Oliver studies a class of portable sacred objects — cemís — that the native peoples of the Greater Antilles treated as living, potent beings rather than as representations. His question is what their scatter across the islands reveals about the ties among the people who owned, inherited, gave away and seized them.

He concentrates on large three-pointed stones, stone heads, stone collars, elbow stones and the face-masks called guaízas, most of them now in museum collections, and reads them against sixteenth-century Spanish accounts, above all the report Ramón Pané compiled for Columbus. Theories of gift exchange and of personhood, drawn largely from Melanesian ethnography, supply the framework.

Along the way Oliver revises the cultural scheme inherited from Irving Rouse and reconsiders what the word Taíno can properly be made to carry, before turning to the struggles over these objects in Cuba, Boriquén and Hispaniola.

Rouse sets out to restore a name and a history to the people Columbus met. They had no collective name for themselves, calling themselves after the islands they lived on; ethnohistorians settled on Taino. He sorts them into Classic groups in Hispaniola and Puerto Rico, Western groups in Cuba, Jamaica and the Bahamas, and Eastern groups in the small islands, arguing that populations at the edges stay conservative and shade into their neighbours.

Against the older view that these were recent migrants from the far headwaters of the Amazon, he brings four bodies of evidence — the Spanish chronicles, archaeology, historical linguistics and biology — to a single conclusion. The ancestors did come from South America, traceable to the Orinoco valley, but the Taino themselves took shape within the archipelago over millennia that radiocarbon dating has made visible.

He also separates them sharply from the Island-Caribs, whose alleged cannibalism colonists inflated to license enslaving them.