Earning the Right to Explore

There is a kind of repair that ultimately works and teaches nothing, or almost nothing. My car had started making a horrible noise on the highway. I took it in a few different times, received multiple overlapping and competing diagnoses, had a few different parts replaced, and at some point highway driving was quiet again.

What fixed it? One of those replacements. Which one? I have no idea. All I wanted was a quieter car.

A quick repair and the full scientific process have very different goals.

Take a basement filling with water. Does anyone with any stake in that basement particularly want someone carefully changing one variable at a time to preserve the epistemic integrity of the occasion? No, no chance. Shut off everything. Plug everything NOW. Need towels. Get the water out.

Later, though, if it's my basement, I might care very much which pipe, behind which wall, is sometimes failing.

And there's another problem: if the problem isn't so painful I can't imagine it happening again, my instinctive level of needing to discover why it was there has probably collapsed. Or at least, drawn way down from where it was. My car now drives fine on the highway? Perfect.

Whether it's an acute problem or a slow motion disaster, the best moment to learn is often the moment when you and nobody around you particularly wants to: afterward. When something's actively going wrong, understanding is partly a luxury. After the immediate wrongness, understanding can feel like divination.

The car drives now. The website is back up and responding. Great, and here comes our powerful new preference for moving on.

And moving on is fine right up until the same thing happens again. Uh oh.

Then somebody says, didn't this happen... already? What fixed it?

Good questions.

For learning's sake, there might be a perverse advantage to being bad at quick repair.

If the basement takes three days to dry out, you're probably going to find that bad pipe. Whereas if somebody knows exactly which valve to turn immediately, where to put the towels, and how to get everything back to normal before lunch, the mystery can survive for years as the basement is mostly dry. It might be hard to imagine someone tolerating a problematic pipe in a basement, but perhaps that person dreads the incoming plumber bill. Oh, and the drywall needs punching out to get at that pipe.

Restarting the router works, which is probably why I've never learned anything that deep about router internals. Ultimately I don't care that much. The internet comes back, I forget the outage, and the little box gets another chance to behave badly in peace.

Some problems survive because they're cheap enough to forgive.

Footnotes

Sometimes this is as it should be. It's fine. I don't need to know what's happening from point A to point Z of every noise my car makes. Some headaches yield to Advil. Works for me.

Because deepening my understanding costs something: time, attention, sometimes money. There are plenty of mysteries where "well, it works now" is as far as I want to go.

Once the thing works again, a repair can call it a day. But even if nothing is technically broken now, somebody suspects there is a better way to do it. A better way to be, or something new to learn.

Exploration is asking the change to tell you something.

Which means you need to know a little to a lot about what you might be changing. If the old thing is held together by unexplained fixes, mismatched definitions, and stuff that "just works if you do this first," whatever you try next inherits all of that uncertainty.

Then the new approach works, or doesn't. Great. Compared to what, exactly?

This is roughly what I mean when I think about earning the right to explore. Perfect understanding would be absurd as a prerequisite. But there has to be enough solid ground underneath the new approach you're taking that stepping somewhere new teaches you something about the step.

Becoming extremely good at fixing the same thing brings with it a funny problem. Eventually the person who knows the workaround is the workaround. Everyone knows who to call. That person picks up the phone and knows exactly what to do. Somebody has become excellent at living with that problem and making it less painful.

An unintended consequence of competence: it can lower all the building pressure to make that competence unnecessary.

Cheap recovery changes the payoff calculation for curiosity. If the immediate problem disappears in thirty seconds of minimal effort, a true diagnosis can feel wildly overpriced. Just unplug it and plug it back in, you don't need a new coffeemaker.

It's just hard to tell if you should keep paying the same tiny bill forever.

First published	2023
Reading time	5 min
Tags	standards, philosophy

Reply

I’d welcome your thoughts on this essay. Send me a note →