BEFORE THE DISPERSAL · RECONSTRUCTED LANGUAGES AND INHABITED LANDSCAPES

Before Bantu Worlds

Before the spread of Bantu languages

Comparing languages allows proposing ancient relationships, but not reconstructing a single route followed by all their communities.

Museum documentation · free public access, no account required.

Editorial visualisation of sound basins and abstract linguistic routes over a real map of central Africa — **EDITORIAL VISUALISATION DECLARED · AI-GENERATED IMAGE · TRACEKEEPERS**
Editorial visualisation · Contextual correspondence
Declared editorial visualisation, based on the cartography and sources of the fact sheet; it is not an exact historical frontier.

History and context

Long before the expansion of Bantu languages, central and western Africa hosted communities with diverse technologies, mobilities and economies. The proposed Grassfields region for proto-Bantu was not an empty starting point nor a homeland with borders.

Historical linguistics compares sounds and vocabularies to reconstruct kinships. Archaeology, palaeoenvironment and genetics formulate complementary questions, but no pot speaks a language and no gene carries a complete cultural identity.

Evidence and interpretation

Documented fact: Hundreds of languages show systematic correspondences that allow reconstructing ancestral forms and ramifications. Interpretation and caution: A linguistic tree simplifies contact, bilingualism, displacement and adoption; it does not equal a biological tree of peoples. Historical transition: The museum’s start is not a departure date, but the gradual convergence of linguistic branches, populations and archaeological assemblages in several corridors. Limit of the transition: Do not automatically attribute all Iron Age pottery or metallurgy to Bantu speakers, nor suppose that agriculture, iron and language travelled together in a single movement.

Catalogue facts

Previous period: Late Pleistocene–≈3000 BCE
Sphere of influence: West-central Africa · forests, savannas and networks prior to proto-Bantu diversification
Threshold element: The proto-Bantu reconstruction: an object that does not exist
Element date: Modern linguistic reconstruction on historical and living languages
Location: West-central Africa · debated region of origin
Material or support: Phonological series, compared vocabularies and phylogenetic models
Bantu means: A branch of Niger-Congo languages and the societies that speak them; not a homogeneous civilisation
Method: Cross linguistics, archaeology, palaeoenvironment, genetics and oral history while maintaining their boundaries
Scale: More than five hundred languages and hundreds of regional histories, not a single demographic march
Caution: Ceramics, iron, cultivation, genes and language do not always move together
Visual credit: OpenAI ImageGen · editorial direction Tracekeepers · original visualisation for the project

A SHARED TOPIC, DIFFERENT MOMENTS

Objects connected through time

Follow a topic across the collections. A shared tag is a way to compare objects, not proof that one culture influenced another.

The order uses the midpoint of each approximate date range; ranges can overlap. Spacing does not represent elapsed time. Swipe, scroll or use the arrows. With the sequence focused, use ← / →, Home or End.

THE CURRENT OBJECT

Its date is not sufficiently established to place it on the timeline. We keep it here, separate from the dated objects.

You are viewing

Modern linguistic reconstruction on historical and living languages

Before Bantu Worlds

Bantu-speaking worlds

Modern linguistic reconstruction; not an individual ancient document or object.

Current record

Languages1 dated object

First third of the 1st century BCE
Hand of Irulegui
Iberians
First third of the 1st century BCE; date of the destroyed dwelling beside whose entrance it appeared.
Read the record

1 / 1In the centre: Hand of Irulegui

Also connected, without a comparable date (5)

These objects are not placed after the others: an unknown date does not mean a more recent object.

49 Classical VoicesIberians · Authors and traditions between the 6th century BCE and 4th century CE
Ba-ntuBantu-speaking worlds · A long linguistic history; an academic term since the 19th century
Clicks and contactBantu-speaking worlds · precolonial contacts with varying chronologies and subsequent continuity
Indian Ocean KiswahiliBantu-speaking worlds · from the first millennium CE; major development in medieval and modern times
Voices from outsideGaramantes · Fifth century BCE–Late Antiquity

You can scroll the first group without JavaScript. The other topics are also available below.

Oral tradition

Current object, without a comparable date: Before Bantu Worlds.

Two songs, missing names · 15 July 1952 · Batwa / Twa histories
Batwa on a colonial record · 1954 · Batwa / Twa histories

Without a comparable date (not part of the chronological order):

Ba-ntu · A long linguistic history; an academic term since the 19th century · Bantu-speaking worlds
Living m’bolon · The present with historical memory · Mande Worlds
Memory in performance · 19th–20th centuries and transmitted practice · Bantu-speaking worlds
Musical bows · Twentieth-century ethnomusicological documentation · San histories
Soma stones · Rigvedic strata; later ritual tradition · Vedic and later Indian worlds
Sosso-Bala · Tradition linked through narrative to the 13th century; the present instrument and its custodianship have their own chronology · Mande Worlds

Archives and documents

Current object, without a comparable date: Before Bantu Worlds.

