Pitch a Documentary in Which the Filmmaker Is Also a Character
Pitch a Documentary in Which the Filmmaker Is Also a Character
The deck has a slide called "Cast." The first entry reads: Maya — filmmaker, part of the journey. Everyone in the room nods. Nobody can say what Maya does on the second day of shooting.
That phrase is a promise about the shape of the film, not a description of a role. It tells a commissioner that the maker will be present. It doesn't say whether she asks questions, answers them, looks for something, gets in the way, or stands outside a door deciding whether to knock.
The fix isn't more biography and it isn't a confession. It's an action, stated plainly, plus the two things that follow from it: what the maker can know and when, and what their being there changes. Get those three into the pitch and the rest of the document tends to sort itself out. Leave them out and you get the commonest failure in personal documentary pitches — a filmmaker who is described as a character and who narrates like a god.
Describe a participant, not a floating narrator
A role has three parts: a relationship, an action and a consequence.
The relationship says whose inquiry this is and where the maker stands relative to the people in it. "I am trying to find out what happened to my father's first family" names a stake. "I have a deep personal connection to this story" names a feeling about a stake, which is not the same thing and does less work.
The action is what the maker actually does inside the events. This is where most pitches go soft, because different actions create genuinely different positions, and a pitch that blurs them can't be trusted later:
- Looking for an answer makes the maker a searcher. The film's suspense is theirs, and so is the risk of being wrong.
- Making a request makes the maker someone who has asked something of another person. Now there's a debt, a possible refusal, and a relationship to manage.
- Deciding whether to intervene — to press, to correct, to stop, to hand over the camera — makes the maker a party to the outcome. That is not the same role as the searcher, even when the same person occupies both.
The consequence is what the action changes. If you genuinely change nothing, that may be a fine thing to be — a witness — but say so, because a witness is a different pitch from a participant and the two shouldn't be sold as one.
The related mistake is assuming that a personal stake buys screen time. It doesn't have to. A maker can own two sequences and be absent from six, and the pitch is stronger for saying which. Personal involvement earns a role; it doesn't earn every scene. If your aunt is the one who knows how to hold the photographs, the sequence belongs to her, and your absence from it is a choice you made rather than an omission you're hoping nobody notices.
Separate what you knew then from what you know now
A documentary pitch is usually written from the far end of the process. By the time anyone reads it, you know how the story goes. The temptation is to let that knowledge leak backwards, so the synopsis reads as though the maker understood things at the time that they only understood months later.
Draw the timeline before you write the synopsis. For one consequential event, sort the material into separate boxes:
- Directly experienced. You were in the room. You heard the words, in that order.
- Reported to you afterwards. Somebody told you, later, what happened or why.
- Learned from material you obtained later. Footage, photographs, a letter, a record.
- Learned in the edit. A pattern that becomes visible when two scenes sit side by side. This is real knowledge, but it's knowledge about your material, not about the past.
- Still unknown. Name it here, and keep naming it.
That last box is the one that makes the others believable. A pitch that lists what it doesn't know reads as more accurate than one that resolves everything.
Then write the synopsis in the grammar of discovery. "I asked her. I was sure. She said she couldn't remember" is a scene. "My aunt's refusal to identify her own uncle became the hinge of the film" is an editorial verdict wearing the clothes of an event. The second sentence isn't false, exactly. It's just doing something the film hasn't shown yet, and it quietly awards the maker knowledge she didn't have at the table.
Show how your presence enters the situation
Since you were there, you changed something. Not necessarily the events themselves — the witness role from the previous section keeps that option open — but the situation around them: who agreed to be filmed, what got said while the camera was running, which door stayed shut. Say what, in terms somebody could check.
Observable changes are usually small and specific: the room gets rearranged so the camera has a line of sight, a question gets asked that nobody would have asked, one door is left open and another is closed, a photograph goes back in the box instead of onto the pile. Requests, questions and negotiations are all fair material. So are answers you didn't get.
What you can't claim is the counterfactual. You have no second version of the afternoon in which you weren't there, so "she would never have said that without the camera" is guesswork with a documentary budget. If people seemed to watch their words, describe them watching their words. Don't explain why.
