Stripping a whole field at Golden Rock reveals a village plan

On St Eustatius, a small island in the northern Lesser Antilles, A. H. Versteeg and K. Schinkel stripped broad areas of ground instead of digging the usual scatter of small test pits. What came up was a village plan. They could see where the large communal houses stood, how the ground between them was used, and which corners of the settlement were given over to particular tasks. None of that layout is visible from a hole a metre square. A test pit tells you that people lived somewhere and what they threw away; opening a whole field tells you how they arranged themselves around each other. The Golden Rock results became the reference plan for domestic architecture in the Troumassoid period, and other projects across the region changed their methods to match. The lesson is not only archaeological. Ask a small question of the ground and it returns a small answer, and the small answer then gets written up as the limit of what those people were doing.

Ghana began around the year 300 under princes described in Arab accounts as white, and their line held power for about seven centuries before Soninke rulers, who were black, took the throne. The new dynasty pushed its conquests into the desert and subdued Berber populations there. By the year 1000 the Arab geographers al Bakri and al Idrisi were describing a state of real political weight and material wealth. Ghana lived on its position between north and south. It was the depot for salt cut from the mines of Tatental in the Sahara, a substance so scarce in central Africa that it sometimes served as money alongside gold and silver, and it was the great market for gold dust and nuggets carried up from the fields south of the Senegal river. That wealth drew raiders. Abu Bakr ibn Umar, the Almoravid ruler who had taken southern Morocco and the whole Sahara, invaded the empire and destroyed it.

Portuguese speakers built their sorcery vocabulary out of a Latin root meaning manufactured or artificial. A feitico was a made thing, a charm or a spell, and feiticaria was the practice of making them. The earliest surviving records of both words are legal, in anti sorcery legislation of 1385 and 1403. The root carries two ideas at once. One is doubt over whether the thing is real or only constructed to deceive. The other is seduction, the swaying of another person against their will. In the late middle ages Portuguese speakers made protective charms from broken mirrors and pieces of hangman's rope. Three centuries afterwards, in Bahia, the same word was applied to objects and illnesses that Africans described in entirely different terms, and the Lisbon Inquisition tried people for feiticaria on the strength of it. Colonial officials used feitico and veneno almost interchangeably, which is how an illness became a poisoning.

Through the fifteenth century Portuguese and Spanish builders produced ships made for the open ocean, and their navigators learned to use the compass and to read the Atlantic's currents and prevailing winds well enough to sail out and get home again. That combination opened the Atlantic islands. Madeira, the Canaries and Cape Verde were reached and settled. Italian merchants, who had financed and run sugar around the Mediterranean for generations, worked with the Iberian colonisers to establish the crop on those islands, grown by enslaved people. The plantation of the Americas was assembled there first, in miniature and offshore. The ships did something else as well. Until then the winds and the distances had kept the Western Hemisphere out of contact with Africa and Eurasia. Once the route was solved that barrier was gone for good, and everything that followed for the people of the Caribbean followed from a set of technical problems being solved in Lisbon and in Seville.