Who were we before anyone wrote it down?

Fossils from Jebel Irhoud in Morocco, dated to around 300,000 years ago, show a mix of modern facial features with more archaic braincase shape. Their describers read them, with finds across the continent, as evidence for a pan-African origin: an emergence of our species that involved the whole continent rather than one region of it. A 2023 genomic model of African population history points the same way, while stressing how uncertain the specific models still are.

The Neanderthal genome, first sequenced in draft in 2010, showed that people with ancestry outside sub-Saharan Africa carry a small percentage of Neanderthal DNA. A finger bone from Denisova Cave then revealed a second archaic population, previously unknown, which contributed ancestry to living people in Oceania and, later work found, in parts of eastern Asia.

The Denisovans were identified from genetics before anyone knew what they looked like. They were first described from their DNA, as the Neanderthals had been from their bones.

Homo naledi, described from a South African cave system in 2015, has a mix of primitive and derived features and is far younger than its anatomy suggested. Homo floresiensis, a small-bodied species on Flores, and Homo luzonensis in the Philippines, named in 2019, both lived until relatively recently. When our own species appeared, and for a long time after, it shared the world with other kinds of human.

Pigment use, ornament, bone tools and blade technology appear in the African record tens of thousands of years before the traditional markers of behavioural modernity in Europe. Whether there was a sudden cognitive transition at all, or a gradual assembly that looked sudden because Europe was studied first, is debated.

Several genetic studies report contributions from “ghost” populations, groups detectable only as statistical signals in the genomes of living people, with no fossil yet assigned to them. One finds that four West African populations carry 2 to 19 per cent of their ancestry from a lineage that split off before Neanderthals and modern humans did. A 2023 analysis argues that a weakly structured stem, two or more African populations linked by gene flow, explains the same patterns without introgression from a deeply divergent archaic group, so whether that lineage existed is itself disputed.

Burial, pigment and ornament survive. Belief, language, kinship and law do not. Whether Neanderthals used symbolic language in anything like our sense is unresolved and may be permanently unresolvable with available evidence.

Every other human lineage ended as a population of its own. Environmental pressure, competition for the same niche and simple demographic chance have all been proposed, and the documented interbreeding adds absorption. The evidence does not currently distinguish between them, and the answer may differ for each population.

Given the documented interbreeding, the disappearance of Neanderthals and Denisovans may have been partly a merger rather than an extinction: small populations absorbed into a larger expanding one, leaving genetic traces but no distinct lineage.

Two lines of work reach the outcome without one. A neutral model, offered as a null hypothesis rather than a finding, gets the replacement from repeated migration out of Africa and species drift alone, with no selective advantage anywhere in it, and over a wide range of parameters it cannot be rejected. Separately, a model of how skills accumulate finds that regional population density, rather than any increase in cognitive capacity, can explain when complex behaviour first appears in the record, which is an argument that larger and better-connected groups keep and pass on more of what they learn.