Boğazköy Archive · Above all 17th–13th centuries BCE · Hittites
Derveni Papyrus · copy conserved towards 340–320 BCE; comments on a previous text. · Thracians
Venus in the codex · Preserved manuscript: Postclassic, probably 13th–14th centuries; earlier knowledge · Maya worlds
Bamum document · 28 April 1922 · Bantu-speaking worlds
Two songs, missing names · 15 July 1952 · Batwa / Twa histories
Batwa on a colonial record · 1954 · Batwa / Twa histories
A map with the right to be closed · 2012–2013 · Batwa / Twa histories
Eleven articles · Adopted April–May 2022; public portal 2025 · Batwa / Twa histories

Without a comparable date (not part of the chronological order):

49 Classical Voices · Authors and traditions between the 6th century BCE and 4th century CE · Iberians
Dance with weapons · Towards 400 BCE. Xenophon's story · Thracians
Kumbi Saleh · medieval occupation; prominent phases between the 11th and 13th centuries · Mande Worlds
Shum Laka · c. 8,000 and 3,000 years before present · Bantu-speaking worlds
The Banquet of Seuthes II · Towards 400 BCE. Anabasis episode · Thracians
The cost of transcription · 19th century–current digital edition · San histories
Translating faith · 16th–18th centuries · Bantu-speaking worlds
Voices from outside · Fifth century BCE–Late Antiquity · Garamantes
Zalmoxis · testimony of Herodotus, 5th century BCE; no biographical date of the divinity. · Thracians

Memory and oral history

Current object, without a comparable date: Before Bantu Worlds.

Without a comparable date (not part of the chronological order):

49 Classical Voices · Authors and traditions between the 6th century BCE and 4th century CE · Iberians
Ba-ntu · A long linguistic history; an academic term since the 19th century · Bantu-speaking worlds
Memory in performance · 19th–20th centuries and transmitted practice · Bantu-speaking worlds
The cost of transcription · 19th century–current digital edition · San histories
Voices from outside · Fifth century BCE–Late Antiquity · Garamantes

A closer look

About 1 min to read · 1 section. Further reading and sources · free access. Editorial review recorded: September 5, 2026.

A plant name that predates the branching

Bostoen, Grollemund and Koni Muluwa examine the reconstruction *-téndé, related to the oil palm. Its distribution is not limited to Bantu languages: it also appears in related external branches, allowing a pre-Proto-Bantu inheritance to be proposed. In some eastern languages, continuators of the term came to designate other palms when the vegetation changed. The example offers a concrete window onto the preceding period: plant knowledge could be inherited and later transformed semantically. The relative antiquity of a word does not need to become an exact migration date.

Research · Comptes Rendus Géoscience · Bostoen, Grollemund y Koni Muluwa, Climate-induced vegetation dynamics and the Bantu Expansion, 2013Access recorded: September 5, 2026.

The region before this culture

Late Pleistocene–≈3000 BCE · West-central Africa · forests, savannas and networks prior to proto-Bantu diversification

Before the 3rd millennium BCE · Documented
Pre-kingdom Landscapes
Diverse communities occupy savannas and forests before an identifiable dispersal of Bantu languages.
≈3000–1500 BCE · Debated
Reconstructed Proto-Bantu
Linguistic comparison proposes an ancestral community and vocabulary, not a text nor a state.
II millennium BCE · Debated
Early diversifications
Western and eastern routes are modelled using evidence still under review.

What changes and what we should not assume

The museum’s start is not a departure date, but the gradual convergence of linguistic branches, populations and archaeological assemblages in several corridors.

Do not automatically attribute all Iron Age pottery or metallurgy to Bantu speakers, nor assume that agriculture, iron and language travelled together in a single movement.

The proto-Bantu reconstruction: an object that does not exist

Modern linguistic reconstruction on historical and living languages · West-central Africa · debated region of origin

Makes visible that the earliest evidence is primarily a relationship between languages, not a single archaeological relic.

Hundreds of languages show systematic correspondences that allow reconstruction of ancestral forms and branches.

A linguistic tree simplifies contact, bilingualism, displacement and adoption; it does not equal a biological tree of peoples.

Phonological series, compared vocabularies and phylogenetic models

Tracekeepers · ImageGen · OpenAI ImageGen · editorial direction Tracekeepers · Original visualization for the project

Bantu means: A branch of Niger-Congo languages and the societies that speak them; not a homogeneous civilisation
Method: Cross linguistics, archaeology, palaeoenvironment, genetics and oral history while maintaining their boundaries
Scale: More than five hundred languages and hundreds of regional histories, not a single demographic march
Caution: Ceramics, iron, cultivation, genes and language do not always move together

Location and territory

REGIONAL AREA

West-central Africa · debated region of originCurrent maximum linguistic envelope and historical corridors; it was not a single political territory nor was it occupied simultaneously. The footprint indicates the indicative centre of the sphere, not a border or an exact point.

Sources and references

Expand this antecedent and consult its sources

Did you enjoy this exhibit?

Your opinion helps us improve the museum. You can change it whenever you like.

Loading your rating…

History and context

Evidence and interpretation

Catalogue facts

Before Bantu Worlds

Hand of Irulegui

A closer look

A plant name that predates the branching

The region before this culture

Pre-kingdom Landscapes

Reconstructed Proto-Bantu

Early diversifications

What changes and what we should not assume

The proto-Bantu reconstruction: an object that does not exist

Location and territory

Sources and references

Did you enjoy this exhibit?