Principle 01 · The root

Data-Identity

An entity exists, in knowledge systems, to the extent that it is structured as a verifiable datum, not as content.

From findability to resolution

For twenty years the question was: how do I get found? It was the right question in a web where the terminal consumer of information was a person, who opened a page and read it. That world is ending. The terminal consumer is becoming a machine, and a machine does not choose a page: it resolves an entity. The new question is not "am I findable?", but "am I resolved?".

To be resolved means the system does not merely find text about you: it finds you — an identified, disambiguated node, with verifiable attributes and explicit relations (structured with standard vocabularies such as Schema.org). The difference is the one between a flyer that talks about a company and the registry that records it. The first is content; the second is datum.

The limit of self-declaration

It is not enough to declare yourself. I can write on every page that I am the foremost authority in my field: to a graph system that is noise, not signal. Self-declared authority is worth zero because no external node can verify it. Becoming a datum means precisely the opposite of self-declaration: it means offering yourself to verification — structure, sources, coherence — so that recognition arrives from the outside, from the network of relations, not from yourself.

Why it is the root principle

Everything else follows from here. If the unit of existence is the entity-datum and not the page, then:

  • authority is built by propagation between nodes, not by accumulation on a single domain;
  • what is asserted must be falsifiable and corroborated, because a non-verifiable datum is not a datum;
  • the recognition obtained must be consolidated over time, because the graph recomputes continuously.

The other three principles are not additions: they are what happens after you accept the first.

The falsifiability condition

This principle is not a matter of style: it produces checkable predictions. If "becoming a datum" matters more than "producing content", then a correctly structured entity on a domain with no history must be resolvable by a generative engine in times that exclude the accumulation of domain authority. It is a prediction that can be refuted: if well-structured entities were never resolved except after months of domain age, the principle would be false.

Field proof

On a new domain, a page published at 06:39 was spontaneously retrieved by Google AI Mode at 08:402 hours and 1 minute later. The interval excludes the accumulation of domain authority and instead indicates inheritance of entity authority from the upstream graph. A single dated event is not a law: it is the first point of a series. See the series and the method →

Limits

The principle does not say that structure is enough. It says it is the necessary condition of entry. Outside its scope remain the cases where an entity is well structured but contradicted by the graph (discordant sources), or where the reference is genuinely ambiguous (strong homonyms). There the other principles come into play. Declaring the limits is part of the method: a principle that explains everything explains nothing.


This principle is the discipline's version of the foundational essay by Paolo Galbiati. The two pages are the same idea on two different entities: the author and the discipline.