Documentary gets written in the edit because nobody could write it in advance

This is less a set of instructions about documentary structure than an argument, and it is worth saying so at the start.

Shooting before knowing

A documentary crew commits to a subject long before anyone can say what will happen, which means filming on the possibility of significance. The resulting shooting ratio is vastly higher than in drama, because there is no script telling anyone which moments will matter. Editors on long-form documentary can face hundreds of hours for a ninety-minute film, all of it needing to be logged and searchable.

The first task is therefore selection rather than shaping, and it can consume months before any structural idea is testable. Directors who describe the edit as where the film is written are being literal rather than modest.

Access decides the subject

What a documentary can be about is bounded by what its makers were allowed to film, and permission is never evenly distributed. Institutions with press offices are easier to film than institutions without, which quietly skews which stories get told at all.

Individuals who consent early can withdraw later, and a withdrawn contributor can remove the spine of a film in a single conversation. Films therefore hedge, gathering more strands than needed so that the loss of any one does not end the project. That redundancy is expensive and is one reason documentary schedules stretch across years rather than months.

Finding a spine in the material

The commonest structural discovery is that a secondary figure filmed as background turns out to carry the film's actual question. Editors test spines by assembling rough versions around different centres and watching which one generates its own momentum.

A spine works when the audience acquires a question in the first fifteen minutes that the material can genuinely answer. Where no such question exists in the footage, the film usually becomes an essay or a portrait rather than a narrative. Neither is a failure, but the shape has to be chosen honestly rather than imposed on material that will not support it.

The three repair tools

Narration supplies information the footage lacks and is the fastest fix, at the cost of putting an authority above the material. Interview supplies the same information in a participant's voice, which is warmer but constrains you to what somebody was willing to say. Archive supplies context and time depth, and it is limited by rights, cost and whether the relevant footage was ever shot.

Most documentaries use all three, and the balance between them is the clearest signal of how confident the film is in its own footage.

A film leaning heavily on narration is often one whose access ran out before its structure arrived.

Where the ethical pressure sits

Every documentary is a construction, and constructing a person from selected footage is a genuine responsibility rather than a philosophical puzzle. Standard practice varies by broadcaster and country, and contributors' rights over their own contributions differ considerably between jurisdictions.

Here is what the choice buys: the recurring hard cases involve reconstruction, sequencing that implies a causation the footage does not establish, and consent that predates the finished film. Editors and producers debate these constantly, and the honest position is that the rules are contested rather than settled. Viewers can watch for the signs: unexplained cutaways during interviews, reconstructions unlabelled as such, and questions edited out of answers.

Watching a documentary structurally

Ask what question the film established early, and whether the material it gathered could ever have answered that question. Notice when narration arrives, since it usually appears exactly where the footage stops carrying the story on its own. Watch how interviews are cut, because a subject cut mid-thought repeatedly is being used as a supplier of lines rather than as a voice.

On the call sheet, consider who is absent, as absence in documentary is almost always about access rather than about editorial choice. Reading a documentary this way makes it more interesting rather than less, which is not true of every kind of scepticism.

The takeaway

In documentary the structure is found, not planned, and the finding is the work.

Popularity and craft are two different measurements, and both of them are real.