Archaeologists call them tells - artificial hills built not by quarrying but by living. As generation after generation built, collapsed, and rebuilt in mud brick, the ground level slowly rose, sealing the past beneath the present. Dia is really three of these sites clustered together: Dia-Shoma, Dia, and Dia-Mara. Dia-Shoma is the eldest and the largest, sprawling across 49 hectares and reaching back to around the 9th century BCE. Settlement at the complex peaked roughly in the 10th century CE, by which time Dia had been continuously inhabited for the better part of two millennia. The neighboring mound of Dia-Mara is a relative newcomer, with occupation beginning only in the 6th century CE - young, by the standards of this remarkable place.

Much of what we know was nearly lost. Beginning in 1980, researchers set out to inventory the archaeological sites scattered across the Inland Niger Delta before erosion, farming, and looting could erase them. The vicinity of Dia was chosen in 1998 as the focus of a long-term Malian-Dutch cultural heritage program led by the Rijksmuseum voor Volkenkunde in Leiden. Their excavations revealed how an early West African city was structured - how its people lived, traded, and buried their dead. Dia sat on a vital artery: the water system that linked the southern savanna to the northern Sahara made it one of the key trading posts of the region, a place where goods and ideas from across the continent changed hands.

Dia is often compared to Djenné-Jeno, and the two do share certain pottery styles - the same rim shapes, the same red slip on their vessels. For a time this looked like proof of a single connected people. But the closer archaeologists looked, the more the differences emerged. The decoration styles diverge. Djenné-Jeno's distinctive funerary urns, in which the dead were buried, are completely absent at Dia - and with 86 inhumations excavated at Dia-Shoma alone, that absence cannot be chalked up to chance. Different burial customs point to different communities, neighbors along the same river who shaped clay in similar ways but laid their dead to rest by their own traditions. The mounds, it turns out, hold an argument as much as a history.

Dia raises a delicate question: who gets to tell its story? Oral history names the Bozo - hunters and fishermen - as the region's first inhabitants, and Dia today is governed by the Soninke rather than the Fulani who dominate the surrounding Macina. But the ground tells a story the spoken record sometimes resists. Excavations at the Shoma cemetery found most of the dead buried in clearly non-Islamic positions, with the few Islamic-style graves confined to the surface and dated only to the 18th and 19th centuries. Terracotta figurine heads from nearby Mara hint at a tradition of honoring ancestors that an exclusionary new faith would later condemn. As Islam took deep root from the 19th century onward, communities had reason to forget older practices - and Dia, whose people cherish the memory of their Muslim ancestors, shows how oral history itself can be quietly rewritten. The mounds remember what the stories have let go.