A hand that points to the page, then steps aside.

Every student has a folder called “to read”, which later became “to read, really” and later still “thesis_final_2”, and somewhere inside it is the paragraph you need. You underlined it one afternoon. You know it was on a left-hand page, somewhere near the end of a chapter, with a pencilled note of yours beside it that said something like “yes!”. You cannot find it.

There has never been so much to read, nor has keeping it ever been so easy, and yet the commonest experience of anyone who studies is still this one: knowing that something exists and being unable to reach it. Our libraries fit in a pocket and behave like that other one, the endless library of hexagonal galleries, which holds every possible book and opens none of them at the right page. Abundance did not solve the problem of finding; it enlarged it until it became invisible, because now the fault seems to be ours.

An archive is never innocent. What is kept, the order in which it is kept and what is left out decide, before anyone else does, what can later be thought. For centuries copyists, librarians and indexes made that decision; today it is made, more and more, by machines that do not show their shelves. When you cannot tell where what you read comes from, you cannot argue with it, and whatever cannot be argued with ends up in charge.

That is why provenance matters so much to us. It is no academic ornament, no footnote that nobody reads, but the least a reading needs in order to be yours: knowing who said it, where, in which edition, on which page and with what degree of certainty, and being able to go and check.

A search engine gives you ten links and a page of advertisements; a chatbot gives you a rounded answer, sure of itself, without a single page number. The first leaves you at the door of somebody else’s building; the second tells you what is inside without letting you in and, now and then, invents the books, the authors and the pages in the same calm voice it uses when it is right.

Both of them flatten. They turn the difference between one text and another (between a 1925 edition and a 1970 translation, between what someone said and what someone is said to have said) into a smooth surface where everything weighs the same. And both, without meaning to, take away the work you most wanted to do yourself: read the passage, distrust it, set it beside another one, and think.

In the margins of old books a small hand with an outstretched finger appears again and again. Palaeographers call it a manicule, and it was for exactly this: to tell some future reader “look here”, without telling them what to think of what they were about to see.

Scholaris wants to be that hand. Ask it something and it gives you back the passage, the book and the printed page (the real one, the one you would put in a footnote) or the exact minute of the interview in which someone said it. If it cannot find it, it says so. It does not invent citations: each one is checked against the text you gave it before you ever see it, and every piece of data carries its provenance in plain view, where it came from, how sure we are of it and who corrected it.

A well-read library reveals relations nobody had written down: two authors who never cite each other and talk about the same thing, a concept that changes its name when it crosses a border, a question that comes back every forty years in different clothes. Scholaris reads your whole library at once and draws those relations (people, works, places, concepts, dates) like a chart of the sky, with constellations that were in no single book and only appear when the books are put together.

The machine draws the lines, always anchored to the page they come from; what they mean, which of them matter and where they lead is for you to decide. We want a structure that multiplies thought instead of closing it, one that works less like a tree, with its trunk and its hierarchy, and more like a rhizome, where any point can connect to any other and every connection is the beginning of an idea that does not exist yet.

It will not write your essay. It will not summarise a book so that you can skip it. It will not decide for you which quotation is the good one or which author is right. Thinking, judging and writing remain yours, and they do as a matter of principle: a thought that has not passed through the body of the person who signs it is too much like a rumour.

For those who read in order to write. For the doctoral student with three hundred PDFs and a supervisor who wants the page number; for the lecturer preparing a class with twenty years of underlining; for the journalist going back to a two-year-old interview in search of one sentence; for the translator, the archivist, the first-year student who doesn’t yet know they will need it, and for anyone who has ever said “I read it somewhere”.