Portuguese Empire

The legal claim came first. The Treaty of Tordesillas, signed on 7 June 1494 and ratified by Castile at Arévalo on 2 July, moved the partition of the non-European world 370 leagues west of the Cape Verde Islands, replacing the papal grants of the previous year with a negotiated meridian. Its purpose was to settle a quarrel about the Atlantic, not about Brazil, which no European had yet seen. Six years later, on 22 April 1500, a fleet of thirteen ships and some 1,200 men under Pedro Álvares Cabral, bound for India by way of the Cape of Good Hope, made landfall on the coast of what is now Bahia. The coast lay on the Portuguese side of the line. Everything that followed rested on that accident of geometry.

What the Portuguese held, for the first thirty years, was not a territory but a shore. They logged brazilwood, left a few convicts behind, built no towns, and lost ground commercially to French and Norman traders who came for the same red dye. Settlement began in the 1530s, and the colony that grew from it was never a single enterprise: it was a set of plantations, mission villages, cattle ranches, mining camps and frontier settlements, held together by a crown that was small, poor in people, and usually far away.

The map year of 1732 shows this America at its colonial height. Gold from Minas Gerais, Goiás and Mato Grosso had pulled settlement far to the west of the 1494 line; the Amazon was tied to Belém; Colónia do Sacramento faced the Spanish bank of the Plate. It does not show the Portuguese empire. Goa, Macau, Mozambique, Angola and the Atlantic islands were older, in places richer, and always more central to the crown's idea of itself. This entry is about the American holding only, and the map's borders are the borders of that holding.

At a glance

Why did Portugal claim a coast it had barely seen?

The Treaty of Tordesillas was a compromise between two crowns that both wanted the Atlantic and could not afford a war over it. Alexander VI's bulls of 1493 had given Castile everything west of a line 100 leagues from the Azores; John II of Portugal refused to accept it, argued the case for a year, and came away with a meridian 370 leagues west of Cape Verde. Neither side could measure longitude at sea, and the treaty said nothing useful about which cape, or which of the Cape Verde islands, the measurement started from. The line was never surveyed. Each crown drew it where it suited, and the practical question — where it crossed the South American coast — stayed open for two centuries and was still being argued in the nineteenth.

On most modern reckonings the meridian falls between about 46° and 50° west. A line at the western end of that range would cross the coast near the Oiapoque river in the north and again at the Gulf of San Matías in southern Argentina; at the eastern end it passes close to São Vicente, the site of the first permanent Portuguese settlement. Either way the whole shoulder of Brazil — the bulge from Maranhão to Espírito Santo — lay in the Portuguese half, and the claim needed no conquest to justify it.

Cabral's landfall is usually described as an accident of wind, and the fleet's own log supports the plain reading: he was steering for India, he swung far west in the south Atlantic either to catch favourable winds or to avoid the calms off Guinea, and he found land he did not expect. A minority tradition, argued most fully by Jaime Cortesão, holds that the crown knew what was there — that the Portuguese had touched the coast on an earlier westward voyage and kept quiet about it, so as not to weaken its bargaining position at Tordesillas. The evidence is thin and indirect, and the argument has never been settled. What is certain is that the crown moved quickly. In 1501 Manuel sent André Gonçalves to report on the new land, and he came back with little to recommend it beyond the trees. In 1502 the king awarded a monopoly of the brazilwood trade to a syndicate of Lisbon merchants, Fernão de Noronha among them, who undertook to send six ships a year and pay the crown a share. That contract, not a settlement, is what the map's start date of 1502 records.

The first decade produced no colony, only a name and a commodity. Cabral's fleet stayed nine days, heard mass, planted a cross, and left two convicted men ashore to make their way among the Tupi — an early and characteristic Portuguese bet on miscegenation as a means of possession. Pero Vaz de Caminha's letter to King Manuel, dated 1 May 1500, is the founding document of Brazilian literature and a poor guide to Portuguese intentions: it is mostly about nakedness, innocence and the ease with which the inhabitants might be converted. The land was called first the Island of the True Cross, then the Land of the Holy Cross, and finally Brazil, after the wood. The name that stuck was the commercial one.

For thirty years the Portuguese presence consisted of feitorias — fortified warehouses of the kind they had already built from Guinea to Malabar — where brazilwood, parrots, cotton and monkeys were exchanged for knives, axes, combs and cloth. The labour was Tupi, at first by barter and increasingly by coercion. French ships worked the same coast from the 1500s and were still doing so in the 1610s; André Thevet's woodcut of 1575, showing Indians carrying logs to a European vessel, is a French image of a trade the Portuguese claimed as a monopoly. Portuguese settlement followed the French threat rather than the wood.

How did a shore of trading posts become a colony?

The decision to occupy the coast was taken in Lisbon in the late 1520s, when the Indian Ocean trade stopped returning what it had, and when French and Castilian ships began to make the Brazilian shore unprofitable as well as insulting. In 1530 John III sent Martim Afonso de Sousa with a squadron and a general instruction to settle. Sousa founded a town on the island of São Vicente, near the modern port of Santos, in early 1532, planted sugar, distributed land and set up the first municipal council in Brazil. It was the first permanent Portuguese colony in the Americas.

Two years later the crown tried to solve the problem of distance by giving it away. Between 1534 and 1536 John III carved the whole coast into hereditary captaincies — fourteen by the usual count, fifteen in some reckonings, depending on how the later grants are numbered — and granted them to men who would defend and develop them at their own expense. The donatários were not the great nobility; they were soldiers, officials, minor nobles and merchants with court connections. Each received a strip of coast running inland to an unspecified western limit, with powers that were almost regalian: to found towns, grant land in sesmarias, appoint magistrates, raise mills, keep a share of the tithes, and enslave Indians in lawful war. The crown reserved the brazilwood monopoly, the royal fifths on precious metals, and the right to divide or revoke.

Most of the captaincies failed. Some were never occupied; some were abandoned after Indian wars; some were bankrupt within a generation. Two worked. Pernambuco, granted to Duarte Coelho, had good soil, a short sea crossing to Portugal and a docile supply of labour, and by the 1540s it had mills, a town at Olinda, and the beginnings of the richest sugar district in the Atlantic. São Vicente survived on the southern coast, poor and far from markets, but it bred a frontier society that would later matter more than any plantation.

The failures pushed the crown into direct rule. In 1548 John III drew up a regimento creating a governor-general with authority over the whole coast, and in 1549 Tomé de Sousa arrived in the Bay of All Saints with a fleet carrying about a thousand people — officials of justice, revenue and war, soldiers, exiled convicts to be used as colonists, converted Jews, and the first six Jesuits under Manuel da Nóbrega. He founded Salvador on 29 March 1549, on the bluff above the bay, and made it the capital. A bishopric followed in 1551; a high court, the relação, began sitting in 1609. Salvador remained the seat of government until 1763 and was, for most of that time, the only city in Portuguese America with the full apparatus of a European capital.

Direct rule did not pacify the colony. The Tupi of the Bahian coast had been allies, trading partners and enemies by turns; under governor Mem de Sá, in office from 1558 to 1572, the Portuguese fought a series of wars that broke the coastal tribes and pushed the survivors into mission villages or into slavery. The Jesuits ran those villages, learned Tupi, and built a second kind of colony inside the first — a network of aldeias that supplied labour to settlers while trying to protect its inhabitants from them. That contradiction ran through Brazilian history until the order was expelled.

The French were the other pressure. In 1555 Nicolas Durand de Villegagnon established a colony in Guanabara Bay, which the Portuguese called France Antarctique, and for a decade it was a real competitor, a Calvinist refuge and a base for the wood trade. Estácio de Sá landed at the foot of the Sugar Loaf on 1 March 1565 and founded Rio de Janeiro; the French were broken by 1567, and the Tamoio who had allied with them were destroyed or enslaved in the years that followed. By the 1580s Portugal held a continuous coast from Pernambuco to São Vicente, with outposts in the Amazon and a capital in Bahia.

What did sugar make of the colony?

Sugar was the reason the colony existed, and the engenho was the institution that shaped it. An engenho was not only a mill but a whole productive unit: cane fields, a water- or ox-driven crushing works, boiling houses, a drying platform, a chapel, the planter's house and the slave quarters, with oxen, carts, woodland for fuel and a landing on a navigable river. The technology came from Madeira and São Tomé, where the Portuguese had learned to grow cane with slave labour in the fifteenth century; the capital came from Lisbon, from Italian and Flemish merchants, and increasingly from the planters themselves.

The industry grew fast and then settled into a shape. Pero de Magalhães Gandavo counted sixty engenhos in Brazil in 1570, fifty of them in the captaincies from Ilhéus northwards; reports made between 1583 and 1585 put the total between 108 and 128, with more than half in Pernambuco. Bahia had eighteen mills in 1570 and between forty and fifty by 1590. By the early seventeenth century an average mill was producing about 6,000 arrobas of sugar a year, and the largest between 8,000 and 10,000 — several times the output of the best Madeiran mills a century earlier.

Labour was the constraint. The first engenhos ran on enslaved Indians, acquired by barter, by capture in "just war", or by purchase from other Indians, and the planters' demand emptied the coastal villages faster than disease alone would have done. African slaves were being imported in quantity by the middle of the sixteenth century — Bahia had between three and four thousand of them by the last decade of the century, when they still made up only a quarter of its slave force. The proportion reversed in the seventeenth century. Brazil became an African colony with an Indian past, and the trade that supplied it was the most profitable single business in the Portuguese Atlantic.

Sugar also tied Brazil to Europe in ways the colony could not control. Prices were set in Lisbon, Amsterdam and London; credit came from merchants who advanced goods against the next harvest; the crown taxed the crop, the shipment and the labour, and granted monopolies on salt, tobacco, soap and the sale of slaves. Planters who owned land, people and mills were chronically in debt to men who owned none of those things. The fleet system, in which convoys (frotas) sailed from Lisbon to Bahia and Pernambuco under escort once a year, was meant to protect the trade from privateers and did as much to concentrate it in the hands of a licensed few.

The colony was never only sugar. Tobacco, grown in Bahia and worth more per pound, became the currency of the slave trade on the Mina coast. Cattle ranches spread up the São Francisco river and into the sertão, feeding the sugar zone and later the mines. Cotton, rice, cocoa, indigo and hides all found markets. But sugar set the pattern — large estates, enslaved labour, production for export, a coastline of ports facing Europe and Africa rather than an interior facing itself — and that pattern outlived the crop.

Why did Portuguese America nearly disappear between 1580 and 1654?

In 1580 the Portuguese crown passed to Philip II of Spain, and for sixty years Portugal and its empire were governed from Madrid within a composite Spanish monarchy. Brazil's boundary with Spanish America ceased to matter to the crown that governed both, the Dutch became enemies rather than trading partners, and Portuguese America was drawn into a European war that had nothing to do with it.

The map reflects this honestly. It draws Portuguese America from 1502 to 1581, again in 1619–1621, and then from 1640 to 1823. The gap is the Iberian Union: during those decades the territory is drawn as part of the Spanish monarchy, because that is what it was. The short 1619–1621 window is a border-data artefact, not a moment of independence.

The Dutch attacked in the middle of that union. A fleet of the Dutch West India Company took Salvador on 8 May 1624 with 26 ships and some 3,300 men; a combined Spanish-Portuguese expedition of 52 ships and about 12,500 men retook it on 30 April 1625. Piet Heyn's capture of the Bahia sugar fleet in 1627 and of the Mexican silver convoy at Matanzas in 1628 paid for a second attempt, and in 1630 a fleet of 65 ships took Olinda and Recife. By 1635 the Dutch held the coast from Rio Grande do Norte to Cabo de Santo Agostinho, the richest sugar district in the world.

The occupation is remembered in Brazil for its culture rather than its politics. Count Johan Maurits of Nassau-Siegen governed from 1637 to 1644 with a court of scientists and painters: Willem Piso and Georg Marcgraf, whose Historia Naturalis Brasiliae appeared in 1648, and Albert Eckhout and Frans Post, whose canvases are the first European images of Brazil drawn from observation rather than fantasy. He built a new town on the island of António Vaz, with bridges, an observatory and a botanical garden, and allowed Catholics, Jews and Calvinists to worship. None of this made the colony Dutch. The settlers remained Portuguese, the planters remained Catholic, and the company remained a business.

When Portugal recovered its independence in 1640, the Dutch and Portuguese found themselves with a common enemy and signed a ten-year truce, which the Dutch used to seize São Luís do Maranhão and, more damagingly, Luanda in Angola, the source of the slaves on which Brazil ran. The Pernambucano planters rose against the Dutch in 1645, in a war they called the War of Divine Liberty, financed largely by a tax on their own sugar. Two battles at Guararapes, in 1648 and 1649, broke the Dutch field army; the last Dutch garrisons surrendered at Recife on 26 January 1654, and the Netherlands formally ceded New Holland to Portugal by the Treaty of The Hague in 1661.

The war destroyed what it won. Pernambuco's output had fallen by 1654 to roughly a tenth of Brazil's total, its mills burned, its slaves dispersed or run away, and the Dutch capital and expertise that had made Recife a sugar city moved to the Caribbean, where English and French planters were only too glad to receive it. Brazil's sugar economy recovered, but it never again had the field to itself.

How did São Paulo's raiders push the frontier past the treaty line?

The interior was taken by men the crown could not control. São Paulo, founded as a Jesuit mission in 1554 on the plateau above the escarpment, grew into a town of Portuguese men, Indian and mixed-race women, and their descendants — the mamelucos — who spoke Tupi more often than Portuguese and lived by raiding. Their expeditions, the bandeiras, went out for captives: to the sertão, to the Jesuit missions of Guairá and Tape on the Spanish side of the line, and eventually to the Amazon.

The Guairá missions were the great crime. Between 1628 and 1630 António Raposo Tavares led a force of 69 whites, some 900 mamelucos and more than 2,000 Indians against the reductions the Jesuits had built among the Guaraní in what is now Paraná, and brought back thousands of captives. A second campaign in 1636–38 destroyed the Tape missions further south. These were the same mission settlements that would later be rebuilt as the Guaraní reductions of Paraguay and Argentina; the Brazilian raids are the reason the reductions were moved and armed.

Raposo Tavares' greatest journey was not a raid but an exploration. In 1648 he left São Paulo with a few hundred men, went west to the Paraguay, north down the Guaporé and Madeira, and reached the Amazon; he returned to São Paulo by way of Belém in 1651, having covered thousands of kilometres of unknown country. The route was not followed again for a century, but it established a claim that later occupation settled.

The Amazon was a second frontier, and a worse one for the Indians. Belém, founded in 1616, was the base for expeditions that went upriver for droga do sertão (cacao, sarsaparilla, turtle oil, spices) and for slaves. The Jesuits tried to regulate the trade and were expelled from Maranhão by the settlers in 1684 for their pains. Estimates of the destruction are wild and unverifiable: a vicar-general of Maranhão claimed on his deathbed that two million Indians had been destroyed in Maranhão and Pará, a figure that cannot be checked and is certainly too high, but the direction of it is not in doubt.

Cattle did more than raiders to hold the interior. Herds followed the São Francisco south and north, ranches spread through Piauí, Bahia and the sertão of Pernambuco, and a cattle frontier that needed few slaves and no capital pushed Portuguese occupation inland at a fraction of the cost of a plantation. When the gold strikes came, the ranches were already there to feed them.

By the early eighteenth century Portugal held, on the ground, a territory far larger than the line of 1494 allowed. The Treaty of Madrid of 1750 accepted the fact: it abolished the Tordesillas meridian as a boundary, replaced it with the principle of uti possidetis — who holds, keeps — and gave Portugal most of the frontiers of modern Brazil, including the Amazon and the west. In exchange Portugal surrendered Colónia do Sacramento, the trading post opposite Buenos Aires that had been founded in 1680 and fought over ever since. The treaty was undone by the Guaraní war that followed the transfer of seven missions, and renegotiated in 1777, but the principle survived. Brazil's modern shape is a Pombaline compromise, not a Portuguese discovery.

