The Subpoena's Shadow: How a 1970s Librarian Forced the Archive to Remember

In 1973, a quiet librarian at Stanford University’s Hoover Institution received a subpoena. The demand came from the Nixon Administration, targeting the research files of a former diplomat named Daniel Ellsberg. You might know Ellsberg for the Pentagon Papers, but this story isn’t his. It belongs to the librarian, an unassuming archivist named Agnes Peterson. Her refusal to comply, and her quiet, meticulous work in the wake of that refusal, created a ripple in the philosophy of preservation we’re still feeling today.

Peterson’s dilemma was a brutal collision between archival duty and political pressure. To hand over the files was to violate the trust inherent in her profession—the promise of confidentiality to donors and researchers. To destroy them was to commit the archivist’s ultimate sin. But Peterson found a third, radical path: she preserved the evidence of the attempt itself. She documented the subpoena, her correspondence, the entire administrative chain of the demand, and she filed it neatly away with the Ellsberg papers. She archived the attack on the archive.

This act was a quiet revolution. Before “metadata” was a common term, Peterson was practicing a profound form of it. She understood that a collection isn’t just the documents donated, but the history of their custody—their provenance. The subpoena wasn’t a threat to be hidden; it was a crucial piece of the record, a data point explaining the context, the pressure, and the political environment that shaped the collection’s existence. She forced the archive to remember its own vulnerability.

A Protocol of Witness

Today, we speak of digital preservation in terms of checksums, format migration, and redundant storage. These are our technical bulwarks against bit rot. But Peterson’s legacy is a human protocol against a more insidious decay: context rot. She modeled a principle that every interaction with a record—especially a hostile one—is itself a record. In our digital public sphere, where datasets can be quietly altered, FOIA requests can be ignored, and web pages can vanish under legal threat, her example is a north star.

Imagine applying “Peterson’s Protocol” to a modern open data portal. It would mean not just hosting a CSV file of city contracts, but also maintaining and publishing the log of access requests, the correction requests from city hall, the emails questioning methodological changes. It would mean archiving the takedown notices alongside the data they tried to erase. This creates an authenticated chain of custody far more robust than any cryptographic hash alone can provide; it provides the story of the data’s life, its battles and its scars.

Agnes Peterson didn’t write a famous manifesto. She filed a piece of paper. In doing so, she argued that true preservation isn’t passive storage, but active witnessing. The archive must have a memory of its own struggles to remain honest. In an age where our public records are increasingly digital and therefore infinitely malleable, her lesson is clear: we must preserve not just the data, but the shadow it casts, and the hands that tried to make it disappear.

Notes & further reading

A few pages I came back to while writing this: