History

The world your story happens against — real events, drawn on the same axis as the Timeline tool, so a chapter set in October 1970 sits directly above what was actually going on in October 1970.

What it is, and what it isn't

A History record is not a Story Bible Event. A Story Bible event is something that happens in the book — it has participants, a place, a spot in the narrative sequence, and you're answerable for it. A History record is furniture: the Grey Cup was played whether or not anyone in your manuscript watched it. Keeping them separate means a research set with hundreds of rows never shows up in the Name Registry, the continuity scan, the relationship graph, or the AI roster — it only ever appears on the timeline, where it belongs.

Building a History

Reached from History… in the Timeline tool's Visual tab (see below) — three ways in, because research arrives three ways.

Import a CSV or TSV

Import CSV… opens a file dialog first, before anything else — choose the file, and only then does the sheet open, already showing what it found. The standard columns are Start Date, End Date, Precision, Year, Strand, Category, Event / Touchstone, Significance, Region, Novel Link, Notes, Source, Duration (Days) and Date Confidence, but close spellings are accepted and only the event name column is required — a file missing Region or Duration imports fine, just blank in those columns.

Dates can be almost anything readable: 1965, March 1965, 1965-03-17, 17 March 1965, 1960s, c. 1965. A row whose date can't be parsed is still imported — it's held and shown in the table with the date text you gave it, just not drawn on the timeline until you fix it.

Add Event

Types one event at a time, in a plain form: what it is, why it matters, its category, its start and end date, and optionally a region, a generation, notes and a source. The same form opens when you edit an imported row afterwards.

Build with AI…

For when you have no spreadsheet yet. This opens a short questionnaire — subject, timeframe, roughly how dense, the parallel strands you want tracked, and so on — and assembles them into a full research brief. Sgeulai does not send this anywhere itself: the brief wants a model with web access and a file sandbox, which the app's own Claude integration doesn't have. Copy, then Open ChatGPT (or Claude — pick which from the small menu beside the button, remembered for next time) puts the brief on the clipboard and opens the assistant's site for you to paste it into and run. What comes back is a spreadsheet in the same standard shape, ready for Import CSV… above.

Which manuscript it belongs to

A research set for a series is rarely one timeline — it's several braided together. Importing asks plainly which manuscript the file is for, with checkboxes to also show it under others, rather than walking through the file's own tag values one at a time. If the file carries its own column tagging each row to a book — Twain's research set calls it "Novel Link" — an optional toggle lets those tags override the choice above wherever they actually name a manuscript in your project; a tag that names nothing in the project falls back to what you chose rather than silently becoming background research seen everywhere.

An event with no manuscript at all — nothing selected, or a project with no manuscripts yet — means "true of the world no matter which book is open," and is drawn under every manuscript's timeline.

Inside the History editor itself, an imported file becomes a named set — export it, and later remove the whole set in one action (Manage ▸ Imported sets) without touching anything you typed by hand. The same menu bulk-allocates a whole set, or a selection of rows, to a different manuscript after the fact.

Where you see it

The Timeline tool

Open Timeline and switch to its Visual tab — the same place the analysis tools describe as "the same events on two clocks." A History band runs beneath the story's own ruler, coloured by category, with a toggle to show or hide it, a key naming what each colour means, and its own zoom and focus controls shared with the ruler above it — so zooming into one chapter narrows both at once, and they never drift out of alignment with each other.

The scene and chapter footer

A compact version of the same band sits at the foot of the editor while you're writing a scene or chapter — what else was going on, right where you're working. It shows a stretch of time centred on that record's own date, with zoom controls (±7 days out to ±20 years) to widen or narrow the window. This record's own place in it is drawn as a solid black bar directly among the coloured History marks, rather than on a separate axis of its own — the point is seeing where you are amid what was really happening, not two different rulers that might disagree. Along the bottom, the earliest and newest dates the window currently shows are labelled directly under the band.

A chapter or scene with no date of its own still gets a window — centred on the story's overall midpoint instead, so zoom has something to narrow toward — but draws no black bar, since there is nothing yet to honestly mark. See below for where that date comes from.

The chapter or scene's own date

The black marker's position, and the footer's default view, come from one shared answer: an Event tied to that chapter or scene takes priority if there is one, otherwise the chapter or scene's own In-Story Date field (in the Story Bible) is used, and a chapter with none of its own can inherit one from its scenes. This deliberately does not fall back to a file's modification date or any other date that isn't part of the story itself.

Typing that date by hand is one way to set it. Find Only What's Missing… can also fill it from the manuscript's own prose, for a chapter or scene that doesn't have one yet — see Reading the manuscript with AI.