A Discipline Already Happened

You don't start this game in an empty field. You start it in the middle of a career that isn't yours: decades of publications, arguments and half-settled debates, produced by scholars who were digging this world's sites while you were still being trained. Some of them are dead. None of them are infallible.

A career, simulated, not a database

The library you inherit isn't a list of facts about the world. It's a body of claims: things a handful of researchers concluded, from what they happened to excavate, filtered through their own specialisations and biases, same as yours will be. An excavation gets simulated for each of them: where they dug, what their institution had access to, what their interests pointed them toward. A researcher who studies warfare digs fortifications, not temples. Their publications get generated from what they actually found, read through their own lens, the same lens that's going to be reading yours.

Where several of them agree, you inherit something close to consensus. Where they don't, you inherit a live argument, one you can eventually join.

Right about most things, wrong for real reasons

The corpus is calibrated, not random. Roughly seven claims in ten hold up under scrutiny. The other three are wrong, and they're wrong for the same handful of reasons real excavation reports are wrong:

• They only dug the fortifications. A culture reads as a warrior society because the researcher who wrote it up never looked at a single kitchen midden.
• Nobody's found the gold yet. "This culture didn't use gold" isn't a lie; it's a report from someone who hasn't found the gold-bearing sites, because nobody has excavated there yet.
• The rare thing was misidentified. Something described as scarce might just have been overlooked, or wrongly labelled, by whoever dug it up first.
• The contact evidence is buried somewhere nobody's touched. "These two cultures never met" can be a perfectly reasonable conclusion from an incomplete dig, and still be false.

None of these are the corpus lying to you. They're the corpus doing exactly what a real discipline does: reporting what it's found, honestly, from wherever it happened to look.

What nobody dug

The gaps aren't random either. Large, urban, monumental sites attract researchers. Modest sites in remote regions get left alone. Burials and temples get excavated years before anyone bothers with a rubbish heap, because a burial makes a better paper. A careful reader can learn almost as much from what a bibliography doesn't mention as from what it does: "nobody's written anything about this culture's burial practices" is itself information, and it's information the corpus is honestly reporting by its absence.

That gap is also the opportunity. Whatever the existing literature hasn't looked at is exactly where there's room to say something nobody else has said.

How bold is your claim?

Once you understand what the corpus already believes, your own claims start to mean something specific. Repeating what's already established is safe and counts for little. Extending it a little further is more interesting. Directly challenging an established position is the highest-risk, highest-reward move available, because you're arguing against people whose reputations are attached to being right. Saying something genuinely nobody has said yet is the rarest move of all, and the hardest to be trusted on.

A player who only ever confirms the consensus will plateau. A player who trusts their own evidence over the published record either leaps ahead of it or falls flat on their face in front of everyone who already had an opinion.