A second copper industry was invented in the mountains of Niger

In the far north of Africa metallurgy followed the sequence textbooks expect. Copper arrived in Egypt in the fourth millennium BCE, as native metal from deposits in Sinai, first for ornament and then in earnest as Egypt entered the Bronze Age early in the third millennium. Slightly later, and two thousand five hundred kilometres away in the Aïr mountains of present day Niger, another people started again from nothing. They most likely spoke a Nilo-Saharan language of the western Sahelian group. The archaeologist Augustin Holl reads the earliest Aïr copper sites, running from 2500 to 1500 BCE, as a long period of experiment with furnace designs, settling by 1500 BCE into a mature and standard technology. Copper then went into trade. Work by the historian Constanze Weise on borrowed vocabulary puts Aïr copper in northwestern Nigeria before 1000 BCE, and by the ninth century BCE the smelting technique itself had travelled a thousand kilometres further west into southern Mauritania.

Vignier III sits near the coast of Haiti and was occupied around 3630 BCE by Casimiroid people, among the earliest populations known in the Greater Antilles. What survives of them is mostly shell. The deposit is thick with it, which is the ordinary signature of a community taking a large part of its food from the shoreline, from the shallows and the reef rather than from farmland. That is a workable diet in this archipelago and it supported people here for thousands of years. I put a shell heap on a Haitian timeline for one reason. The country is usually introduced through 1804, or through 1492 if the account reaches back that far. Occupation of this ground runs about five and a half thousand years, and the first people to live on it were feeding themselves from the same water that later carried the ships. Haiti has a prehistory, and it has a physical address.

Phoenician traders founded Utica on the North African shore in 1100 BCE, and by that date their colonies were already scattered along the coast of what ancient writers called Libya. Herodotus described the people of that coast as a confederation of families of several origins. The populations there had formed through the mixing of indigenous black tribes with tribes that came in from Asia, so the northern edge of Africa was a meeting ground long before Rome or Islam reached it. This belongs on a Haitian timeline because Haitians descend from peoples whose history was declared empty by the men who sold them. A coast with dated cities, named traders and recorded confederations is the opposite of a blank. The Africa that supplied captives to Saint Domingue two and a half thousand years later was never a continent without a record, and the Phoenician settlements are among the earliest fixed points anyone can put on its map.

Omotic languages are spoken today across southern Ethiopia. They were once spoken much further north. The evidence is a set of Omotic loanwords sitting inside the Agaw languages of northern Ethiopia, which can only have entered by contact, and the contact has to predate 2000 BCE. That puts Omotic speakers across a large stretch of the Ethiopian Highlands four thousand years ago, well beyond the territory their descendants hold. What moved them was the expansion of Cushitic speaking peoples, which pushed the Omotic communities southward over the following millennia until the map came to look as it does now. The words stayed behind. Borrowed vocabulary is durable in a way that settlements are not, and a handful of Omotic terms embedded in Agaw speech is the surviving trace of a linguistic geography that has otherwise been erased. Ehret works this way throughout, on the principle that languages remember migrations which left nothing at all in the ground.