How to Organize Sources for a Documentary with a

The Documentary Research Problem

A documentary is built from evidence — interview subjects, archival footage, documents, photographs, data, and published research. The research phase of a documentary can span months or years and generate thousands of individual sources. The editing phase requires being able to find any source, assess its reliability, and connect it to specific claims in the narrative.

Documentary filmmakers who don't organize their research systematically encounter the same problem at edit: they remember that someone said something crucial in an interview three years ago, or that there was a newspaper clipping that established a key date, but they can't find it. They know the source exists; they just can't locate it when it matters.

A knowledge system for documentary research organizes sources not just for storage but for retrieval under deadline — organized by narrative thread, subject, time period, and evidentiary function, with annotations that capture why each source matters and how it fits into the story.

---

Documentary Source Types

Documentary research involves six types of sources, each requiring different organization:

1. Interview subjects: People who will be or have been interviewed on camera. Their biographical information, prior statements, areas of expertise, and relationship to the story.

2. Archival footage: News footage, personal video, institutional archives, documentary film archives, surveillance footage. The question is always: where does it exist, who holds the rights, and how do we access it?

3. Documents: Written records, government documents, corporate records, court filings, internal communications, letters, diaries. These establish facts, dates, and context.

4. Published research and journalism: Academic papers, news coverage, investigations, books. These provide established facts, expert context, and narrative background.

5. Visual materials: Photographs, maps, graphics, statistical visualizations. Non-moving visual evidence used in the film.

6. Expert sources: People who provide context and analysis but may not be primary interview subjects. Scientists, historians, analysts.

---

Setting Up the Knowledge System

Collections structure

Create a primary Collection: "Documentary: [Film Title]"

Sub-Collections organized by narrative thread and source type:

By narrative thread (customize to your story structure):

• "Thread 1: [Act 1 subject]" — all sources relevant to the first narrative arc
• "Thread 2: [Act 2 subject]"
• "Thread 3: [Act 3 subject]"
• "Background Context" — context that informs the whole film but isn't thread-specific

By source type:

• "Sources: Interview Subjects" — profiles of all interviewees
• "Sources: Archival Footage Leads" — potential footage with access and rights notes
• "Sources: Documents" — key documents
• "Sources: Published Research" — academic and journalistic research
• "Sources: Expert Sources" — potential expert interview subjects

Tags for documentary research

By source type:

• interview-subject — a person to be or who has been interviewed
• archival-footage — a footage source
• document — a written document
• published-research — academic or journalistic source
• visual-material — photograph, map, or graphic
• expert-source — analyst or specialist for context

By status:

• potential — identified as a possible source; not yet confirmed
• confirmed — confirmed as a source; interview scheduled or footage identified
• completed — interview conducted or footage obtained
• declined — source declined to participate
• inaccessible — footage or document exists but cannot be accessed (rights, classification, etc.)

By evidentiary function:

• establishes-fact — proves a specific fact in the narrative
• provides-context — background and context
• counterpoint — a perspective that challenges the main narrative
• emotional-anchor — powerful emotional element for the film
• visual-evidence — visual proof of a claim

---

Organizing Interview Subjects

The interview subject profile

For each potential or confirmed interview subject, create a profile in the "Interview Subjects" sub-Collection:

SUBJECT: [Name]
Status: [potential / confirmed / completed / declined]
Primary relevance: [why this person is important to the story]
Thread(s): [which narrative threads they appear in]

BIOGRAPHICAL BACKGROUND
Current role/position:
Past relevant roles:
Connection to the story: [how they're involved or what they witnessed]
Prior public statements on this topic: [links to prior interviews, written statements, testimony]
Online presence: [LinkedIn, website, social — useful for research before interview]

ACCESS
How to contact:
Who introduced us / connection:
Current relationship to the filmmaker: [cold outreach / warm introduction / prior relationship]
Willingness to participate (estimate): [unknown / likely / uncertain / unlikely]

INTERVIEW PLANNING
Topics to cover in interview:
Questions to ask:
Documents to show them: [what specific materials to reference or show]
Sensitive areas: [things that may require careful handling]

