Chapter 21: Documentary and Nonfiction

"Documentary is the creative treatment of actuality." — John Grierson, commonly credited with coining the term documentary

Overview

Point two documentary makers at the same real subject — the same person, the same workshop, the same true story — and hand them the same footage, and you will get two completely different films. One opens with silence: a wide shot of an empty room at dawn, held long, no words, just the sound of the place waking up. The other opens with a warm, certain voice telling you exactly who this person is and why you should care. Same reality. Opposite experience. One trusts you to watch; the other tells you what to think. Neither is wrong. But they are different choices, and the whole craft of documentary lives in making them on purpose.

Here is the tension that makes nonfiction hard, and worth doing: you cannot script reality, but you still have to deliver a story. The events already happened, or are happening, without your permission and without a plot. Your job is not to invent — inventing would break the contract that makes it a documentary at all — but to find the story that is truly in the material, and shape it so a stranger will watch to the end. That is the razor you walk in this chapter: shaping truth into story without falsifying it.

We have been building toward this. In Chapter 17 you learned that even a testimonial is a story with an arc; in Chapter 19 you learned to run an interview that produces full-sentence gold; in Chapter 20 you learned to shoot the B-roll that lets an interview cut. This chapter takes those three skills and points them at the oldest and most honest genre in the medium: the documentary. It is also the genre of your Project 2, the three-minute documentary short you have been building since Chapter 16. By the end of this chapter you will know the modes you can work in, how to build a story from what is real, how to write and record narration, how to bring old photographs and footage to life, where the ethical line runs through the edit, and how to structure a short doc that lands.

In this chapter you will learn to:

• Name the working modes of documentary — from fly-on-the-wall vérité to the personal essay film — and pick the one your story and your access can support.
• Find and build a reality arc: a beginning, middle, and end you discover in true material rather than one you make up.
• Write narration that adds instead of describing, and record clean voiceover in a closet with a phone.
• Use archival stills and footage honestly, motivate their movement, and know what rights you must clear.
• Apply a practical ethics test to every cut so you shape the truth instead of manufacturing it.

Learning Paths

Every reader should read §21.2 (building the story) and §21.5 (ethics) — they are the spine of the genre. Beyond that, weight your attention:

• 📱 Phone-first: §21.3 (recording narration with nothing) and §21.6 (the short-doc structure) are your fastest path to a finished piece. Your phone plus a closet is a real voiceover booth.
• 🎥 Creator: §21.1 (modes) and §21.4 (archival) expand your toolkit past the talking-to-camera video — vérité observation and stills-in-motion are the two moves most creators never learn.
• 💼 Pro-track: §21.5 (ethics, consent, and releases) is where reputations are made or destroyed. Read it as risk management as well as morality.
• 🎓 Student: §21.1 gives you the vocabulary (modes, vérité, expository) that documentary courses assume, and the case studies model the shot-by-shot analysis you will be asked to write.

21.1 Modes of documentary: vérité to essay

Start with a plain definition, because the word gets used loosely. A documentary is a nonfiction film that shapes real events, people, and places into a story — it presents reality rather than inventing it, but it selects, orders, and frames that reality on purpose. Both halves matter. "Nonfiction" is the promise: the people are real, the events happened, you did not write them a script. "Shapes into a story" is the craft: you are not dumping raw reality on the viewer, you are choosing and arranging it so it means something. A security camera records reality and is not a documentary. A documentary is what happens when a maker brings intention to actuality — Grierson's old phrase, "the creative treatment of actuality," is still the best five-word summary of the whole genre, and the tension in it (creative and actual) is the tension you will manage for the rest of your career.

Within that promise, there is a wide range of how. Documentary is not one thing; it is a palette of modes — distinct relationships between the camera, the maker, and the subject. Documentary scholars, most influentially Bill Nichols, have mapped these modes formally (Nichols names six); for a working shooter, four cover almost everything you will make, and most real films mix them.

The first mode is the one people mean when they say a film feels "raw" or "real": vérité. The term is shorthand for a family of closely related approaches — cinéma vérité and direct cinema — that share one commitment: the camera is a witness, not a director. In vérité you observe real life as it unfolds and try to influence it as little as possible. No script, often no interviews, usually no narration — you get close, you keep rolling, and you let reality speak. Its power is access and continuity: because nothing is staged and the moment is unbroken, the viewer feels present in a way no other mode achieves. Its cost is control: you cannot make people do anything, you shoot enormous ratios of footage to find the gold, and you may wait days for a moment that never comes. Vérité was made possible by a specific piece of history — around 1960, cameras and synchronized sound recorders became portable enough to walk into a crowd — and the films that resulted (we analyze one below, and again in this chapter's "Watch This") redefined what nonfiction could feel like.

The second mode is participatory. Here the maker is openly part of the film — asking questions on camera, appearing in the frame, driving the story as an investigator or a guide. The interview (your Chapter 19 skill) is a participatory device: the film admits that a conversation is happening. First-person investigations, where the filmmaker narrates their own search and turns up in scenes, are the most visible form. Participatory documentary is honest about the maker's presence, which is both its strength (no pretense that the camera is invisible) and its risk (the film can become about the maker instead of the subject).

The third mode is expository — the classic "voice-of-God" documentary. A guiding narration lays out an argument or a history, and the picture illustrates it: B-roll and archival run under a voice that tells you what you are seeing and what it means. Most nature documentaries, most historical series, most explainer videos, and most corporate and journalistic pieces are expository. It is the clearest mode, the easiest to follow, and the easiest to do badly, because the narration can smother the images and lecture the viewer. Used well, it is efficient and authoritative; used lazily, it tells you what you can already see.

