Engineers trying to warn people 10,000 years from now about buried nuclear waste realised that no language could be trusted to last that long. So they imagined warnings that would not depend on words alone: fields of jagged concrete thorns, hostile earthworks and monuments designed to feel dangerous before they were understood. The message they wanted to send was almost anti-monumental: this is not a place of honour, and nothing valued is buried here.

The problem was not how to write a warning sign. It was how to write one for people 10,000 years from now, after every language on it may have changed beyond recognition.

That was the strange design problem behind one of the most discussed documents in nuclear-waste history: the 1993 Sandia National Laboratories report Expert Judgment on Markers to Deter Inadvertent Human Intrusion into the Waste Isolation Pilot Plant. The report was prepared for the Waste Isolation Pilot Plant, or WIPP, the U.S. Department of Energy’s deep geologic repository for defence-generated transuranic waste near Carlsbad, New Mexico.

WIPP is not a high-level civilian spent-fuel repository. According to the DOE’s own description, it isolates defence-generated transuranic waste 2,150 feet underground in an ancient salt formation, 26 miles southeast of Carlsbad. The disposal problem still stretches beyond ordinary institutional memory. The site had to be considered across a regulatory horizon of 10,000 years.

That made the warning problem almost absurd in scale. A sign can rust. A symbol can change meaning. A language can die. A monument can become attractive simply because it looks like a monument.

The warning could not rely on words alone

The Sandia report did not assume that one inscription, in one language, would be enough. Its authors, K.M. Trauth, S.C. Hora and R.V. Guzowski, described an expert-panel process built around redundancy: multiple components, multiple communication methods, multiple levels of message complexity and materials with little recycle value.

The basic logic was clear. The site should say more than one thing, in more than one way, because no single message could be trusted to survive the full span.

At the simplest level, the markers needed to say that something made by humans was present. At the next level, they needed to communicate danger. More detailed levels would explain what was buried, why it was dangerous, where the danger was located, and how future people should avoid disturbing it. The later Permanent Markers Implementation Plan described Level II, III and IV messages combining text, pictographs, diagrams, maps, geologic cross-sections and records held in archives.

Language was still part of the system. It was just not allowed to be the whole system.

The anti-monument

The phrase most people remember from the Sandia work is the proposed warning that “This place is not a place of honor.” The line is famous because it reverses the usual purpose of a monument. Most monuments say: remember this, value this, come closer. The WIPP marker problem required the opposite message.

Do not admire this. Do not dig here. Do not assume there is treasure below.

That is why some of the proposed concepts have the emotional force of bad dreams. The report and its later public life are associated with designs such as spike fields, jagged concrete forms, menacing earthworks and landscapes that appear wounded or repellent. The aim was not beauty. It was a kind of pre-linguistic refusal.

In ordinary architecture, danger signs are usually added to a structure. In the WIPP marker concepts, the structure itself was asked to behave like a warning.

Why fear was not enough

The idea sounds simple at first: make the place look frightening and people will stay away. The report was more careful than that. Fear can attract as well as repel. A future society might read a large engineered landscape as a tomb, a temple, a fortress, a place of punishment or a sign of hidden value.

This was the central tension. To mark the site too weakly might allow future drilling or excavation. To mark it too impressively might invite exactly the curiosity it was meant to prevent.

The report’s answer was not to rely on one emotional cue. The marker system would need layers: a hostile-looking landscape, durable markers, buried messages, pictographs, scientific diagrams, records in archives and repeated information at different scales. A person who could not read the language might still understand that the place was artificial and dangerous. A person with technical knowledge might be able to reconstruct the more detailed warning.

The WIPP compliance material later described passive institutional controls in similar terms. The EPA expected the DOE to describe markers that would warn future generations about the repository’s design and contents, including the presence and hazards of radioactive waste. Those markers were to be as permanent as practicable with current technology.

The official plan became more practical

The jagged landscapes are the part that travelled furthest in public imagination. They are also the part that needs the most care in retelling.

Those designs were concepts in the broader marker discussion, not simply a finished construction plan. The later WIPP permanent-marker planning focused on a more practical system: large surface markers, small subsurface markers, a berm, buried storage rooms, a hot cell, an information centre, archival records and multilingual messages. The 2004 implementation plan described messages in English, French, Spanish, Arabic, Russian, Chinese and Navajo, with the hope that the markers might serve partly as “Rosetta Stones” for future populations.

That does not make the thorn fields irrelevant. They exposed the problem with unusual clarity. The most memorable proposals were attempts to communicate before language, to make the surface of the Earth itself say that the danger was below.

The actual marker programme is less like a warning from myth and more like a long-lived information network.

The real problem was memory

The WIPP marker work is often treated as an odd corner of design history, and it is odd. But the problem underneath it is ordinary enough: every society leaves behind things that outlast its explanations.

Nuclear waste makes that mismatch harder to ignore. The hazard is not visible in the way fire or flood is visible. It does not announce itself to the eye. A future person standing above a buried repository might see only land, stone, weathered structures and signs in dead languages.

The Sandia report did not solve that problem in a final way. It said plainly that the effectiveness of markers would likely decrease with time, depending on who encountered them, why they were intruding and what level of technological development their society had reached. That honesty is part of what makes the document worth reading.

A warning meant for 10,000 years cannot behave like an ordinary sign. It has to survive erosion, reuse, misunderstanding, curiosity, translation failure and the human habit of treating old things as meaningful.

The message was not only that danger was buried there. It was that meaning itself might not survive long enough to protect anyone.