INTERVIEW NOTES (after interview)
Date interviewed: [date]
Key statements: [most important things they said]
On-camera moments: [particularly strong or usable footage from the interview]
Follow-up needed: [questions not fully answered; areas to explore further]
Other subjects they mentioned we should speak with:

Tracking the subject network

Documentary subjects often connect to each other — they were in the same meeting, they know each other, they're contradicting each other's accounts. The Connections graph in WebSnips can map:

• Subject A was present when Subject B's event occurred (corroboration)
• Subject A and Subject C have conflicting accounts of the same event (contradiction)
• Subject D was introduced to the filmmaker by Subject A (network chain)

Visual mapping of these relationships often surfaces connections that aren't obvious from reading profiles sequentially.

---

Organizing Archival Footage

The footage lead system

Finding archival footage is often the most labor-intensive research task in documentary work. A source identified early in development may not be accessed until deep in post-production.

For each potential footage source, create a capture:

FOOTAGE: [Description of what the footage shows]
Date of original footage: [when it was filmed/recorded]
Relevant to: [which narrative threads or specific claims]

ACCESS
Who holds this footage: [network / archive / individual / unknown]
Archive name and contact: [specific institution or person]
Status: [potential / contacted / confirmed access / license obtained / inaccessible]
Licensing requirements: [rights holder / fee estimate / restrictions]
Digital or physical: [digital file / physical tape / unknown]
Resolution/format: [HD / SD / specific format if known]
Reference number (if archived): [catalog number if applicable]

CONTENT NOTES
What's visible in this footage: [description of content]
Specific moments of interest: [timecode or description if you've reviewed it]
Quality assessment: [excellent / good / usable / marginal]

EVIDENTIARY FUNCTION
What this footage establishes or supports: [which claims does this visual evidence support?]
Can this be substituted if unavailable: [yes — with / no — critical footage]

Footage types requiring different access strategies

Broadcast news archives: Major networks (ABC, CBS, NBC, CNN, BBC) have licensing offices that handle access. Local news stations vary — some have digital archives; many have nothing. Historical news footage often requires working directly with networks or through licensing agencies like Getty Images, AP Archive, or Reuters.

Government and institutional footage: Government agencies may have footage subject to FOIA (Freedom of Information Act) requests in the US. Court footage, Congressional hearing footage, and military footage have specific access processes.

Personal and home footage: Sometimes the most powerful footage is personal — home videos, phone recordings, personal photos. This requires identifying the individuals who have it and building the relationship to obtain access and rights.

Social media footage: User-generated footage from YouTube, TikTok, Instagram. Rights are held by the original poster; YouTube's Content ID system has specific processes; some creators will license footage directly.

---

Organizing Documents

Document capture and annotation

Documents establish facts — dates, communications, agreements, official records. Their evidentiary function depends on authenticity and chain of custody, which is important to track.

For each key document:

DOCUMENT: [Title and description]
Date of document: [date it was created/signed/issued]
Author/source: [who created it]
Original location: [where this document physically exists]
Obtained from: [how we got it]
Status: [copy / original / certified copy]

CONTENT
What this document shows: [specific factual claims this document establishes]
Key passages: ["relevant verbatim excerpt"]
Thread relevance: [which narrative threads this document supports]

EVIDENTIARY FUNCTION
What claim in the narrative does this establish: [specific]
Can this fact be established another way: [backup sources if this document can't be used]
Authenticity questions: [any questions about this document's authenticity or provenance]

ON-SCREEN USE
Can this document be shown on camera: [yes / with redactions / unclear / no]
Format for on-screen display: [full document / specific pages / selected passages]

Sensitive and restricted documents

Many documentary investigations involve documents that require legal review before use. Capture notes on any document with legal questions:

• Classified or confidential designations
• Documents obtained under NDA that may have restrictions
• Documents in ongoing litigation
• Documents that could constitute a privacy violation if used without authorization

Note these explicitly so the film's legal review doesn't surface them as surprises late in production.

---

The Narrative Evidence Map

A documentary argument is made through the accumulation of evidence. The narrative evidence map connects each claim in the film's argument to the sources that establish it.

