Before the advent of DNA testing, scholars relied on shipping records that listed the African ports from which slaves were exported to determine where in Africa the African-descended population of the United States originated. But these lists were quite limited because they noted only the port of departure and not the actual community from which the enslaved were taken.
Advancements in DNA analyses, along with African shipping records, have revealed that African Americans do not have roots in the entire continent. A relatively small number of African groups supplied the lion’s share of the ancestral African population.
In fact, three large regions of Atlantic Africa were the major contributors to the slave trade: Upper Guinea, including the modern countries of Senegal, Mali, Gambia, Guinea, Sierra Leone and Liberia; Lower Guinea, including the southern portions of Eastern Ivory Coast, Ghana, Togo, Benin and Nigeria; and West Central Africa, which encompassed mostly the western portions of the Democratic Republic of Congo and Angola. In all, these regions made up only about 15 percent of Africa’s total area, all on the Atlantic side of the continent.
People were once skeptical of claims made by early DNA ancestry-tracing services that they could identify a subject’s “tribe” or “ethnicity” in Africa; the available data didn’t seem to sustain such claims. But new ways of calculating ancestry from the genome and larger African samples can make determining ethnic identifications more accurate.
The Language Connection
Today, speaking a common language is the primary way to identify an African tribal or ethnic affiliation. Since African languages are quite stable and reports of these languages demonstrate that there has not been any large population movement within the slave-exporting region of Africa in the past 400 years or so, it should be relatively easy to match modern ethnicities or tribes with those of the slave-trade era. However, the names of these languages and ethnic groups have changed over that period. For example, in 1767 a German missionary named Christian Georg Andreas Oldendorp did a survey of slaves living in the Danish West Indies to try to determine which languages should be used for evangelical purposes. The Danish West Indies received slaves from the same shipping route that North America used.
Oldendorp, calculating ethnicity by language, listed 30 apparently different languages (his terminology sometimes makes it unclear where political and where linguistic units divided), and he provided vocabulary for 26 of these languages, which allows us to be certain of the modern equivalent.
In the Americas, Africans were most likely to form social units with other people who spoke their language, even if they might belong to different political units; in Africa their identity was more likely connected to a political unit. Their rulers collected taxes, demanded service (including the military service that resulted in their enslavement) and rendered justice, while neighboring polities might well be hostile even if they spoke the same language.
People collecting information about identity in America were likely to choose linguistic units, while those commenting on it in Africa were more likely to focus on political units. This created an interesting paradox: The names of African “nations” in America often did not match exactly with the names of “nations” in Africa.