Three clans, one night, 264 years without agreement

Where they actually come from

The seven survivors of the Salvaterra Collapse never agreed on what they saw that night in 1761. Over decades — not by decree, but through accumulated, minuted disagreement — those divergences hardened into three positions. This isn't social-feature decoration. It's institutional archaeology: each clan is, literally, one specific survivor's testimony, stretched across more than two centuries by people who kept disagreeing for the same original reason.

Índice comes from the survivor who measured before he ran — by his own account, he stopped to count how many "points of wrong light" he saw, even while terrified, because counting was the only gesture that made sense to his hands that night. His line, recorded decades later when he was already an old man:

I don't know what I saw. I know I saw seven. That's all I can swear to, and it's more than any of the others can.

Índice today keeps the Registro and treats the question "what does it mean" as the same question that killed fourteen people.

Vigília comes from another survivor who swore, until he died, that whatever crossed over looked at him before it left. The other six never confirmed it, and he never backed down. The line that became the clan's war cry, word for word, still in use:

Índice says there's no message. I asked how he knows. Did he measure it?

Raiz wasn't born from that night — it was born the next day, weeks later, when the survivors found that one of the missing had left behind, hidden in the abandoned headquarters, a small creature he'd been feeding for months without ever reporting it. The founding line, from an anonymous cultivator addressing the archivists of his own time:

You spent a hundred years cataloging visitors. I spent ten learning how to feed one.

Why I resisted giving one of them a win

The easiest mistake to make when writing three factions is letting one come across as objectively wiser than the other two — because the author, without noticing, agrees with one of them more. I wrote and rewrote the clans' exchanges several times until each one had the best possible version of its own argument, and none of the three can beat the others in a fair argument. Índice measures and distrusts meaning. Vigília accepts every number Índice has and disagrees about exactly one thing: that the door opened by chance. Raiz offers coexistence instead of vigilance or measurement, and answers accusations of sentimentality with specimen survival numbers that are, honestly, better than the other two's.

None of the three is right. None is wrong. It's real philosophy of science dressed up as a game faction — positivism, theology, and naturalism arguing about the same night for over two centuries, with no new data to break the tie. That's intentional: the day one of the three definitively wins is the day the Ordem stops needing peer review — and the whole institution exists, mechanically, so that day never comes.

What this actually changes in the game

In the endgame, the three clans attack your final submission from three different angles: Índice attacks your method, Vigília attacks your interpretation, Raiz attacks your ethics. That's not narrative decoration glued on afterward — it's the same disagreement structure from 1761, applied mechanically to the last thing you do in the game. You don't convince all three to agree. You survive all three lines of scrutiny at once, exactly like the whole Ordem had to learn to do after the disaster.

A technical detail that's also a statement of principle

On the implementation side, clan differs by data, not by code — meaning choosing Índice, Vigília, or Raiz never triggers a different system branch, a different combat rule, a mechanical advantage hidden in the game's engine. It's the same engineering decision that keeps the clan choice from accidentally turning into a disguised buff: if the difference lived in code, sooner or later one of the three would end up, without anyone consciously deciding it, a little stronger than the other two. Living only as data — text, cosmetics, dialogue context, angle of attack in the endgame — mechanical neutrality is guaranteed by the architecture itself, not just by my promise to keep things balanced.

What I still haven't finished calibrating

I still haven't decided, in numbers, exactly what choosing a clan changes in practice beyond identity and narrative — because the rule I most need to protect here is "this is not a buff," and every progression system I sketch out risks turning into a disguised buff if I'm not careful. There's also a tension I've deliberately left unresolved: the Ordem forbids individual heroism and never names a founder, but it does have, in at least one current leadership figure, a face and a name — and nobody inside the fiction itself has resolved that contradiction yet. That should bother people. It stays that way for now.