The case of the disappearing source

Imagine that a post announces, “Scientists have discovered that listening to a particular sound doubles creativity.” It includes a dramatic picture and a thousand reposts. A second account copies the claim. A video creator repeats it. A third website says the discovery is now “widely reported.”

You look for the study. None of these accounts links to it. One cites the other, and the second cites the first. That is repetition, not independent confirmation. A convincing chain of retellings can still lead back to an unverified assertion.

In Lesson 001 we distinguished observation from inference. In Lesson 002 we learned to test explanations. Here we ask: Who first made this particular claim, what did they actually establish, and where can someone else examine it?

1. Primary does not mean automatically correct

A primary source is material close to the event or question under study: an original interview, speech, photograph, letter, dataset, laboratory report, meeting record, or firsthand testimony. A secondary source interprets, compares, contextualizes, or summarizes primary material. A reference work may synthesize still more sources.

Which category applies depends on the question. A newspaper article can be primary evidence of how a newspaper covered an event, yet secondary evidence for what caused that event. A firsthand witness can misremember. An official record can contain errors. A research paper can have limitations. “Original” tells us where to investigate; it does not stamp a claim as true.

The Library of Congress provides guidance on working with historical primary sources and interpreting point of view, audience, and context. We will use its Getting Started with Primary Sources resource as one of our reference points.

2. Every claim has a route

Think of information as a message traveling through a series of hands. Someone observes or creates a record. Another person selects a sentence. Another quotes it. A headline compresses it. A social account rewrites the headline. Eventually the retelling may feel more familiar than the original.

For every important claim, draw a simple path:

A source should be attributable with enough detail to find it again: a creator or responsible organization, the exact document or work, a publication or event date when known, and an accessible location. When pages change, an archived or stable record can help preserve the trail.

3. Read sideways before reading deeper

When you reach an unfamiliar site, you need not decide whether to trust it using only its own “About” page. Lateral reading means leaving the page briefly to investigate who operates it, what other credible sources say about it, and whether its specific claims can be corroborated independently.

1. Identify the publisher. Who owns or runs the site? Is the author named? What expertise or direct access is relevant to this particular claim?
2. Search outside the site. Look for independent reporting, institutional records, corrections, funding disclosures, and a history of reliability.
3. Return to the specific claim. Does it link to the actual evidence? Does that evidence say what the headline says?
4. Compare genuinely independent accounts. Three websites copying one press release are not three independent confirmations.
5. Keep your judgment provisional. A credible publisher can make a mistake, and an unfamiliar publisher can publish an accurate item.

The Digital Inquiry Group, which grew from the Stanford History Education Group, offers Civic Online Reasoning resources that help learners evaluate online information. A web address ending in “.org,” an impressive design, or a polished biography is not by itself a reliable credibility test.

4. Context changes a quotation

Imagine a headline that says: “The mayor confirmed the bridge is unsafe.” The full meeting transcript might instead read: “If an inspection finds that the bridge is unsafe, we will close it immediately.” Removing “if” changes a conditional statement into a claim of established fact.

The preceding mayor and bridge quotation is invented for teaching. It is not a statement by an actual public official.

When a screenshot, clip, or quotation feels decisive, ask what occurred immediately before and after it. Check whether it was edited, translated, written as satire, or originally posed as a hypothetical. Dates matter too: an accurate report about a past situation may mislead when presented as current.

5. Search for the evidence, not just the phrase

Searching for the headline alone can return many copies of the same headline. Improve the search by adding clues that lead toward original material. Try combinations of a quoted phrase, a named author, an institution, a date, or a publication title. Search within a relevant primary-source collection when one exists.

• Historical claim: Search a library, archive, contemporaneous newspaper collection, or official record; compare dates and authorship.
• Scientific claim: Look for the actual study, its authors, sample, measurement, limitations, publication date, and whether it was reviewed or later challenged.
• Government or policy claim: Locate the enacted text, official announcement, meeting record, or public dataset; distinguish proposals from adopted decisions.
• Quotation or image: Seek the longer recording, original photograph, credited creator, first publication, or complete transcript.
• AI-generated statement: Ask for sources and check them independently. A confident answer or plausible-looking citation is not proof that a real document exists.