What did gold do to the colony?

Gold was found in the interior of what became Minas Gerais between 1693 and 1695, by Paulistas hunting Indians, and the news did what no royal proclamation could do: it moved people. Within a decade tens of thousands had left the coast for the highlands, from Bahia, Pernambuco, Rio and Portugal itself. The rush created famine in the south and inflated prices everywhere; between 1697 and 1701 two famines struck, and coastal merchants were accused of holding food back to sell it in the mines at three or four times its price.

The mining country was lawless by design: it lay in no established captaincy, its population was mixed in origin and colour, and the Paulistas who had found the gold claimed the right to allot it. The outsiders — emboabas, a term of obscure origin — fought them for it in 1709 in a short, vicious war that ended with the Paulistas expelled and the crown, for the first time, in control of the district. Minas Gerais became a captaincy of its own in 1720, and the revolt of Vila Rica that year, against the crown's attempt to impose smelting houses and a mint, was put down with executions.

Gold moved the centre of the colony. The strikes at Cuiabá in 1719 and Goiás in 1725 opened the far west; the discovery of diamonds in the Serro do Frio in 1729 created a second, stranger economy, in which the crown closed the district to free miners and worked it through contract companies under guard. By the 1730s the mining districts had become the colony's fastest-growing economy and the crown's chief fiscal interest, and in 1763 the capital moved from Salvador to Rio de Janeiro, closer to the mines and to the trade that carried their output to Lisbon.

The crown's share was the fifth, the quinto, and it was evaded on a scale that made the tax a running grievance. Smuggling was general, connived at by officials, and impossible to stop along a frontier of thousands of miles; the crown's answer was to convert the fifth into a fixed quota per district and, when the quota was not met, to impose a derrama, a forced levy on the whole population. That instrument, used in Minas Gerais in the 1780s, was the immediate grievance behind the conspiracy of 1789.

Gold did not replace sugar. In 1760 Brazil's exports were valued at 4.8 million milréis, of which sugar was 50 per cent and gold 46; the mining economy enriched the crown and the merchants of Rio, but the plantations still employed more people and fed more of the Atlantic trade. What gold did was to make the interior Portuguese, to create a free population of mixed descent that outnumbered the slaves in the mining districts, and to tie Brazil to Britain: perhaps half to three-quarters of the gold that reached Lisbon in a good year — around 1,200 arrobas, or 38,400 pounds avoirdupois — was shipped on to London to pay for English goods.

Who actually governed Portuguese America?

Portuguese America was governed by a great many authorities and no single one. At the top stood the crown, advised after the 1640s by the Overseas Council in Lisbon, which read the correspondence, issued the ordens régias, appointed the governors and quarrelled with them by return of post. In Brazil stood the governor-general — viceroy after the title was attached to the office in the seventeenth century — with authority over the captaincies, the treasury and the militia, but without the means to impose his will on a captaincy a thousand miles away.

Below him the captaincies were a patchwork. Some were royal, governed by a capitão-mor or a governor appointed from Lisbon; some were still hereditary, in the hands of descendants of the sixteenth-century donatários; some, like Minas Gerais and Goiás, were created in the eighteenth century as the mining districts filled. Maranhão and Grão-Pará were a separate state from 1621, with its own governor and its own direct line to Lisbon, and the Amazon was administered apart from Brazil for most of the colonial period.

The towns governed themselves, in a fashion. The câmara — the municipal council — was elected by the "good men" of the parish, which in practice meant the planters, merchants and officials who owned the place. The câmaras set prices, regulated trades, taxed locally, petitioned the crown and, in moments of crisis, acted as a government. They were also the instrument by which a local oligarchy kept the governor at arm's length.

