# Chapter 6 · The Camera Separates Us

## Standing in the Same Room

A rehearsal can crowd actors, directors, camera operators, and spectators into the same space. Yet the finished film always gives later viewers only part of that space. Some people enter the frame; some remain as voices; some have their position indicated only by another person’s gaze offscreen. Once filmed, their shared presence already takes on different degrees of visibility.

This determines how I understand the camera’s participation.

The person filming can approach someone, select their face, and follow their movements. In doing so, they may lose someone else’s reaction. Getting closer does not necessarily mean understanding more. A close-up gives weight to a face while pushing its surrounding conditions offscreen. A hesitant expression, detached from the request it faces, can easily be taken as a character trait. A statement detached from the arrangement that brought it forth can likewise be read as a self-contained confession.

Moving the camera farther away does not automatically solve the problem. A wide shot can preserve several people’s positions while making the smallest gestures between them difficult to discern. Including many people in one image does not distribute their importance equally. Light and shadow, front and back, and distance from the camera all help determine who is easier to see.

These are choices cinema has to make every day. I want to begin with such choices when speaking about truth. The frame need not contain the entire event, but it must give the fragment it keeps a reason to stand. The more clearly a shot seems to show someone, the more necessary it is to understand what has been excluded to make that clarity possible.

In an image of the public event in Shoot Self with You, spectators, a transparent partition, and lighting equipment enter the same field of view. A light stand occupies part of the image, spectators are on the other side, and bodies share the space with the objects that keep the event running. The performance does not fill the entire frame; the space supporting it is also retained.[6-1]

But showing equipment is only a beginning. Equipment does not do the film’s reflecting for it. It has to enter into relations with people’s positions, words, and choices to become more than another object at the event. Seeing who faces the performance area, who stands beside the partition, and what blocks another part of the view gives us concrete conditions of viewing that we can question.

What interests me is allowing the conditions of viewing to enter what is viewed. A spectator can recognize whom a person is addressing and also realize that they are seeing this person through someone else’s choices. A film’s honesty need not depend on the disappearance of the person filming.

## A Sentence Belongs to More Than One Face

Sound unsettles the borders of the frame. The image gives us only what is before us, while sound can bring in another space. Hearing someone without seeing them, and seeing someone without knowing whether they are listening, are two different ways of being with others.

In Shoot Self with You, near the dialogue about embracing and temperature, the image shifts from a wide shot of the bathing pool to a close view of bodies on a bed.[6-2] The wide shot places the bodies within their surroundings; the closer view gives weight to faces, arms, and cushions. As these two scales meet, viewing shifts from the space the people occupy to parts of their bodies and nearby objects.

The dialogue makes temperature a condition, while the close view gives weight to bodily shapes and positions. An embrace thus appears as more than a declaration or a wish: it also involves postures and distances that have to be coordinated. Sound and image may not show the continuous completion of the same action, yet together they can present intimacy as something that needs to be worked out. Cinematic thought can arise from such connections.

This is also why I do not want to understand offscreen speech merely as missing information. Some voices need a speaker to be found; others have an effect precisely because they cannot immediately be fixed to a face. A voice may be speaking of its own experience, quoting a character, or allowing someone who has left the frame to keep affecting what is happening. Identifying every sound source can sometimes prevent a mistaken attribution; at other times, it erases a distance the film has deliberately kept.

The value of that distance depends on what it allows us to compare that could not previously be seen together, and how it changes the meaning of a statement. An incidental relation between sound and image may gain force in the finished film; deliberately manufactured ambiguity may simply make recognition harder.

Someone can say “I,” but the film can bring that statement before people who were not originally present. The words still have a source; the range of people receiving them has changed. The director has to take responsibility for that change. It also gives viewers a chance to hear the statement again from a different position.

## How Much Must Remain for Us to See?

Once an experience enters the editing room, another difficulty arises. For those who took part, every gesture has memories before and after it. Viewers do not have those memories. Keeping an entire piece of footage cannot also give them what the camera never recorded.

Creators can easily overestimate how self-evident their material is. That moment mattered so much—how could anyone fail to see it? But the work has to make clear what, exactly, mattered. A silence may come from waiting, exhaustion, or simply equipment that is not yet ready. If the relevant conditions all remain outside the film, prolonging the silence does not automatically bring viewers closer to the experience of those involved.