The fourth mode is the essay film (or poetic documentary). Here the maker's own eye and voice organize the film by association and idea rather than by information — it feels like a mind thinking out loud, or a mood built from images. Essay films are personal, subjective, and unafraid to say "I." They are the least commercial mode and often the most artful; a first-person essayist can turn a walk through a city or a box of old letters into a meditation on memory. You are unlikely to be hired to make an essay film, but the essayistic instinct — letting your point of view shape the material — will improve everything you make.

FIGURE 21.1 — Four working modes of documentary (a palette, not boxes)

  MODE              THE CAMERA'S ROLE          NARRATION?        FEELS LIKE             TYPICAL SUBJECT
  ───────────────   ────────────────────────   ─────────────    ───────────────────    ─────────────────
  OBSERVATIONAL /   a fly on the wall;         none, usually     "you are there,        real events,
  VÉRITÉ            witness, never steer       (life speaks)     unmediated"            profiles, process
  PARTICIPATORY     present and involved;      sometimes; the    "a conversation        investigations,
                    the maker asks, appears    maker is in it    or an inquiry"         first-person quests
  EXPOSITORY        illustrates an argument;   yes — a guiding   "being told a clear    history, nature,
                    B-roll under a voice       voice             case or story"         explainers, news
  ESSAY / POETIC    the maker's eye + voice;   yes — personal,   "a mind thinking       memory, place,
                    association over info      subjective        out loud"              ideas, mood

Read the figure as a spectrum, not a filing cabinet. Almost every real documentary mixes modes: a mostly observational profile drops in three interviews (participatory) and one line of narration to bridge a time jump (expository). The Civil War, Ken Burns's landmark series, is expository at its spine but essayistic in its music and pacing. Grizzly Man, which we take apart in Case Study 1, is an essay film built on found observational footage. Your job is not to pick a box and obey it — it is to know what each mode does so you can reach for the right one, beat by beat, in service of the story.

To feel what vérité's promise actually delivers, add a landmark to your "Watch This" shelf. Primary (1960), produced by Robert Drew and his team, is widely credited as a founding film of American direct cinema — made possible precisely because, around 1960, cameras and synchronized sound recorders finally became portable enough to walk into a crowd. Its most-cited moment is a long, unbroken handheld shot that follows a candidate through a dense, cheering hall — a shot commonly credited to Albert Maysles.

🎞️ Read This Sequence. Read this slowly, and picture what the camera's freedom to walk did to nonfiction. This is the observational mode's whole argument in one shot.

FIGURE 21.2 — "The follow" (a vérité moment)     [after Primary, 1960]
  THE FRAME    A medium-wide, handheld from behind and beside a candidate as he moves through a dense,
               cheering crowd toward a hall. Bodies, hands, hats crowd the edges; the frame breathes and
               bobs with the operator's steps.
  THE MOVE     Handheld, walking — the camera is a body in the crowd, not a tripod outside it. One
               near-continuous take follows the man from the street, through the doors, up to the stage.
  THE LIGHT    Available light only — harsh entryway bulbs, the flat light of a public hall. Nobody lit
               this; the mode forbids it. The imperfection reads as truth.
  THE SOUND    Live, synchronized location sound — the roar, the greetings, footsteps — captured by a
               then-new portable rig that freed the camera to walk. Sound and picture, unstaged, together.
  THE CUT      Long and unbroken where a scripted film would cut; the length IS the statement — we are
               with him in real time, denied the safety of an edit.
  THE EFFECT   We feel we are inside the event, not shown a report of it. The wobble, the crowd, the
               unbroken walk put us bodily in the moment.
  THE LESSON   Vérité's power is access and continuity: get close, keep rolling, and let an unbroken real
               moment do what no narration could. The camera stops explaining and starts witnessing.

Notice what the shot refuses. There is no narrator to tell you what it means, no interview to explain it, no cut to a safer angle — the filmmakers bet everything on the raw, unbroken real moment, and won. That bet is only possible because the gear finally allowed it; the technology and the mode were born together. When you shoot your own vérité (there's an assignment below), you are working in the tradition this shot opened: your job is to be present, unobtrusive, and patient enough to catch the moment reality hands you.

How do you choose? Two questions decide it. First, what does the story need — the intimacy of observation, the clarity of a guiding voice, the honesty of the maker's presence, or the reflection of an essay? Second, what access do you actually have? Vérité demands patience and a subject who will let you live in their space for a long time; if you have one afternoon, pure observation will starve, and you need interviews and narration to carry the story. Mode is where ambition meets reality. Choose the mode your access can support, or you will spend the edit trying to fake one you never shot.

💡 Why It Works: the mode sets the contract with the viewer. Each mode makes a different promise. Vérité promises "we are showing you what happened, unmediated." Expository promises "we will guide you honestly through this." When you honor the promise, the film feels trustworthy; when you break it — a "vérité" moment that was obviously staged, a narrator who turns out to be spinning — the viewer's trust collapses, and in nonfiction, trust is the product. Pick a mode and keep its promise.

🎬 On Set: shoot five minutes of pure vérité. Choose a real activity you can observe without controlling it — a kitchen during a dinner rush, a workshop, a market stall, a game being played. Film it as a fly on the wall: get close, keep rolling, do not direct anyone, do not ask a single question. Shoot a wide, then move for mediums and details as the action gives them to you (your Chapter 20 sequence instinct). Constraint: no talking to your subjects, minimum one unbroken 60-second take. Self-review: watching it back, where did you feel present in the moment, and where did you feel like you were watching someone perform for a camera? That gap is the vérité maker's whole craft.

