A Jamaican village settled in a swamp and lived off clams

Sweetwater in Jamaica stands on a dune inside hardwood forest with swamp around it, which is not where Meillacoid people usually built. The pottery is Meillacoid throughout, mostly plain, and the few decorated pieces match the Montego Bay style found elsewhere on the island. The food remains set it apart. Where the nearby site at Paradise ate more widely, the deposit here is dominated by two bivalves, Lucina pectinata and Codakia orbicularis, animals of the muddy low oxygen ground the surrounding swamp provides. The tools follow the diet. Scrapers made from bivalve shell are the commonest implement recovered, and they are missing altogether from contemporary sites such as Bluefields Bay. So a community placed itself in difficult wet country, worked out what could be taken from it, and built the toolkit that job required. Sites a few kilometres apart on one island were living in different ways at the same moment.

Antam Goncalves sailed to western Africa in 1441 and opened his expedition by seizing a single man and a single woman. A companion told him that by carrying off those two captives he would allow Prince Henry to learn something about these people. The chronicle of the voyage was written by Gomes Eannes de Zurara, who returned again and again to the pleasure the prince took in each arrival of captives and to the hope of taking more. The pair matter because of what they were for. A man and a woman together do not make a sample. They make a demonstration that the coast could supply people who would go on producing more people, and that an owner's holdings could grow through the reproduction of the people he owned. The Portuguese traffic in African captives that opens here reached the Caribbean within a lifetime and eventually delivered most of the ancestors of Haiti.

Morne Cybele occupies La Montagne, the steep limestone plateau that rises to 275 metres on La Desirade. The occupied core is small, somewhere between 200 and 300 square metres. The height buys shelter from the strong currents and unpredictable winds that make the surrounding water difficult, and the northern cliffs make the site hard to approach from that side, which cuts both ways for the people living on it. A shell of Cittarium pica taken from the deposit was radiocarbon dated to the middle of the fifteenth century, so the community was in place a few decades before Columbus sailed. That date is what makes the site useful. It shows what people in the eastern Caribbean were doing immediately before the arrival that would define everything written about them afterwards. Morne Souffleur, on the same plateau, belongs to the same pattern. Small, high, difficult to reach, and chosen on purpose.

Morne Souffleur was found by walking La Desirade systematically rather than by chance, and what the survey turned up closely resembles Morne Cybele on the same high plateau. Both sites sit where the elevation shelters them from the currents, both are protected by the steep cliffs along the northern shore, and both look out over the channels canoes had to use. One site of that kind can be put down to local preference. Two of them, matching each other, make a pattern, and the pattern points to a deliberate move onto marginal ground late in the pre-Columbian period. The people who lived there had a view of everything passing to the north and east of Guadeloupe. Within a few decades of their settling, ships of a kind nobody on that plateau had seen would come up the same channel. The vantage they chose for their own reasons faced the direction the change arrived from.