A Black-Box View of Life

A couple of years ago I decided to get into badminton. I wanted a teams sport involving exertion of my brain as much as my muscles, because why not use both? For some reason badminton gave me the right kind of vibe. I found a friendly club nearby and started training with them every week in gruesome 4-hour sessions. The sport is very fun, and I'm now able to complete doubles matches with only moderate levels of embarrassment. It's great in and of itself, and I recommend to anyone living near a court to give it a try.

But I enjoyed this period if learning even more because I could observe something I hadn't observed in a long time: myself learning a new physical skill from scratch.

When I started, I thought I could get ahead by grokking the needed skills with raw brainpower. Since then, I've learned to play much better than before... and I still have no idea how any of it works. My body just does the right thing a little more often than before.

The thing that most helped correct my silly expectations for the learning process was seeing little kids (even seven or eight years old) who are enormously better than me at the game. They just do all the movements right, and it looks effortless. Even assuming that they've been at it longer than me, they don't have the physics training, the celestial mechanics research background, or the (flimsy) "adult street smarts" that I have. They presumably got there without a trace of intellectual labor.

It looks like the learning process, at its core, is something more basic and fundamental—although not necessarily easier—than rational understanding. Of course it involves repetition, but only the special kind of repetition we call trial and error.

As mundane as that sounds, it's quite magical if you think about it. A sport like badminton is squarely in the realm of physics and spatial relationships, yet you don't need to know Newton's laws, nor the Navier-Stokes equations, nor even trigonometry in order to master it. At the most fundamental level, we achieve this by treating things like black boxes—mystery machines—and tinkering with them to see how they respond.

The child who tries to open a door has to manipulate the handle (the input) so as to produce the desired movement at the latch (the output); and he has to learn how to control the one by the other without being able to see the internal mechanism that links them. In our daily lives we are confronted at every turn with systems whose internal mechanisms are not fully open to inspection, and which must be treated by the methods appropriate to the Black Box.

— W. Ross Ashby, Introduction to Cybernetics

But what about things where all the pieces are in full view, like badminton or bicycles? Although they may not be literally boxes, the principle is the same.

At first we are apt to think that a bicycle is not a Black Box, for we can see every connecting link. We delude ourselves, however. The ultimate links between pedal and wheel are those interatomic forces that hold the particles of metal together; of these we see nothing, and the child who learns to ride can become competent merely with the knowledge that pressure on the pedals makes the wheels go round.

— Ashby, ibid.

Treating something as a black box is as much about the interesting outcomes a certain kind of tinkering will produce as it is about the bits that can be safely ignored.

I learned my current flimsy badminton skills not through mental problem-solving but curious tinkering, in the form of swinging my racket in all possible ways near a shuttlecock until I found the ways that led to slightly less ridiculous results. Here my racket, my body, and the shuttle are the mysterious boxes I'm trying to figure out.

Black-box tinkering is also how you learn to program, sculpt statues, grill sausages, and beat a videogame. You provide some input in the form of muscle movements, and the output is a more or less desirable outcome from the black box.

Black-box tinkering is how this scientist has learned to "design" snowflakes in this documentary segment:

In this case, Dr Libbrecht was able to build a scientific understanding of how snowflakes work, but that isn't strictly necessary to the objective of designing pretty crystals: give a group of 9-year-olds access to Libbrecht's instruments and dials and they'll quickly learn to farm snowflakes in the shapes of any Minecraft character you want.

In this sense, everything can be a black box—everything is there ready to respond in one way or another to the inputs you give it.

What is being suggested now is not that Black Boxes behave somewhat like real objects but that the real objects are in fact all Black Boxes, and that we have in fact been operating with Black Boxes all our lives.

— Ashby, ibid.

But tinkering isn't the only thing you can do with a black box, and the same mindset is useful far beyond the topic of learning. It's also how we create anything at all.

There is no hard evidence of how string instruments were first invented, but the leading hypothesis is that, one day, a primitive (let's say) woman realized that plucking the string of a hunting bow makes a funny sound. I imagine she would have much enjoyed the pastime for about ten minutes, after which her band members would have threatened to exile her for the monotony. But some others would have been curious enough to try with their own bows, and soon they would have found that strings of different lengths, and of different thicknesses and materials, emit different sounds. The variety is interesting! So why not tie several different strings to the same piece of wood? And why not add an empty turtle shell, since it seems to amplify the sound so nicely?

Or, anyway, something like that must have happened in one or more places around the world: people without the least education in music theory and resonance physics who tinkered with and modified black boxes until they got the results they liked.

And a lot of tinkering and modifying followed...

...and followed...

...and followed.

A similar story has unfolded for most tools and technologies ever invented by humanity. We seem to be extremely interested, as a species, in plucking and kicking things and cobbling them together to assess the sounds (literal or metaphorical) they will make.

Maybe black-box thinking goes even further.

Text functions as a "whole", an aggregate, in other words a black box. The reader has the right to take that, process it, and chew it as they prefer. If the author were to process and chew it before it reaches the reader, the meaning of the text would be greatly damaged. ... There is nothing more inconvenient, for a novelist, than beginning to analyze oneself.

— Haruki Murakami, Novelist as a Vocation (translation mine)

Perhaps works of art are especially meaningful as black boxes—resonance devices tuned by one person for a specific "timbre" or quality, but capable of responding differently to the inputs of each observer.

That is not to say that we should never "open" the black boxes and try to understand what's going on inside. If you can learn how the smaller black boxes inside work and interact, it'll be easier to combine and re-combine them to produce new desired effects. Science is all about that, and so is the even more ubiquitous process of creating mental models.

At the same time, we need to learn the limits of those more theoretical approaches. No matter how much neuroscience and psychology you study, you can't fully divine how a "Person" black box will respond to your next utterance. And, when researchers in 1935 introduced a poisonous toad to the "Australian Ecosystem" black box, hoping to eliminate a nasty beetle, they couldn't foresee that the result would be devastating ecological effects on many local species (except the beetle) over half the continent.

Sometimes the inside of the box is too complex, too interconnected to model or explain with any accuracy. Sometimes it's not even clear what to consider as part of the black box or not—how to draw a boundary around it.

Often it's a combination of both complexity and naive boundaries—the Australian researchers thought they were inputting the toads in a small black box containing only their sugarcane crops and the nasty beetles, and (apparently) they believed that the two species would neatly eliminate each other and disappear like +1 and -1. Sometimes tinkering based on oversimplified mental models can be very costly.

My process of learning badminton seems easy in comparison, but the same principles apply. I've redrawn the boundary in my head to include more than just racket and shuttlecock: my body is part of the black box, too, and also all the other players on court, and environmental factors like air temperature and currents. The complexity of finely coordinating movements for a stroke, which involves calibrating and timing hundreds of muscles, means that my analysis will never be enough. Not to mention the recursive psychological interactions with my companions and opponents. Studying the theory might help a little bit. Just playing the game and seeing what works sounds more promising.

This is the reality we face every day, and we're pretty accustomed to its more mundane manifestations. But working with the complex phenomena, from ecosystems to markets to other brains, doesn't come natural to us. The black box lens helps you debug these processes. It makes you ask sharp questions like should I tinker or should I open? and are you talking about the same black box I'm talking about? ●