Pitch a Film Retrospective Through a Curatorial Argument, Not a Filmography

Two documents arrive on a programmer's desk in the same week. The first lists fourteen films by one filmmaker, each with a year, a running time, and the same one-line credit. It is entirely accurate. The second lists four films, states a question, and says why the four appear in the order they do. Only one of these documents has asked anyone to decide anything. The second one has.

A filmography answers what did this person make. A program answers what happens when you watch these works together, in this order. Those are different questions, and only the second one needs a screening to settle. Your proposal is judged on the second.

A filmography answers a different question

Start by writing the question your selection lets a viewer investigate. Not the theme of the filmmaker's career — the thing a person in the room could actually turn over while watching. Make it a sentence someone could disagree with. "A portrait of an artist" is not a question; nobody can argue with it. "Who is a performance for when the audience that was supposed to be there never arrives" is one, because three films can give three different answers and a viewer can prefer one.

A career chronology can serve a question perfectly well. What it cannot do is stand in for one. "Complete retrospective" promises coverage; a program promises an encounter. Both are legitimate projects, but a programmer reading your document needs to know which one is on the table, because only one of them has a reason to be in this order rather than any other.

Here is a program to work with. The films are invented, and so is the filmmaker; nothing below reports a real production or anyone's actual intentions.

Three works by one hand. Solo (2022) spends an hour with a single performer running the same passage in an empty room. Interruption (2017) watches an ensemble rehearse while the rehearsal keeps being broken into — someone comes in, a scene restarts, a conversation happens that is not the scene. Aftermath (2014) is set in the room once the audience has gone: the performance is over, the people are still there, and there is nothing left to play to.

The question the selection proposes: who is a performance for when its expected audience is absent? The program's bet is that each film presses that question at a different point, and that the differences — not the resemblance — are why all three belong.

Give each work a job

Once you have the question, every work needs a distinct job in answering it. A film can introduce the question, complicate it, contradict an answer the program has already offered, or change what the question means. What a film must not do is repeat a job another film already holds.

The test is removal. Take one title out and ask what the viewer can no longer see. If the answer is "nothing important," the film is decoration. If the answer is the same for two titles, you have either a distinction to sharpen or a selection to cut.

Apply it here. Solo introduces the question in its purest form: an audience is not merely absent, it was never expected. A performance is happening for one person. Remove Solo and the program starts with an audience already somewhere in the frame, because Interruption is a rehearsal toward a performance that will have one. The question loses its zero point — the version of it that has no crowd in it at all.

Interruption complicates. Here an audience is anticipated but not present, and the people who actually appear are not that audience. The film pulls apart two things the question usually keeps together: an audience and whoever happens to be in the room. Remove Interruption and the program jumps from no audience to a finished one, skipping the distinction between who was expected and who showed up.

Aftermath changes the tense. The audience came and went; the performance already had its recipients. Remove it and the program stays inside preparation, asking what a rehearsal is for without ever asking what it was for. Three positions, then: Solo is not here, Interruption not yet, Aftermath no longer. They are close enough to belong to one program and far enough apart that none of them answers the question the way the others do.

Watch for the weak version of this pitch, which claims all three films are "about performance" and lets that sameness carry the argument. It makes each work redundant and gives a programmer no reason to keep three instead of one. The stronger move is not to find a theme they share but to find what each one does that the others cannot.

Two running orders, two arguments

The same three films make two different programs, depending on which question the order asks the viewer to hold.

Suppose the films were made Aftermath (2014), then Interruption (2017), then Solo (2022). Career order plays them that way. The viewer meets a post-performance room first, returns to a rehearsal second, and lands on a late, solitary film last, which will read as a retreat or a summing-up — the artist alone at the end. The gain is that you see a body of work change over eight years, and the question becomes "how did this filmmaker's thinking about an audience move?" The loss is that the question stops being about performance and becomes a story about a person. Solo arriving last reads as a conclusion rather than as a third position.

Now play them Solo, Interruption, Aftermath. The order tracks the life of a performance — private, interrupted, past — rather than the life of a career. The gain is that the question stays the subject; each film is the next beat in one arc, and the viewer is carried along it. The loss is real: a viewer who does not read the program note will assume the films were made in the order they are shown, and Solo, the most recent work, will be read as an early one. That misreading can change what the film seems to be doing.

Notice that the second order is the exact reverse of the career. Some pitches would take that as proof of sophistication, as if nonchronological sequence were somehow more authored. It isn't. The reason to reverse here is specific: this order follows the object under study, not the author. That is the only justification that survives a programmer's question, which will be "why this order and not the obvious one?"

Answer the question by naming what the first screening makes the viewer expect and what the second does to that expectation. Solo sets up an empty room and a performance with no recipient. Interruption then removes the privacy: now there are people, and they are the wrong people. Aftermath removes the performance. Each screening has to change something the previous one established, or the order is just a shuffle.

Let the context earn its place

An introduction, a note, or a discussion is part of the encounter you are proposing, and it should have a job as specific as the films do. Ask what question it opens, and what a viewer needs before a film rather than after it.

