The Mountain They Chose

That last sentence is the entire uncomfortable content of Casey Carter's documentary To Use a Mountain, and it takes a feature film to deliver because we are extraordinarily good at not knowing it.

Carter goes to the places that made the shortlist. Heflin, Louisiana, hollowed out after the mill closed. Richton, Mississippi, a predominantly Black town where families have held the same land for generations. Hanford. And Yucca Mountain in Nevada, which is Western Shoshone and Paiute land, and which won.

He describes the government's own files as hundreds of pages of statistics and maps with no human dimension anywhere in them. That is not a criticism of the paperwork. That is a description of what the paperwork was for.

Here is the part you already suspect and would rather not say out loud: nobody in that process had to want this outcome. Not one person had to harbor a single hostile thought about the Western Shoshone. The criteria did the work. Sparse population. Geological stability. Economic distress, reframed as communities that might welcome the jobs. Willingness to host. Each one defensible standing alone, each one the sort of thing a reasonable analyst would put on a list.

Run them together and you have built a search algorithm for places whose objection costs nothing. Nobody wrote put it on Indian land. They wrote low population density and let four hundred years of American history handle the addressing.

And then somebody signed it. By name. Out loud.

That's the part a tidy version of this argument leaves out, and it turns out to be the best evidence in the case. In December 1987 Congress amended the Waste Policy Act — the amendment is known in Nevada, without affection, as the Screw Nevada Bill. Three sites were still under active study: Deaf Smith County, Texas; Hanford, Washington; Yucca Mountain, Nevada. Congress struck two and left one.

Why those particular two survived was never a mystery, and nobody troubled to make it one. A member of the House-Senate conference committee explained it to Nevada congressman James Bilbray in plain words: the Speaker of the House was Jim Wright, of Texas, and the House Majority Leader was Tom Foley, of Washington. That was the analysis. Nevada had four votes in Congress and nobody in leadership.

So the criteria did not make the final call. Power did, in a room, with names attached.

Which looks like it breaks the argument and actually completes it. The criteria machine was never built to eliminate the signature. It was built to pre-cook the menu. By the time Congress reached the last step, every remaining option was the same kind of place — sparse, poor, already dispossessed, structurally unable to retaliate — so choosing among them cost nothing. The laundering happened upstream, in the shortlist. The knife-work at the end was cheap because the criteria had already narrowed the field to communities whose objection had no price.

Nobody had to sign is the wrong sentence. By the time anyone signed, signing was free. That's the achievement, and it is a much better piece of engineering than an anonymous crime.

The Guardian's interview lands on a word for all this, and the word is violence. Watch what happens when you swap it in. Environmental risk is a weather report — it has no author, it merely occurs, and the appropriate response to it is mitigation. Environmental violence has a subject, an object, and a verb. Someone did something to someone. That's the whole difference, and it's why the second phrase gets called inflammatory by people who found the first one perfectly neutral. Neutral language isn't neutral. It's language that has already decided who doesn't have to answer.

But violence isn't quite right either, and the way it falls short is the most interesting thing here. Violence implies a perpetrator — a hand on the instrument, an intent you could put in front of a jury. What we are looking at is a machine that produces the same result while requiring almost none of that. So the word isn't inflammatory. It's nostalgic. It reaches back toward a world where harm on this scale needed somebody to want it, and it asks us to go find that person. We do not have a word for authorless harm. That gap isn't a deficiency in the dictionary. It's a large part of why the thing keeps working.

You run a version of this. Not at 77,000 tons — at kitchen-table scale, several times a week.

It's not that I don't want to see you, it's that the schedule doesn't allow it. We had to let him go — the numbers made the decision. That's just how the system works. The passive voice is not a stylistic preference. It is load-bearing infrastructure. It is the specific piece of equipment that lets a person make a choice and then stand next to that choice with the posture of a witness.

But be exact about the difference, because collapsing it is how an argument like this degrades into a cheap indictment of whoever happens to be reading. The kitchen-table version is retail: one person, one choice, an alibi they half-know is an alibi. There's a flicker in there. The siting process is wholesale — an architecture that manufactures alibis in advance and distributes them so thinly that every individual disclaimer is sincere. No analyst was lying. That's the point. Nobody had to fake anything, which is exactly why it scales and the retail version doesn't.

The retail version is the one you have access to, though. It's where the flicker still is.

And before this becomes a story about them — the waste is real. It exists because electricity was used, by people, including you. The mountain is downstream of a light somebody left on. There is no version of this where the material vanishes and no version where its location isn't a moral fact. The question was never whether a decision had to be made.

Carter's real accomplishment is grammatical. He takes a process engineered to have no subject and puts one back in the sentence. Somebody chose the mountain. And the reason naming them doesn't satisfy isn't that they're hiding — Congress did it in the open, and the conference committee said why out loud. It's that by 1987 the choosing had been made so cheap that the chooser barely explains anything. The expensive part happened years earlier, in a list of criteria, each one defensible, that nobody remembers writing.

So the question to carry out of the theater isn't who decided. It's who wrote the criteria — and the next time you hear yourself say a process decided, that's the one to ask. Processes don't assemble themselves. Somebody chose what would count. That choice is always older than the outcome, always more boring, and always the one that did the work.