A court calls Bunel sieur, and the title is the evidence of his race

The earliest document placing Joseph Bunel in the colony is a ruling ordering a merchant of Port-au-Prince to pay four thousand three hundred and seventy-two livres to a captain from Le Havre. Girard's interest is in one word. The honorific was normally reserved for white men, while men of colour were designated by a formula meaning the person called such and such, and the distinction persisted in legal papers even after racial inequality was abolished in 1792. In his 1807 will Bunel described himself as the legitimate son of a sieur and a dame, both terms reserved for whites. A law of 1773 required free people of colour to take African-sounding names, and Bunel is a French one. Add his profession, since plantation managers and merchants were overwhelmingly white, and his birth at Pont-Audemer in Normandy in March 1753 to parents of the minor provincial bourgeoisie, and Girard concludes the case is effectively settled. Where this account departs from the standard one: Against McCullough, Egerton, Wills, Levine and Santoro, Girard argues Bunel was white and that the claim he was the first man of African descent to dine with an American president rests on a chain of citations none of which states his race.

Pennsylvania freed any enslaved person whose owner brought them into the state and settled there for six months. Colonists arriving from Saint Domingue met the deadline and then signed the newly freed person to an indenture, until the age of twenty eight for a child, or seven years for anyone over twenty one. Most of the black refugees were teenagers, so the term ran for most of their working youth. Between 1791 and 1795 some 284 slaveholding colonists released 442 people, and nearly all went straight into indentures which the former owners then sold for cash to farmers who wanted labour cheaper than white wages. Of 659 people released from bondage in the city between 1789 and 1810, forty five walked away outright. One colonist tried to beat the clock by moving her slaves across the river to Burlington after five months and three weeks; the abolition society argued six lunar months of twenty eight days and won.

The first population figures Oostindie treats as reliable date from 1789 and give a total of almost twenty-one thousand. Of these, 12,864 were enslaved, 3,714 were free people of colour and 4,410 were white, the white group itself split by a census into 2,469 Protestants, 1,095 Sephardic Jews and 846 servants of unstated religion. Oostindie's point is that this was an atypical Caribbean colony: no tropical export staple, a free trade harbour, and an unusually large free non-white population. Roughly a hundred thousand Africans had been landed on Curaçao across the Dutch period, mostly between 1667 and 1730, with the last imports in the 1770s, so the enslaved population was already heavily locally born. By the time slavery ended in the Dutch West Indies in 1863 the enslaved number had fallen to 5,498.

The engravings of the revolution are remarkable for their narrowness. Where the visual record of the French and American revolutions follows events year by year and sometimes month by month, that of the Haitian Revolution fixes on two moments at Cap Francais, the rising of 1791 and the fighting of 1793, and then jumps to the independence campaign of the early 1800s. Many prints were made in Europe by people who had never crossed the Atlantic, whose aims ranged from selling an arresting picture to damping enthusiasm for abolition to attacking the French revolutionary government. As a group they show black violence against white residents and property and pass over the brutality of white men toward people of African descent. White draws the political consequence. Readers in Europe and America learned of abolition and of black troops defeating British invaders in text, but what they saw was the port on fire. Where this account departs from the standard one: White argues the Haitian Revolution was made into a race war partly by objects, against accounts that treat the racial framing as a straightforward description of events.