When Did Humans Reach the Americas?

For most of the twentieth century, American archaeology ran on a single confident number: 13,000 years. That was when the makers of Clovis points — distinctive fluted stone spearheads found across North America — supposedly walked south through an ice-free corridor between the retreating Laurentide and Cordilleran ice sheets, becoming the first Americans.

“Clovis first” was more than a hypothesis; it was a gatekeeping standard. Claims of older sites were routinely rejected, and careers suffered for making them. Today that model is finished, and the current picture is both older and stranger.

How Clovis-first collapsed

The decisive blow was Monte Verde in southern Chile, excavated by Tom Dillehay from 1977. Waterlogged peat preserved wooden tent stakes, hide fragments, seeds, medicinal plants and a child’s footprint — a settlement dated to about 14,500 years ago. And it is 16,000 km south of the Bering Strait. Whoever lived there had ancestors who arrived considerably earlier.

Monte Verde was attacked for two decades. In 1997 a delegation of prominent sceptics visited the site and publicly accepted it. Once one pre-Clovis site stood, others followed: Paisley Caves in Oregon, where human coprolites containing human mitochondrial DNA date to 14,300 years; Page-Ladson in a Florida sinkhole, where a mastodon tusk shows butchery marks alongside a stone knife at 14,550 years; Cooper’s Ferry in Idaho, with stemmed points and hearths at about 16,000.

White Sands: the footprints that reset the clock

In 2021, a team led by Matthew Bennett reported human footprints in the dry lake bed of White Sands National Park, New Mexico — adults and children walking along the shore of ancient Lake Otero, in some places alongside mammoth and giant ground sloth tracks. Radiocarbon dates on seeds of the aquatic plant Ruppia cirrhosa in the track-bearing layers gave ages of 23,000 to 21,000 years.

That is not pre-Clovis. That is before the Last Glacial Maximum, when the ice sheets were at their fullest extent and the ice-free corridor did not exist.

The criticism was immediate and legitimate: aquatic plants can take up dissolved ancient carbon from groundwater, making them appear older than they are — the “hard water” or reservoir effect. So the team went back and dated the same layers two more ways: radiocarbon on terrestrial pine pollen, which cannot absorb old dissolved carbon, and optically stimulated luminescence on quartz grains, which measures when a grain last saw sunlight. The pollen returned 23,000–21,000 years. The OSL returned 21,400–18,000. Three independent methods, converging. The 2023 paper in Science is titled, pointedly, “Independent age estimates resolve the controversy.”

What the genetics say

DNA tells a compatible but not identical story. Native American ancestors split from their closest Siberian relatives around 23,000–25,000 years ago, then spent thousands of years genetically isolated — the Beringian Standstill, a period of separation in or near the now-submerged land bridge of Beringia. The population then split into two branches, sometimes called Ancestral Native American northern and southern lineages, roughly 16,000–15,000 years ago, before spreading rapidly south.

Two individual genomes anchor this. USR1, an infant from Upward Sun River, Alaska, dated 11,500 years, belongs to a previously unknown “Ancient Beringian” population that stayed north. Anzick-1, a 12,800-year-old Clovis-associated infant burial from Montana, is directly ancestral to a large share of Indigenous peoples in Central and South America.

Genetics and White Sands are not in flat contradiction, but they are in tension: if people were in New Mexico at 22,000 years ago, either the founding split was earlier than the genomes suggest, or an earlier group left footprints and no surviving descendants. Both are live possibilities.

Which route?

The ice-free corridor has fallen out of favour. Sediment cores studied in 2016 showed that although the corridor opened geologically around 15,000–14,000 years ago, it was a barren, lake-choked landscape with no bison, no vegetation and nothing to eat until roughly 12,600 years ago — too late to be the first route south.

The leading alternative is the coastal route, sometimes called the “kelp highway”: boats and shoreline foraging down the Pacific margin, exploiting kelp forest ecosystems rich in fish, shellfish and marine mammals. It explains the early dates in Chile and the speed of the southward spread. Its central problem is that the relevant coastline is now under 100–120 metres of water, drowned by post-glacial sea level rise, taking most of the evidence with it.

Still-open questions

• Chiquihuite Cave. A Mexican site claimed in 2020 to hold stone tools at 26,500 years. Sceptics argue the “tools” are naturally fractured rock. Unresolved.
• Population Y. Some Amazonian groups carry a faint genetic signal with affinities to Australasian populations. Its source has never been explained.
• Where are the artefacts? White Sands has footprints but no stone tools. A 22,000-year-old occupation should have left camps. The search for them is the central problem in the field right now.

Where it stands

The defensible summary today: humans were in the Americas by at least 16,000 years ago on the strength of well-accepted archaeology, and very probably by 21,000–23,000 years ago on the strength of the White Sands footprints and their independent dating. The Clovis culture is not the first American culture — it is a later, spectacularly successful continental phenomenon that spread across an already-inhabited land.