Originally written in French. Translated by AI — the meaning has been preserved, not the prose.
Main idea
A reading note stays a prisoner of its book. It keeps what the book said, in the order it said it, and what you retain six months later is a vague recollection — "that author said something about that subject" — which forces you to reopen the source to recover the idea. Knowledge is then filed by origin, and every new reading produces one more note sitting beside the previous ones.
Filing by concept flips the axis. The unit is no longer the book but the idea, and a single idea can take in what three books, two conversations and one project have brought to it. The note stops being "the note from chapter 3" and becomes an available building block.
The gain is therefore not one of tidiness but of accumulation: two readings on the same subject settle in the same place instead of living in two notes that ignore each other.
Why it matters
This explains why some bases grow denser while others merely grow bigger. A base filed by origin expands linearly with the sources read, without any of its entries becoming any more solid; a base filed by concept gets better at what it already holds.
It also changes what you do with a new source. The question is no longer "where do I file this reading?" but "what does it confirm, qualify or limit in what I already think?".
Nuances and limits
The source axis keeps a function. Traceability, exact quotation, going back to the original all assume you know which text a given thing came from — origin does not disappear, it merely stops being the filing principle.
And the shift carries a cost paid on every reading: recognising that an idea is already there assumes you know your own base, which a reading note never demands.
Open questions
- How do you recognise that a new source brings a qualification to an existing note rather than a distinct idea, without rereading the whole base?