Atlas insight · Research for writers
All insightsHow to build a research dossier for a novel.
You need to know whether a letter could arrive before your character leaves town. The folder holds maps and railway timetables, but no answer. A research dossier should be organized around questions like this: details that affect what happens in the manuscript and what its characters can know.
Begin with the manuscript’s pressure points.
Do not begin with the century, country, or broad subject. Begin with the moments where the story must become credible. How would a message travel? What could a physician know in that year? What would a room smell like after the machinery started? Which law, ritual, material, route, or social hierarchy constrains the character’s choice?
These questions create a finite research agenda. They also reveal the difference between background knowledge and story bearing detail. Background prevents obvious error. Story bearing detail changes action, voice, timing, motive, or consequence and deserves deeper support.
Use primary sources for texture and secondary sources for structure.
A letter can supply a form of address. An advertisement can give a price; a manual can show how a tool was used. Primary sources preserve details that later accounts may omit. The Library of Congress and national archives are useful starting points for locating them.
Use secondary scholarship to place those details in context. One vivid source may describe an exception rather than common practice. Record whether an account is contemporary or recalled later, and whether it represents personal experience or an institution’s stated position.
Build the dossier around story systems.
A single subject folder is rarely enough. Organize around systems that create constraints: time and travel, money and exchange, medicine and injury, work and craft, law and authority, domestic life, language and register, belief and ritual, geography and weather, material culture, and communication. Each system can contain verified details, uncertainties, visual references, and consequences for the plot.
These questions interact. A travel time can affect an alibi, while a communication delay changes what a character knows. Keep the connected details together so a revision to one does not leave the others inconsistent.
Separate historical fact, plausible construction, and invention.
The dossier should label what is directly documented, what is reasonably inferred from adjacent evidence, what is disputed, and what the author has invented. This protects the manuscript from accidental certainty and gives the author freedom to break from the record deliberately rather than unknowingly.
Plausibility is not a weaker category when it is honest. Many private experiences leave no surviving record. A well supported construction may combine known physical conditions, documented institutional practice, and a character specific situation. The label allows the author and later fact checker to see how the scene was built.
Track what each character can know.
Historical and speculative fiction often fails through information, not atmosphere. A fact may be true in the world and still unavailable to the character. The continuity system should track discovery, rumor, secrecy, expertise, memory, and the source through which knowledge moves.
A character knowledge ledger can sit beside the master chronology. For each consequential fact, record who knows it, when they learn it, how reliable the information is, and what they believe it means. This prevents later revisions from giving characters impossible knowledge and creates opportunities for suspense rooted in the world rather than imposed upon it.
Preserve provenance without burdening the prose.
The manuscript should feel inhabited, not footnoted. The working system can carry the citations so the final prose does not have to announce its research. Record the source, page, archive, date, image rights status, quotation limits, and the specific detail supported. For web material, retain an access date and an archived or stable location when possible.
Rights matter as well as accuracy. A public record may be accessible without every image, edition, or transcription being free to reproduce. The dossier should separate a source used for background from material intended for quotation, illustration, or publication and flag questions for rights review.
Organize the dossier for revision
The final dossier may include a research brief, annotated chronology, world rules, location sheets, character knowledge ledger, terminology guide, visual references, unresolved questions, and a source register. The order should follow the writer’s actual revision process. A novelist working scene by scene needs a different interface from a technical author verifying claims chapter by chapter.
The best dossier remains alive through revision. New details can be added without breaking the structure, contradictions are surfaced, and decisions are recorded. It should reduce the time spent rediscovering research while making it easier to see where the story is asking for a better question.
Research gateways
These public institutions explain the value of primary sources and provide practical starting points for historical and literary research.
An introduction to original documents and objects as the raw materials of historical research.
Library of Congress literary research guideAn introduction to literary collections and methods for finding relevant material.
National Archives research startA practical path from defining a topic to locating records and planning deeper archival research.