Cotton weaving was invented at least three times with three different plants

Loom weaving is the mechanical line of technological descent, and cotton is where it shows most clearly at a global scale. The machine looms of modern industry evolved from methods that begin in the middle of the Holocene. People invented cotton weaving independently in at least three parts of the world and possibly four. The proof lies in the plants. Each set of inventors domesticated a cotton species native to their own region: Gossypium barbadense in Peru, Gossypium hirsutum in Central America and Mexico, and across India and Africa the two species Gossypium herbaceum and Gossypium arboreum. Nobody was carrying seed between these regions. The pattern is the one that already holds for ceramics and for metallurgy. Communities far apart, with no contact, set out on the same trajectory of invention thousands of years ago and arrived at the same transformative technology by their own routes, working with whatever grew where they lived.

The site at Banwari Trace in Trinidad holds one of the earliest human occupations known anywhere in the Caribbean, dating to around 5000 BCE. Archaeologists label the period the Lithic or early Archaic Age, and it marks the point at which mainland populations began moving out into the island chain. What survives is mostly stone tools and refuse, the residue of people who hunted and gathered. That is thin material, and it still answers real questions: how the first islanders handled environments none of their ancestors had seen, and how they adjusted as sea levels kept changing under them. Everything that follows in the region, thousands of years of indigenous settlement, agriculture and exchange across the whole archipelago, begins from these deposits. Banwari Trace is where the human record of the Caribbean starts, and it sits about six and a half thousand years before the arrival that most chronologies treat as the beginning.

Genetic work on donkeys points to two separate domestications, one among communities in Somaliland and another farther north along the Red Sea Hills. People in both places would have been speaking early languages of the Cushitic branch of Afrasian, and within that branch linguists can follow a word distinguishing the domestic donkey from the wild one back to roughly 5000 BCE. The linguistic date agrees with the genetic date, and both agree with the archaeology, where domestic donkeys show up in Egypt after 4000 BCE. Three independent lines of evidence converging is what makes a case of this depth solid. Two African communities, neighbours in the Horn, took the same animal in hand at about the same time and set off consequences that reached transport, trade, warfare and the organisation of states across the ancient world. The road from Somaliland to Sumer is one of the least told routes in world history.

One animal went out from Africa to the rest of the world and changed how goods move. Modern contempt for the donkey obscures the scale of that. Centuries of Western usage have turned the animal's name into a term of abuse for stupidity, and the abuse has done real damage to the record. The donkey was the first multipurpose beast of burden in history. It carried loads and it carried riders. In the Middle East people harnessed it early to wheeled vehicles. Long distance trade in goods depended on owning donkeys, and so did the first wheeled transport anywhere. The uses people found for it very likely supplied the model for domesticating and working horses and camels later on. Every chariot in Egypt, every camel caravan crossing Arabia, every cavalry charge in the ancient world descends conceptually from an animal that Africans brought under care thousands of years before a horse carried anyone.