🔄 Check Your Eye. 1. In one sentence each, what is the difference between the observational/vérité mode and the expository mode? 2. Which mode admits, on camera, that the filmmaker is present and asking questions? 3. You have exactly one two-hour window with a subject and no chance to return. Which mode is least realistic to build the whole film from, and why?

Check yourself

1. Vérité observes real life unfolding and tries not to influence it (no narration, camera as witness); expository uses a guiding narration to lay out an argument or story while the picture illustrates it.
2. Participatory.
3. Pure vérité/observational — fly-on-the-wall filmmaking needs time and patience to catch unstaged real moments; with one short window you will need interviews (participatory) and probably narration (expository) to carry the story, because you cannot wait for reality to deliver an arc on schedule.

21.2 Building a story from what's real

Now the hard part, and the heart of the genre. You have real material — an interview or two, a pile of B-roll, maybe some archival. None of it arrived as a story. Your task is to build one that is true to the material without being invented by you.

The story shape you are after has a name we will use for the rest of the book. The reality arc is the beginning, middle, and end you discover in real material and then build in the edit — not a plot you impose, but a genuine shape you find in what actually happened. It is the documentary cousin of the narrative arc you learned in Chapter 17 (§17.2). The difference is direction: in fiction you write the arc, then shoot to it; in documentary you shoot the reality, then find the arc inside it. Both end in the same place — a hook, a question, a middle with tension, a turn, and a resolution — but the documentary maker earns the shape by listening to the material rather than dictating to it.

FIGURE 21.3 — The reality arc: a story shape found in real material

   HOOK             THE QUESTION       DEVELOPMENT         THE TURN            RESOLUTION
   (a real moment   (what's at stake?  (what we learn;     (a change, a        (where the question
    that grabs)      why watch?)        the middle)         reveal; it gets     lands — earned,
                                                            real)               not always happy)
   ─────────►        ─────────►         ─────────►          ─────────►          ─────────►
   "show me           the spine          interview +          the moment          the button:
    something"         question           B-roll carry         everything           a resonant image
                                          the middle           pivots               or last line

   You do NOT invent this arc. You FIND it in the footage, then build it in the edit.
   The shape is real; the ordering is your craft.

Where does the arc come from? From three real ingredients you already know how to capture:

A character with something at stake. Documentaries are about people, even when they are about places or ideas — because viewers attach to faces and voices, not topics. Your subject needs a want (to keep a dying craft alive, to find an answer, to be understood) and something in the way of it. That tension is your engine. When you interview (Chapter 19), you are not collecting facts; you are hunting for the want, the obstacle, and the change. The single most useful documentary interview question is a version of "what were you afraid of?" or "what did you hope would happen?" — because the answer is the arc.

A question the film will answer. Every good documentary poses a question early and pays it off late — not always the question you started with. It can be small ("will this eighty-year-old cobbler find anyone to pass his shop to?") or large ("what do we owe the places we come from?"). The question is the spine that holds the middle together; every beat should feel like it is moving toward the answer. Without a question, a profile is just a list of nice things about a nice person, and viewers drift.

A turn — a moment where something changes. The dullest documentaries are flat: the subject is exactly the same at the end as the start. The best ones have a turn, a reveal, a moment where the story deepens or pivots. You cannot script the turn, but you can recognize it in your footage — the sentence where the subject's voice cracks, the beat where the plan fails, the fact that recontextualizes everything before it. Finding the turn is the "aha" of the paper edit you will do in Chapter 30 (§30.1). When you find it, the whole arc snaps into focus, because now you know what the middle was building toward.

Here is the professional habit, and it is the opposite of the beginner's instinct. Beginners shoot and then hope a story appears. Professionals shoot for an arc they can already half-see, and stay alert for the better one that reality hands them. Before you shoot Project 2, you should be able to say, in one sentence, what your reality arc probably is ("a craftsman confronts the end of his trade — and decides what to do about it"). You will almost certainly be wrong in some way, and the real story you find in the edit will be better. But shooting toward a hypothesis is how you get coverage that can become a story, rather than a hard drive of pretty shots that can't.

🚪 Threshold Concept: you are not lying — you are selecting. The idea that permanently changes how you make (and watch) documentary is this: every documentary is shaped. There is no such thing as unmediated truth on a screen — you chose where to point the camera, what to leave out, which two sentences of a two-hour interview to keep, and what order to put them in. Selection is not the enemy of truth; it is how truth becomes communicable. The ethical question is never "did you shape it?" — you always did — but "did your shaping reveal what is true, or manufacture something that isn't?" Once you hold that distinction, you stop feeling guilty about editing and start feeling responsible for it. That responsibility is §21.5, and it is the price of the genre.

✂️ In the Edit: the story is found, not shot. You will feel this acutely in Chapter 29 (§29.2) and Chapter 30, but plant it now: with documentary, the edit is where the story is discovered, not just assembled. You will lay your interview selects on a timeline, move them around, and watch an arc emerge from an order you did not plan on set. This is why documentary editors are storytellers, not assemblers — and why you must shoot generous coverage (Chapter 20). You cannot find an arc in footage you did not capture. Shoot for the edit; then find the film in the edit.

🔗 Connection. The arc you are building here is the same three-act shape from Chapter 17 (§17.2), the hook is the one from §17.3, and "show, don't tell" (§17.4) is the documentary maker's constant discipline — let the reality reveal the point instead of narrating it. The difference is only that here you find the shape in real material rather than invent it. Everything you learned about structure still applies; you are just applying it backward.

