Lexicon
The working vocabulary
Terms used in the open register and on this site.
- SAID vs DONE
- Two kinds of fact. SAID is what a named person or office stated (a speech, post, briefing). DONE is what a named office executed and left a trace of (an order, signature, filing, strike, print). The register marks each row as one or the other, and the gap between them is the thing being measured.
- Grade V
- Verified: the row matches a primary or official trace.
- Grade I
- Inference: reasoned from verified rows, labelled as inference.
- Grade TK
- To come: the claim is attributed but the primary trace is not yet in hand. It holds open until the falsifier resolves it.
- Grade UNVERIFIED
- Circulating, with no trace found yet.
- Grade REFUTED
- Checked against the trace and found false.
- Falsifier
- The stated observation that would prove a row wrong, written down at the same moment as the claim. If that observation happens, the row is marked falsified and the fired falsifier is logged in public in the register. Every row names one, so every claim can be checked against the world.
- Retract
- A filed correction. The original row stays visible, struck through, with the note explaining what changed.
- Theatre
- The arena a row belongs to: force, Hormuz, energy, markets, money, files, powermap, narrative, religion, softener, optics and others.
- Gazette
- An official trace: a government notice, an order, a filing, a named official on record. A gazette reprices a band; a caption does not.
- Tape
- Market prints read against filed bands.
- Band
- A price level filed in WINDOW.md with a reference print and a stated reversal (for example Brent Dec ≥ 100, reversed by a Friday weekly close under 100). Bands expire with the weekly window.
- Window
- The seven-day working file (WINDOW.md) that holds live bands. It resets on its expiry date; expired windows are archived.
- Hole
- A documented absence: a search that should return an official trace and returns nothing. Holes are logged with the query, a mundane and a charged phrasing, the count found and who made the hole (for example the gazette or an API), and are closed when a row fills them.
- System-preserver
- The test of what the powers that be attempt: when a shock lands, do named offices act to keep the existing arrangement running (calm the tape, defend the index, soften the story) instead of resolving the cause? The tracker logs VIX, breadth, volume and Brent on shock days against stated falsifier and confirm triggers.
- Blame menu
- A rotating list of causes offered by an office for an outcome that never names that office's own policy. Logged with a falsifier window like any other claim.
- Softener
- Language or framing from an office or outlet that lowers the felt cost of a DONE without changing it, such as elevating a denial over a document.
- Selection-beneficiary
- Who gains if a given line is believed, logged as inference beside the claim.
- Ladder
- How far a row has climbed toward a full trace (for example from a caption, to an attributed count, to a named person).
- World entropy
- On this site (method of Nov 14 2026): a 0–100 index of world-side harm. It is 70% verified DONE rows from the last 30 days describing strikes, attacks, seizures, shutdowns, bans, sanctions or deaths, plus UKMTO incident rows, and 30% market stress read against fixed bands. Rows are scored against a fixed baseline, so a sustained rise shows as a rise. Full method
- World restoration
- On this site (method of Nov 14 2026): a 0–100 index of world-side repair. It counts verified DONE rows describing a court ruling, a release, a ceasefire holding, an official record corrected, or accountability, on the same fixed baseline as world entropy. Our own work is not counted.
- Our register line
- The grey line on the curve: our own grading quality (share graded V and share not unverified or refuted). It is kept apart from both world lines.
- Method change
- On Nov 14 2026 the curve moved from market percentiles plus our grading shares to world-side event counts. Earlier points were recomputed under the new method, and the change is marked on the chart.