“The first principle is that you must not fool yourself - and you are the easiest person to fool.”
Richard Feynman
“All models are wrong, but some are useful.”
George Box
In 1847, in the First Obstetrical Clinic of the Vienna General Hospital, women were dying of childbed fever at a rate that everyone could see and nobody could explain. A Hungarian doctor named Ignaz Semmelweis had the run of two maternity wards and he noticed something you weren’t supposed to notice. The ward run by doctors killed roughly one mother in ten. But the ward next door, run by midwives, killed only about one in twenty-five. It was the same building and same disease, but less than half the death rate, and the only difference was who did the delivering.
So naturally he chased it. The big break came when a colleague, Jakob Kolletschka, cut his finger during an autopsy and died of something that looked identical to childbed fever under the microscope. Semmelweis put it together: the doctors were walking from the dissection room to the delivery bed without washing, carrying what he called ‘cadaverous particles’ on their hands. He ordered handwashing in a chlorinated lime solution and mortality on the doctors’ ward dropped to match the midwives’ inside a couple of months, and he had the monthly figures on paper to prove it.
But the profession rejected him. And they did so with dark offence, instead of merely because they had devised a better experiment, which would have been fair. The suggestion that respectable physicians were the ones carrying death from bed to bed was an insult, and worse than an insult, it made no sense. This was two decades before Pasteur and Lister. There was no germ, no invisible agent riding on a hand, no framework in which a clean-looking hand could be a murder weapon. So the numbers were real, the explanation was correct, and yet both got thrown out together, because the frame had nowhere to file them. Semmelweis was pushed out, grew erratic, and in 1865 was committed to an asylum where guards beat him; he very ironically died two weeks later of an infected wound, which is a detail I have never been able to read without wincing. He was right the whole time. But being right bought him nothing, because the people who needed to follow the argument were carrying a picture of the world in which the argument could not be true.
That is the subject of this essay.

Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.