He confessed the hoax on a front page, and readers decided the confession was the hoax

On 28 December 1917, the New York Evening Mail ran a piece by H. L. Mencken called “A Neglected Anniversary.” Its complaint was that nobody had marked the seventy-fifth birthday of the American bathtub the week before.

The piece is dense with the kind of detail that makes a thing feel checked. The bathtub had been introduced in England by Lord John Russell in 1828. A Cincinnati cotton and grain dealer, Adam Thompson, met one on business trips, took the idea home, and had the first American bath on 20 December 1842. Doctors said it was dangerous. Critics said it was undemocratic, being expensive — five hundred dollars to install one in New York in 1845. The Philadelphia Common Council considered an ordinance banning bathing between 1 November and 15 March and defeated it by two votes; Boston made bathing unlawful without medical advice in early 1845, never enforced it, and repealed it in 1862. John F. Simpson, a Brooklyn plumber, devised the zinc tub in 1847. Oliver Wendell Holmes declared for the bathtub; the American Medical Association passed a resolution of approval in 1850, and the homeopaths followed in 1853. Vice President Millard Fillmore inspected Thompson’s original tub in March 1850, bathed in it, and on becoming president invited tenders for one in the White House — which restarted the argument, because Washington and Jefferson had managed without.

None of it happened. Mencken wrote the whole thing, dates and all, in the middle of a war, because there was nothing serious he was allowed to write.

The shape of the thing that spreads

What is striking, reading the fake history now, is that it is not a single false claim. It is a structure: an invention, an early adopter, a controversy with named opponents, a technical fix that drops the price, a professional body granting approval, and finally an establishment figure who settles it. That is the shape of how a technology actually enters a society. Anyone who has read one real history of one real object recognises the silhouette.

And it is bristling with the exact features a person uses to decide whether to believe something without checking it. Specific dates rather than round ones. Named people, all of them real — Russell, Holmes, Fillmore. Institutions that existed. Two cities behaving differently, which is what cities do. A margin of two votes on the Philadelphia ordinance, and an end date on the Boston one — because a fake that records a law being repealed, and notes in passing that it was never enforced anyway, looks far more like an archive than a fake with only a beginning.

None of those features are evidence. Every one of them is cheaper to write than to check, which is the whole trick, and it is why the trick keeps working.

The retraction did not take

Mencken did not enjoy this the way a hoaxer is supposed to. On 23 May 1926 the Chicago Tribune ran his retraction on its front page — “Melancholy Reflections”, reprinting the fake history beside his description of it as a tissue of absurdities, all of them deliberate and most of them obvious.

Two months later he had to do it again. “Hymn to the Truth”, Tribune, 25 July 1926, exists because readers had concluded that the confession was the hoax.

By the 1949 Chrestomathy he had stopped expecting it to work:

The success of this idle hoax, done in time of war, when more serious writing was impossible, vastly astonished me. It was taken gravely by a great many other newspapers, and presently made its way into medical literature and into standard reference books. It had, of course, no truth in it whatsoever, and I more than once confessed publicly that it was only a jocosity … Scarcely a month goes by that I do not find the substance of it reprinted, not as foolishness but as fact, and not only in newspapers but in official documents and other works of the highest pretensions.

Here I have to correct my own essay, because the sentence about medical literature and reference books is the load-bearing one and Mencken is the only witness for it. Every claim about how far the hoax travelled before 1926 — the learned journals, the works of reference — traces back to his own 1926 and 1949 accounts of himself. I could not find one dated pre-retraction encyclopedia entry. The man who invented a history of the bathtub is the sole source for the history getting into the history books.

What is documented is everything after the front page. Hans Zinsser put it in Rats, Lice and History in 1935, a bestseller. A former New York City health commissioner broadcast it in a radio address in 1936. Mencken’s own paper, the Baltimore Sun, printed it in 1942. President Truman told it in a Philadelphia speech in 1952. A Washington Post travel column ran it in February 2004.

That is the stronger claim and the one I can actually stand on: not that a falsehood outran a correction that never quite caught it, but that it kept going for seventy-eight more years with the correction on record, on a front page, from the author, twice.

The reason is not that people are gullible. It is that the two things travel by different roads. The false history is a good story that improves any paragraph about hygiene, presidents, or American progress. The retraction is an anti-anecdote: it is only useful to someone who already has the false version and is about to use it. It has no reason to be forwarded, quoted, or filed. A correction can only reach a reader who is already holding the error, and by the time they are holding it, they got it from somewhere that did not carry the correction.

Worse, once it is in a reference book, the citation loop closes. The next writer does check. They check the encyclopedia, and the encyclopedia checked the newspaper, and the newspaper checked nothing. Three sources agreeing can be one source repeated, and from the inside those two situations are identical.

What it costs to notice

I keep a list of my own errors, and the pattern in it is not that I believed outrageous things. It is that I believed things wearing the costume above: a figure that was specific, a source that looked institutional, a number that arrived already rounded to a plausible-sounding precision. The question that actually works is not how many places say this but what does this figure cite. Repetition is not corroboration. Repetition is the failure mode that looks exactly like corroboration.

Mencken’s joke is the cleanest experiment anyone accidentally ran on this. Every condition favoured the truth. The author was named, alive, prolific, and willing to confess. The claim was about a domestic object anyone could have questioned. The falsehood had no constituency, no money behind it, nobody who benefited.

It won anyway, to 2004 at least, because it was better shaped for travel.

Two honest endings, and I owe both. The first: the true version is why the fake one worked. The White House had piped water and a bathing room under Andrew Jackson in 1833, and hot-and-cold fixtures under Pierce in 1853; Fillmore’s real plumbing footnote is a toilet. Mencken did not invent a bathtub out of nothing, he wrote a slightly wrong version of a real timeline, which is the hardest kind to feel wrong.

The second: nearly every account of this hoax, mine included, draws its list of repetitions from one hub — the Museum of Hoaxes. Several sources agreeing may be one source paraphrased back, which is precisely what this essay is about, and I am not outside it. And by now the hoax is better known as a hoax than as history. The correction did win. It took about seventy years, which is the actual measurement, and the reason to write the number down rather than the word.