Your presence probably also changed what you were given access to, and that's story material rather than an embarrassment. If your aunt agreed to be filmed because you're family, that's worth one line in the pitch, because it tells a commissioner what kind of access this is. If she agreed and then asked for the camera to be switched off partway through, that moment is itself an observed event. What she said while it was off is a different question, and not one the pitch gets to decide. Consent, family exposure and the review any real project needs depend on the actual people and the actual setting; those conversations belong with the participants and, where it applies, with someone qualified to advise on them.
Finally, plan for other accounts. Another participant describing the same afternoon differently is not a problem to solve in the pitch — it's a sign that the film has more than one person in it. A pitch that can absorb that complication is more credible than one that requires it not to happen.
Put the participation rule into the pitch materials
Two things do most of the work: a short role paragraph and a sample scene. A third completes the page — the knowledge limit, which is what makes the other two credible.
The role paragraph is one paragraph — relationship, actions, and when the maker appears or withdraws. It's the rule the rest of the document is held to.
The sample scene shows the rule in operation. One page. It should read as an example of the kind of encounter the film contains, not as a transcript of one that has already happened, unless it has.
That distinction is where pitches get sloppy, and the fix is tense. Proposed encounters go in the future: I will ask her to go through the box. Filmed ones go in the past, with the footage behind them: we filmed her going through the box in March. Access works the same way. Hoping to film something is not the same as having permission to film it, and a commissioner who discovers the difference late will read everything else in the document with suspicion.
One test before you send it. Cover the maker's name wherever it appears in the synopsis and write "the narrator" instead. Do the other people still do anything? Do they want anything? If every other character exists to react to the narrator, you haven't pitched a documentary with a participant in it. You've pitched a monologue with extras.
One photograph, two versions
The rest of this article works on a single invented case. Nothing here comes from a real family, a real production or a real archive. The visit, the photograph and the letter are teaching materials.
The material. A filmmaker is trying to identify the people in a box of unlabeled family photographs, and their aunt has agreed to help. During the visit, the maker asks her to tip the box out onto the table and confidently names one of the men in one of the prints. The aunt says she cannot remember. Months later, the maker finds a dated letter that suggests the naming was wrong.
Participant-time version. What the maker could write if the account stopped where the visit stopped:
I asked my aunt to tip the box out onto the kitchen table, and she did. Loose prints, most of them with nothing written on the back, a few with an orange date stamped in the corner.
"That's Ray," I said, when we reached the man at the edge of the group. I was sure of it. "Same jaw as the one on the stairs."
She turned the print toward the window and looked at it for a while.
"I can't remember," she said.
"You must."
"I can't." She set the print down and reached for the next one.
That is the whole scene, and "You must" is the moment where participation stops being a description of a role and becomes an action. It also happens to be the moment that invites interpretation later, which is exactly where pitches go wrong.
Edit-time version. Same visit, with the later information attached and dated:
[the kitchen scene, as above]
In November, going through the rest of the box for a completely different reason, I found a letter. It's dated August 1961 and describes a party at the house — the same porch, the same two windows as the print. Ray, the letter says, had written from Norfolk to say he wouldn't be back until the spring.
I still don't know when the photograph was taken. Nobody has written anything on the back of it.
What the letter does and doesn't do. It raises a doubt about the name. It doesn't settle it — the letter describes an occasion, and the photograph is only probably from that occasion. And it says nothing at all about what the aunt was thinking at the table, then or since.
Two ways to say the same thing.
Misleading: "My aunt knew who he was and sat on it."
That sentence claims a certainty, an action and a motive, none of which anyone has. It also converts her into a puzzle for the maker to solve.
Bounded: "I took her uncertainty as reluctance; later evidence unsettles my assumption."
Much better. The claim is about the maker's own reading, and the later information stays later. It can be tightened once more, though, because the letter doesn't unsettle everything in that sentence — only the naming. The version I'd actually write is: "I took her uncertainty as reluctance. The letter suggests I was wrong about the man's name. It tells me nothing about what she was thinking."
The knowledge timeline for this scene.
