Zettelkasten for Fiction Writers: Connect Ideas to Drafts

A Zettelkasten for fiction writers is a collection of small, linked notes that connects observations and research to characters, settings, and possible scenes. Use it to explore relationships between ideas, then choose which ideas belong in your draft. Tools like fenbs can keep the approved story choices and next steps available when you work with different AI assistants.

You do not need to turn your novel into a database. Start with a handful of notes about a scene you actually want to write. The useful result is a clearer dramatic choice, such as why a character refuses help, rather than a larger collection of neatly labeled fragments.

How do you start with a scene you want to write?

1. Choose one live question, such as why your protagonist stays in a town she wants to leave.
2. Write separate notes for an observation, a character motive, and a possible consequence. Give each a descriptive title.
3. Link the notes and add a sentence explaining the connection. A shared keyword alone is not a reason.
4. Open a scene note that points to the useful ideas. Write what the character wants, what blocks her, and what changes.
5. Draft the scene in your manuscript. Return afterward to mark which ideas you used, changed, or rejected.

For a hypothetical coastal mystery, your observation might be that an empty ferry terminal feels different after dark. A second note says the protagonist associates departures with abandonment. Linking them suggests a scene in which she misses the last boat deliberately. The relationship produces a decision you can dramatize.

Keep this experiment small enough to finish in one sitting. If you spend the entire session deciding on labels, simplify the system until you can get back to the scene. The wider second-brain collection covers ways to keep knowledge useful beyond a single draft.

What belongs in a useful fiction note?

Give each note one clear purpose. A character biography can remain a longer reference document, while a linked note explores one tension inside it. “Mara wants forgiveness without confessing” gives you something to test in several scenes. “Mara” alone tells you little about why you saved the note.

Write whether a detail is observed, invented, or researched. A remembered conversation is not evidence about how an entire profession behaves. Preserve a source and a page or section reference for research, and distinguish your interpretation from the source’s actual claim.

How do you make links that generate story ideas?

Write the relationship beside the link. “This pressure could expose Mara’s fear of abandonment” gives you a reason to reopen the other note. “Related” leaves the creative work undone. Useful relationships include causes, contradictions, repeated images, reversals, and facts that rule out an easy escape.

In Obsidian, you can connect notes using its documented internal-link formats↗ (opens in a new tab). In another app, use its ordinary note links; on paper, use an identifier written on both cards. Choose a method you can maintain without stopping to rebuild your whole collection.

Try linking one character motive to two incompatible outcomes. If Mara accepts help, she risks discovery; if she refuses, someone else takes the risk. Keep both possibilities available while exploring. A link records an avenue to consider, not a promise that the finished novel must follow it.

How do linked notes become a manuscript?

Make a short assembly note for the chapter or scene you are drafting. Collect only the links that affect that passage, arrange the intended events, and write the emotional change in a sentence. Your manuscript remains the place where those choices become continuous prose.

An outline can sit alongside linked notes. Purdue OWL’s outlining guidance↗ (opens in a new tab) describes its usefulness for organizing plot threads and character traits. Use the outline for sequence and the linked collection for connections that may cross chapters or projects.

After drafting, update only what changed the story. If the scene establishes that Mara cannot swim, record that choice where you keep approved continuity. Leave abandoned possibilities in the idea collection with a clear label. Otherwise an attractive unused option may return later disguised as an established fact.

What should an AI assistant read before helping?

Give an assistant the current scene goal, relevant notes, approved continuity, and the limits of the request. Ask it to identify tensions or suggest alternatives before requesting prose. A small, selected packet makes it easier for you to see whether the suggestions respect what you already chose.

State explicitly that proposals remain proposals. For example: “Mara cannot swim is approved. Her fear of boats is only a possibility. Suggest three scene turns without changing the approved detail.” The distinction matters more than whether the assistant can see every note you have ever written.

The guide to portable AI memory explains the broader handoff problem; an AI brain gives that retained work a home. With fenbs, record the choices you approved in Decisions and rules and save the next writing step through its shared work record.

How to do this with fenbs

1. Create a free account and add a project for the story you are drafting.
2. Add a context note with the premise, point of view, relevant source locations, and current scene goal. Include only material needed for the work.
3. Record your approved continuity and writing constraints in Decisions and rules. Keep undecided scene options clearly described as suggestions.
4. Connect your preferred assistant through the integration guides, and ask it to read the project context before making suggestions.
5. Before stopping, have the assistant save Where we left off: the scene attempted, your accepted changes, and the question to address next.

fenbs does not import your note collection or replace a manuscript editor. If you write alone without AI and your existing notes preserve continuity well, that setup may be enough. For more on the wider method of retaining reusable ideas, read what an AI second brain is.