The hash in the newspaper: how we prove something existed on a given date
19 October 2026
From the medieval notary to the postmark, and from a 1991 paper to a weekly notice in The New York Times: a look at how we have tried to prove a document existed on a specific date, and what a timestamp really proves (and what it doesn't).
A Date Is Worth More Than a Signature
When there is a dispute over who came first — who wrote a letter, who bought a plot of land, who invented something — the decisive question is rarely "who did it?" but "when?". A signature can be imitated; a date, with nobody to certify it, can be made up. That is why, for centuries, people have looked for ways for someone, or something, to vouch that a document existed at a specific moment.
The Notary: Dating as a Profession
The oldest solution is human: a trusted person who witnesses the act and records it in writing. Ancient Rome had tabelliones, professionals who drafted private documents, and in medieval Italy the notarial profession took firm shape, with Bologna as its great school. In Spain, the Notary Act of 1862 still shapes the system today: the notary keeps a master copy of the document in a protocol register, so the date doesn't depend on the parties' word but on an archive held by a third party.
The key to the system is exactly that: you don't trust the document, you trust whoever keeps it.
The Postmark and the "Poor Man's Copyright" Myth
In 1840, the United Kingdom launched the Penny Black and, with it, a uniform postal service in which every letter was date-stamped. A persistent legend followed: if you mail your manuscript to yourself and never open the envelope, you will have proof it existed on that date. This is the so-called "poor man's copyright".
It doesn't work the way many people think. The U.S. Copyright Office itself has stated that the law makes no provision for this kind of protection and that it is not a substitute for registration. A sealed envelope is, at best, a piece of evidence that is easy to challenge: nothing in principle prevents it from having been opened and resealed, or from having contained something different all along.
1991: Two Cryptographers and a New Problem
Digital documents made the problem worse. A file can be copied, edited and rewritten without a trace; its "creation date" is just a field anyone can change. In 1991, cryptographers Stuart Haber and W. Scott Stornetta published a paper in the Journal of Cryptology titled "How to Time-Stamp a Digital Document".
Their idea had two ingredients. The first is the hash: a digital fingerprint of a file, a short string of characters that changes completely if a single comma is altered and from which the original document cannot be reconstructed. The second is chaining: each timestamp includes the fingerprint of the previous one, so forging one would require redoing every later link in the chain.
A Notice in The New York Times
To make sure that not even the company issuing the timestamps could tamper with them, Haber and Stornetta founded Surety and turned to a delightfully analog solution: periodically publishing a summary of all the fingerprints in a printed newspaper. According to a report in Vice, Surety has published a weekly hash in the classified section of The New York Times since 1995, under the heading "Notices & Lost and Found". We could not confirm that it still does so today.
The logic is elegant: nobody can rewrite thousands of copies already printed and distributed around the world. The paper acts as a public, unalterable witness.
From There to Bitcoin
This line of work was not just an academic curiosity. When Satoshi Nakamoto published the Bitcoin white paper in 2008, three of its eight references were works by Haber and Stornetta, as the same report notes. The blockchain is, in essence, the 1991 idea with one fundamental change: instead of a company and a newspaper, thousands of computers watch each other.
What a Timestamp Proves (and What It Doesn't)
Precision matters. A timestamp proves that a specific file existed before a given moment. It does not prove that whoever presents it is the author, that the content is true, or that any legal right follows from it. A notary, a property registry and a digital timestamp answer different questions, and confusing them is the usual source of misunderstandings.
That is why the useful question about any record is always the same: who keeps it, what exactly does it claim to prove, and what happens if its keeper disappears?
A Closing Thought: Trust Has a Date Too
From the notary's protocol to the hash in the newspaper, the history of records is the history of how we distribute trust: first in a person, then in an institution, then in a chain no one controls alone. None of these solutions is perfect or answers everything; each, in its own way, tries to turn a date from an opinion into a fact that others can check.
Sources / Fuentes: Haber & Stornetta, "How to Time-Stamp a Digital Document", Journal of Cryptology, 1991; Vice, "What Was the First Blockchain"; Nolo, "Poor Man's Copyright" (citing the U.S. Copyright Office).