Jo Nagai of Kobe, raising Asian swallowtails, Papilio xuthus, since early childhood, noticed released butterflies returning to him and asked whether they remembered being caterpillars. He wrote to Martha Weiss, the Georgetown entomologist whose 2008 study had shown the effect in moths, and over two years adapted her protocol for home use: lavender oil as the odor, a mild pulse from a muscle-therapy device as the shock, comfort monitored, a Y-maze for the test. The documented results: trained caterpillars became butterflies that overwhelmingly avoided lavender, roughly 80 percent, the first demonstration of the effect in butterflies, and the offspring of trained butterflies, never shocked themselves, also avoided it. He presented the 33-page report at the International Congress of Entomology, Kobe, 2024, where Weiss flew in to meet him, and Radiolab's Signal Hill podcast documented the collaboration as Caterpillar Roadshow.
The folk core of the story is the observation that started it: hand-raised butterflies returning to the boy who raised them, the kind of datum the field would file as anecdote and the child filed as a hypothesis. The tradition's own moral is the method's humility: he asked the professor instead of concluding, simplified instead of exaggerating, used the gentler stimulus, and translated his own report, the whole project being the archive's favorite kind of experiment, one where the rigor and the kindness are the same variable. Weiss's documented verdict: he's a real scientist, and he's figuring out new stuff.
The claim: the experiment happened because the question sounds naive, does the butterfly remember the caterpillar, and a hundred years of entomology mostly left it to moths and to a single Georgetown lab, the field's maturity functioning as its own filter, the archive's Reducing Valve applied to a discipline: the professionals knew the soup model, so the question went unasked where it mattered, until a child who did not know what was impossible ran it in his house. Verdict: the archive's verdict is the collaboration's own: the child had the question, the professor had the protocol, and the union of the two is the entire history of science in one correspondence. The butterflies remembered. The offspring remembered. The field, per the congress, applauded. The archive files the finding and the lesson together: the veil between caterpillar and butterfly leaks, the veil between generations leaks, and the veil between amateur and professional, per the postage records, was never real. Write the letter.
If a ten-year-old's home rig settled a hundred-year-old question, what is the laboratory actually for, and what else is waiting on a curious child?