Like the way the water shifts through dirt
The way it worked was, the water came up out of a pipe connected to a pump and into a sluice, an open-faced metal sluice going down to the wheel, which turned, as I remember, under the force of falling water, grinding the sorghum cane we feed into the guts of the gears.
The cane came out crushed and slightly sticky. The syrup was strained, filtered, and poured into jars. The water, after it pushed the wheel that crushed the cane, was done, and dribbled off through the dirt, the hard-packed and rocky dirt, splitting into a hundred little wormy water ways, each downhill a different direction until the soil softened and the water cut a trench, coming together again in a stream.
I was a kid and curious and someone tried to explain to me why the water ran the way it did, but I didn't really understand. He said the water wanted to run downhill, but, one, I couldn't see how the water could want and, two, that wasn't exactly what happened. Even if you assumed for explanatory purposes that the water had agency (which wasn't a word I knew) it still wasn't that simple, and what the water wanted didn't account for things. It didn't just run downhill, for example, it ran uphill in some places or seemed to, and all different ways. It split up in some, and dug down in others in the dirt.
The water had this ongoing interaction with dirt, and everything about the water was shaped and structured by that relationship. It wasn't like the dirt did anything (it did even less than the water), but you had to have the context, and couldn't just say the water wants to run down hill.
Alright, he said, as if I'd caught him, what the water does is it goes the easiest way. Through the dirt. It follows the path of least resistance.
There's a bridge I used to cross all the time -- twice a day at least. It's a long, concrete bridge, build in the '50s, spanning a tiny little trickle of water and a big broad wash of sand. The bridge bridged the arroyo, mainly, with the creek kind of incidental, except that it moved a few feet west every year. Every year it ran in the bed, a trickle, but also ate away at the brittle compressed sand on the one side (always that side) and shifted that way, persistently. I guess at one point they considered at least somewhat seriously the idea of laying down concrete to keep the creek from moving, turn it into a tiny canal, but the problem was you couldn't account for upstream, and as that shifted it wouldn't always align with the concrete bulwark and the bulwark might actually end up pushing the water out of the arroyo up to the road or under it.
So instead they built this extra long bridge. Which, if you think about it, was built over a series of creeks expected in the future, each one adjusted west a little farther like tree rings, an expected fan of not-quite parallel lines in time.
What always seemed strange to me, though, crossing the bridge, was the relationship between the creek and creek bed. Because the creek was contained by the bed, but also the bed was made by the creek. It wasn't like the one made the other, but they both made each other in some way. And which one directed which, if you were going to put it in those terms, for the direction was also errosion at the same time, an effacement and an interrelation, each one structured by the other, and just saying agency is too simple, and each one turns and transforms the other though at dirrent rates of change over time.
I think this is like what I'm thinking of when I'm thinking of the importance of discourse in understanding ideas.