Before Solo, a viewer who knows the program's question can watch an empty room as an answer. A viewer who doesn't will watch it as a blank, and may spend the hour waiting for the film to start. That is the whole test: information that changes how the next ninety minutes are read belongs in front; information that only confirms what the film already showed belongs behind it, if anywhere.

Now separate two things that look alike in a proposal. "A post-screening conversation about who counts as an audience" is a proposal; you can offer it without anyone's agreement. "The filmmaker will attend" is a claim about a person's calendar. It is true only if it is confirmed, and a programmer will treat an unconfirmed attendance claim as a warning about everything else in your document. Propose the discussion. Report the attendance separately, and only once you have it.

Program notes function a little differently, and they vary enough that this piece isn't going to quote one for you. When you read the note attached to a program you're pitching, mark which relationships it states outright and which you are bringing yourself. If the note says the films are joined by their shared setting, and your pitch says they are joined by three different positions on an absent audience, that is fine — but it is your reading, and it should not be attributed to the note or to the filmmaker. The distinction matters when someone asks why a work was chosen, and the honest answer is "because of what it does in this sequence," not "because the note said so."

Biography earns its place last. Knowing that a filmmaker once ran a company might explain why a rehearsal film was shot in a working room, and if that fact changes how a viewer reads the interruptions — as familiar rather than staged, say — it belongs. If it only adds color, it is a substitution for the curatorial argument, and it will read as one.

Separate the proposal from the logistics

A program proposal makes two kinds of statement, and they should never blur. The first kind is the argument: these works, this question, this order. The second kind is a status report, and its job is to be boring and precise.

For status, "unknown" is a complete answer. It is not a soft way of saying "unavailable," and it is not a polite way of saying "secured." A programmer reading the following is being told exactly what is true and nothing more.

All three titles are invented, which is why every row reads the same way. The point is the format. A real proposal would fill these in as facts become facts, and would leave "not established" in place wherever it is still true. Nobody has ever lost a screening because a proposal said it didn't know something yet. Proposals get refused for saying "print available" when nobody checked.

When a work is essential and turns out to be unavailable, say what the program loses before you name a substitute. If Interruption cannot be shown, the program loses its middle term — the distinction between the audience that was expected and the people who actually turned up — and shrinks to two positions instead of three. That loss is worth stating plainly. Then, if you propose something else, check that the substitute does the same job rather than merely fitting the theme. A replacement that also complicates the question is a fix. A replacement that introduces it is a different program.

The venue is its own status report, and the requirements are the venue's, not a standard. Deadlines, required materials, technical specifications, and the format they expect for a proposal all differ between organizations, and they change. Read the current published requirements for the place you are actually approaching, rather than a template you found from another one.

Your proposal is complete when a programmer can say back three things: why these works belong together, why they appear in this order, and what remains to be arranged. Get those three sentences out of someone who has read the document, and you have a pitch. If the only thing your document can explain is what the filmmaker has made, you still have a filmography — which is a perfectly good list, and not yet a program.

Frequently asked questions

What question should a retrospective program propose?

It should propose something a viewer in the room could investigate and disagree with, not just a theme. 'A portrait of an artist' cannot be argued with, while 'who is a performance for when the audience that was supposed to be there never arrives' can be answered differently by three films. The selection should let a viewer turn that question over.

How can each film earn its place in the program?

Each work needs a distinct job in answering the question: introduce it, complicate it, contradict an answer, or change what the question means. Use the removal test: take one title out and ask what the viewer can no longer see. If removing two titles yields the same answer, either sharpen the distinction or cut. A program that claims all films are about performance makes each work redundant.

How should a programmer justify the running order?

Name what the first screening makes the viewer expect and what the second does to that expectation. In the invented example, career order plays Aftermath, Interruption, Solo and makes Solo read as a retreat or summing-up; the order Solo, Interruption, Aftermath tracks the life of a performance rather than a career. Reversing chronology is not inherently more authored; the reason must follow the object under study, and the order should be explained by why it is not the obvious one.

When does context, such as an introduction or discussion, belong in a retrospective proposal?

Information that changes how the next ninety minutes are read belongs in front; information that only confirms what the film already showed belongs behind it, if anywhere. Separate a proposal from a claim: 'a post-screening conversation about who counts as an audience' is a proposal, while 'the filmmaker will attend' is a claim about a calendar and should be reported only once confirmed. Biography earns its place last, only if it changes how a viewer reads the work.

How should a proposal handle uncertain logistics and unavailable works?

Use a status report where 'unknown' or 'not established' is a complete answer, not a soft way of saying unavailable or secured. Do not write 'print available' if nobody checked. If an essential work cannot be shown, state what the program loses before naming a substitute, then check that the substitute does the same job rather than merely fitting the theme. Venue requirements differ and change, so read the current published requirements for the place being approached.

Further reading

• Pitch a Community Screening as an Event, Not Just a Film Link — If the program is one evening in a room, this shows how to describe its run, name the division of labor, and ask for one answerable decision.