Cite Documentary Research Without Turning the Deck Into an Academic Paper
Cite Documentary Research Without Turning the Deck Into an Academic Paper
Keep the claim on the page where it matters. Give it a short, readable source note. Put the full reference somewhere the reader can actually reach, with a locator that gets them to the relevant passage rather than a repository's front door.
That is enough of a system to begin with. A documentary deck does not need to become a bibliography with pictures. It does need to distinguish what someone remembers, what a document records, and what the filmmaker concludes from either.
The difficult part often comes before citation design: checking that the source says what the sentence says. No footnote can turn eighteen advertised screenings into eighteen sold-out performances.
Start with the sentence you are asking someone to believe
Consider this deliberately fictional pitch for a documentary about a neighborhood cinema:
In the summer of 1998, Harbor Cinema stayed open past midnight every night. Its late program sold out eighteen times. Surviving publicity captures the moment the neighborhood discovered its second home.
The accompanying research page lists an interview, a program card, and an annual bulletin. It looks adequately sourced. It is not.
We have created all three records for this exercise. Harbor Cinema, its staff, the documents, and their contents are invented. The files are included with this article so the citation paths can be inspected. They are teaching material, not research into an actual cinema.
The interviewee, former projectionist Mara Ives, remembers finishing after midnight on some nights. She explicitly says she cannot supply a count. A surviving program card in the fictional collection advertises one 11:15 p.m. screening. The annual bulletin counts eighteen scheduled late screenings during June through August and explicitly excludes attendance information.
These sources support three different kinds of statement. The interview supports an attributed recollection. The card supports an observation about an announcement. The bulletin supports a bounded count of scheduled events. None establishes sellouts or the neighborhood's collective feelings.
Before choosing a type size for the source notes, rewrite the pitch:
Former projectionist Mara Ives remembers finishing after midnight on some nights. A June program card advertises an 11:15 p.m. screening. The cinema's annual bulletin lists eighteen scheduled late screenings that summer. The film will ask who those late nights were for.
The last sentence is a proposed inquiry, not a historical conclusion. It does not need a source that proves an answer the filmmaker has not found. The first three sentences do need sources, and now each has something specific to support.
Notice that the revision does not become timid. It becomes more useful. We know what the filmmaker has found and what remains to be investigated. The cinema can still matter deeply to people without the deck pretending to speak for an entire neighborhood.
Make a short note lead to a complete record
For the fictional slide, the compact notes could read:
A. Ives interview, April 6, 2026, transcript ¶2.
B. June 1998 program card, EX-ARC-01.
C. Annual Bulletin 1998, p. 4, “Late program.”
Place A after the recollection, B after the program-card sentence, and C after the count. A note at the end of the entire paragraph would leave the reader to guess which record supports which assertion.
The full source entries belong in a research appendix or an accompanying source sheet. That sheet should preserve the identifiers needed to recover the material. For archival documents, the National Archives' catalog guidance includes the document title and date, series, record group, repository, catalog-description URL, and access date.[^1] That is guidance for its records, not a mandatory style for every pitch deck. Its useful lesson here is that a repository name alone is not an item reference.
For our example, the full entries lead to:
A — Interview. Mara Ives, interview by Ellis Chen, fictional interview date April 6, 2026, EX-INT-01, transcript paragraph 2. Open the teaching transcript at the relevant paragraph.
B — Archival object. Late Program, June 1998, Harbor Cinema Working Collection, Printed Programs series, EX-ARC-01. Open the original fictional object record.
C — Published record. Harbor Cinema, Harbor Cinema Annual Bulletin 1998, fictional publication date February 1999, p. 4, “Late program,” EX-PUB-01. Open the text-only teaching facsimile.
These are real relative links to invented material, not invented links to supposedly real archives. In an actual deck, use the actual item's identifying information. Keep the short note legible and make the longer route work. There is no benefit in a perfectly formatted reference that points to the wrong edition.
Give interviews a locator that belongs to the material
An interview can exist as a recording, a transcript, a published excerpt, and a set of research notes. Those versions are related, but they are not interchangeable.
Our teaching interview has a transcript and no recording. Its locator is paragraph 2. Giving it a timecode would invent precision: there is no audio at that timestamp to check.
For an actual recorded interview, identify the speaker, interviewer, session date, recording or transcript version, and the passage used. When the authorized reviewer can access the recording, a timestamp range can take them to the relevant exchange. For a paginated transcript, a page and paragraph or line reference may be more useful. If the transcript omits a qualification audible in the recording, resolve that difference before relying on the excerpt.
The Oral History Association recommends contextualizing excerpts, citing the location of the full oral history, and providing access aids such as transcripts or time-tagged indexes. It also cautions users against manipulating a narrator's words.[^2] Those principles support a clear route from the sentence to its context; they do not require you to publish every recording in a pitch package.
Here, paragraph 2 matters because Ives's uncertainty is part of the answer. Quoting only “I finished after midnight” would remove both “some nights” and her refusal to provide a count. Even a working link would not repair that distortion.
Keep the attribution in the main sentence when it changes what the reader is being told. “Ives remembers” cannot disappear into an appendix if the sentence otherwise presents her recollection as an established operating schedule.
Do not ask an object to establish more than it contains
The program card is a different problem. Its printed date and advertised start time are inspectable within the fictional record. They support a statement about what was announced.
They do not establish that the show happened, that it started on time, that it was full, or that anyone stayed afterward. The phrase “program subject to change” makes the distinction especially plain, but the distinction would exist without it.
