What this is
One structured record per topic, holding seven things: the people the topic turns on, the texts that establish its timeline, the timeline itself, a flag on the events that are hinges, the commentary on the church's part in it kept separate by camp, the stories worth telling, and a plain statement of why the topic matters. The site is generated from those records — nothing here is hand-maintained, so a correction to a record propagates everywhere it appears.
The evidence base
Historical claims are drawn from the Mormon Corpus: roughly 49,600 works and 800,000 indexed passages, including the Joseph Smith Papers in full text, the Journal of Discourses complete, General Conference back to 1880, full runs of the early Latter-day Saint and dissident periodicals, the Journal of Mormon History, Dialogue, Sunstone, BYU Studies, FAIR, Interpreter, the Gospel Topics essays, Saints, and a named list of monographs.
The rules the research followed
Six of them, and they are load-bearing. Coverage questions go to the corpus manifest, never to search — a book the corpus does not hold still returns every place it is cited, which reads exactly like a hit. Contested questions are reported by perspective label, never merged. Perspective balancing stays on. Primary documents are found by title, date or exact phrase, because a document is named rather than summarised. Asking for more sources means excluding the ones already seen. And a first-query timeout is a cold start, not a failure.
Two traps are worth naming because they produce confident wrong attributions. FAIR's rebuttal pages reproduce another author's book at length in order to answer it, so a passage there may be the critic's words carrying a faithful-apologetic label. And the B. H. Roberts Foundation archive — the largest collection in the corpus — is an annotated index of one-paragraph citation records, superb for locating a document and never quotable as one.
What the source lists mean
Every source record says whether the corpus holds the text or only reaches it through quotation inside something else. That distinction is the difference between a citation you can verify and one you are trusting an intermediary for, so it sits on the card rather than in a footnote. Sources are ordered by evidentiary weight: primary, then contemporary, then documentary editions, then official, scholarly, apologetic and critical treatments.
Verification
An adversarial pass ran over the finished records: twenty checks across nine topics — six verbatim quotations relocated in the corpus, six ‘held in full’ claims tested, four dates and figures, two perspective labels, and two claimed absences. Seventeen passed. Three produced corrections, all applied: a Gospel Topics essay wrongly marked as held, an Edmunds–Tucker date that disagreed between two topics, and one source over-labelled as primary. All six sampled quotations matched the corpus exactly.
What is not here yet
No Mormon Stories episode numbers. The transcript database is a separate system and was not queried for this build, and a guessed episode number is worse than none — so each topic carries search leads for it instead, ready to be resolved into citations.
The corpus also has real gaps, recorded per topic rather than papered over. Collected across every topic in this build:
