Girard reads the papers of a single estate, the Bréda plantation at Haut-du-Cap, and finds that flight from it was roughly ten times more common than the runaway notices in the colonial press would indicate.

The gap is widest for collective departures, the temporary walkouts long dismissed as something less than real marronage. These show up neither in the newspapers nor in the plantation's accounting records, yet on this estate they were both the most frequent form of flight and the one most likely to succeed.

If other plantation studies turn up the same pattern, Girard suggests, the long-running Debien-Fouchard argument about the links between marronage and the revolt of 1791 would need rethinking. His proposal is to set aside the division between small and large marronage and look instead at group flight, whose revolutionary potential has been undervalued.

Girard opens on an 1802 Guadeloupe census of hundreds of people described as belonging to the State: legally free under the abolition of 1794, yet catalogued with a price beside each name, called negre while planters were addressed as citizens, and reenslaved within months. The document poses his question. When France had outlawed slavery and the trade in captives, what should one call the traffic that plainly continued?

The traffic persisted in substance while the word was avoided. Ships under French captains supplied Saint Domingue out of Kingston; exiles took the people they owned overseas instead of freeing them; cultivators were kidnapped and sold in Santo Domingo; privateers landed recaptured Africans in Guadeloupe; officials shipped prisoners of war to Central America and the United States.

To sort cases Girard proposes three tests, asking whether a movement was voluntary, whether unpaid forced labour followed, and whether profit was the motive, and reads the era's shifting vocabulary as both cover and evidence.

Haitians honour Jean-Jacques Dessalines as a founder devoted to his countrymen's freedom; white contemporaries called him a butcher. Girard sets both aside for a ruler who took much of his outlook and policy from the Atlantic plantation system he grew up inside.

Archival work in France, Britain and the United States yields an uncomfortable record. Dessalines urged France to arrest fellow black revolutionaries, hesitated for years over breaking with France, and kept close relations with some white Frenchmen even after the massacres of 1804. He negotiated widely with neighbouring states to hold on to colonial trade links, worked to keep the sugar plantations whole though he had himself been enslaved — probably by Toussaint Louverture's son-in-law — imposed a strict feudal order on former slaves, tried to import African labourers, and drew on American and European as well as African sources.

Interest, in Girard's reading, drove Dessalines as much as the race and ideology that dominate the historiography.

Joseph Bunel served Toussaint Louverture and then Dessalines as a diplomatic and commercial envoy; Marie Bunel, born Fanchette Estève, spent her adult life trading in Cap Français and later Philadelphia. Girard reconstructs both careers from French, British and American archives.

He was a white Frenchman, she a free black Creole, but the argument is that ambition and money shaped their paths more than race did. After years there as a merchant and plantation manager, Joseph took an administrative role under Louverture from 1798, negotiating terms of trade and non-aggression with Britain and the Americans while the Quasi-War ran, before serving as paymaster. That proximity to the regime got him deported when the Leclerc expedition arrived.

Marie, already prosperous before 1791, kept trading straight through the revolution; closeness to Louverture and Henri Christophe never became political engagement. Persecuted with her husband, she rebuilt her business in Philadelphia and returned to Haiti only in 1810, after which the record thins out.

Every 18 May Haiti marks the day Dessalines cut the white band out of the tricolor and Catherine Flon stitched what remained into a flag. Girard finds almost nothing in the archives to support the scene, and reason to doubt even its date and location.

Madiou, writing in 1847, gave the first detailed version and placed the design earlier, prompted by Petion after the French captured a rebel color and published a pamphlet claiming the insurgents had no thought of independence. Ardouin, more skeptical, kept the order and dropped the swordplay. Later writers preferred the theatrical telling and added to it, Flon included, in ways that track whether the author wished to credit Dessalines or Petion.

The archival record yields a less expected finding. Haiti's flag and its other emblems borrow heavily from French revolutionary iconography, and through most of 1803 the leadership presented itself as the defender of emancipation rather than as the enemy of white people.

Toussaint Louverture governed Saint-Domingue as a virtually independent ruler from late 1798 until early 1802, yet his record as a statesman has drawn far less attention than the slave revolt that raised him. Girard works through British, French, Spanish, Cuban and American archives to argue that a literature built on United States policy papers has left Louverture receiving other people's decisions rather than pursuing designs of his own.

Those designs were neither the export of revolt nor a straight march to independence. He declined to attack the slave colonies around him, courted London and Washington for trade and for the arms he needed against Rigaud in the south, and stopped short of severing the tie to France. A planter and slave owner himself before 1791, he bound the survival of emancipation to his own survival in office.

The four-way manoeuvring that resulted, among Britain, France, the United States and Saint-Domingue, held for years before collapsing in 1802.

Girard examines what white, black and mulatto women did during the final two years of the war that ended French rule, 1802 to 1804. The assumption that wartime narrowed women's part in events does not survive the evidence.

Desirable women served as tokens of a man's standing in colonial society, and some traded on that position to extract political favors. Women of color contributed through logistics, espionage and at times combat; those convicted of assisting the rebel army were put to death for it.

White women occupied a different place. Because they were taken to be essential to the colonial order itself, the forces under Toussaint Louverture and Jean-Jacques Dessalines made them targets alongside other civilians, and after the rebel victory the surviving white population was killed.

Girard centers his account on Jean-Bertrand Aristide's political career, the calculations he made, the concessions he accepted, and the long negotiation he conducted with the Clinton administration. Rather than place one leader in the right, Girard weighs what each man needed and the moral claims each pressed against the other.

The method gives both sides substance. Aristide and Clinton appear with their constituencies behind them, and the bargaining between them reads as consequential rather than procedural. At the book's core sit the confrontation itself and the foreign-policy machinery running between Haiti and the United States.

Around that core Girard also sketches Haiti as it stood at the time of writing, setting out with unusual clarity the tensions that built toward the violence of recent years and Aristide's removal from office.