a flood mound 3 kilometres from Djenné, Mali

Djenné-Djeno

From about 250 BCE to the 14th century (a city before the trans-Saharan gold texts)

Three kilometres from the living town of Djenné, a mound in the flood of the Bani was occupied from about 250 BCE until the 14th century. At its height, between about 750 and 1100 CE, it covered more than 33 hectares. Historians had dated the city to the thirteenth century, a child of the gold trade across the Sahara.

Why this is extraordinary

Djenné-Djeno means old Djenné. It is a mound in the floodplain of the Bani, about 3 kilometres from the town with the famous mosque, and it was lived on from about 250 BCE to the 14th century. Excavation in 1977–1981 found the span continuous. Between about 750 and 1100 the settlement passed 33 hectares and the mound, mapped in 1977, stands up to about 6 metres. Iron and stone in the first layers were carried in from outside the delta. Around 400 CE the levels have copper. African rice is there early. People were farming, fishing, and trading on a flood island long before the written city.

The older story, read from when the name turns up in texts, had Djenné appear with Timbuktu in the mid-thirteenth century, a market for gold headed north. The heritage file still describes that later boom, in the 9th or 10th century, under merchants tied to northern demand. Both clocks are in the record. The mound’s radiocarbon is the earlier one. Calling 33 hectares a city is the remaining argument, because the place has no surviving stone monument. The excavators’ point is that mud towns will not look like a palace skyline, and that the urbanism is the cluster of mounds, the craft, and the food. A village that grew is the other name for the same layers. The living town has been inhabited since the 11th century. The old mound goes quiet in the 14th. They overlap, and they are not the same pile of earth.

The seated figure in the photograph is labelled Jenne-jeno, 900 to 1400. The head is missing. A label is not a trench, and a lot of delta terracotta was collected without one. It belongs to the later centuries of this culture’s art, not to the third century BCE. Prayer direction at the modern mosque is a qibla, a bearing toward Mecca, and it is the later town’s. No solstice is published for the mound. The mound sits where the flood does not cover. What this changes: the Inland Delta had a large settlement centuries before the caravan story. If that settlement is a city, urban life here is local and early. If it is a big village until the gold market, the dates still start at about 250 BCE, and the word is the part left to argue.

Why it matters

If the mound is right, towns in the Inland Niger Delta grew up on rice, fish, and regional trade centuries before Muslim merchants were supposed to have founded them. If the early dates are a thin scatter and the “city” is only the later centuries, the old story survives and the mound is a village that got large.

Digging from 1977 to 1981 found a continuous occupation across that whole span. The heritage evaluation describes an early phase of fishing and rice, and a later market phase. Iron and stone in the first levels were brought in from outside the delta. Levels around 400 CE already have copper, read as coming from Saharan sources. African rice is in the early deposits. The mound was mapped in 1977 at 33 hectares and stands up to about 6 metres high. In the 14th century the old mound was abandoned for Djenné, which itself had been lived in since the 11th century. The evaluation’s trade paragraph puts the gold traffic in the 9th or 10th century and ties the town’s merchants to demand from the north. The excavation’s point is that the imports start earlier than that paragraph. The figure in the photograph wears a museum label of Jenne-jeno, 900 to 1400. That is the later half of the mound’s life. A terracotta in a vitrine is not a wall. A great many delta figures were dug without a record. This one is evidence of a labelled object, and the label is part of what you have to weigh.

Photo tour

1 photograph, unedited. On a phone, swipe the row. Choose one to look closely.

No openly licensed photograph of the mound’s trenches is used. The modern mosque of Djenné is a different building and a different century.

What is still unexplained

These sit beside the consensus account, with the same visual weight. None of them is marked solved.

Open A town before the caravans The older historical reading, from the dates of the written mentions, had Djenné rising with Timbuktu in the mid-thirteenth century as a town of the trans-Saharan trade. Oral tradition of an eighth-century foundation was set aside. The mound’s radiocarbon says iron-using people were here in the third century BCE, and that the place was already large between 750 and 1100, before the thirteenth-century texts. Copper at about 400 CE is the awkward import. It is earlier than the gold-for-the-north story and it still comes from far away, so “local growth” and “long-distance goods” are not opposites. The catch is the word city. The excavators argue that this urbanism does not look like a walled capital with monuments. There is no stone palace. The evidence is density, craft, rice, and trade, across a cluster of flood mounds, not a single skyline. A sceptical reading can call 33 hectares a big village until the monuments turn up. The monuments may be mud, which does not stand. The modern mosque of Djenné is a twentieth-century rebuilding of a later town. It is not this mound, and it should not be used as a picture of 250 BCE.

Open No sky on the mound The living town’s mosque is oriented for prayer, which is a direction toward Mecca, a religious bearing of the Islamic centuries. The old mound was abandoned in the 14th century, and the evaluation places Islam’s hold at Djenné at the end of the 13th century. A qibla is not a solstice, and it belongs to the later town if it belongs anywhere. No azimuth of a house or a wall at Djenné-Djeno is published in the sources used here. Flood mounds are sited on the ground that stays dry. That is hydrology. The terracotta’s posture, seated, arms along the legs, has been given ritual readings in the delta’s art. A rite is not a star. This room does not add one.