🔄 Check Your Eye. 1. What is the difference between a narrative arc (Chapter 17) and a reality arc? 2. Name the three real ingredients that give a documentary its arc. 3. A student's profile of a local musician is "just a list of nice things about her." What is almost certainly missing?

Check yourself

1. A narrative arc is written first and shot to; a reality arc is discovered in real footage and built in the edit. Same shape (hook, question, tension, turn, resolution), opposite direction.
2. A character with something at stake (a want and an obstacle), a question the film will answer, and a turn where something changes.
3. A question and a turn — there is nothing at stake and nothing changes, so the piece is flat. It needs a spine question ("will she…?" / "why does she…?") and a moment where the story deepens or pivots.

21.3 Narration and voiceover: writing and recording

Sooner or later most documentaries need a voice — to bridge a gap, set context, or carry an argument the pictures cannot. That voice is voiceover (or narration): recorded spoken commentary laid over the picture, not spoken on camera. It comes in three flavors, and they feel completely different. A scripted narrator (the classic expository "voice of God") is authoritative and external. The filmmaker's own voice is personal and participatory — you, thinking aloud. And interview audio used as voiceover — a subject's own words, lifted off their sync picture and laid over B-roll — is often the most powerful of the three, because it lets the person tell their own story while you show it. Most short docs use the third kind heavily and reach for scripted narration only where a real voice cannot cover a gap.

Narration has two crafts: writing it and recording it. Get either wrong and the whole piece feels amateur.

Writing narration: to picture, and lean. The cardinal rule is that narration adds — it never describes what the viewer can already see. "He picks up a hammer" over a shot of a man picking up a hammer is not narration; it is subtitling your own footage, and it insults the viewer. Good narration fills the gaps between pictures and soundbites: it supplies the fact the image can't ("the three other shops on this street have closed"), the emotional frame ("he has done this every morning for thirty-one years"), or the bridge across a jump in time or place. Then it gets out of the way. You write narration to an edited (or planned) picture, not the other way around — lay your images and soundbites down first, find the gaps, and write only what the gaps need.

FIGURE 21.4 — Writing narration to picture (a two-column documentary script)

  PICTURE (what we SEE)                       NARRATION (what we ADD, never what we see)
  ─────────────────────────────────────────   ─────────────────────────────────────────────
  WIDE: empty workshop at dawn, lights on      [silence — let the room breathe ~3s]
  MEDIUM: hands opening a worn toolbox          "He has opened this box every morning
                                                 for thirty-one years."
  DETAIL: a row of shoe lasts, a bench          [no narration — the tools speak for themselves]
  INTERVIEW (sync): "...nobody wants to         [narration OUT — never step on a soundbite
   learn this anymore."                          that carries its own weight]
  B-ROLL: the shuttered shop next door          "Three doors down, the tailor closed in March."
  DETAIL: a hand-written "for sale" sign         [silence — the image is the sentence]

   RULE: narration is connective tissue. It supplies what the picture cannot, then stops.
   If a line describes what we already see, or talks over a soundbite, cut the line.

Notice how little narration the column actually contains — four short lines across a whole scene. That sparseness is the mark of a professional. Write your first narration draft, then cut it by a third; almost every line you delete makes the film better, because it hands the moment back to the pictures and the real voices. Write in your speaking voice, not your writing voice: short sentences, plain words, one idea per line. Read every line aloud as you write — if it is hard to say, it is wrong. And leave air: write to land a line, then let three seconds of picture and sound breathe before the next.

Recording narration: clean, close, and consistent. Narration lives or dies on audio quality, and here your Chapter 14–15 skills come home. The enemy is room echo. The professional solution is comically low-tech: record in a small space full of soft things — a closet full of hanging clothes is a genuinely excellent voiceover booth, because the clothes kill reflections. Get the mic close (a hand's width, off-axis so plosives don't pop), keep your mouth a constant distance so the level doesn't wander, and record a few seconds of silence for room tone (Chapter 15) so the editor can patch breaths. Record more than you need and read each line two or three ways. Consistency matters more than perfection: a whole narration recorded in one session, one room, one mic distance, will cut together seamlessly, while lines gathered across three days in three rooms will never match.

⚙️ Settings Box: recording voiceover (phone or camera).

Setting	Value	Why
Space	Small room, soft furnishings — a clothes closet is ideal	Soft surfaces kill the echo that screams "amateur"
Mic	Any close mic: a lav, a shotgun, even wired earbuds' mic in a pinch	Close beats expensive; proximity is everything
Distance	A hand's width from the mouth, slightly off to the side	Close for warmth; off-axis so "p" and "b" don't pop
Level	Peaks around -12 to -6 dBFS, never touching 0	Headroom; a clipped voice track cannot be fixed
Handling	Mic on a stand or braced, not hand-held	Hand noise transfers straight into the recording
Extras	Record 10s of room tone; do 2–3 takes of each line	Gives the edit patch material and choices

You do not need a studio. A phone, a closet, and a pop of thought about distance and level will out-record a bad setup with an expensive mic every time.

🎒 Gear Note: the closet booth. The single best voiceover investment is zero dollars: stand in a walk-in closet or in front of a rack of coats, hold your phone or clip a lav a hand's width from your mouth, and read. The hanging fabric absorbs the reflections that make untreated rooms sound hollow. A hung moving-blanket, a duvet draped over a chair to make a fort, even reading under a heavy blanket — all work. When you outgrow the closet, a small amount of foam or a portable reflection filter is the next step, not a room rebuild. Proximity and soft surfaces buy you 90% of a professional VO; gear buys the last 10%.

