The problem with famous last words, as a category, is that most of them are invented. Not all. But most. A person is dying, usually surrounded by a small number of witnesses, and the recording of what they said is left to memory, to emotion, to the circumstances of the moment, and eventually to the person writing the biography decades later who wants the final line to have a certain kind of resonance.
What this means, in practice, is that the "famous last words" we have inherited are a mix of three different things: words that the person probably actually said (often prosaic, sometimes embarrassing), words that were said but were cleaned up by posterity, and words that were invented entirely and attached to the person because they suited the narrative of their life. Sorting these out is, if you care about the texture of the past, one of the small pleasures of archival work.
I have been collecting last words for about four years, in a notebook I keep on my desk. What follows is a small sample, eight entries, arranged not by fame but by the interesting question each one raises. For each, I have tried to note what we actually know, what was probably added later, and what the discrepancy reveals.
Oscar Wilde, 1900
The received quotation: "Either that wallpaper goes, or I do."
What probably happened: Wilde did make a remark about the wallpaper of his room in the Hôtel d'Alsace in Paris during his final illness. The exact phrasing is disputed. The earliest recorded version, written down by a friend some weeks after his death, is close to the received one but not identical. The actual final utterance before his death, according to the priest who attended him, was probably a prayer-related mumble rather than the witty epigram.
What this teaches: Wilde's biographical legend required him to exit on a witticism. Biographers and friends supplied one, using materials (the wallpaper remark) that were available from the preceding weeks. This is common. The "last words" become a retrospective construction, not necessarily dishonest, but shaped to fit the shape of the life.
Humphrey Bogart, 1957
The received quotation: "I should never have switched from Scotch to martinis."
What probably happened: Bogart's wife Lauren Bacall was with him in his final days. The "switched from Scotch to martinis" line, according to her own memoir, was something he said in the weeks before he died, as part of his general commentary on dying, and not specifically on his deathbed. His actual last words, to her, on the night he died, were more tender and less quotable. Bacall did not record them.
What this teaches: the discrepancy between "last words" and "words said near the end" is larger than most quote collections admit. A memorable line from three weeks before death is almost always preferred, for the posthumous record, over an intimate exchange in the final hours. The public version is the polished one.
A memorable line from three weeks before death is almost always preferred, for the posthumous record, over an intimate exchange in the final hours.
Pancho Villa, 1923
The received quotation: "Don't let it end like this. Tell them I said something."
What probably happened: Villa was assassinated by gunfire in his car. He was hit multiple times and died very quickly. There is no credible witness who reports him saying anything coherent in the few seconds between the first shots and his death. The quotation appears to have been invented by a journalist writing up the assassination, who thought the line was funny and resonant given how image-conscious Villa had been in life.
What this teaches: some famous last words are flatly fabricated, and the fabrication sometimes becomes canonical because it is better than anything the person actually said. There is no simple solution to this. The historian can only note the invention and move on.
Beethoven, 1827
The received quotation: "Applaud, friends, the comedy is finished."
What probably happened: the source is a second-hand account written down more than a decade after Beethoven's death. There is no contemporary witness who recorded such a line. The composer was very ill in his final days and probably said little that was coherent. One witness does report him raising his fist during a thunderstorm on his last afternoon, which has been variously romanticized by biographers.
What this teaches: for historically distant figures, the "famous last words" are often reconstructions from the 19th-century biographer who wanted to give the Great Man a suitable exit. The late 18th and early 19th centuries, in particular, produced a lot of this, the Romantic biographical mode required dramatic deathbeds, and supplied them where the original material was thin.
For any of these quotations, the most reliable filter remains the Oxford Dictionary of Quotations and, for the Roman-era material, the Perseus Digital Library at Tufts. I've also found that Project Gutenberg has the full text of most of the contemporary biographies I cited above, for readers who want to check the disputed cases against the source material themselves. Related material on recovered letters and correspondence is collected in Three Letters the Recipients Never Opened.
Julius Caesar, 44 BCE
The received quotation: "Et tu, Brute?"
What probably happened: Shakespeare wrote this line in 1599. Caesar, according to the earliest Roman sources that mention his death at any length (Suetonius, a century and a half later), either said nothing at all, or said "You too, child?" in Greek (kai su, teknon?) to Brutus. The Greek line is itself disputed. Some sources have him silent. Some have him covering his face with his toga. The Shakespeare line, which is Latin, is a dramatic invention that has completely displaced whatever the historical Caesar may or may not have said.
What this teaches: a line invented for the stage can, if it's good enough, become the canonical "last words" of a historical figure. This is an unusual outcome but not a unique one. The power of a well-written dramatic line, over time, exceeds the claims of any historical source.
James Joyce, 1941
The received quotation: "Does nobody understand?"
What probably happened: this is, by the standards of this list, unusually well-sourced. Joyce's nurse recorded the line as one of the last things he said before slipping into the coma from which he did not recover. The context was confusion about something in his hospital room, it was not, as sometimes presented, a grand philosophical lament. It was, essentially, the muttering of a dying man who was briefly lucid and frustrated about an immediate physical situation.
What this teaches: even well-sourced last words can be misrepresented. The line "Does nobody understand?" is authentic. The interpretation that Joyce was asking a cosmic question about the comprehensibility of his work is a later gloss. The actual context was small and bodily. We have, consistently, romanticized the circumstances of literary deaths in ways that the people doing the dying would have found, probably, embarrassing.
Kathleen Ferrier, 1953
The received quotation: "Now I'll have eine kleine Pause."
What probably happened: this one appears to be real. Ferrier, the British contralto, was dying of cancer in a London hospital. She spoke German; she had just been given morphine; she made the remark to her nurse. It was recorded at the time and has stood up to subsequent scrutiny. It is also, I think, one of the genuinely perfect last lines, a musical joke from a singer, a rest in place of a cadence, precisely the right scale of humor for the moment.
What this teaches: occasionally, despite everything, the famous last words are real and are beautiful. These are rare. When they happen, they deserve to be celebrated rather than retouched. I'd put Ferrier's line at the top of the list of last words I believe and love.
Charles I, 1649
The received quotation: "Remember."
What probably happened: Charles I was executed on a scaffold in front of a large crowd, and his last words were recorded by multiple independent witnesses. He delivered a long final speech, of which several versions exist and which largely agree on the substance. His final utterance, spoken quietly to his chaplain just before putting his head on the block, was the single word "Remember." Its meaning has been debated for nearly four centuries.
What this teaches: sometimes the genuine last word is a small, enigmatic thing, and the mystery is the point. Charles was asking someone to remember something. We don't know what. The gap is the interesting part of the archival record. I am content to leave it open.
Eight entries. Half authentic, half cleaned-up, embellished, or invented. This ratio is roughly what you'll find if you go through any larger collection of famous last words with a skeptical eye. The wise thing to do, when you encounter one, is to treat it as literature first and history second, which is, I think, what most of these lines are. Literature, applied retroactively. The Caesar we have is Shakespeare's. The Wilde we have is the biographers'. The lives were real. The exits were edited.