The charts under Research draw the corpus. These are findings about it — each a short claim with the evidence behind it, computed from the same index at the moment you load the page. They are observations about the edition and its vocabulary, not arguments about the history of the Latin East, and each says what it cannot establish.
The same party, twice
Where one party name sits verbatim inside a longer one — the reconciliation work the database has never had.
126 names affectedThe shape of the record
How few parties account for half of every recorded relationship, and how long the tail of one-off appearances runs.
78 parties are half of itActs that do two things
Transaction types that share an entry far more often than chance would put them together.
29.2% carry several typesWhere the record thins
Entries by decade across the whole span, and what that says about coverage rather than about history.
173 entries in a median decadeAsk me a research question
A question drawn from the corpus, with the figures it rests on and a link to the entries behind it. A place to start, not an answer.
65 in the pool
.jpg)
.jpg)
,_roi_de_Jérusalem_-_sceau.jpg)