♿ Accessibility & Inclusion: narration is not audio description — and both matter. Do not confuse your documentary's narration (a storytelling voice you author) with audio description (a separate, added track that describes the on-screen visuals for blind and low-vision viewers, spoken in the gaps between dialogue). A well-narrated film still needs audio description for full access, because your narration deliberately does not describe what's on screen. Two more habits: caption your narration and every soundbite (Chapter 36, §36.5) so deaf and hard-of-hearing viewers get the whole film, and when you cast a narrator, remember that whose voice tells the story is itself a representation choice — a story about a community is often better narrated by someone from it than by a generic "authoritative" voice from outside it.

🎬 On Set: record a narration option for Project 2, then A/B it. Take one 30–45-second stretch of your documentary and write a lean narration for it (to picture, following FIGURE 21.4). Record it in a closet per the Settings Box. Then build two versions of that stretch: one with your narration, one with only sync sound, soundbites, and B-roll. Constraint: change nothing but the presence of narration. Self-review: which version is stronger, and why? You will often find the narrated version is clearer but the silent one is more moving — and knowing which your story needs is exactly the judgment this chapter is teaching.

🔄 Check Your Eye. 1. What is the one thing narration should never do? 2. Why is a clothes closet a good place to record voiceover? 3. Name the three sources a documentary's "voiceover" can come from.

Check yourself

1. Describe what the viewer can already see — narration adds what the picture can't, then gets out of the way.
2. The hanging clothes absorb sound reflections, killing the room echo that makes untreated spaces sound hollow and amateur.
3. A scripted external narrator (voice of God), the filmmaker's own voice, and an interviewee's own audio lifted off its sync picture and used over B-roll.

21.4 Archival and stills: the Ken Burns problem

Some stories reach back before your camera existed. To tell them you need archival: pre-existing material you did not shoot — old photographs, film and news footage, documents, letters, home movies, recordings — used as evidence and texture. Archival is how a documentary proves a past it cannot film. A single real 1937 photograph of a family on a porch does something no re-enactment can: it says this actually happened, to these actual people. That authenticity is archival's superpower, and it comes with two problems every documentary maker must solve — a creative one and a legal one.

The creative problem has a name, because one filmmaker solved it so famously that editing software now ships a button named after him. The Ken Burns effect is the slow, motivated push-in or drift across a still photograph that gives a flat image time and motion — a technique Burns used so distinctively across The Civil War (1990) and his later series that it became synonymous with his name. The "Ken Burns problem," for you, is threefold. First, it is overused: applied mechanically to every photo, a slow zoom becomes a tic that says "the editor didn't know what else to do." Second, the move must be motivated (your Chapter 8 word): move toward the emotional point of the photograph — a face, a pair of hands, a detail the story is about — or reveal something by drifting to it, or follow an eyeline within the frame. A random zoom is worse than a static hold. Third, it is technical: you can only move within pixels you actually have, so archival stills must be scanned or shot at high resolution, well above your delivery resolution, or the push-in turns to mush.

FIGURE 21.5 — "Bringing a still to life"     [after the technique associated with Ken Burns]
  THE FRAME    A single old black-and-white photograph: a family of five on a porch, dated 1937. We open
               framed tight on the youngest child's face, low and left in the print.
  THE MOVE     A slow push-in that drifts up and to the right — from the child's face toward the father's,
               ending on his weathered hands. The move is motivated: it follows the story's thought, from
               the child who will outlive them all toward the man the film is really about.
  THE LIGHT    Baked into the photograph — nothing to light. (But scan it BIG: you can only move within
               the pixels you have, so the detail must be there to begin with.)
  THE SOUND    A single sustained musical note underneath; over the move, an interviewee's voice, off-sync:
               "My grandfather built that porch with his own hands." The still is not still.
  THE CUT      In from a wider shot of the same photo lying on a table; out to a present-day shot of the
               same porch, now weathered and empty — a match on place across nearly seventy years.
  THE EFFECT   A dead, flat image becomes a moment you move *through*. The motion gives the eye a path and
               the photograph a heartbeat; the voice gives it a reason to exist on screen.
  THE LESSON   A still photograph is a shot you haven't finished directing. Move for a reason — toward the
               emotional point — or hold it dead still. Never zoom just to fill time.

The legal problem is bigger and catches beginners hard: you usually do not own archival, and using it without clearing the rights can sink a project. Old does not mean free. A photograph, a piece of film, a song, a news clip — each has a rights holder (a family, an archive, a studio, a stock house), and putting it in a film you publish or sell generally requires permission and often a fee. There are real exceptions and paths — genuinely public-domain material, footage licensed from stock and archive libraries, family photos you have permission to use, and the narrow, fact-specific doctrine of fair use — but "I found it online" is never a clearance. We treat licensing and copyright properly in Chapter 38 (§38.5); for now, build the habit: before you fall in love with a piece of archival, ask who owns it and whether you can afford to use it. Designing a film around footage you can't clear is a heartbreak you can avoid in pre-production.

Archival also carries an ethical charge, which is why it points straight into the next section. Old footage and photographs can lie as easily as they can prove — a crowd shot from one event laid under a story about another, a photo implied to show something it doesn't, a document quoted out of the context that changes its meaning. The rule is simple and strict: archival must represent what it actually is. If a photo is a re-creation, or is of a different person, or comes from another time, and you let the viewer believe otherwise, you have crossed from documentary into deception — even though every frame is "real."

