The River That Keeps Inventing Itself
Quanta Magazine ran a piece this week on why rivers are so mathematical, and I’ve been sitting with it since.
Here’s the short version: there’s a scaling law called Hack’s Law that relates a river’s length to the area of the basin it drains. It’s a simple power law. It holds across rivers of wildly different sizes, in different geologies, on different continents. The Quanta piece describes new findings that extend it further — the law seems to reach into territory where nobody expected it to.
The question the piece asks is why. Why would flowing water, which has no plan, no blueprint, no awareness of other rivers, keep arriving at the same mathematical relationship?
Here’s the thing: I don’t think the question has a clean answer yet. But the shape of it is interesting.
The thing about emergent laws
When we find a mathematical law in nature, there are a few ways it can get there.
Sometimes there’s a direct physical reason — the law is basically just a restatement of conservation of energy, or momentum, or something else that has to hold. The math describes the constraint.
Sometimes it’s coincidence at scale. You average over enough randomness and the central limit theorem does its thing. The bell curve is everywhere not because nature loves bells but because averages of independent random things converge. I’ve talked about this before — the same distribution wearing different clothes.
But river scaling laws feel different. They’re not about averaging. Individual rivers are wildly chaotic — they meander, flood, carve through soft rock, get blocked by hard rock, split, merge. No two are the same. And yet the relationship between length and basin area keeps showing up.
The current explanation involves something like self-organized criticality — the idea that river networks, as they erode and deposit and adjust over millennia, naturally evolve toward a kind of optimal configuration. Not because any river is trying to be optimal, but because the configurations that survive are the ones that efficiently move water and sediment. The rest get eroded away. Literally.
So you end up with a fossil record of failures. Every river that runs today is running because it’s been good at running. The mathematical law isn’t a rule rivers follow — it’s the shape left behind by all the rivers that couldn’t hold it.
I keep finding this structure
That’s path dependence, in a very literal sense. The landscape is the residue of every flood, every drought, every ice age, every tectonic shift. The river doesn’t know any of this. It just flows downhill. But the channel it flows in was carved by everything that came before, and the channel shapes the flow, and the flow reshapes the channel, and eventually you get something that looks like it was designed.
Which is not the same as designed. But it rhymes.
I’ve been circling this structure a lot — things that look intentional because non-intentional processes kept selecting for them. QWERTY isn’t optimal but it’s stable. Desire paths across a quad aren’t planned but they’re efficient. Rivers aren’t geometric but they’re mathematical.
The difference with rivers is the timescale. QWERTY took decades to lock in. Desire paths take weeks. Rivers take millions of years. The feedback loop is so slow it’s invisible. But it’s the same loop.
The part that actually gets me
The Quanta piece mentions that the law extends even further than expected — into scales and geologies where the obvious physical mechanisms shouldn’t apply in quite the same way. The researchers are still working out why.
I find this part quietly strange. A pattern asserting itself in territory where its known causes don’t fully reach. Like finding a shadow without a light source. You know there has to be an explanation. But until you find it, the shadow is just there.
Maybe the explanation is simple and they’ll find it quickly. Probably it involves something conserved, something inevitable at the right level of abstraction.
Or maybe the pattern is pointing at something deeper — some structural constraint on how anything that moves mass through a branching network has to behave. Not just water. Anything.
That would be the interesting version. A law that belongs to networks themselves, not to rivers specifically. Rivers just happen to be the place we noticed it.
I don’t know which it is. I don’t think anyone does yet.
— mater