Penloom History · Notes

How to Check a History Fact Before You Repeat It

You have seen the format. A dark, astonishing fact, delivered in forty seconds over stock footage, with no source and a comment section full of people saying they had no idea. Some of those are true. A striking number are a nineteenth-century embellishment, a mistranslation, or a novel that people forgot was a novel.

The problem is that the false ones are better stories. They have been polished by a century of retelling — the details that did not land got dropped, the ones that did got sharper. Accuracy has no such editor. So the tidiest, most quotable version of a historical event is, on average, the least reliable one.

Here is the method for telling them apart. It is the same one used to build every volume we publish, and it takes about fifteen minutes per claim.

1. Find the earliest telling you can reach

Trace the claim backwards. The video quotes a blog, which paraphrases a popular history book, which cites a Victorian anthology, which cites an anecdote a traveller wrote down long after the fact. That last one is the actual evidence. Everything above it is repetition.

You are looking for the gap between when the thing happened and when someone first wrote it down. A record made the same week by someone present is a different category of thing from a colourful account written three hundred years later, even if both appear in a footnote. When the trail dead-ends at a source written long after the event, with nothing behind it, you have found a story rather than a record.

Two practical moves: search the distinctive phrasing rather than the topic, because a fabricated detail travels verbatim and you can watch it hop from page to page. And check whether the earliest version even says what the modern one says — claims routinely acquire numbers, names and motives on the way down.

2. Ask what kind of record it is

Not all primary sources carry the same weight, and the difference is usually about what the document was for.

• Administrative records — municipal minutes, court transcripts, parish registers, ships' logs, ledgers, tax rolls. These are the strongest, because they were created for a bureaucratic purpose by people with no interest in a good story. Nobody embellishes a burial register.
• Contemporary correspondence and diaries. Strong, but written by someone with a point of view, and often for a reader.
• Chronicles and published accounts. Useful, and frequently written to entertain, flatter a patron, or make a moral point.
• Later retellings. Evidence of what people believed later, not of what happened.

This is why the 1518 Strasbourg dancing episode survives scrutiny when so many similar stories do not: it is not preserved only in folklore, but in the city's own municipal records — the dull, official paperwork of a council dealing with a problem. The bureaucracy is the proof.

3. Count your sources properly, because most of them are the same source

"Two independent sources" is the working standard, and almost everyone counts it wrong. Five books that all cite the same 1897 anthology are not five sources. They are one source and four echoes, and the fact that they agree tells you nothing at all — of course they agree, they are copies.

Independence means the two accounts do not derive from each other. A council record and a letter written by an unrelated witness are independent. Two modern documentaries drawing on the same standard reference work are not, no matter how reputable each is.

So: open the footnotes. If both of your sources point at the same third thing, you have one source, and your real job is to evaluate that third thing.

4. Distrust the too-perfect detail

Specific, vivid, quotable details are the first things storytellers invent and the last things records preserve. Be sceptical when a claim includes:

• A dying sentence. Famous last words are almost always reconstructed, frequently by someone who was not in the room.
• An exact count of something nobody was counting. Crowd sizes, casualty figures in chaotic events, numbers of victims — these were often estimated later, and the roundest number is usually the invented one.
• Interior thoughts and motives. A record can say what someone did. It cannot say what they were thinking, unless they wrote it down.
• An ironic coincidence. The detail that makes the story land is the detail most likely to have been added to make the story land.

None of these prove a story false. They tell you which specific part to check first.

5. Separate the event from the explanation

A great deal of history is well documented in its what and completely contested in its why. The records show something happened; the cause has been argued for centuries and remains open.

Popular retellings collapse those two together and present the currently fashionable explanation as though it were part of the evidence. It is not. When you repeat one of these, say which part is documented and which part is interpretation — and if you cannot tell the difference in the version you read, that is a strong signal about the version you read.

The same logic applies to the great disappearances. That an expedition set out in 1845 and did not come back is a matter of record. Precisely what happened afterwards has been reconstructed from fragmentary evidence and revised repeatedly, and the honest version says so.

6. Label the disputed part instead of deleting it

The temptation, once you find that a beloved detail is shaky, is to drop it and tell the clean version. Resist it. The disputed part is usually the most interesting part, and the disagreement between historians is itself real, checkable history.

The move is to flag it in the telling: the records confirm this; that part is contested; this last bit appears first in a novel. It costs one sentence, it makes you the person who can be trusted next time, and it means nobody who repeats you gets embarrassed.

Being the one who says "that part is disputed" is not the buzzkill role. It is the only version of this that survives being checked — and on the internet, everything eventually gets checked.

The short version

1. Trace the claim to its earliest telling, and see how big the gap is.
2. Ask what the document was created for. Bureaucracy beats anecdote.
3. Check that your two sources are not the same source twice.
4. Check the vivid details first — last words, exact counts, motives.
5. Separate what happened from why people say it happened.
6. Say which part is disputed, out loud, in the telling.

Do this once and you will never look at a viral history clip the same way again. Do it habitually and you will find the true stories are stranger than the invented ones anyway — they are just less tidy, which is exactly why nobody polished them into fiction.