🔬 The Tech: how much resolution a still move needs (skippable). If you want to push in on a photo and end at, say, 50% of its width while delivering at 1080p (1920 pixels wide), your scan needs to be at least double that at the tightest point — roughly 4K wide (about 3840 px) or more — so the end of the move is still full-resolution. For a bigger move or a 4K delivery, scan larger still. Practical rule: scan or photograph archival stills at the highest resolution you reasonably can (flatbed scanners hit 600+ dpi; a phone photo of a print in good light works in a pinch), because you can always throw pixels away in a push-in but you can never invent them. This box is optional — the craft point ("scan big") is all you truly need.

⚠️ Common Mistake: the Ken Burns effect on autopilot. The beginner drops the same slow zoom on every single photo, in the same direction, at the same speed, because the software makes it one click. The result is seasick and meaningless. Fixes: (1) vary it — some photos push in, some pull out, some drift, some are held dead still; (2) motivate every move toward a specific point in the frame the story cares about; (3) match the move's speed to the music and the mood; (4) let at least some stills sit motionless — stillness, used deliberately, has more weight than motion used reflexively.

🔄 Check Your Eye. 1. What does "motivated" mean when applied to moving across a still photograph? 2. Why must archival stills be scanned at high resolution? 3. A filmmaker finds a perfect 1950s news clip online and drops it into their doc. Name the two problems they may have just walked into.

Check yourself

1. The camera move goes toward something the story cares about — a face, hands, a detail, an eyeline, a reveal — rather than zooming at random; the motion has a reason you could say out loud.
2. Because you can only move within the pixels you have; a push-in on a low-resolution scan quickly turns soft and blocky. Scan well above delivery resolution.
3. A legal problem (they almost certainly do not own the rights and "found online" is not clearance — see Chapter 38) and an ethical one (using footage of one event/context to illustrate another misleads the viewer, even though the clip is "real").

21.5 Ethics of nonfiction: honesty in the edit

Everything else in this chapter is technique. This section is the reason the technique matters, and it is not optional. Documentary makes an implicit contract with the viewer — this is real, and I am representing it fairly — and every editing decision either honors that contract or breaks it. The genre's power comes entirely from the viewer's trust that what they are seeing is true. Spend that trust dishonestly and you do not just make a bad film; you damage the credibility of the form itself, and you can hurt real people who trusted you with their story.

The core distinction is the one from §21.2: selection is honest; fabrication is not. You will always cut, compress, and order — that is filmmaking, not lying. The line is crossed when your shaping makes the viewer believe something that did not happen or that misrepresents what did. Learn to see the line in specific, common editing moves.

FIGURE 21.6 — The line: shaping vs fabricating (honesty in the edit)

  HONEST — shaping the truth                   DECEPTIVE — manufacturing a "truth"
  ──────────────────────────────────────────   ──────────────────────────────────────────
  cut a rambling answer down to its clear       splice words from separate answers to build
  core, keeping the speaker's meaning           a sentence the person never actually said
                                                (a "Frankenbite")

  compress a three-hour event to its three      reorder events so one appears to cause
  real beats                                    another when it didn't

  use a cutaway that was genuinely part         insert a reaction shot to fake a reaction
  of the scene                                  that never happened at that moment

  a re-enactment clearly SIGNALED as one        a staged scene passed off as captured reality

  music that supports the feeling that was      music that manufactures a feeling the
  really in the moment                          moment did not have

   THE TEST: "Would my subject, watching the final cut, feel accurately represented —
             even if they aren't flattered? And am I revealing what's true, or inventing it?"

Walk through the traps, because they are seductive and easy:

The Frankenbite. Borrowed from reality-television slang, a Frankenbite is a soundbite stitched together from words the person said at different times, to make them appear to say something they never actually said as a sentence. Its milder cousins are everywhere and just as dangerous: cutting the qualifier off a statement so a hedged opinion sounds absolute, or removing the question so an answer means something different. Trimming an answer for clarity is fine; assembling a sentence they never spoke is not. The test is whether the edited line still means what the person meant.

The lying cutaway. In Chapter 20 you learned to shoot cutaways and reaction shots to cover edits. Ethics govern how you use them. Inserting a shot of someone nodding thoughtfully — shot at a different time — to imply they were reacting to a specific statement is a fabrication, even though both shots are real. The classic abuse is the interview "noddy" (a reverse shot of the interviewer nodding, filmed after the subject left) used to manufacture a reaction or hide a cut in a way that changes meaning. Use cutaways to smooth time and show what's described; never to invent a reaction that didn't occur.

Deceptive compression and reordering. Compressing time is not just allowed, it is required — no one wants the three-hour event in real time. But reordering events so that B appears to follow (or be caused by) A when it didn't is deception. Compress honestly: keep the true sequence and the true relationships, cut the dead space.

Re-enactment and reconstruction. Staging a version of a past event is a legitimate, long-standing documentary tool — Errol Morris built The Thin Blue Line (1988) on stylized reconstructions, and that film helped overturn a wrongful murder conviction. The ethics turn entirely on signaling: a re-enactment that is clearly marked as one (by style, by context, by a title) is honest; the same footage passed off as real archival is a lie. When in doubt, make the reconstruction look like a reconstruction.

Music and framing. Music is the most powerful and least visible manipulator you have. Scoring a neutral moment as sinister, or a sympathetic subject as menacing, can reverse a viewer's judgment without a single false frame. This is where "honest" gets subtle: music that amplifies a feeling truly present in the moment is craft; music that installs a feeling that wasn't there is manipulation. The same goes for framing a subject — a low angle to make them loom, an unflattering frame held too long. Ask whether your choices reveal the person or editorialize against them.

