Seven Knives in the Back

Everything I described in the previous eight chapters may be wrong.

Wrong as a whole, from beginning to end: a beautiful, coherent, and convincing construction that has nothing to do with reality.

That’s a normal state of affairs — that’s how science works.

In 1847, the Hungarian doctor Ignaz Semmelweis discovered that if obstetricians washed their hands before delivering babies, mortality from puerperal fever dropped from 18% to 2%. He published the data, and his colleagues rejected it — for a reason that had nothing to do with the quality of the data itself: it followed from the data that doctors were killing their patients with dirty hands, and that was unbearably inconvenient. Semmelweis was fired, sent to a psychiatric hospital, and at the age of forty-seven died there from a beating by orderlies. Twenty years later Lister and Pasteur proved germ theory, hand-washing became standard, and Semmelweis was rehabilitated posthumously.

The “Semmelweis reflex” is the tendency to reject data that contradict what you already believe. Not out of stupidity, but out of self-preservation: if the data are right, then all your previous work is wrong, and no one wants that outcome.

I tell this story to remind you: a proof is data plus the willingness to see it. Rejection and acceptance are merely two additional hypotheses on top of the data, and the truth, in the end, is decided by an experiment. (I’m not about to compare myself to Semmelweis — arrogance isn’t my thing; besides, so far I’m still alive.)

Ninety percent of beautiful hypotheses are wrong, and the remaining ten are incomplete; both kinds are useful, because errors move science along no worse than discoveries. Ptolemy was wrong about the Sun, and his error drove astronomy for a thousand years; Newton was wrong about space, and his error still flies into space.

If I’m wrong, I want to be wrong productively. And for that one has to look honestly at what might not be right.

---

Objection one: pareidolia.

“You’re seeing patterns that aren’t there. The brain and the Universe ‘look alike’ — so what? Clouds look like elephants, toast like Jesus. Structural similarity isn’t yet proof of a shared mechanism.”

Answer. Fair — but only as it applies to visual similarity. Vazza and Feletti compared the mathematics: spectral analysis, graph topology — and in those quantities the brain and the Universe turned out to be more similar to each other than to other network systems (clouds, trees, turbulence). That is already data, and the game of clouds and elephants doesn’t apply here.

But. The data do not prove the cause of the similarity. Perhaps before us is a property of self-organization in general; perhaps a fundamental principle; we don’t know for sure. The honest answer is: the similarity is real, the explanation is still hypothetical.