For scholarly references, Crossref Metadata Search can help locate publication metadata and DOI records. A DOI is a persistent identifier, not a guarantee that the paper is accurate, peer-reviewed, or applicable to a given claim.

6. The attribution trail: claim, source, evidence, limitation

Good attribution is more than placing a link after a paragraph. A reader should know which statement comes from the source and where your own interpretation begins. This works for schoolwork, reporting, art criticism, history, and technical research.

Example: “The Library of Congress explains that primary sources are original records from the period under study (Library of Congress, Getting Started with Primary Sources, accessed October 22, 2026). This helps us locate firsthand material, but it does not mean every firsthand account is complete or accurate.”

When the source itself is missing, the responsible form is simple: “I have not located the original source, so I cannot verify this specific assertion.” Uncertainty is a finding, too.

7. Your first source investigation · 10–15 minutes

FIELD ASSIGNMENT · NO ACCOUNT OR PAYMENT REQUIRED

Choose one of these routes. Do not share private correspondence or personal details. The point is to practice tracing a claim, not to make a public accusation about a person or institution.

Route A — Find the original

Choose a harmless public statement you have encountered, such as a claim about a museum artifact, scientific discovery, historical date, or public event. Search for the earliest clearly documented source you can access.

1. Copy the claim exactly, placing quotation marks around any exact words.
2. Record where you first encountered it and its visible date.
3. Search for a named original author, primary record, or dataset.
4. Open the longer document and identify what it actually supports.
5. Find one genuinely independent second source if available.
6. Write a three-sentence conclusion separating verified information, interpretation, and remaining uncertainty.

Route B — Investigate the evidence in this lesson

Visit the Library of Congress primary-source guide. Record the institution, page title, URL, and the key distinction it makes between primary and secondary sources. Then visit the Digital Inquiry Group and find a resource about evaluating online information. Compare what each resource teaches without claiming they are the same publication or represent independent verification of every other claim.

8. Practice case: a viral quotation

Fictional classroom scenario: An image shared online reads: “The university has banned imagination in every science class.” Beneath it is the logo of a fictional university. A popular account reposts the image and writes, “Confirmed by the university itself.” No original notice is linked.

You search the fictional university’s official announcement archive and find a note saying: “A particular exam prohibits the use of generative writing tools; students may still use their own diagrams, analogies, and creative explanations throughout coursework.”

Compare the claim and evidence. The viral image changes a specific exam rule into a sweeping claim about imagination. The logo does not prove the screenshot is authentic. An actual university-wide ban on imagination has not been established in this invented scenario.

---

Reflection: Does the source survive the journey?

You will not always locate the first record. Some original material is missing, inaccessible, oral, disputed, or protected. Sometimes a claim belongs to a personal, cultural, or symbolic tradition rather than an empirical research question. The task is not to erase meaning, but to avoid manufacturing certainty or misattributing a source.

When you preserve a specific origin, another person can retrace the route and disagree with your interpretation without losing the record. When you identify what remains unknown, you keep open the possibility of correction. Attribution is how knowledge remains accountable across retellings.

Takeaway: Follow the claim backward. Identify the creator. Check the original. Read beyond the excerpt. Compare independent evidence. State the limit. Cite the trail.

Completion: You have completed Lesson 003 when you can produce a source trail and a one-sentence claim that accurately distinguishes evidence from interpretation. There is no quiz score, certificate, or enrollment requirement.

References and further learning

• Library of Congress — Getting Started with Primary Sources: original records, context, perspective, and classroom inquiry.
• Digital Inquiry Group — Civic Online Reasoning: educational approaches to evaluating online information and lateral reading.
• Crossref Metadata Search: discovering authorship, publications, and DOI metadata for scholarly works.

← Review Lesson 002 · Explore Nova Academy Lessons · Visit the Quantum Atrium

Nova Academy is a free public learning resource within the Electric Icarus Project. These introductory lessons are non-accredited and do not award professional qualifications. Original Electric Icarus interpretations are distinguished from independently established findings.