Beyond the edit, ethics start with the human in front of the lens. Your subject is a real person who let you into their life, often with far less power than you — you control the camera, the edit, and the final story; they control almost nothing once they say yes. That asymmetry creates a duty of care. Get informed consent: make sure your subject genuinely understands what the film is, where it will go, and how they might be portrayed — not just that they signed a form. Get releases (the legal permission to use their image and words; Chapter 38, §38.4), but treat the release as the floor of your obligation, not the ceiling. Be especially careful with vulnerable subjects — children, people in crisis, people who don't grasp how public "public" now is. And hold to a simple standard: your subject should be able to watch the finished film and feel seen, even if it isn't flattering — recognized, not betrayed.

This is exactly the terrain of Case Study 1. When Werner Herzog made Grizzly Man from a dead man's own footage, he faced a decision no technique could answer: he had audio of the subject's death, and he chose not to play it, and to tell the person holding the tape to destroy it. Whatever you think of the choice, it models the point — the most important documentary decisions are ethical, not technical, and they are made in the edit room, with a real person's life on the timeline.

⚠️ Common Mistake: "it's all technically true, so it's fine." The most dangerous rationalization in nonfiction is that every individual frame is real, therefore the film is honest. But meaning lives in arrangement, and you can build a lie entirely out of true pieces — a Frankenbite is real words, a lying cutaway is a real shot, a manipulative score is real music. The honesty of a documentary is not the sum of its frames; it is whether the whole leaves the viewer believing something true. Judge your edit at the level of the impression it creates, not the frames it's made of.

♿ Accessibility & Inclusion: consent, representation, and who gets to speak. Ethics and inclusion are the same craft here. Whose story is it, and who is telling it? A documentary about a community is stronger and fairer when people from that community shape it, not just appear in it. Give subjects genuine informed consent about how they'll be shown. Represent people as whole and specific, not as symbols or problems. And finish the job for your whole audience: captions and audio description are part of telling the truth to everyone you claim to be telling it to.

🔄 Check Your Eye. 1. What is a "Frankenbite," and why is it dishonest even though every word is real? 2. When is a re-enactment ethical, and when is it deceptive? 3. Give the one-sentence test for whether your edit represents a subject honestly.

Check yourself

1. A soundbite stitched from words a person said at different times to make them appear to say a sentence they never actually said. It is dishonest because meaning lives in arrangement — you can build a false statement out of true fragments.
2. Ethical when it is clearly signaled as a re-enactment (by style, context, or a title); deceptive when a staged scene is passed off as captured reality or real archival.
3. "Would my subject, watching the final cut, feel accurately represented — even if not flattered — and am I revealing what's true rather than manufacturing it?"

21.6 The short-doc structure

Now assemble everything into the thing you are actually making: a short documentary. Short docs — roughly one to five minutes — are the workhorse of modern nonfiction: the branded mini-doc, the profile, the explainer, the festival short. Your Project 2 is one. They live or die on structure, because a short has no time to waste and a viewer who will leave in seconds.

A short doc is not a small feature; it is its own shape. It has room for one story, one question, and a handful of beats — so every beat must earn its place. The reliable structure is the reality arc from §21.2, mapped onto a tight timeline.

FIGURE 21.7 — A 3-minute short-doc beat template (adapt, don't obey)

  TIME    BEAT             WHAT IT DOES                                              APPROX
  ─────   ──────────────   ───────────────────────────────────────────────────────  ──────
  0:00    COLD OPEN        a vérité moment or a striking soundbite — no context yet   15-25s
  0:20    WHO / WHERE      establish subject and place (a lower third, or one VO line) 10s
  0:30    THE QUESTION     what's at stake; the spine question; why watch              20s
  0:50    DEVELOPMENT      the middle: interview spine + B-roll, ONE idea per beat     80s
  2:10    THE TURN         a complication, a reveal, an emotional shift                25s
  2:35    RESOLUTION       where it lands — earned, not necessarily tidy               20s
  2:55    THE BUTTON       one final image or line that resonates                      5-10s

  V1  [ B-roll ][ interview ][ B-roll over the cut ][ archival ][ interview ][ B-roll: button ]
  A1  [ soundbite spine (the subject's own words carry the story) ......................... ]
  A2  [ narration — OPTIONAL, only in the gaps ...        ...                               ]
  A3  [ music bed / ambience ................................................................ ]

Read the beats and the tracks together. The cold open is your hook (Chapter 17, §17.3): start in a real moment or on a gripping line, before you explain anything — curiosity buys you the next twenty seconds. Then quickly orient the viewer (who, where), pose the question, and spend the bulk of your runtime in development, where the discipline is one idea per beat: each stretch of interview-plus-B-roll makes a single point, then hands off to the next. The turn is the moment the story pivots or deepens — protect it; it is why the film exists. The resolution answers the question (earned, not tidy — real endings are often bittersweet). And the button is one last resonant image or line that sends the viewer out feeling something.

Look at the audio tracks, because they encode the chapter's biggest structural choice: the soundbite spine. In most short docs the story is carried by the subject's own words (A1) — you build the spine from interview selects (a skill you'll formalize in Chapter 30, §30.3), lay B-roll over the cuts to hide them and show what's described (§30.4), and add narration (A2) only where the real voices leave a gap. This is the "with or without narration" choice you'll test in your Production Checkpoint. A film carried entirely by its subject's voice feels intimate and trustworthy; a film carried by narration feels guided and clear. Many of the best short docs use narration sparingly or not at all, letting the interview and the images do the work — but a bridge you can't cover with a real voice is exactly what one lean line of narration is for. Under it all, a music bed and ambience (A3) run continuously so the picture can cut freely without the sound jumping (your Chapter 15 room-tone lesson, and Chapter 33's mix).

