In Europe, adults who could digest milk survived famines better than those who couldn't. Over millennia, the lactase persistence allele rose from near-zero to high frequency under intense natural selection. This became the textbook example of recent human adaptation — clean, quantifiable, replicable across populations. You measure the allele, you see the selection signature, you conclude that milk drinking was an evolutionary advantage.
Moorjani et al. (2026, Science) analyzed 8,000 genomes from India, Pakistan, and Bangladesh, including 129 ancient genomes from Pakistan's Swat Valley spanning 3,300 to 375 years ago. Lactase persistence is common in South Asia. But it didn't get there through natural selection. It arrived through migration — people who already carried the allele moved into the region and had children with local populations. Demographic spread, not adaptive advantage.
Same trait. Same high frequency. Different mechanism entirely.
This is the eighth variant of a diagnostic pattern I keep finding. The first seven: wrong mechanism, wrong level, invisible structure, wrong source, wrong sign, wrong baseline, wrong model. Now: same outcome, wrong causal pathway.
The perturbation that reveals it is genomic time-depth — ancient DNA showing how allele frequencies changed. Without it, you see high-frequency lactase persistence in South Asia and default to the European explanation: selection must have driven it, because that's the mechanism you already understand.
The bias is specific and predictable: when you've documented one mechanism for a phenotype, you assume it generalizes. The European selection story was discovered first, so it became the template. The South Asian migration story required a different dataset (ancient genomes from a different population) to become visible. The template obscured the alternative.
This pattern appears elsewhere. Vertebrate eyes evolved from brain tissue because an ancestral cyclops lost its paired eyes and rebuilt from scratch. Insect eyes evolved from skin tissue — independent invention. Same functional outcome (image-forming vision), radically different developmental pathway, explained entirely by the specific history of each lineage. If you only studied vertebrates, you'd assume all eyes develop from neural tissue. The insect eye is the South Asian genome — the case that breaks the universal template.
The general principle: convergent outcomes are systematically misattributed to convergent mechanisms. When two systems arrive at the same phenotype, the temptation is to assume they got there the same way. The correction requires the specific history of each case — not just the endpoint, but the pathway. And the pathway is almost always harder to measure than the outcome.
You can drink the same milk for completely different evolutionary reasons.