The Waste Isolation Pilot Plant (WIPP) near Carlsbad, New Mexico, is a salt mine 655 meters underground where the US has been burying plutonium-contaminated waste since 1999. The salt is geologically stable — it flows slowly, sealing the waste in over time. The facility is designed to last 10,000 years without significant leakage.
But 10,000 years is longer than any civilization in recorded history. The Egyptian pyramids are about 4,500 years old. The oldest continuously operating institution on Earth (the Roman Catholic Church) is about 2,000 years old. Our nuclear waste will be dangerous for roughly four times the entire span of written history.
So in the 1980s, the Department of Energy assembled a panel of experts — linguists, architects, anthropologists, astronomers, science fiction writers — with a single question: how do you build a message that lasts 10,000 years?
The answer, they discovered, might be to not rely on a message at all.
WIPP already has its markers. The centerpiece is a 25-foot granite monolith at the site entrance, inscribed with a text drafted in 1993 by a government contractor named Maureen Kaplan. It reads:
Something about the repetition, the broken syntax, the denial of honor — it reads less like a government document and more like scripture found in a cave. That was intentional. The panel wanted a text that would feel wrong, that would make a future visitor uneasy. The message they're really sending is: this place isn't like other places. Go somewhere else.
The Sandia National Laboratories report on marker design (1993) proposed a layered communication strategy. Each layer assumes the previous ones may have failed:
Something man-made is here. Doesn't say what — just makes the site feel different from the surrounding landscape. Massive earthworks, carved stone, geometric patterns visible from the air.
Something is wrong here. Don't dig. Scary faces, spikes, menacing symbols. The kind of thing that makes you want to leave even if you can't read it.
What is here, when it was made, why it's dangerous. In multiple languages. Multiple writing systems. Also: pictograms, diagrams. Assume the reader can't read your language but can interpret a picture of a skull with a radiation symbol.
Full records: maps, figures, schematics, chemical inventories, engineering plans. The complete knowledge needed to understand and manage the site. Stored in a buried room, on durable materials, in multiple formats.
Each level is surrounded by the next — you don't even get to the text unless you've already gone through the stuff that makes the ground feel wrong first.
One design specified that the information building itself would be shaped to produce “dissonant and mournful” whistling sounds when wind passed through it — so that even blind or illiterate visitors would feel the site was wrong.
The skull and crossbones is the best example of semiotic decay the panel encountered. When sailors in the 17th century flew a Jolly Roger, it meant “we are about to kill you.” Today it means “Halloween decorations” or “fun pirate theme.” The same symbol, same design, opposite reading in under 400 years. What will it mean in 10,000?
This is the core problem: every semiotic system degrades. Languages die (Sumerian, Etruscan, likely English eventually). Writing systems are forgotten (Linear A, Proto-Elamite, Rongorongo). Symbols are reassigned (“swastika” went from good-luck charm to genocide symbol in one generation). Institutions collapse (Rome lasted ~1,000 years; the US has made it 250).
The backup plan, therefore, is to stop relying on symbols entirely and instead make the ground itself feel wrong. Hostile geology as a language. A landscape that says “go away” without using any words at all.
The WIPP panel was unusual not just for its subject but for its composition. It included a linguist (Thomas Sebeok, who had worked on zoosemiotics — animal communication), a logician, an astronomer, a science fiction writer (Gregory Benford), an anthropologist, and an architect. The problem forced them to think about communication the way no one had before: not across space, but across time. Not to a different culture, but to an unknown one.
Sebeok's “atomic priesthood” was the most controversial proposal, but also the most honest about the problem: any message that survives 10,000 years will require an institution to maintain it. The Catholic Church is the closest model we have. He proposed building something like it deliberately, with a panel that replaces itself every generation, maintaining the knowledge and the rituals. The rituals matter because they carry meaning even when the language dies — a ceremony that says “we are doing this because we remember something important” outlasts the specific fact that was being remembered.
Umberto Eco, who wrote about the proposal in The Search for the Perfect Language, had a different worry: that an atomic priesthood would eventually forget what it was supposed to protect and begin venerating the waste itself. A religion with real, glowing gods.
The best argument against all of this comes from the panel members themselves: the design assumes a future civilization sophisticated enough to drill 655 meters underground but not sophisticated enough to figure out what it's drilling into. If you can build a borehole, you can probably run a spectrometer. Your radioactivity detectors will tell you what you've found before your archaeologists can read the warnings.
So the real audience isn't an advanced society. It's a civilization that has collapsed and partially regressed — one that can still dig, but can no longer read or measure. The markers are for the people who come after the fall.