How do historians date the origins of xiangqi?
These answers restate the rules implemented by our open-source engine, which follows the WXF laws and is regression-tested against the official 110-diagram repetition casebook. Where an answer refers to a specific position, that position lives on the linked reference page, where the build verifies it against the engine before publishing.
Direct answer
Historians combine three kinds of evidence: textual mentions of a game called 象棋, surviving manuals and literature that describe the game in detail, and internal evidence from the game's own contents, most usefully the cannon, which cannot be older than cannons themselves. Where all three converge, in the Song dynasty, the dating is solid.
Why internal evidence is the most reliable of the three
Texts can be ambiguous or later copies with uncertain dates, but a piece that captures by jumping a screen has to postdate the weapon it is modelled on. That single fact rules out any claim that the modern game, cannon included, is far older than the Song period, regardless of what any individual text says. See when the cannon appeared.
Why the three kinds of evidence do not always agree
A name can outlive the game it once described, and a text can mention "象棋" without describing rules in enough detail to know whether it is the modern game. This is exactly why when xiangqi was invented has to separate name continuity from game continuity rather than trusting the earliest mention of the word.
Related questions
Is there a single document that settles the question?
No. The Song-dynasty dating rests on multiple sources converging, not one decisive text.
Could earlier evidence still turn up?
In principle, yes, and historical dating is always provisional. Nothing currently known moves the modern game earlier than the Song period.
.