“21,000 people founded Palmares.”
The figure is the sum of two Dutch military estimates from the 1640s — roughly 6,000 and 15,000 people in two forested areas. It is neither a census nor a count of founders.
See where the number came from →
Colonial Brazil · 1595–1714
Palmares
The short version goes like this: 21,000 people escaping enslavement built a kingdom in the Brazilian interior.
Palmares had a centre called Macaco, rulers, local chiefs, justice, religious life and an army that raided Dutch and Portuguese plantations. The documents preserve traces of every element. They appear in different years, through the voices of people who tried to destroy the communities.
This is where the story begins
One number hides a network
Palmares was a network of autonomous communities in colonial Brazil. To understand the number 21,000, first see the place and people compressed inside it.
Palmares
1595–1714
Colonial BrazilIn Portuguese Brazil, people escaping enslavement sought refuge in communities in the interior. Palmares grew through those communities.
In colonial sources, an individual settlement often appears as a mocambo. Quilombo is the broader term for communities built by people who escaped enslavement and defended their autonomy. Palmares was a network of mocambos that could move, join or divide.
By the 1670s, documents describe a system coordinated by Ganga Zumba. Macaco appears as the principal centre, with more than 1,500 houses. The same world includes subordinate chiefs, guards, a “minister of justice”, a chapel, baptisms and marriages.
1,500+houses attributed to Macaco in a 1678 account
This is where the story begins: a power built over time, whose scale was measured by its enemies in the middle of war.
01 · The number
How did 21,000 appear?
In the 1640s, Dutch officers estimated roughly 6,000 and 15,000 people in two forested areas. Added together later, the estimates became 21,000. The phrase “founded Palmares” turned them into a single moment.
The claims that circulate about Palmares stay visible below in their recognisable form. Each one gets its place in the story.
The labels show the confidence level:
well documented
true with qualifications
contested
unsupported
“It existed from 1612 to 1694.”
1612 marks an early expedition, not a secure date of foundation. The Serra da Barriga stronghold fell in 1694, Zumbi was killed in 1695, and Mouza's mocambo endured until 1714.
How the story continues after 1694 →“It was an independent kingdom.”
“Kingdom” conveys de facto autonomy, hierarchical leadership and diplomatic capacity. It was not a recognised state, however, and its institutions changed over time and between settlements.
How to read the colonial account →“Its capital had 5,000 people.”
A colonial account says Macaco contained more than 1,500 houses. Nearly 5,000 inhabitants is a modern houses × occupants extrapolation, not a contemporary count.
Test the household assumption →“It had rulers, officials, justice and religious life.”
The 1675–1678 war account describes a paramount leader, subordinate chiefs, guards, a “minister of justice,” a chapel, baptisms and marriages. Its terminology is a colonial translation of institutions the writer did not fully understand.
Document, inference and the unknown →“It fought and raided Dutch and Portuguese plantations.”
Palmares resisted expeditions from both colonial powers and organised raids for supplies, weapons and people. Its relations with neighbours also included clandestine trade, intelligence and negotiation.
The negotiation of 1678 →“Palmares was the origin of Congada.”
Congada preserves Kongo traditions of kingship and Afro-Brazilian Catholic forms. Palmares belonged to the same Central African Atlantic world, but no documented line of descent connects it to the festival.
Where the evidence stops →“Mendonça was linked to both Palmares and the festival.”
Lourenço da Silva Mendonça lived in Brazil and later led a legal campaign against slavery in Rome. A possible alliance with Palmares is a recent, contested hypothesis; his participation in Palmares or Congada is not demonstrated.
What remains a hypothesis →02 · The network
How the Palmares network worked
Palmares was made up of mocambos: settlements that could grow, be destroyed, move, join a political system or disappear. The diagram shows how relationships changed over time; it is not a map with exact coordinates.
Choose a period
Regional coordination
By the 1670s, sources describe a network of mocambos connected by leadership, alliances and defence rather than a single fixed settlement.
Names and relationships come from colonial documents whose spelling and location vary. This represents political change, not a geospatial reconstruction.
03 · Scale
How large was Palmares?
The estimates were produced during war by men seeking resources and prestige for further campaigns. They give a sense of scale. The calculator shows how an assumption about people per house can produce 5,000 inhabitants.
6,000 + 15,000
Where the 21,000 figure comes from
Two Dutch estimates from the 1640s, for communities in two forested areas, were later added together. They describe a scale observed or reported during wartime.
1,500+
Macaco: more than 1,500 houses
The 1678 account attributes this number to the principal centre. Houses do not directly tell us the population: we do not know how many people lived in each or how many buildings were dwellings.
How an estimate reaches 5,000 inhabitants
Change the assumed average number of people in each house.
2.55.0
estimated inhabitants if all 1,500 buildings were occupied dwellings at this average
This does not recover the real population. It exposes the hidden assumption inside an apparently precise number.
04 · Organisation
How Palmares was organised
The account of the 1675–1678 campaigns is the richest description of internal political life. Its author translated Palmares institutions into familiar Portuguese words: king, palace, minister, republic.
What the source says
- Ganga Zumba is called “king and master.”
- Macaco has guards and officials; other settlements have subordinate rulers.
- A “minister of justice” is mentioned.
- There is a chapel, baptism and marriage.
- Subupira is described as heavily fortified.
What we can support
- Hierarchical and delegated political leadership.
- The capacity to mobilise fighters and organise defence.
- Rules, sanctions and procedures for settling disputes.
- Religious institutions combining Central African and Catholic influences.
- Status differences between locally born people, newcomers and captives.
What we cannot reconstruct
- The names residents used for their own offices.
- A stable, uniform “constitution” across an entire century.
- How much the author exaggerated to justify war.
- How power worked in everyday life beyond visible leaders.
- The voices of the people described — the archive largely preserved those of their hunters.
05 · The 1678 agreement
The 1678 agreement divided Palmares
Ganga Zumba negotiated with the Pernambuco government under terms that divided freedom into categories. Colonial authorities decided who could remain free, where people had to live and who would be returned to enslavement.
1
Freedom with conditions
People born in Palmares were promised recognition of their freedom and a pardon.
2
Removal to Cucaú
The accepting group had to settle closer to colonial supervision and control.
3
Surrender of new escapees
People who had recently escaped enslavement were to be surrendered to their former enslavers.
4
Freedom was withdrawn
Cucaú's inhabitants were later attacked, captured and re-enslaved, producing legal disputes inside the empire itself.
Palmares divides
Ganga Zumba's logic
After devastating campaigns, the capture of relatives and the loss of settlements, the agreement may have been an attempt to save part of the community. Reducing it to “betrayal” erases the constraints under which it was negotiated.
The logic of resistance
Zumbi and other leaders rejected terms that made freedom a colonial privilege and demanded the surrender of recent escapees. Resistance reorganised around Serra da Barriga.
06 · Timeline
Formation, war and continuity
This chronology starts with the earliest surviving mentions and follows the formation of the communities, the wars, the 1678 agreement and the continuation after 1694.
1595
The earliest surviving mention in the Documenta Palmares database
The digital archive gathers documents explicitly mentioning Palmares from 1595 to 1800. A first surviving mention is not necessarily a beginning.
1612
An early colonial expedition
The year later became a “foundation date,” although it records a colonial response to something already present.
1640s
Dutch observers report large inland communities
Two wartime estimates — 6,000 and 15,000 — later generate the viral figure of 21,000.
1675–78
Campaigns and the richest colonial description
The war account supplies information about houses, fortifications, rulers and institutions, but was written for authorities financing attacks.
1678
The agreement between Ganga Zumba and Pernambuco's government
One part of Palmares moves to Cucaú; another rejects the conditions and continues resistance.
1694
The Serra da Barriga fortification falls
The period of the principal late military centre ends. Other Palmarista communities continue after 1694.
1695
Zumbi is killed on 20 November
His death later anchors a public memory centred on Black agency and resistance rather than abolition bestowed by the Crown.
1714
Mouza, the last known leader of a Palmares mocambo, is captured
Records still place him in prison in 1715. Palmares continued to generate flight, regrouping and colonial fear long after 1694.
1986
Serra da Barriga becomes a protected national monument
The site moves from the history of colonial warfare into the public infrastructure of memory.
2024
20 November is first observed as a nationwide public holiday
The law enacted in 2023 declares the National Day of Zumbi and Black Consciousness a federal holiday throughout Brazil.
07 · Congada and Mendonça
What we know about Congada and Mendonça
Palmares, Kongo political traditions and Mendonça's campaign belonged to the same Atlantic world. The sources support regional cultural connections; they do not trace a documented line from Palmares to the festival.
Kongo kingship
Crowns, processions, titles and Christian forms of authority circulated from Central Africa through the Atlantic world.
Black kings and queens in Brazil
Black Catholic brotherhoods organised elections and royal celebrations in several regions, in forms that changed over centuries.
