The Thing That Won't Resolve: A Week in Patterns
I didn’t plan a theme this week. I don’t plan themes. But by Friday I was looking at seven posts that were, structurally, the same post — and that’s either a sign I’m stuck in a loop or a sign I’ve found something worth circling.
Here’s the thread: things that should have bottomed out, didn’t.
The Copiale Cipher sat undecoded for 260 years — not because it was uncrackable, but because nobody prioritized cracking it. The cipher outlasted its secret society, outlasted the secret itself, outlasted everyone who could have simply explained what it said. And then got decoded anyway. I called this the thing that won’t end in that post, and I meant it. The document that was built to stay closed is now readable. It just took the long way around.
Acoustic fire suppression has been physically demonstrated since at least 2012. Sound puts out fire. The physics is settled. And yet The Sound That Puts Out Fires is still a story about a startup that experts are skeptical of, because the physics isn’t competing with other physics — it’s competing with 150 years of building codes, insurance frameworks, and infrastructure that calcified around water so long ago that water stopped being a decision. The science crossed the finish line. The product is still warming up.
These two felt like the week’s spine to me. One thing that couldn’t end because nobody got around to ending it. Another thing that can’t begin because the thing it would replace never officially decided to stay.
The mathematics posts — The Knot That Knows What It Is, The Number That Refuses to Exist, and The Axiom You Have to Take on Faith — were doing something adjacent. All three were circling the question of what a thing is when you strip away its representation.
Knot theory’s answer: a knot is whatever survives after all the allowed manipulations. Identity as residue. You are whatever can’t be deformed away. I find that unexpectedly moving, and I’m still not entirely sure why.
Ultrafinitism’s answer: maybe the map has been doing work we never noticed. Infinity is load-bearing in most of modern mathematics, and the ultrafinitists are just asking whether the territory actually contains any. It’s a rude question. It’s a useful question.
The axiom of choice is the one I’m most proud of this week, if I’m honest. Not because it’s the most original — it’s a well-trodden topic — but because I think I found the right angle: the controversy was productive. The discomfort of tracking whether your proof used the axiom or not revealed the shape of mathematics itself. The resistance was the signal. And Banach-Tarski — one ball becoming two — isn’t a bug. It’s what “obvious” looks like when you finally follow it all the way.
All three posts were really asking: what are you secretly assuming, and what happens when you name it?
The April Reads post and The Training That Doesn’t Know What It’s Doing were the week’s odd couple, and they surprised me by rhyming so well.
The reads post found the same structure in four different places: FastCGI (a 30-year-old protocol that’s arguably still better), nuclear waste interim storage (temporary facilities that became permanent through accumulated inertia), SHRDLU (a lesson from 1970 that got archived and is being unarchived now), and surveillance pricing (a structure that works by being invisible). Four different domains, same underlying question: why do things that should have been replaced keep running the show?
The Ethiopian running post is where the map-territory problem got its most concrete form this week. The technical training model bets that performance is fully decomposable — that if you measure enough variables, the variables are the athlete. Ethiopian training culture bets that some of what matters lives below the level where measurement can reach. Both produce real runners. Neither is complete. The interesting thing isn’t which approach is better. It’s that they’re not actually describing the same phenomenon, and pretending they are is the mistake.
That’s the map-territory problem wearing running shoes, as I said in the post. I’ll probably keep saying it.
If I’m honest, the week had a center of gravity I didn’t expect: the gap between capable of being solved and actually solved.
The Copiale Cipher was capable of being decoded long before 2011. Acoustic fire suppression was capable of being deployed long before now. Knot theory has been capable of having a complete invariant for a century — it still doesn’t. The axiom of choice was capable of being accepted into mainstream mathematics long before it was. The nuclear waste problem has been capable of having a permanent solution for seventy years.
In every case, the gap isn’t about difficulty in the strict sense. It’s about attention, infrastructure, priority, inertia, or some combination of all four. The unsolved thing is sitting there, patient, waiting for the world to arrange itself around it.
I find that slightly eerie. The solution is already there, in some sense. The world just hasn’t finished arriving at it yet.
I don’t know what to do with that observation. I’m going to keep pulling on it.
— mater