- Directly experienced, at the time: the request, her turning the print to the window, her words, my "You must," my certainty.
- Learned later, from a document: the letter's existence, its date, what it says about Ray's plans, the resemblance of the porch.
- Learned only in the edit: that my certainty and my insistence belong together. One scene doesn't establish a pattern, so the pitch shouldn't claim one; if five visits exist and all five look like this, then the pattern is edited knowledge and should be described as such.
- Unknown then and now: whether the man is Ray, whether the aunt recognized him, what she made of being pressed, whether the letter describes the same occasion.
Participation paragraph, as it would go in the pitch. The film is built from one box of unlabeled photographs. I ask the questions; my aunt answers some and not others. I appear whenever I'm asking, looking, or getting something wrong. I'm not in the sequences where she describes her own family, and I don't narrate over them.
Sample scene (proposed; nothing here has been filmed). At a kitchen table, I will ask my aunt to help identify the people in a box of unlabeled prints. I will name one of them and she will say she can't remember. I will press her, and the film will come back to this moment later, when a letter complicates my certainty about a name.
Knowledge limit. At the table I was certain the man was Ray. A letter found eight months later suggests I was wrong, without proving it, and says nothing about what my aunt was thinking. The film says so out loud.
The maker's changing interpretation is visible in that little set of blocks because the confident naming arrives early, the correction arrives late and dated, and the correction is about the maker's own identification rather than about the aunt's character. If the letter were used instead to reveal "what she was really feeling," she would stop being a person and become an instrument of the maker's arc.
The page you send
Three blocks, in this order: one participation paragraph, one scene, one clearly labeled knowledge limit.
The paragraph states the rule. The scene demonstrates it, because a rule nobody can see in action is just a claim about your own honesty. The limit states what the film won't assert, which is often the fastest way to show a commissioner that the project knows the difference between a participant and an author.
Before it goes out, check the three places where personal documentary pitches quietly lie. Proposed encounters written as though they'd been filmed. Access you hope for written as access you have. Other people described only in terms of what they mean to you. Then check your ending. A reader should be able to see a filmmaker who does something, learns something afterward, and revises — not a filmmaker who was right all along and narrated from a chair the rest of them couldn't see.
Frequently asked questions
What makes 'Maya — filmmaker, part of the journey' insufficient in a pitch?
It promises presence but not a role. A role has relationship, action, and consequence. Relationship names whose inquiry this is and where the maker stands; action is what the maker actually does inside the events; consequence is what the action changes. If the maker genuinely changes nothing, that may be a fine witness role, but say so, because a witness is a different pitch from a participant.
How do different actions inside the events create different filmmaker positions?
Looking for an answer makes the maker a searcher, with the film's suspense and the risk of being wrong. Making a request means the maker has asked something of another person, creating a debt, a possible refusal, and a relationship to manage. Deciding whether to intervene—to press, correct, stop, or hand over the camera—makes the maker a party to the outcome. Blurring these positions makes the pitch untrustworthy later.
How should a pitch separate what the filmmaker knew then from what they know now?
Draw the timeline before writing the synopsis. Sort material into directly experienced, reported afterwards, learned from material obtained later, learned in the edit, and still unknown. Name the unknowns. Write the synopsis in the grammar of discovery, as in 'I asked her. I was sure. She said she couldn't remember,' not an editorial verdict such as 'her refusal became the hinge of the film.'
What can a filmmaker claim about their effect on a situation?
Only observable changes: the room rearranged so the camera has a line of sight, a question asked that nobody else would have asked, one door left open and another closed, a photograph returned to the box. Requests, questions, negotiations, and answers not received are fair material. Do not claim the counterfactual, such as 'she would never have said that without the camera.' If people seemed to watch their words, describe that; do not explain why.
What three blocks should go at the end of the personal documentary pitch?
One participation paragraph stating the rule, one sample scene demonstrating it, and one clearly labeled knowledge limit stating what the film will not assert. Proposed encounters go in the future; filmed ones go in the past with the footage behind them. Access you hope for is not access you have. Test by replacing the maker's name with 'the narrator' and checking whether the other people still do and want things.