Location and dating

Location: A toguere, a flood mound, in the Bani basin about 3 kilometres southeast of Djenné. The coordinate is the mound, about 13.89 north and 4.54 west. Sister mounds at Hambarketolo and elsewhere are part of the same cluster.
Period label: Current consensus — open to revision About 250 BCE to the 14th century CE. The largest extent is about 750 to 1100. Djenné the living town is inhabited from the 11th century. The gold-trade story that put the foundation in the 13th century is the one the mound challenged.
How this dating was done: Radiocarbon from the 1977–1981 excavations. Imports, rice, and iron for what the dates are dating. The museum figure’s 900–1400 is its label, not a new assay.
What could change it: A published section that ties the copper to a single hearth, rather than to “levels dated to about 400,” would lock the early trade. A house plan with a measured axis would be the first sky claim. None is used here.

Published ranges, drawn as bars. A bar is a claim about a span, not a point.

Occupation

Greatest extent

250 BCE500 CE1400 CE

Occupation. Continuous occupation in the heritage evaluation and the excavation summary. Not one sample’s error bar.
Greatest extent. When the settlement exceeded 33 hectares.

What is actually there

The mound is up to about 6 metres high and was mapped at 33 hectares in 1977.
Iron and stone in the earliest levels were imported into the delta.
Copper appears in levels dated to about 400 CE.
African rice is in the early occupation.
The labelled terracotta is 900–1400 on its museum card, and headless.

Open questions

Question	Status
Is 33 hectares a city if it has no surviving monument?	Open. The excavators say urbanism here will not look like a stone capital. Mud does not last.
Does copper at about 400 CE mean the Sahara trade is already running?	Open. The metal is foreign. The gold texts are later. One import is not the whole caravan.

The current narrative

Where publication stands today. It is the best-supported reading, and it is open to revision. It is not a verdict.

The current narrative — open to revision

A regional town from the third century BCE

Djenné-Djeno was a large settlement from about 250 BCE to the 14th century, fed by rice and fish and already trading beyond the delta by about 400 CE. It was not founded by the thirteenth-century gold caravans.

What supports this

Radiocarbon from the 1977–1981 dig spans that occupation, and the peak size, over 33 hectares, falls between about 750 and 1100.
Imported stone and iron at the start, and copper around 400 CE, are goods the delta does not produce.
Rice in the early layers ties the growth to a local staple, not only to luxury gold.

What pushes back

Written Djenné is late. Archaeology and texts can be about neighbouring places, the mound and the later town, and the evaluation itself says the living town was inhabited from the 11th century while the mound goes quiet in the 14th.
“City” without monuments is a definition. Density is not a palace.
The terracotta in museums, including the one pictured, often lack a trench. They illustrate a style. They do not date the first occupation.

What would settle it

It is the centre because the dates and the imports were excavated on the mound, and they are earlier than the thirteenth-century story. It does not need every museum figure to be from a recorded pit.

Emerging narratives

Other published readings. Each may be right. None is adopted here unless the evidence carries it.

An emerging narrative

A town of the gold trade, the early layers a village

The real city is the later market, 9th century and after, and the third-century BCE occupation is a hamlet that the word urbanism has stretched.

What supports this

The heritage text still describes a market boom with the gold trade of the 9th or 10th century. That phase is not imaginary.
Oral tradition of an eighth-century start, which historians doubted, is closer to the trade centuries than to 250 BCE. Doubt about traditions cuts both ways.
Thirty-three hectares at the peak is a late first-millennium size. The early mound may have been much smaller. The excavation says it grew.

What pushes back

Iron-using occupation in the third century BCE is a date, not a metaphor. Calling it a hamlet has to say what size the early layers actually are, and the published point is that growth was early.
Copper around 400 CE is already long-distance. The trade does not wait for the 9th century even if the gold texts do.
Abandonment in the 14th century for a town that had existed since the 11th shows two sites, not one late foundation.

What would settle it

It stays because the gold-trade paragraph is in the same heritage file as the 250 BCE date, and they have to be kept distinct. It is not the centre for the beginning.

Should the current view be reevaluated?

Open to revision

Occupation from about 250 BCE and a peak over 33 hectares are excavated results. Whether that is a city, and how much of the Sahara trade is already in the copper of about 400 CE, are the open edges. The museum figure is a labelled terracotta of the later centuries, not the mound. No sky alignment is claimed for the old site.

What drives this

Radiocarbon occupation from about 250 BCE to the 14th century, peak size after about 750.
Imported iron and stone at the start, copper around 400 CE, rice in the early layers.
A thirteenth-century foundation story that the mound’s dates are earlier than.

This is the museum’s own weighing of the evidence. It is not a visitor vote, and it is separate from the slider.

Weigh it yourself

Anonymous. Nothing here is stored except on this browser. Your lean does not change the assessment beside it.

A regional town from the third century BCE Unsure A town of the gold trade, the early layers a village

If the contested reading is right

What would follow if a reading that is not the consensus turned out to be true, and what could settle it. This explains the stakes. It does not adopt the reading.

The early mound is already urban

West African towns would be a first-millennium development on local staples and regional trade, not a side effect of the later gold caravans. Other flood mounds in the delta would be part of that city-region, not empty until the texts.

Why it is still worth a test

House-by-house dates for the first centuries, so the size at 250 BCE is a measurement rather than the start of a long span.

Sources

ICOMOS evaluation, Old Towns of Djenné UNESCO, the 250 BCE to 14th century occupation
The Inland Niger Delta before the Empire of Mali Journal of African History, the 33 hectares and the copper around 400 CE

Image credits

A seated figure labelled to Jenne-jeno, about 900 to 1400, in a museum. The head is gone. Applied discs cover the limbs. Many Inland Delta terracottas reached collections without a trench record, so this label is a provenance, not a photograph of the mound. . Daderot, Public domain. Shown unedited.

← Mali World map