Hidden History · Lesson 06 · Michigan, 1969 → Washington, 1974
Thousands of people chased one rumor for three weeks. Two reporters built one case for twenty-six months. The difference was not intelligence or effort. It was a rule.
A caller got through to Russ Gibb, a disc jockey at WKNR. He said Paul McCartney was dead, the Beatles had hidden it, and their records contained the proof.
The caller told Gibb to play a track backward. Gibb did it live. His signal reached tens of thousands of people at the same time from a voice many heard every day.
Within about three weeks, newspapers and television stations across the country were running the story.
Detroit disc jockey whose broadcast helped launch the rumor nationally.
Two days later, a University of Michigan student named Fred LaBour published a piece announcing McCartney's death and listing “new evidence.”
LaBour says he invented most of the clues as satire. His original readers had context. When stations and other papers repeated the piece, the context fell away. What reached another city sounded like “the Michigan Daily reported,” not “a student wrote a joke.”
The article gave people a shared checklist. From then on, thousands were not independently discovering clues. They were checking the same list.
Michigan Daily student writer who later said he invented most of the familiar clues as satire.
People in Cleveland and Sacramento found the same clues without knowing one another. That looked like corroboration because the results matched.
The missing step was tracing where each person got the checklist. Agreement counts as independent only when the sources could have gotten it wrong separately.
Walk the signal backward.
At about two in the morning, five men were arrested inside the Democratic National Committee offices in Washington's Watergate complex.
The Democratic National Committee runs the Democratic Party's national campaign operation. Breaking into it during a presidential election was not an ordinary office burglary. The men carried cameras and electronic surveillance equipment, not things they could sell.
The next morning it was a small story. It took twenty-six months for the case to help end a presidency.
If information came from a source who would not be named, the Washington Post did not run it until a second source, independent of the first, confirmed it.
Two people who heard the same third person are one source and an echo. A person who agrees after you reveal what you already know may only be agreeing with you.
The rule killed stories reporters believed were true. That cost was the point. A standard matters because it can stop you from doing what you want.
Group statements by origin, not by speaker.
The popular picture is a shadowed parking garage and a source in a trench coat. The actual work was campaign filings, phone records, payrolls, and repeat interviews.
A record holds still. A phone log shows the same call every time you look. That makes a witness more useful because the question can become exact: why did this number call that number on this date?
On July 16, 1973, White House staffer Alexander Butterfield answered a routine Senate question and confirmed that a taping system had recorded the president's conversations. Nobody had started the question looking for a taping system. Thorough work made room for the answer.
Do not begin with “Do I believe this?” Ask what finding would end the claim, then see whether that finding exists.
A living Paul McCartney should have ended the death rumor. He was alive, photographed, and interviewed. The rumor absorbed that evidence.
The Watergate case could have died if financial and communication records failed to connect the burglars to the campaign. Investigators obtained the records. They showed the opposite.
Same three questions. Both cases.
Three weeks. Great energy. One shared checklist. Nothing that survived contact with a living man.
Twenty-six months. Records, repeated interviews, and a rule that refused to count echoes as confirmation.
Two independent sources. The method is simple. Following it when a good story is waiting is the hard part.
| term | what it means here | where it happened |
|---|---|---|
| source | A person or record information came from, counted by origin. | Two events in Michigan. |
| independent | Two sources that could have failed separately. | The network trace. |
| corroboration | Independent support for a claim, not repetition. | What the rumor looked like but lacked. |
| downstream sourcing | Doing real work on somebody else's list and mistaking it for a new investigation. | Checking LaBour's clues. |
| pareidolia | Finding a definite pattern in unclear input, especially after being told what to expect. | Backward audio. |
| two-source rule | Requiring a second independent confirmation for anonymous information. | The Post standard. |
| falsifiability | Whether some possible finding could end a claim. | The disproof test. |
| primary record | A document created at the time that does not change between interviews. | Finance and phone records. |
| burden of proof | The job of establishing a claim, held by the person making it. | All three method questions. |
| the quick version | the record supports | check |
|---|---|---|
| Fans discovered the famous clues. | Fred LaBour says he invented most of the shared list as satire, and later readers checked it downstream. | Section 2 |
| So many people found them that something had to be there. | The branches trace to the same Michigan origins. Repetition is not independence. | Section 3 |
| A magazine proved the rumor false and ended it. | Disproof was available while belief continued. The rumor faded later. | Section 7 |
| Watergate was solved by one man in a garage. | Records, repeat interviews, newsroom rules, hearings, courts, and tapes built the case. | Sections 4–6 |
| Two sources means a story is true. | It means the source structure meets a publishing standard. That is not certainty. | Section 5 |
| The reporters were smarter. | The difference was a procedure that separated an origin from an echo. | Section 8 |
The receiver, identity markers, network, and records on this page are native HTML explanations, not archival photographs or facsimiles. The page deliberately avoids a station-by-station spread chronology where accounts differ.