In 1895, two Belgian lawyers sitting in a Brussels office made a decision that would consume the rest of their lives. They would create a complete index of every piece of recorded knowledge that existed — every book, article, patent, photograph, and map — in a single searchable system.
Not a library catalogue. A universal index of everything.
Their names were Paul Otlet and Henri La Fontaine. La Fontaine would win the Nobel Peace Prize in 1913. Otlet would die in 1944, largely forgotten, in a city occupied by the same forces that had ransacked his life's work four years earlier. At its peak, the system they built held roughly 12 million index cards in a building in Brussels.
We now call what they imagined the internet.
A Crisis Made of Paper
By the 1880s, scientific and commercial publishing had exploded in a way that genuinely frightened serious thinkers. The number of scientific journals had grown from a handful at the start of the century to over 9,000 by 1900. No single scholar could keep up with their own field, let alone adjacent ones. Important discoveries were being duplicated, lost, or ignored simply because nobody knew they existed.
To Otlet and La Fontaine, this was not merely a logistical annoyance. It was a civilizational problem. If knowledge could not be found, it might as well not exist. And if the same problems were being solved over and over by researchers who couldn't find each other's work, humanity was spinning in place.
Their solution was radical in its simplicity: standardize everything onto 3×5 index cards, one fact or reference per card, filed under a modified version of Melvil Dewey's decimal classification. Then get every library, archive, and institution in the world to contribute to a single unified collection, housed in Brussels.
By 1912, they were receiving around 1,500 queries per year via mail and telegraph — from researchers, governments, and corporations — and sending back answers from their card files. It was, in every functional sense, a reference search engine. You wrote your query, they searched, they replied. The average turnaround was a few days.
The Building in Brussels
By the 1920s, the Mundaneum occupied a substantial portion of the Palais du Cinquantenaire in Brussels. Visitors described room after room of wooden cabinets, each drawer packed with hand-typed cards. There were classification numbers written in colored ink, cross-reference cards, and subject hierarchies reaching several levels deep.
Otlet's team had found ways to include not just bibliographic references but summaries, translations, photographs, and assembled dossiers on specific topics. Some entries linked to other entries across subject areas — across disciplines that normally had no institutional reason to connect. From any distance, these look exactly like hyperlinks.
"The Mundaneum was not just a collection," wrote Alex Wright in Glut: Mastering Information Through the Ages. "It was a complete epistemological system — an attempt to define what knowledge itself is, and how it should relate to other knowledge."
The scale was the thing. And Otlet was never satisfied with it. He kept designing extensions, refinements, new classification layers. He was constitutionally incapable of thinking the project was finished.
The 1934 Vision
By the time Otlet published his Traité de Documentation in 1934, he had moved well beyond indexing. He had begun imagining the network that all this information required.
He called it the réseau mondial — the world network. He described workstations where users could search all human knowledge remotely, receive documents on a screen, annotate them, and send responses to other users anywhere in the world. He wrote about "electric telescopes" through which a person could consult the world's holdings from their own home. He imagined social dimensions: profiles, correspondence, the ability to see what others had found relevant.
He was 66 years old. He had been working on this for 39 years. He was not writing science fiction. He was writing a technical proposal — in the same genre as a building specification or an engineering brief.
In one passage he described something close to the Wikipedia model: a document that multiple contributors could edit and refine, always current, drawing on collective expertise. He used the phrase "collective brain." He meant it literally.
The specificity is what makes this remarkable. Not just "a machine that finds information" but the precise architecture of linking, annotation, remote access, and collaborative authorship. He had worked through the design problems.
What Was Lost
Germany invaded Belgium in May 1940. By that August, the occupying administration had evicted the Mundaneum from its building in Brussels to make room for an exhibition of contemporary German art. The 12 million cards were moved to storage in Mons, roughly 50 miles southwest. Some were destroyed. Others were scattered in the chaos of wartime redistribution.
Otlet spent the last years of his life in poverty, trying to salvage what he could from the dispersed collection. He died on December 10, 1944 — four years before the Manchester Baby ran the first stored-program computation, and forty-five years before Tim Berners-Lee submitted his proposal for the World Wide Web at CERN in Geneva.
The Mundaneum sat in storage until the late 1960s, when a Belgian archivist named W. Boyd Rayward rediscovered it and began the slow work of restoration. It eventually became a museum in Mons. In 2012, Google — finding the parallel too resonant to ignore — partnered with the museum and funded digitization of the collection. The paper index was finally being incorporated into the actual network Otlet had designed.
The Part We Keep Getting Wrong
The easy version of this story is "Paul Otlet predicted the internet," and leave it there as a curious footnote.
But that framing misses the specific quality of his vision. Otlet wasn't just imagining a technology. He was imagining a moral project. He and La Fontaine believed — with genuine conviction — that if all human knowledge could be made universally accessible, the material conditions for war would dissolve. The League of Nations, which La Fontaine helped design in the early 1900s, was part of the same impulse. The Mundaneum was not just an information system. It was a peace project in index card form.
He may have been wrong about that particular theory of history. The internet arrived and did not end war. But the depth of his conviction is what made the Mundaneum something more than an ambitious library scheme. He was trying to change what it meant to know something — and to whom that knowledge belonged.
Most of us assume the internet's inventors knew what they were building only in retrospect, that the full implications took decades to become visible. Otlet knew exactly what he was building. He described it with precision. He just didn't live to make it.
That gap — between clarity of vision and the moment the world catches up — is where most of the interesting work in history actually happens.