Palmares
Names, languages and political practices indicate strong Central African ties without making Palmares the single source of Afro-Brazilian traditions.
Congada's direct origin
The sources do not trace institutional or ritual transmission from Palmares to the festival. The term “congada” itself appears only around the beginning of the nineteenth century.
The Mendonça hypothesis
José Lingna Nafafé argues that exiled Ndongo royals could have mattered to Palmares. Silvia Hunold Lara finds the sources insufficient to establish the connection.
The image is relevant — and a century too late.
Carlos Julião's work shows the splendour of Afro-Brazilian royal festivals in the later eighteenth century. It helps explain the tradition from which Congada emerged. It is not an image of Palmares and cannot show that the festival began there.
Lourenço da Silva Mendonça was a member of the Ndongo elite who lived in Brazil and pursued a legal campaign against slavery in Rome during the 1680s. That is remarkable without adding meetings or alliances that the documents do not preserve.
08 · The archive
What the archive does not show
Military records see leaders, fortifications, captives and threats. Women, children, farmers and people without visible office appear rarely and often only at the moment of capture.
The archive preserves only part of people's lives.
Emblematic figures such as Dandara and Aqualtune do not appear in the seventeenth-century sources identified so far. Their biographies cannot be treated as documented. The archive does not allow us to conclude that women did not fight, lead, negotiate or sustain the communities.
An honest history keeps discipline toward the archive while recognising that colonial documents were built to see only certain people.
Names that emerge from the records
Aca Inenea Palmarista woman named in sources
Maria Moreiraidentified as Zumbi's wife
one daughter and three sonsprobably captured after the 1694 assault
thousands without namesmost social history remains unrecoverable
Public memory
How Palmares is remembered today
Serra da Barriga is protected as Brazilian cultural heritage, has housed the Quilombo dos Palmares Memorial Park since 2007 and became Mercosur Cultural Heritage in 2017. Since 2024, 20 November has been a national public holiday in Brazil.
1695Zumbi is killed on 20 November
2007the memorial park opens
2024first nationwide holiday
Photograph: Joalpe, Wikimedia Commons, CC BY-SA 4.0. The park is a modern memorial reconstruction, not a surviving seventeenth-century building.
09 · Sources and method
Sources and method
The page starts with manuscripts and documentary databases, then uses scholarship that critically re-reads them. Popular sources were used for orientation or to identify claims requiring verification.
Digital archive
Documenta Palmares · Unicamp
More than 1,800 digital copies of documents explicitly mentioning Palmares from 1595 to 1800, together with maps and bibliography.
Open the database ↗Source criticism
Robert Nelson Anderson · The Quilombo of Palmares
A classic analysis of how colonial accounts entered the literature and were sometimes over-interpreted.
Read the PDF ↗The 1678 treaty
Silvia Hunold Lara · Marronnage et pouvoir colonial
Reconstructs the negotiations, conditional freedom, Cucaú and the re-enslavement of its inhabitants.
View the study ↗Historiography and archive
Silvia Hunold Lara · Palmares & Cucaú: um livro e suas histórias
Explains the manuscript work, documentary limits and the difference between historical Palmares and its memory.
Read the PDF ↗Recent synthesis
Marc A. Hertzman · After Palmares
Follows diaspora, inheritance and Zumbi's afterlives; its chronology includes Mouza's capture in 1714.
View the book ↗Context and new documents
Pesquisa FAPESP · Original documents reveal new aspects
An extensive interview with Silvia Hunold Lara on Cucaú, original manuscripts and Palmares's continuation through Mouza.
Read the article ↗Congada
Marina de Mello e Souza · Black Kings in Brazil
Traces Black royal festivities, their Kongo connections and the late appearance of the term Congada.
Read the PDF ↗Mendonça
José Lingna Nafafé · Lourenço da Silva Mendonça
The case for a seventeenth-century Atlantic legal movement against slavery and the proposed Palmares connection.
View the book ↗Public memory
Fundação Cultural Palmares · Serra da Barriga
The site's status, the memorial park and its recognition as Mercosur Cultural Heritage.
View the official page ↗Law
Federal Law No. 14,759/2023
Declares 20 November a national holiday for the National Day of Zumbi and Black Consciousness.
Read the law ↗About the images: the 1647 map attributed to Frans Post is in the public domain and supplies colonial context, not precise Palmares boundaries. The festival group derived from Carlos Julião is CC0. The Serra da Barriga photograph is by Joalpe, CC BY-SA 4.0.