Cutting therefore has constructive work to do. It gives a place to distinctions previously submerged in a mass of activity, allowing one statement to meet the next statement that actually changes its use. Editing can also make an action appear more isolated, more certain, even more like someone’s essential nature. Keeping and cutting both shape judgment; neither is inherently innocent.

Discussing the making of L’Amour fou, Rivette distinguishes the unfolding of events during filming from their organization in editing. The shoot can give time to things not yet known, while the work still has to choose among the material that remains.[6-3] This matters to my own practice. Opening up the event does not require me to surrender every judgment in postproduction as well.

How much time I give a piece of footage should depend on how the relations within it change. If another attempt makes an earlier instruction cease to work, lets someone else take over the questioning, or leads viewers to revise their understanding of the previous action, that duration is doing work. If each repetition merely confirms what is already known, keeping it may chiefly satisfy the creator’s reluctance to leave the event behind.

Nor can this judgment be reduced to cutting repetition. Some repetition works precisely by accumulating pressure, making a request that was initially easy to accept increasingly difficult to bear. Viewers’ fatigue can become material for the work. But the film still needs to give that fatigue an object: what is being repeated, who has to continue, and who can call a halt? A general demand that viewers endure a long duration merely passes responsibility for form on to them.

We can imagine how a request cut off as soon as an answer appears, and one that keeps the revisions and further questions after the answer, would make two different scenes. The first emphasizes the reply; the second lets its meaning continue to change. In comparing these possible edits, the key question is what the additional time actually changes.

I do not ask film to make experience smooth. It can be difficult to understand, offer resistance, and make us hesitate. The difficulty needs to lie in the difficulty of the experience itself, while form must make that difficulty approachable. Between disordered material and a difficult work lies a choice that cannot be skipped.

## A Stranger Sits Before the Screen

Once the film is finished, viewers occupy a different position from those who took part in rehearsals. A later viewer cannot alter the rules of the time, interrupt at that moment, or change what has already happened. They can pause, watch again, compare what comes before and after, or reject the interpretation the work wants them to accept.

This distance has capacities of its own. A reaction that had to be made quickly at the event can be reconsidered in the film; words drowned out by a louder voice may gain weight when seen again. A viewer’s understanding need not reproduce any participant’s experience. If a film can only stand by saying “you had to be there,” it still has not given enough to strangers.

Rancière’s critique of the idea of the passive spectator helps me preserve this capacity for viewing. Sitting and watching does not mean doing nothing; understanding also involves selecting, connecting, and translating.[6-4] Film cannot prescribe where these connections must end for the viewer. Creators can offer relations, but cannot reclaim every possible understanding in advance.

This does not mean that form can settle for vagueness. Precisely because viewers have judgments of their own, a film should give them something they can judge. A participant’s statement needs enough surrounding material for its circumstances to be discerned. An abrupt transition needs traces that invite inquiry. A duration that refuses to end needs gradually to reveal the differences it retains.

I want strangers to be able to enter P4 without first having to agree with it. A work should make its choices available, giving both engagement and objection something concrete to address. The camera separates those who were present from those who come later, and also leaves them a form through which to meet again.

## Sources

[6-1] Shoot Self with You, fixed 4K master, frame at 02:32:25.000. The objects discernible in the foreground are lights and light stands; the general term “camera equipment” used in older materials is not retained. For the image source and sampling time, see the list of plates at the end of the book.

[6-2] Two frames from the fixed master, at 01:02:20.000 and 01:02:20.500; dialogue reference: V10 human-confirmed Chinese dialogue and speaker-role ledger, May 6, 2026. This passage discusses only the change in images shown by adjacent samples. It does not establish a word-level edit point or whether the embrace was completed.

[6-3] Jacques Rivette, excerpt from an interview in Cahiers du cinéma, no. 204, 1968; see [MoMA’s programme notes for L’Amour fou](https://www.moma.org/calendar/events/7329). This chapter draws only on his distinction between events during filming and the organization of the finished film.

[6-4] Jacques Rancière, Le spectateur émancipé, La fabrique, 2008. See the [excerpt from pages 18–29 in the materials for the Strasbourg conference “Penser le spectateur”](https://labo-laps.com/wp-content/uploads/2013/05/colloque-penser-le-spectateur.pdf), on the spectator’s acts of selecting, comparing, and interpreting. This chapter brings that question to the viewing of films.