Renaming a character mid-manuscript.

Every writer eventually meets this problem.

You're deep into a draft — twenty, thirty chapters in. And somewhere between chapters eighteen and twenty-two, it hits you: the character you've been calling Elizabeth doesn't feel like an Elizabeth anymore. She should have been Anne all along. The name shift is small on the page; the cleanup is not.

You've referred to her by name dozens of times across the manuscript. You've tagged those paragraphs to a character record. You've defined her relationships, cross-referenced her to a location, and given her a profile with backstory and motivation. And in your story bible, Elizabeth and Anne are already linked — when you first noticed the two names belonged to one person, you set Elizabeth up as an alias of Anne so search and the presence map would treat them as one. But an alias keeps both names alive, and you've now decided the book should only ever say Anne.

The prose itself is the easy part. Everything the prose points to — the tags, the relationships, the cross-references, the presence data — is the hard part.

Find and Replace only changes the letters.

Word can turn every "Elizabeth" into "Anne" in about two seconds. That's the surface change. But in a continuity tracker, the surface change is not enough on its own.

After a text-only rename, you still have:

• A story element called Elizabeth, kept alive as an alias of Anne — so the name still lives in your bible even though it no longer appears on the page.
• Every paragraph you tagged to the Elizabeth record still points at Elizabeth — not at Anne.
• Relationships defined on Elizabeth (mother of, colleague of) that belong on Anne.
• Cross-references pointing to Elizabeth, orphaned from the current text.
• An Anne record that Elizabeth points to as an alias — but with its own separate tags, relationships, and profile.
• A Character Presence Map that still carries Elizabeth as a tracked name in every chapter she was tagged in.

The alias keeps search working across both names — but you don't want both names anymore. You want one record, named Anne, with everything folded into it. That's a merge, not an alias — and rebuilding it by hand (retagging every paragraph, migrating relationships, deleting the ghost record) is exactly the ledger work a continuity tracker is supposed to save you from.

The rename-and-merge workflow.

In NovelContinuity, collapsing one name into another is a short, deliberate sequence — and the tracking data follows the text.

What the merge carries with it.

A rename that ignores the connective tissue is a rename in name only. The merge is the point where the change lands everywhere it needs to land.

• Paragraph tags. Every paragraph tagged to Elizabeth is retagged to Anne. Search now finds them under the correct name.
• Presence Map. Every chapter where Elizabeth was tagged now shows up under Anne. The map reflects the manuscript as it actually reads.
• Aliases. Any alias linked to Elizabeth (Liz, Beth, Miss Farrow) rolls onto Anne. Search expansion keeps working across every variant.
• Relationships. Anyone Elizabeth was defined as the mother, sister, colleague, or rival of — those relationships now belong to Anne.
• Cross-references. Every XRef pointing to or from Elizabeth is rewritten to point at Anne.
• Profile fields. Backstory, motivation, arc, age, occupation, physical description, speech patterns — each field on Elizabeth is merged into Anne, with your choice of which value to keep when both records had content.

Nothing gets lost. Nothing gets orphaned. The prose changed; the tracking data followed it.

When rename-and-merge earns its keep.

The Elizabeth-to-Anne example is the classic case, but it isn't the only one.

• Two records that should have been one. Sometimes the AI captures a character under a formal name in one chapter and a familiar name in another before you've defined the alias. Merging the two consolidates them into a single canonical record.
• Consolidating a minor character. A walk-on you named in chapter three turns out to be the same person as a walk-on you named in chapter fourteen. Merge them; keep the one name.
• Renaming a location. The fictional town you called Cranby in early chapters became Cranbury when the map got drawn. If both records exist, merge collapses them without you having to retag a single paragraph.
• Correcting a scanning mistake. The AI split what should have been one entity into two — a canonical and something the scan misread as a separate person. A rename-and-merge is faster than reviewing each suggestion individually.

The common thread: any time two element records need to become one, merge is the operation that carries the data with you.

This is different from an alias.

Aliases and merges solve related problems, but they are not the same operation.

An alias links different names for the same entity while keeping both surface forms visible in the manuscript. Robert is still called Bob in dialogue and Robert in narration; the alias makes sure searching for either finds every paragraph where either appears. Two names, one record, both names still on the page.

A merge collapses two element records into one because they should have always been one. Elizabeth becomes Anne; Elizabeth's record ceases to exist. One name, one record, everything the retired record carried moves onto the surviving one.

The two can meet in the middle. If you started by aliasing Elizabeth to Anne — keeping both names on the page — and later decide the book should carry only one, you remove the alias and merge instead. That's the exact path in the workflow above: an alias, once you're sure, becomes a merge.

The features involved.

The workflow above touches four parts of NovelContinuity. Find and Replace makes the text-level change. Aliases are where you started — and removing the alias link is what frees the record to be renamed. Manage Elements is where the rename happens. Rename-and-merge is the safety net that catches you when the new name is already in use — and turns what could have been an hour of manual cleanup into a single click.