The Burden

In 1847, Ignaz Semmelweis reduced childbed fever mortality from ten percent to one percent by requiring doctors to wash their hands in chlorinated lime solution. He could not explain why it worked — germ theory had not been articulated — but the mortality data were unambiguous. The medical establishment rejected the finding. The implication that doctors' hands were killing their patients was intolerable. Semmelweis was dismissed from his position and died in an asylum in 1865.

The evidence was sufficient. The claim was too far from the model.

In 1912, Alfred Wegener proposed that continents move. He had coastline matching, fossil distribution across oceans, geological continuity between South America and Africa. He did not have a mechanism — he could not explain what force moves a continent. The evidence was rejected not because it was weak but because the claim required physics that did not yet exist. Wegener died on a Greenland expedition in 1930. Seafloor spreading provided the mechanism in 1962. Plate tectonics formalized the theory in 1965. Fifty years of available evidence, waiting for a framework that would make it admissible.

In 1948, Barbara McClintock published her discovery of transposable elements — genetic sequences that move position within the genome. She presented at Cold Spring Harbor. The audience did not understand. Jumping genes violated the fixed-genome model so completely that the finding was not rejected so much as ignored. "I was startled when I found they didn't understand it," she said later. Molecular biology confirmed transposons in every organism. The Nobel came in 1983, thirty-five years after the discovery.

In 1982, Stanley Prusiner proposed that certain diseases are caused by misfolded proteins with no nucleic acid. An infectious agent that violates the central dogma — self-replicating without DNA or RNA. He coined the term "prion" and was called a heretic. The resistance was partly principled: if a protein alone could transmit disease and replicate its structure, it undermined everything known about heredity and infection. The Nobel came in 1997. The heresy became a textbook chapter.

In 1984, Barry Marshall drank a petri dish of Helicobacter pylori to prove it caused gastric ulcers. He developed gastritis within days. The medical establishment had attributed ulcers to stress and excess acid for decades. Koch's postulates required demonstration in a human host, and no ethics board would approve deliberate infection. Marshall infected himself because the distance between his claim and the prevailing model was so large that only his own body could close it. The Nobel came in 2005.

The pattern is not that science resists new ideas. Every field has mechanisms for incorporating new findings — peer review, replication, conference presentation. The pattern is that the evidence threshold is not fixed. It rises with the distance between the claim and the accepted model.

A finding consistent with the current framework needs ordinary evidence: a well-designed experiment, clean data, peer review. A finding that extends the framework needs stronger evidence: replication, multiple independent confirmations, a plausible mechanism. A finding that contradicts the framework needs extraordinary evidence — and sometimes the only extraordinary evidence available is drinking the culture yourself.

This is not irrational. The more evidence supporting the current model, the more evidence required to overturn it. Sagan's razor cuts both ways: extraordinary claims require extraordinary evidence, which means the most important discoveries are the hardest to establish.

The paradox is temporal. The evidence required to prove a paradigm-shifting claim is often not available at the time the claim is made. Wegener had coastlines but not seafloor spreading. Semmelweis had mortality data but not germ theory. McClintock had genetic observations but not molecular biology. Each needed a future technology or framework to make their existing evidence legible. They were not wrong too soon. They were right in the wrong vocabulary.

And the threshold interacts with credibility in a way that compounds the problem. The longer you work on a claim the establishment rejects, the less credible you become. Semmelweis grew bitter and polemical. Wegener died in the field without vindication. McClintock retreated to her lab and worked in isolation. The social cost of maintaining an unpopular position erodes the standing needed to have the position heard. The threshold rises; the capacity to meet it falls.

Marshall solved the problem by making the evidence impossible to ignore. He put the claim in his own stomach. Prusiner solved it by outlasting the resistance — fifteen years of work before the evidence accumulated past the threshold. McClintock solved it by not solving it: she continued working, and the field eventually caught up. Wegener and Semmelweis did not solve it at all. They were vindicated posthumously, which is to say: the threshold fell after they were no longer present to clear it.

The same quality of evidence that would be celebrated if it confirmed the model is dismissed when it contradicts it. The evidence does not change. The burden does.

Nothing in this structure is fixable without breaking something else. Lower the threshold and you admit more false claims. Raise it and you delay more true ones. The asymmetry is not a bug in scientific practice. It is the cost of having a coherent model at all. Any framework that can organize knowledge must also resist disorganization. The resistance is the framework functioning correctly. It functions correctly against true claims and false ones with equal force, because it cannot tell them apart in advance.

The burden is not proof. Proof they had. The burden is the weight of the model that has to move before the proof can land.