Treponema pallidum has three subspecies. One causes syphilis. One causes yaws. One causes bejel. They are classified as separate diseases with separate clinical presentations, separate transmission routes, and separate geographic distributions. The classification implies that the three diseases are the natural categories — that treponemal disease comes in three kinds.
Majander, Guzmán-Solís, and colleagues (Science, January 2026) recovered a 5,500-year-old T. pallidum genome from a Middle Holocene hunter-gatherer skeleton at the Tequendama I rock shelter in Colombia. The ancient strain — designated TE1-3 — does not belong to any of the three modern subspecies. It sits outside all of them, a sister lineage to the entire group. It diverged from the common ancestor of modern T. pallidum approximately 13,700 years ago — more than twice as long ago as the three modern subspecies diverged from each other, which happened roughly 6,000 years ago.
The ancient strain is not an ancestor of syphilis. It is not an early form of yaws. It is a fourth thing — a lineage that split off before the categories we use to describe treponemal disease even existed. Whatever disease it caused, we don't have a name for it, because the disease no longer exists and the lineage that caused it left no modern descendants.
The structural observation: the categories syphilis, yaws, and bejel are recent. They represent the current survivors of a diversification event that occurred only six thousand years ago. Before that diversification, there was a different landscape of treponemal disease — one that included lineages we cannot reconstruct from living organisms because those lineages are extinct. The taxonomy we inherited doesn't describe the history of the pathogen. It describes the present state of the pathogen, viewed as though the present state is the only one there has ever been.
This is the paleontological version of a familiar problem. When we classify disease by its modern forms, we impose a snapshot on a process. The snapshot captures what survived. What didn't survive isn't absent from history — it's absent from the snapshot. The 5,500-year-old skeleton in Colombia is a window through the snapshot into the version that didn't make it.
The genome survived in bone because T. pallidum spreads through the bloodstream and infects skeletal tissue. The bone preserved the DNA. The disease the DNA caused left no other trace — no clinical tradition, no name, no historical record. The pathogen's strategy of infecting bone is what preserved the evidence; the pathogen's extinction is what erased everything else. The same mechanism that made the disease severe enough to leave skeletal evidence is what made recovery of the genome possible after the disease itself disappeared.
The deeper point is about the relationship between classification and survival. We have three categories of treponemal disease because three lineages survived. If the Colombian lineage had survived and yaws had gone extinct, we would have a different taxonomy — and we would call it natural. The number of categories is a count of survivors, not a count of kinds.