Justice was slow, expensive and personal. The relação of Bahia, sitting from 1609, heard appeals from the captaincies. Below it, ouvidores and juízes de fora travelled their districts, and the crown judges were often the planters' creditors, neighbours or relatives. The Portuguese crown never solved the problem of governing a colony 6,000 kilometres away with officials who had to live among the people they governed, and it never really tried: it relied on negotiation, on the church, and on the fact that most people had no access to law at all.

The eighteenth century brought reform and made the machine heavier. The Marquis of Pombal, in effective control of Portuguese government from 1750 to 1777, created two monopoly companies — Grão-Pará and Maranhão in 1755, Pernambuco and Paraíba in 1759 — to finance the trade of the north and northeast, abolished the fleet system in 1765, ended the legal power of the missionaries over Indians by the law of 6 June 1755 and the Directorate of 1757, extended to all Brazil in 1758, and expelled the Jesuits from the Portuguese dominions in 1759, seizing their estates and selling them. He also moved the capital to Rio and tightened the tax machine.

Pombal's reforms were designed in Lisbon for Lisbon's benefit, and they did not work as intended. The companies raised revenue and angered the merchants; the Directorate put Indian labour into the hands of crown officials who abused it more systematically than the Jesuits had; and the expulsion of the Jesuits removed the only institution that had offered the Indians an alternative to slavery. Brazil's export earnings in 1777 were half what they had been in 1760.

What was Brazil's place in the Portuguese empire?

Brazil was the western half of a Portuguese Atlantic system whose other half was Africa, and the two were joined by ships carrying people. Bahia traded directly with the Bight of Benin and with Angola; Rio de Janeiro with Benguela and, after 1648, with Luanda, which a Portuguese expedition from Rio recaptured from the Dutch. The sugar and tobacco of Brazil paid for the slaves of West Africa, the slaves worked the sugar, and the sugar paid for the slaves: the classic triangle, with the addition that Portuguese ships carried much of the cargo at every leg.

The scale was enormous. The customs house at Bahia recorded 90,809 Africans imported in the twenty years from 1728 to 1748; contemporaries reckoned 10,000 to 12,000 a year from Whydah alone, and another 6,000 to 7,000 from Angola distributed among Rio, Pernambuco and Bahia. Bahia's slave population was about 147,000 in 1819, roughly a third of the captaincy's 500,000 inhabitants, and about half the population of Salvador itself.

The colony's other function was to buy Portuguese goods and sell Portuguese goods, and the crown tried to keep the trade in Portuguese hands. It failed. English merchants had been forced into the Brazil trade by treaty in the seventeenth century and never left; by the eighteenth century much of the colony's import trade was financed from London, and the gold that paid for it went the same way. The Methuen Treaty of 1703, which gave English cloth preferential access to Portugal, completed the arrangement: Portugal sold wine and bought manufactures, and Brazil paid for both.

The empire's Asian and African possessions outranked Brazil in the crown's imagination for two centuries. Goa was the capital of the Estado da Índia and the seat of a viceroy; Portuguese Ceylon supplied cinnamon; Mozambique and Angola supplied slaves; Macau supplied the China trade. Brazil was the colony that paid, but it was also the colony that had no viceroy of its own until 1640 and no university at all — the crown refused to raise the Jesuit college at Salvador to that status, on the advice of Coimbra. Portuguese America was rich, useful and intellectually dependent, and the crown preferred it that way.

The relationship changed only when the empire's centre moved. In November 1807 a French army crossed into Portugal, and the prince regent, the future John VI, took the extraordinary decision to move the court itself across the Atlantic rather than surrender or submit. Some 10,000 people — the queen, the prince, the nobility, the treasury, the archives and the government — embarked at Lisbon on 27 November and sailed two days later under British escort.

What did slavery make of Brazilian society?

Brazil was a slave society, not merely a society with slaves. By the end of the colonial period about a third of the population was enslaved, and in the sugar parishes of the Recôncavo the proportion reached 70 per cent. The slaves were Africans and their descendants, brought in a trade that never stopped growing until the 1820s; they worked the cane, the mill, the boiling house, the gold washings, the cattle ranch, the town house and the wharf. A large engenho might hold 150 or more; the Benedictine and Jesuit estates held hundreds; a small farinha farm might hold two.

