For a long time many European scholars insisted that Africa south of the Sahara had no history worth studying. The German philosopher Hegel, in lectures published in 1837, waved the continent away in a sentence: Hegel claimed that Africa "is no historical part of the World; it has no movement or development to exhibit." As late as 1963 a famous Oxford historian told a television audience that, apart from the story of Europeans in Africa, there was little there for historians to study. Such claims rested less on evidence than on an assumption: that history means written documents, and that people without libraries have no past. African history as a modern field grew up to answer that assumption, and it did so by widening what counts as a source.

Written sources and their limits

Africa was never a continent without writing. Egypt had hieroglyphs, Kush developed its own Meroitic script, and Ethiopia wrote in Ge'ez, which the Ethiopian Orthodox Church still uses in worship. After the spread of Islam, scholars across the Sahel and the East African coast wrote in Arabic, and many also wrote their own languages in Arabic letters. Ajami is the writing of African languages such as Hausa, Fulfulde and Swahili in Arabic script. In the 1830s a Vai man in what is now Liberia, Momolu Duwalu Bukele, invented a new script for his language from scratch.

Outsiders also wrote about Africa: Arab geographers, Portuguese sailors, missionaries and colonial officials. Their accounts are precious, but each was filtered through the writer's purposes and prejudices, and many writers never left the coast.

Oral tradition

Most African societies kept their past in spoken form: king lists, genealogies, origin stories, praise poems, epics and proverbs, often in the care of trained specialists. The Belgian historian Jan Vansina argued in the 1960s that oral traditions can be tested like written documents: by comparing versions, asking who told them and why, and checking them against other kinds of evidence.

Testing reveals typical distortions. Traditions usually remember the founding of a people and the last few generations in detail, while the stretch in between shrinks or vanishes. Vansina called the thinly remembered period between origins and recent times the "floating gap." Traditions also serve present needs, such as justifying the claim of a ruling family to its throne. And they change through contact with print. Historians call it "feedback" when an oral tradition absorbs material from written sources such as the Bible, school textbooks or an earlier historian's book, so that what sounds ancient may be recent.

Archaeology and dating

Archaeology recovers what nobody wrote down: house floors, iron furnaces, pottery, glass beads, burned seeds and animal bones. For Africa its value grew enormously after a breakthrough in physics. Radiocarbon dating, developed by Willard Libby in the late 1940s, lets archaeologists date charcoal and other organic remains. From the 1950s onward, sites with no written record at all could be placed in time, and whole chapters of history opened up.

Language, genes and climate

Languages are archives too. Historical linguistics compares related languages to reconstruct the vocabulary of their shared ancestor. If the descendants of an ancient language all share words for goat, yam and canoe, its speakers probably kept goats, grew yams and fished. Genetics adds another layer by tracing ancestry and movement in the DNA of living and ancient people. Ancient samples are scarce, because heat and humidity degrade ancient DNA, so far fewer old genomes survive from the tropics than from cold regions.

Building a field

African scholars led much of this work. The Nigerian historian Kenneth Onwuka Dike, whose 1956 study of trade and politics in the Niger Delta combined oral evidence with archives, later led the University of Ibadan. In 1964 UNESCO launched the General History of Africa, eight volumes written mostly by African scholars and published between 1981 and 1993. Today the field's working rule is simple to state and hard to practice: no single kind of source is enough.

Worked example: the bronzes of Igbo-Ukwu

In 1938 a man named Isaiah Anozie was digging a water cistern in his compound at Igbo-Ukwu, in southeastern Nigeria, when he struck pieces of decorated bronze. Excavations by the archaeologist Thurstan Shaw in 1959 and 1960 found much more: the burial of an important person seated on a stool, a store of ritual objects, and bronzes cast with astonishing fine detail by the lost-wax method. The site also held tens of thousands of glass and stone beads.

No document mentions Igbo-Ukwu, so each kind of evidence answers a different question. Radiocarbon dates from the site point to around the ninth or tenth century AD, earlier than many scholars had expected for such work. Chemical analysis suggests the metal came from local ores, so the casting was a local achievement. Some of the beads, however, appear to have come from far away, possibly Egypt or beyond, which implies links to long-distance trade across the Sahara. Igbo oral traditions about Nri, a kingdom whose priest-kings held ritual rather than military authority, suggest who the buried person might have been, though that link remains a hypothesis. Together they reveal a wealthy, skilled and connected society that a history built only on documents would miss.