Articuli Vitales
Read an entire encyclopaedia.
Free of charge. No advertising, no analytics, nothing to buy.
The dream
I have always wanted to read an entire encyclopaedia. Not just a few articles, not just the ones that are easy or interesting, but all of them. Finish an entire encyclopaedia the way you finish a novel.
When I was a child, I would take down the Εγκυκλοπαίδεια Δομή, or the Encyclopædia Britannica, start at Aa, Aachen, Aaron, and at some point, inevitably and quite predictably, burn out completely. And then, of course, printed encyclopaedias were no more, and the online ones were endless. The dream was dead. But, lo and behold! The good people of Wikipedia have collected a number of vital articles, in layers of 10, 100, 1,000, 10,000 and 50,000. That's a solid goal. A thousand articles is a real overview of human knowledge, a base to branch out from and keep learning.
The solution: Vital Articles
Wikipedia is the world's encyclopaedia now, with its millions of wonderful articles but also its mistakes, its biases, its omissions, its misunderstandings. Nothing and nobody is perfect, never was, never will be. Furthermore, it's unwieldy, its size orders of magnitude larger than what any single human can digest. Good thing its editors have collected the most important articles in layers. Vital Articles is their list of the articles an encyclopaedia would be incomplete without: not the most popular, not the longest, the ones whose absence would be a hole, the articles everything else is built upon. All this in layers: the 10 most abstract, the 100 crucial, the 1,000 most important, then 10,000, then 50,000.
Let's focus on the 1,000 articles. That is a finite thing. A shelf you can see the end of. It is the encyclopaedia I intend to read, and this site is the room I am building to read it in. With a beautiful reading space, the ability to bookmark, to add notes, to highlight, and at the end a small test, just to make sure I really did the work, something that puts a stamp on the article.
Principia
First principles
10 articles
Summa
The sum
100 articles
Orbis
The world
999 articles
What you get, free of charge
All of it, for nothing, with no version of it that costs money. I am building this because I want to use it, and one more reader costs me nothing worth counting. Anyway, what right do I have to put a price on something that uses other people's effort?
-
A reader worth sitting in
Themes, typefaces, control over text size, column width and justification. Every article opens on an illuminated initial recreated from a missal printed in 1488. Make yourself at home. A library is a safe space.
-
Bookmark it away
A ribbon stays on the paragraph you stopped at, and the home page offers you the thing you were in the middle of rather than something new. Never get lost.
-
Passages, and what you thought of them
Select any passage to keep it, and write a note on it. The notes are set in the margin, beside the lines they belong to, the way you would have written them in a book you owned. They collect in one library, searchable, across everything you have read.
-
The test
Five questions an attempt, drawn from a bank written against the article itself, a mix of easy and hard, at least one of which cannot be answered from the opening paragraphs. Four of the five to pass the first sitting, three on every attempt after, each failure putting the article behind a wait. A question that is wrong or unfair can be flagged from the page it appears on.
-
Meaningful progress
Three levels, counted separately: Principia, Summa and Orbis. A finished article cannot be lost — not by an article leaving the list, not by the question bank being rewritten underneath you.
-
Your data, and the leaving of it
Everything you have written is yours to take: one button exports your completions, passages and notes as a file. Delete the account and it is deleted, not deactivated. There is no analytics, no advertising and no third party to hand any of it to — see the privacy policy.
On relying on AI and Wikipedia
All the articles are from Wikipedia, and all the quiz questions are AI generated.
The text is Wikipedia's, written by volunteers and licensed so that anyone may adapt it, provided they credit the authors. Every article here links back to its source and to the revision history that names everyone who wrote it. Nothing is scraped, everything was collected according to Wikipedia's rules. Not one sentence of an article is rewritten, shortened, summarised or improved. The only changes are stylistic, changes in the layout of the article.
The questions are written by a Large Language Model, checked by a second one against the article, and screened by a third for political bias. I could not have written thousands of them by hand, and I need them myself, to take the test like anyone else.
Using LLMs is not without its own problems: most were trained on work taken without the consent of the people who made it. That is a real cost, and I don't think it is fully paid off just because the output is useful. What I can do is keep this small and free. No subscription, no ads, nothing resold. Nothing here turns that cost into a profit. The most legitimate usage of modern AI is to enable what makes us most human, and I believe that learning is one of the most noble human causes.