Doctrinal Event-Response Predictor.
How Beijing responds to a U.S. export control, a Taiwan crisis, or a financial sanction is not random. The rhetorical framing, the escalation tier, the policy lever they reach for, and the time window from event to response are each shaped by doctrinal continuity that runs back to Mao 1949. This paper codes fifteen verified doctrinal anchors across Mao, Deng, Jiang, Hu, and Xi; builds a record of nine past event-response pairs (2018–2026); and uses the pattern to predict how Beijing will respond to events that have not yet resolved. The current live forecast: the PRC response to the May 2026 talent-curb expansion will be framed as “self-reliance,” at the sectoral-policy tier, with talent restriction as the primary lever, within roughly 257 days. The resolve-by date is 7 February 2027.
What we've actually found.
Xi 2019 hardened the Taiwan use-of-force formulation in three documented steps.
The wording moved from “make no promise to renounce” in the 2019 Xinhua text, to “do not renounce” in the CSIS Interpret translation, to “will not renounce” in the 2022 white paper. The shading from passive restraint not waived to active retention of the force option maps exactly onto the post-Pelosi escalation.
The PRC response template is stable on frame and tier, unstable on lever.
Seven of seven U.S. export controls since 2020 landed on self-reliance at tier 4. Taiwan incidents land on sovereignty at tier 5. But the policy lever oscillates between subsidy, retaliation, and now talent restriction — and that’s where doctrinal retrieval has to do real work.
The May 2026 talent-curb expansion was retrievable from the doctrinal record six months in advance.
The lever was novel (talent restriction had no prior in the 2018–2024 pair set), but the doctrinal lineage — Xi’s 2018 CAS/CAE self-reliance keynote plus the 2024 New-Quality Productive Forces essay — sat at the top of the retrieval index well before the event landed.
The full doctrinal corpus, the nine historical event-response pairs, the live prediction with cited passages, the leave-one-out backtest against three baselines, and the open questions all live inside the v0 founding paper.
Published pieces.
Each card is a standalone read — a substantive finding or a release. Engineering propagation and reconciliation passes live in the Changelog below.
What we caught when we walked every Chinese-policy quote back to its primary source.
Eleven of the twelve anchor passages now carry primary-URL verification against PRC official mirrors and Western translations of record. Six translation discrepancies surfaced in the process; the December 2024 BIS event-response pair was enriched with MOFCOM's retaliatory export ban.
A falsifiable-prediction instrument cannot rest on paraphrase. This is the audit pass that turns the corpus from working summary into verified record — and it caught a doctrinally significant divergence between PRC official English and Western translations of Xi's 2019 Taiwan speech.
The three Xi-era texts no Beijing forecast can ignore — now loaded.
We loaded Xi's 2017 19th Congress political report, his 2021 common-prosperity speech, and the 2024 Third Plenum communique into the corpus. The doctrinal record now runs from Mao 1949 to mid-2024 without a missing decade.
These three texts are how Xi himself frames the modern playbook — the New Era, the redistribution turn, and the deepening-reform agenda. Predictions about Beijing's response to anything China-tech in 2026 now route through the words Xi actually used.
“Will be reunified” vs “will surely be reunified”: how translation reshapes Beijing's Taiwan position.
We surfaced the two doctrinally heaviest translation discrepancies — both in Xi's 2019 Taiwan address — on the live page, and made every variant queryable from code rather than buried in markdown.
The English the world reads is not always the English the PRC publishes. The use-of-force formulation hardened in three documented steps from 2019 to 2022; the difference between 'will be reunified' and 'will surely be reunified' is the difference between resolve and teleology. Doctrinal forecasting that ignores the divergence reads only half the signal.
Predicting Beijing's next move from sixty years of Party doctrine.
The founding paper. Sixty years of PRC doctrine (Mao to Xi) treated as a retrieval corpus; nine historical event-response pairs as the eval set; one live ex-ante prediction logged for the May 2026 talent-curb expansion, resolving 7 February 2027.
Most China-watching is post-hoc. This is the first published, ex-ante, structured forecast on how the PRC state will respond to a specific event — with the frame, escalation tier, policy lever, and time window all named in advance, all scored against the realized response twelve months later.
Propagation, reconciliation, engineering.
Smaller passes — page/code reconciliation, schema work, engineering — that aren't publishable on their own but are recorded here for traceability.
- 31 May 2026Details →Page and Python reconciledThe verbatim-verification pass updated the page but had not fully propagated into the Python pipeline. Closed the gap so a fresh pipeline run reproduces what the page claims rather than reverting it.