one.

This depression spiral is a type of convergent feedback. Each of the sentences feeds off of the previous one in a big circle, and they all point in the same direction – downwards. In contrast to the previous example, if you were to follow this chain of reasoning all the way to the bottom… well, your head would probably also explode, though for entirely different reasons.

Reflexivity: Less of a Mouthful Than “Lobianness”

There’s been this idea skirting around in the back of my head over the last five or six blog posts I’ve written. It’s incessant and invades a lot of my waking-hour ruminations as well. Absolutely every time I ponder about minds, it pokes its head out, begging for attention.

As an aside, there’s a useful metaheuristic to be found here: ideas that keep coming back that you continually don’t give any thought to are probably things that you should give some thought to. Just a thought.

The concept is thus: if you believe something to be true (and have a causal link to it being true), it’s way more likely to be true. The archetypal example here is confidence. You’re confident if and only if you believe that you’re confident. This reeks of a feedback spiral to me.

One day I finally resigned to this concept and decided I’d better chunk it by slapping a name on top of it, and thus being better able to reason about it. My first attempt was “Löbianness”, you know, since it closely mirrors Löb’s theorem – math which describes how to prove things you can’t prove, but can prove that you can prove. Which is to say that it affords us a starting place for begin tackling things which are self-referential.

However, I’ve learned to do my research before coining things. This can’t be a novel concept – it shows up too often in psychology for that – so I did a bunch of reading. Lo and behold, psychology indeed already has a word for it: reflexivity: the degree to which something’s cause and effect can bend and influence one another.

Another way of putting that is that the reflexivity of X is a measure of how strong a feedback loop exists between {X} and {“X”}, and wow does it seemingly pop up everywhere. But again, let’s not jump ahead too quickly.

The Magical First 20 Hours

I’ve written about him before, and you can bet your booties that I’ll write about him again, but there’s a man named Josh Kaufman who writes some pretty fantastic books. One of the books in particular is called The First 20 Hours in which Josh talks about how fucking amazingly good you can get at things in a short amount of time. 20 hours, to be precise, if you didn’t get that from the title =)

The secret is essentially this: do some research on what you want to learn, get a clear picture of what “being good” means to you, and then practice it intelligently for 20 hours. That’s it! That’s the whole book, but described in much more detail and with lots of case studies of cool things Josh learned by this method2.

Why does this work? Well, for the most part, people don’t do things at which they’re not very good. Most people give up on things immediately if they aren’t very good at them. By precommitting to practice for 20 hours, you’re ensuring that you have 20 hours more experience than most people at whatever you’re trying to learn. That turns out to be a surprising amount of time to practice things.

The research beforehand is to ensure that you know how to critique your own efforts, so that all of your deliberate practice isn’t for naught. Pretty convincing argument, right?

Unfortunately, there is a large subset of people who all reply along the same lines when I’m done pitching the book at them. Negatively. The most common replies are:

Unbeliever: “No way. It takes 10,000 hours to get good at something. Everybody knows that. What, are you stupid or something?”

Response: Usually taking the inside view leads to faulty reasoning, but it actually saves the day here. Think about all of the things you’re good at. Have you spent 10,000 hours of any of them? Probably not.

Unbeliever: “Well, maybe. Some skills are learnable, but there is no way to get good at X in 20 hours.”

Response: You might be right about that, but remember, we’re not trying to get good at things. We’re trying to get “good” at things, which is to say, on a subjective bar set by us!

My favorite response of all time, however, was:

Believer: “Okay, I’m in! Let’s try it!”

Response: I like you.

This is entirely the right attitude for most things in life, however I don’t think the person in question (let’s call this person Barbara to protect the innocent) had her heart in the right place. I get the distinct impression that Barbara, whether she explicitly knew it or not, was skeptical of the claim and decided to get in on the sweet 20 hours action in an attempt to prove that it doesn’t work.

It didn’t work for Barbara. After 20 hours, she hadn’t acquired skill X. She came back to me and said “we’ll, I tried it, and it turns out that learning X is impossible in 20 hours.”

That’s not to say that Barbara was deliberately sabotaging herself or the effort whatsoever, except that this mindset of “knowing” it won’t work will hold you back. You’ll be afraid of investing yourself fully if you “know” that your efforts will be all for nothing. You’ll leave yourself a line of retreat: “oh, it’s okay that I failed. I knew it was impossible to begin with.”

Barbara’s problem isn’t that she was being subversive, it’s that she had inadvertently switched into quotation mode. She wasn’t {learning skill X in 20 hours}; she was {trying to learn skill X in 20 hours}.

If you want to {make a million dollars}, then {trying to make a million dollars} is just buying a lottery ticket, even though you know it’s not going to actually work. But hey, at least you can say that you tried, and really, isn’t that all anyone can ask?

Skill development, and goal completion in general, it would seem, has a high degree of reflexivity; it’s only works if you believe it will work. If not, you’ll be doing nothing more than going through the motions, that is, if you can even be bothered to do the motions whatsoever.

Fear and Loathing of Beeminder

While I’m on a roll of shamelessly plugging things, Beeminder is a tool that I think is even more powerful than rapid skill development. It generalizes the notion of precommitment as a means of doing things you find unpleasant. It’s the only reason I maintain a reasonably-regular blogging schedule.

The idea is as follows: you make a quantifiable goal on the Beeminder website. You give it your credit card, and that’s it. If you ever find yourself below the progress goal you set initially, the website charges an ever-increasing amount of money to your credit card. The service is free as long as you don’t screw up.

Now is that a fantastic business model or what?

As displeased as I was with the response to the First 20 Hours, people’s reaction to Beeminder is much, much worse.

“That doesn’t sound good for me. I’d lose all of my money.”

Response: … (the ineffable sound of a palm and face meeting in silent sorrow)

If you think about this for like a second, what these people are saying is that they don’t expect the things the goals they profess to have to be accomplishable. If you aren’t willing to put $5 on the line as a bet to yourself that you’re going to do the things you say you’re going to do, your prospects aren’t looking that great.

The reason Beeminder doesn’t sound good to these people is because it would force them pony up or accomplish their goals. Fun, not-very-surprising-fact: the people who do sign-up, for the most part decide to accomplish their goals rather than lose money unnecessarily for stupid reasons.

You can argue there’s a selection bias here, in that the people signing up are those who anticipate it working. And yes, you’re right; there is, but if you’re seriously considering that as a knock-down argument, apparently you weren’t paying very much attention to me in the previous section.

Accomplishing things has a high degree of reflexivity WHICH IS REALLY COOL BECAUSE IT MEANS YOU CAN GET A SIGNIFICANT HEAD START ON ACCOMPLISHING THINGS JUST BY DECIDED TO ACCOMPLISH THEM.

Yes, I know that you know that already. But knowing it isn’t the same as doing it. Knowing doesn’t count for jack. Doing things is where the magic happens.

Beeminder does magic. Go sign up for it right now, and work a little magic of your own. Things only happen if you make them happen, and you’re the only one who can do that.


  1. Just kidding!↩︎

  2. Programming, playing go, windsurfing, ukulele, yoga, touch typing↩︎