Walter Gehring's lab showed in 1995 (Halder, Callaerts, Gehring, Science) that deleting Pax6 deletes a fly's eyes, and that expressing it on a leg, wing, or antenna grows an ectopic compound eye there, complete and functional. They then induced eyes using the mouse Pax6 gene: a vertebrate switch correctly building an insect eye across half a billion years of divergence. Gehring's 1996 Genes to Cells review proposed the monophyletic hypothesis, all eyes descend from one prototype organ, with squid Pax6 as further support, and mapped the cascade of roughly 2,500 genes the single switch commands.
The folk arc: the eye was creationism's flagship, Paley's watchmaker, irreducible complexity, and Darwin's own confession that the eye evolving by natural selection seemed, quote, absurd in the highest possible degree. The quote-miners always stop there; Darwin's next paragraphs argue gradual evolution from light-sensitive spot to camera eye. Pax6 supplied the receipt: one toolkit, remixed into every optical design on Earth. Gehring liked to close lectures by noting Darwin was right again.
The claim, run in both directions: design advocates take a single conserved eye-switch as evidence of a single designer; the stranger reading takes it as evidence that biology is modular beyond any design story, one gene, and eyes grow on legs. Same photographs, opposite mythologies.
If one gene builds every eye, did the eye evolve once, and does the most famous anti-evolution argument collapse?