The Web of Knowledge
I thought I knew the history of Wikipedia inside and out. Then I read Giota Alevizou’s The Web of Knowledge, and I found myself honestly surprised. I learned things about Wikipedia and its role in the knowledge ecosystem that had a major impact on my career, and that I was completely unaware of.
2026 is Wikipedia’s Jubilee year, and there are a few books published retelling the history of Wikipedia. But this book goes further; it discusses the role of Wikipedia in the wider knowledge infrastructure, and it encompasses the history of (English) digital encyclopedias as a whole, covering in particular the journeys of Britannica and Encarta from the 1990s on, in addition to telling the history of Wikipedia.
The book starts with a foreword by Jimmy Wales, the founder of Wikipedia, in which he closes by discussing the impact AI might have on Wikipedia and the knowledge ecosystem. A timely topic that the book will pick up.
I have to admit that I had a bit of a hard time getting started with the book, particularly regarding its positioning within its theoretical frameworks. But once we started to dive into the topic, I enjoyed it quite a bit and learned a few things.
The book covers very competently the pre-history of digital encyclopedias, starting well before Diderot, covering the ‘Great Books’ of the past, the Summae, and also Otlet and Dewey and their plans to index and organize the world’s knowledge, and Nelson’s Memex, all the way to Lenat’s Cyc. Giota describes knowledge organization and the encyclopedia as parallel threads weaving through the centuries.
For me the biggest surprise was right in the origin story of Wikipedia, where both threads were tied together. A lot has been written and said about Jimmy Wales’ motivation to start an online encyclopedia, including his mother buying an encyclopedia when he was 3 years old. When Jimmy was 30 years old, in 1996, he founded Bomis, a web portal for user-generated webrings and adult content. Critics often raised the second part, but Giota’s book focuses on the first: to structure the topics of the webrings would be helpful. And what better solution for that than an encyclopedia?
This was around the same time that Yahoo started becoming the main portal to the Web, and that predecessors of DMOZ came around. Back then, Yahoo wasn’t a search portal, but a directory. Jimmy’s intuition that an encyclopedia would be a great complement or substrate to organize the webrings would prove extremely foresighted, and became one of the foundational ideas for my own career.
That’s why I was so surprised to read about it in this book. So much that I actually had to discuss this with Jimmy and he, unsurprisingly, confirmed the story.
An encyclopedia is “a systematically organized repository of knowledge”. Already in the early 20th century, Giota describes how “the encyclopedia serves as a means for the semantic structuring of (digital) information spaces”.
I want to expand on this point: Jimmy’s intuition repeatedly bore fruit in the years that followed. Ironically, to the best of my understanding he never used it to improve the organization of webrings by Bomis. In 2005, Markus Krötzsch, Max Völkel and I described the Semantic Wikipedia idea, where we suggested using Wikipedia articles as anchors for topics.
In 2007, Metaweb’s Freebase also took Wikipedia as the central semantic structure to build around, which later became the Google Knowledge Graph, considerably improving Google Search by using “things, not strings”. DBpedia, started in the same year, was also lifted off from Wikipedia and called itself a Nucleus for the Web of Open Data.
In 2012 we kicked off Wikidata to explicitly provide and entangle a machine-readable knowledge graph, firmly based on Wikipedia. Abstract Wikipedia is now a continuation of this idea, working toward a more comprehensive knowledge representation than Wikidata did, towards the more ambitious goal of allowing many more people to share knowledge across language barriers.
And although I have been working on this for more than 20 years, I did not know that using an encyclopedia as the foundational structure of the semantic space was Jimmy’s motivation all along.
But back to the book. For me, that insight alone was worth the prize of admission. And Giota doesn’t stop there. There is the widespread myth that “Wikipedia killed Britannica”, and maybe some of us know that this doesn’t align with events. Britannica was already in a steep decline when Wikipedia was founded, and the reason was Microsoft’s Encarta. And whereas I roughly knew this story, Giota really lays out the narrative with thorough evidence drawn from her own material collected over decades. She had already interviewed people involved with Britannica and Encarta in the 2000s, and used this material to lay out a fascinating story, about Britannica’s turbulent organizational history and Microsoft’s ambitions to use Encarta as the promotional vehicle to get a PC into every household: because for the price of a Britannica, you not only get an encyclopedia for your children, but you get a lot more value.
One interesting point is about finances: Giota mentions that in 1989, Britannica had $627 million in revenue. That’s almost $1.6 billion in today’s dollars.
That is roughly one order of magnitude larger than the budget of the Wikimedia Foundation. While Britannica is available only in English, Wikipedia is available in more than 300 languages.
Giota also reminds us that Encarta premiered multimedia capabilities and functionalities from which Wikipedia could still learn. But just as Encarta eroded Britannica’s financial base, Wikipedia was able to replace Encarta. Microsoft wouldn’t even put up a fight: the PC had already won the marketplace, and whether Encarta or Wikipedia was the encyclopedia, what did it matter to Microsoft?
The book outlines the history of how encyclopedic authority moved from a small group of exalted individuals to a crowd in which everyone can participate. It reminds us again how littered with accidents this history was, and how fragile the current situation is. The wide deployment and usage of AI is destabilizing a situation in more than one way, and the book outlines what we may lose if we are not careful. We must take our role as stewards of the encyclopedia seriously – I am reminded of Umberto Eco’s wonderful quote: "After all, the cultivated person's first duty is to be always prepared to rewrite the encyclopaedia."
Links:
- Website about the book by King’s College London
- A review of the book by Matthew Vetter in the International Journal of Communication
| Previous entry: Centillion to the power of a centillion, or a googolplex? | Next entry: None |