The demography was brutal and is the clearest evidence against the old claim that Brazilian slavery was mild. A slave population that could only be sustained by importation, with a sex ratio of two men to one woman among adults and an African-born majority, is a population that is not reproducing itself — a population being worked to death and replaced. Infant mortality was high, life expectancy at birth low, and the law offered almost nothing: a slave's testimony was worth less than a freeman's, and a master's power to punish was limited in principle and hardly at all in practice. The crown legislated intermittently about food, rest and Sunday labour; the brotherhoods and the courts did what they could; the results were meagre.

The Church was part of the system and a limit on it. The Jesuits, the Benedictines, the Carmelites and the Franciscans all owned slaves; the lay brotherhoods — the Rosary, the blacks of the Rosary, the browns — gave enslaved and freed people a corporate identity, a burial fund and a saint, and were the only institutions in colonial Brazil in which Africans and their descendants held office. Afro-Brazilian religion, in the form of calundus and later Candomblé, survived inside and alongside Catholicism.

Resistance was continuous. Slaves ran away in small numbers constantly and in large numbers occasionally; the quilombos — maroon settlements — dotted the margins of the plantations, close enough to raid for supplies and recruits. The greatest was Palmares, in the forests of Alagoas and Pernambuco, a kingdom of many thousands with a capital of 220 houses, a church, four smithies and a council house, ruled in the seventeenth century by a king called Ganga-Zumba. The Dutch failed to take it; the Portuguese spent years trying, and only broke it in 1694–95 with a Paulista army under Domingos Jorge Velho.

Freedom existed and grew. Manumission was common, especially in the towns; by 1821 the free people of colour of Minas Gerais were 40 per cent of the captaincy's population, and Bahia was probably similar. The free poor — pardos, pretos forros, poor whites, caboclos — made up a large share of the population in most places and had almost no share in government. The colonial order was not a simple two-caste system; it was a graded hierarchy in which colour, legal status, legitimate birth, occupation and wealth all counted, and in which the top was tiny.

Indian slavery was the other half of the story, and it ended differently. The crown forbade the enslavement of Indians in 1570 and repeatedly afterwards, always with exceptions for "just war", cannibalism and ransom, which the settlers used to enslave whom they pleased. Pombal's law of 6 June 1755 abolished Indian slavery in Grão-Pará and Maranhão and declared Indians free; the Directorate that followed put them under crown officials, and the abuses continued until it was abolished in 1798. By then the Indians of the coast and the great rivers were largely gone: killed by disease, war and labour, absorbed into the mixed population, or pushed into the interior.

How did a colony become the seat of an empire?

The transfer of 1807–08 was the single most important event in the colony's history. The court arrived at Salvador on 22 January 1808, and the prince regent's first acts undid the colonial system: the ports of Brazil were opened to the trade of friendly nations, manufactures were permitted, and a bank, a printing press, a medical school and a military academy followed. Rio de Janeiro became the capital of a European monarchy, with an opera house, a royal library and a court of some thousands of place-seekers. Brazil's population was about three million in 1807.

On 16 December 1815 Brazil was elevated to the status of a kingdom, in the United Kingdom of Portugal, Brazil and the Algarves, a device to give the Braganzas a throne in the Americas that could sit at the Congress of Vienna as an equal of the European powers. The colony had become a co-monarchy, and the Brazilians had begun to taste the experience of governing themselves. They liked it. When the Liberal Revolution of 1820 in Portugal produced a parliament at Lisbon that wanted to restore Brazil to colonial subordination, the Brazilian elite resisted, and the prince regent, Pedro, stayed.

Discontent had been building for longer than the court's stay. The Inconfidência Mineira of 1789, a conspiracy of magistrates, poets and army officers in Vila Rica against the derrama and Portuguese rule, was betrayed before it began; its leader, the dentist Tiradentes, was hanged and quartered in Rio in 1792, and the others exiled. A republican revolt in Pernambuco in 1817 was crushed more bloodily.