You could write:
A June 1998 card advertises a late feature at 11:15 p.m.
You could not responsibly turn the same card into:
By June, Harbor's packed midnight screenings had become a neighborhood ritual.
The second sentence adds attendance, a repeated pattern, and a social interpretation. Those may become defensible through other research. They are not hiding somewhere inside the card's citation.
This is also where a source note and an image credit separate. The source note tells the reader which item supports the assertion. A credit identifies the material displayed and its credited source or creator. A permission record addresses whether and under what conditions the material can be used. Do not treat one field as a substitute for all three. Where use requires rights or privacy review, keep that review distinct from the citation check.
If the deck shows a substitute image for an unavailable original, label the substitution. Do not attach the original item's citation to a look-alike and allow the reader to believe they are seeing the evidence itself.
Put the count's boundaries next to the number
The annual bulletin's eighteen is particularly easy to misuse because a number looks settled. The record, however, defines the unit: scheduled screenings with advertised starts at or after 11 p.m., during June through August.
Change “scheduled screenings” to “performances” and you imply they occurred. Change it to “sold-out nights” and you invent ticket-sales evidence. Change “June through August” to “1998” and you expand the period.
The revised slide keeps the important qualifier in the sentence: eighteen scheduled late screenings. Its fuller source entry supplies the page and section. The reader does not have to open the appendix to discover that the confident headline means something smaller.
You do not need to reproduce the entire bulletin definition beside the number if the shorter wording preserves its meaning. You do need to retain any boundary whose removal would change the claim. This is an editorial judgment about the sentence, not a contest to fit the most metadata into the smallest corner.
If a source's wording is itself ambiguous, do not decide silently on the interpretation most favorable to the pitch. Attribute the figure as reported, explain the relevant ambiguity, or leave the number out until it can be clarified.
Test the path the recipient will receive
Check the delivered artifact, not just the working document. Start with one claim in the exported deck. Follow its short note to the complete entry, open the authorized source, and find the passage. Then read enough around it to see whether the sentence has preserved its limits.
For the fictional example, the three paths terminate at paragraph 2, item EX-ARC-01, and the “Late program” passage. The supplied files make this a repeatable exercise: a reader can inspect the exact wording behind each rewritten claim. A simple automated check can confirm that those files and anchors exist. It cannot decide whether the claims are fair summaries; that requires comparing the prose with the records.
An actual review may stop at an access boundary. A password-protected interview is not made public merely because the deck contains a link. Identify who may inspect it and through what authorized route. Do not promise unrestricted availability or put sensitive material into a public folder just to make a footnote convenient. The Oral History Association's access guidance treats restrictions agreed with participants as part of the access decision.[^2]
A restricted source may be legitimate evidence, but the restriction affects how much independent checking the recipient can do. State that limit when it matters. “Confidential interview; authorized review by arrangement” is more honest than a public-looking link that cannot deliver its apparent promise.
Finally, remove the machinery the reader does not need. You may have many background readings that shaped the investigation. They do not all need to sit beside a sentence they do not directly support. Keep the research record; select the references that carry the deck's actual claims.
The finished page can remain a page about a cinema: a projectionist remembering the last bus, a printed late show, a small program whose audience the film has yet to meet. Three modest notes keep those details attached to what is known. They leave the filmmaker free to ask the next question without pretending it has already been answered.
Sources
[^1]: U.S. National Archives and Records Administration, “Frequently Asked Questions (FAQs)”, “How do I cite a record in the Catalog?” Inspected September 19, 2026 (UTC).
[^2]: Oral History Association, “Oral History Best Practices”, “Interviewing” §2 and “Access & Use” §§1–3. Inspected September 19, 2026 (UTC).
Frequently asked questions
How can a documentary deck cite research without becoming an academic paper?
Keep the claim on the page where it matters, give it a short readable source note, and put the full reference where the reader can reach it with a locator to the relevant passage. The deck does not need to become a bibliography with pictures; it needs to distinguish memory, document record, and filmmaker conclusion.
Why does the Harbor Cinema example need rewriting?
The original pitch claims every night past midnight, eighteen sellouts, and the neighborhood discovering its second home. The supplied fictional sources support only an attributed recollection of some late nights, one advertised 11:15 p.m. screening, and a bulletin counting eighteen scheduled late screenings from June through August while excluding attendance information. The revision attributes each claim to what it can support.
What is the difference between a source note, an image credit, and a permission record?
A source note tells the reader which item supports the assertion. A credit identifies the material displayed and its credited source or creator. A permission record addresses whether and under what conditions the material can be used. They are separate fields, and one should not substitute for all three.
How should an interview be cited when it exists as recording, transcript, excerpt, or notes?
Those versions are related but not interchangeable. The locator should belong to the material used: a paragraph or page for a transcript, a timestamp range for an accessible recording, or another appropriate reference. If a transcript omits a qualification audible in a recording, resolve that difference before relying on the excerpt. Do not invent a timecode when no audio exists at that timestamp.
What limits should stay beside the annual bulletin's eighteen count?
The bulletin defines eighteen scheduled screenings with advertised starts at or after 11 p.m. during June through August and excludes attendance information. Changing 'scheduled screenings' to 'performances' implies they occurred; changing it to 'sold-out nights' invents ticket-sales evidence; changing the period to 1998 expands it. If wording is ambiguous, attribute the figure as reported, explain the ambiguity, or leave the number out until clarified.