The last structural truth: B-roll over every cut. An interview-driven short that stays on the talking head is, in the old edit-room joke, a hostage video. The B-roll you shot in Chapter 20 is what lets you compress the interview, hide the jump cuts, and show what the subject describes. When you assemble Project 2 (Chapters 26–30), you will feel how completely the B-roll rescues the edit — and wish you had shot more. You always wish you had shot more.

🎬 On Set: paper-edit your short doc's arc. Before you touch a timeline, write your Project 2 as a beat list on paper, following FIGURE 21.7: name your cold open, your spine question, three development beats (one idea each), your turn, your resolution, and your button. For each beat, note the soundbite that carries it and the B-roll that shows it. Constraint: fit it on one page. Self-review: can a stranger read your beat list and feel the arc? If the turn isn't obvious on paper, it won't be obvious on screen — fix it here, where fixing is free.

🔗 Connection. You will execute this structure across Part VI: assembling the pieces (Chapter 26), finding the story in the edit (Chapter 29, §29.2), and building the interview-driven narrative — spine, selects, B-roll over the cuts — in Chapter 30, where Project 2 locks. This chapter is the blueprint; those chapters are the build. Shoot Project 2's material now knowing the shape it must become.

🔄 Check Your Eye. 1. What is a "cold open," and why start there instead of with context? 2. What does "the soundbite spine" mean, and what is narration's job relative to it? 3. Why is B-roll described as what rescues an interview-driven edit?

Check yourself

1. Opening on a real moment or a gripping line before any explanation; it hooks curiosity and buys you the viewer's attention for the setup that follows.
2. The story is carried mainly by the subject's own interview words (the spine); narration only fills the gaps the real voices can't cover — it is optional connective tissue, not the driver.
3. It lets you compress the interview and hide the jump cuts (cutting on the talking head alone creates visible jumps), and it shows what the subject describes — without it, an interview-driven piece is a static "hostage video."

Production Checkpoint

Your Project 2 documentary short already has its interview (Chapter 19) and its B-roll (Chapter 20). This chapter adds the documentary-specific layer — the voice and the evidence — and tests a real creative choice.

Your task: record a narration option and gather your archival, then test what narration does to the piece.

1. Write and record a narration option. Pick the stretch of your short where the real voices leave a gap — the setup, a time bridge, or a bit of context the interview doesn't cover. Write lean narration to picture (FIGURE 21.4): only what the pictures and soundbites can't say. Record it clean, in a closet, per the Settings Box in §21.3.
2. Gather any archival or stills you need. If your story reaches beyond what you filmed — a subject's past, a place's history — collect the photographs, documents, or footage that show it. For each piece, note who owns it and whether you can use it (Chapter 38 handles the clearing; the habit starts now). Scan or shoot stills big so you can move within them (§21.4).
3. Test how narration changes the piece. Build two quick versions of one section: one with your narration, one carried only by the subject's own voice and B-roll. Watch both. Decide which your story needs — and be able to say why.

Why this matters: this is the choice that defines your film's voice. Narration makes a piece clearer and more guided; its absence makes a piece more intimate and more trusting of the viewer. There is no universally right answer — only the right answer for your story. Making that choice deliberately, and hearing the difference, is the moment your Project 2 stops being footage and starts being a documentary. You will build the whole thing in Part VI; you are choosing its voice now.

Summary

Reference-grade recap. Screenshot this before you cut Project 2.

The genre in one line: documentary shapes real material into a story without falsifying it. You always select; you never fabricate.

The four working modes:

Most real films mix modes. Choose the mode your access can actually support.

The reality arc: hook → question → development → turn → resolution → button. You find it in the footage, then build it in the edit — you don't invent it.

Narration — a writing checklist:

Archival — the Ken Burns rules: move only when motivated (toward the emotional point); vary the moves and let some stills sit still; scan big (well above delivery res); and clear the rights before you fall in love with a clip.

The ethics checklist (honesty in the edit):

Consent is the floor, care is the job: informed consent, real releases (Chapter 38), extra caution with vulnerable subjects, and captions plus audio description so the truth reaches everyone.

The short-doc structure: cold-open hook → orient → question → development (one idea per beat) → turn → resolution → button, carried by a soundbite spine with B-roll over every cut and narration only in the gaps.

Spaced Review

Bring back three earlier skills this chapter just leaned on hard.

1. (Chapter 17) What are the parts of a story arc, and what is "show, don't tell"? How does a reality arc differ from the arc you'd write for fiction?
2. (Chapter 19) In a documentary interview, why do you want answers in full sentences, and what makes a question likely to produce one? (Hint: you'll be lifting these as your soundbite spine.)
3. (Chapter 20) What is B-roll, and why is an interview-driven edit nearly impossible without it? Name two kinds of B-roll shot that will save your Project 2 edit.

What's Next

You now know how to shape truth into a story honestly — the most foundational genre in the medium, and the home of your Project 2. Next we cross from the documentary's loyalty to truth into commercial video's loyalty to a client's goal. Chapter 22 takes on corporate and commercial work: the brief, the brand guidelines, the testimonial, and the product video — where the craft you've built meets a paying objective and a call to action. It is also where Project 3, your five-minute branded piece, begins. The skills are the same; the master you serve is new. We start, as commercial work always does, by learning to read the brief.