John VI returned to Portugal on 26 April 1821, leaving Pedro as regent. The Lisbon cortes then tried to reduce Brazil to a set of provinces, and Pedro, in a series of decrees in 1822, refused. On 7 September 1822, on the bank of the Ipiranga brook outside São Paulo, he declared Brazil independent; on 12 October he was acclaimed emperor, and on 1 December crowned. The war that followed was fought in Bahia, Maranhão, Pará and Cisplatina and ended with Portuguese troops withdrawn. Portugal recognised independence in 1825, in exchange for a payment and a Brazilian promise to take on part of the Portuguese debt.

Independence did not end the colonial order. The Empire of Brazil kept the monarchy, the senate, the established church, the plantation and slavery, which survived until 1888. What changed was the direction of the flow: the taxes, the offices and the patronage that had gone to Lisbon now went to Rio, and the Brazilian state inherited the boundaries, the institutions and the social hierarchy that three centuries of Portuguese rule had made.

How have historians argued about Portuguese America?

The oldest tradition, from Francisco Adolfo de Varnhagen in the nineteenth century, told the story as the construction of a nation by heroic discoverers, governors and bandeirantes, with the Portuguese crown as the author of Brazil. That view survived into the twentieth century in schoolbooks and official history, and elements of it still shape how Brazilians tell the story.

Two critiques of it shaped the twentieth century. The São Paulo school, and Caio Prado Júnior's Formação do Brasil contemporâneo of 1942, described the colony as an exporting appendage of Europe, organised around the plantation and slavery, whose interior was empty of meaningful social life — the "sense of the colony" as a machine for producing commodities. Gilberto Freyre's Casa grande e senzala of 1933, published earlier and read differently, argued that Portuguese colonisation had been unusually intimate and adaptive, producing a mixed society rather than a segregated one; the argument became the ideology of lusotropicalism and, in the hands of the Salazar regime, a justification for empire in Africa.

The postwar generation worked from the archives rather than from theory. C. R. Boxer's The Golden Age of Brazil (1962) and The Portuguese Seaborne Empire (1969) established the empirical standard; Stuart Schwartz's Sugar Plantations in the Formation of Brazilian Society (1986) took a single region, Bahia, and reconstructed its economy, demography, labour and social relations in a detail that had not been attempted before; A. J. R. Russell-Wood and Dauril Alden widened the frame to the empire, and Leslie Bethell's Cambridge History of Latin America set the colony in a hemispheric account. John Hemming's Red Gold (1978) did for the Indians what Schwartz did for the slaves, and made the destruction of the indigenous population a central rather than a decorative part of the story.

Two arguments continue. The first is about the South Atlantic: whether Brazil is best understood as part of a single Lusitanian system that includes Angola and the Bight of Benin, as Luís Felipe de Alencastro has argued, or as a colony whose Atlantic trade was always subordinate to European demand. The second is about the frontier: whether the interior was conquered by the crown and its bandeirantes, as the older literature has it, or whether it was occupied by a mixed population of settlers, refugees, missionaries and Indians whose society the state followed rather than led. The map cannot settle either question. It can only show where, in 1732, the Portuguese flag was flying and where it was not.

Common myths about Portuguese America

Timeline of Portuguese America

What the map shows next

At 1580 on the map the Portuguese coast runs from Pernambuco to São Vicente with a scatter of Amazon outposts, and the interior is still Indian country. At 1700 on the map the gold strikes have begun to pull settlement inland, but the line of Tordesillas still nominally bounds the colony to the west and the south.

At 1789 on the map the captaincies of Minas Gerais, Goiás and Mato Grosso reach far beyond the old meridian, the capital has moved to Rio, and the mining towns where the Inconfidência was plotted are the second city of the colony. At 1820 on the map the map shows a territory that is still Portuguese and already Brazilian: the court is in Rio, the ports are open, and within two years the heir will declare the independence that ends the entry.