For each major claim or scene in your narrative structure, create a note:

CLAIM/SCENE: [Brief description]
Type: [factual claim / witness account / contextual background / expert analysis]

EVIDENCE FOR THIS CLAIM:
  Primary: [strongest source — document / footage / interview]
  Secondary: [backup sources — what establishes this if the primary isn't available]
  Corroboration: [additional sources that confirm from different angles]

COUNTEREVIDENCE:
  What contradicts or complicates this: [sources that challenge the claim]
  How we address the counterevidence: [how the film handles conflicting accounts]

VISUAL ELEMENTS:
  Footage: [what archival footage supports this scene]
  Documents: [what documents can be shown]
  Other visual: [photographs, maps, graphics]

INTERVIEW SUPPORT:
  Who speaks to this: [which subjects address this in their interviews]
  Key quotes: [most important things they said that support this scene]

GAPS:
  What evidence I still need to establish this: [what's missing]
  Action to close the gap: [what I'm doing to find it]

The narrative evidence map reveals:

• Claims that are well-evidenced (multiple independent sources, visual evidence)
• Claims that are thin (one source; no corroboration; no visual evidence)
• Claims where the counterevidence is significant (how will the film handle this fairly?)
• Visual coverage gaps (scenes with no archival footage)

---

Worked Example: Organizing Research for an Environmental Documentary

The scenario: A filmmaker is producing a feature documentary about contamination at a manufacturing site. The film has three narrative threads: the history of the facility, the discovery of contamination, and the legal battle. Research is ongoing over 18 months.

Interview subjects tracked (16 total):

• 4 current and former facility workers — profiles built, 3 confirmed, 1 declined
• 3 community members affected — 2 interviewed on camera
• 2 scientists (contamination experts) — 1 confirmed, 1 potential
• 2 lawyers (plaintiff and defense) — plaintiff confirmed, defense declined
• 2 regulatory officials — 1 confirmed, 1 on record publicly only
• 3 company executives — all declined; using public record materials

Archival footage sources (9 leads):

• Local TV news archive (regional station): 12 segments from 2 major contamination events identified; access pending licensing discussion
• Company promotional videos from 1980s: held by local historical society; copying rights cleared
• EPA inspection footage: FOIA request submitted
• Community meeting footage: obtained from organizer (personal footage)
• Health department hearing footage: obtained from state health department

Key documents (23 captured):

• EPA inspection reports 2018-2024 (public record; downloaded and captured)
• Company internal memo about known contamination risk (obtained via whistleblower; legal review pending)
• Court filings from the civil case (public record)
• Community health study (academic research)

Narrative evidence map finding: Act 2 "discovery" thread has strong interview coverage (4 witnesses) and documentary evidence but thin visual coverage. Gap identified: no archival footage of the moment of public disclosure. Action: seeking personal phone video from community meeting attendees.

---

Key Takeaways

1. Organize sources by narrative thread, not just by type: when writing and editing, you're looking for evidence for a specific scene or claim, not browsing by category.
2. Document access and rights information for every archival footage lead immediately: footage that becomes critical in post-production with unclear rights is a production crisis; track rights from first identification.
3. Build the narrative evidence map early: thin claims (single source, no visual evidence, no corroboration) identified in pre-production can still be researched; identified in the editing room, they become narrative problems with no time to fix.
4. Capture interview subject profiles before the interview, not just after: understanding a subject's background, prior statements, and connection to the story before you speak with them produces significantly better interviews.
5. Note counterevidence explicitly: documentaries that acknowledge and engage with contradictory evidence are more credible than those that don't; knowing the counterevidence lets you address it fairly.

---

Conclusion

Documentary research is the foundation that every creative decision in a film rests on. A feature documentary may take 3-5 years from inception to release; the research conducted in year 1 needs to be accessible and useful in year 4. A knowledge system that organizes interview subjects, footage leads, documents, and published research by narrative thread and evidentiary function — with annotations that capture why each source matters and how it supports specific claims — produces a research library that serves the full arc of production. The filmmaker who can retrieve any source in 30 seconds and who knows precisely which claims need more evidence is the one who produces films that hold up to scrutiny.

To go deeper, check out AI Knowledge Management in 2025.