The origins of African farming lie in the northern
half of the
continent. One development that may have taken place very early is the
appearance of domesticated cattle in the Sahara, then a more benign
environment than it is today; in the eastern Sahara this may go back to
around 7000 B.C. This could have been an independent domestication. On
the Mediterranean coast to the north of the Sahara, it was the Near
Eastern farming package that spread, though at a somewhat later date.
In the savanna belt to the south of the Sahara, domesticated animals
were introduced from the north, but in a region of summer rainfall the
plants had to be local domesticates; here the establishment of farming
seems to have taken place within the period 4000-1000 B.C. Farther
south, in the tropical forest, a pattern of agriculture appeared that
was centered on the cultivation of domesticated yams; but as in
Amazonia, the chronology remains murky. Thus the appearance of farming
in Africa shows some clear instances of diffusion from the Near East,
but often it is hard to decide between stimulus diffusion and
independent development. By contrast, an aspect of material culture in
which the African savanna had clear priority over the Near East was the
development of pottery around 9000 B.C. (as against about 7000 B.C. in
the Near East).
The story of the spread of farming to southern Africa
is a much more
recent one. Here there is no sign of independent domestications, with
or without external stimulus. In one way this is surprising: the
savannas of eastern and southern Africa are the only part of the world
in which the wild herbivores of the late Pleistocene still survive on a
significant scale. But for whatever reason, none of them were suitable
for domestication. Instead, domesticates entered the region from the
north. Two distinct processes were involved: the colonization of the
south by farmers from the north, and the adoption of northern farming
by hunter-gatherer populations already in the south.
The first process is the better known. The evidence
for it comes from
two sources, archaeology and linguistics. The archaeological record
shows that by the second century A.D. there was a farming population
between Lake Victoria and the East African coast that also practiced
iron metallurgy. This population then spread rapidly down the east
coast of Africa, reaching Natal in less than two centuries. It stopped
there, presumably because its crops were adapted to the summer rainfall
of the tropics, as opposed to the winter rainfall of the south coast.
Gradually this or similar populations moved into the interior of
southern Africa, including eventually the west; but this last part of
the process is very little known.
The linguistic evidence consists in two points. First,
virtually all
the languages spoken by the black populations of southern Africa are
closely related, belonging to a single subfamily known as Bantu
languages. This is quite unlike the situation farther north in sub-
Saharan Africa: there the language map of black Africa shows no such
homogeneity, and extensive groupings of languages are much harder to
establish with confidence. Second, there is good reason to place the
homeland of the Bantu languages in this northern region, more precisely
in the northwest near the Atlantic coast. It is here that Bantu itself
is most deeply differentiated and that the non- Bantu languages most
closely related to it are found. Taken together, these points strongly
suggest that the Bantu languages entered southern Africa from the north
in the relatively recent past, perhaps within the last couple of
thousand years.
So the arrival of the iron-using farmers and the spread
of the Bantu
languages look like two sides of the same coin. But this conclusion,
though it may be irresistible, does raise a problem. The archaeological
evidence agrees with the linguistic evidence in suggesting a movement
from north to south; but whereas archaeology points to an origin in the
northeast, linguistics points to the northwest. How do we bring them
together? We can reduce the gap a little by noting that an iron- using
culture appears to the west of Lake Victoria earlier than it does to
the east. But to bridge the rest of the gap, we are reduced to positing
an early eastward migration of Bantu speakers in the direction of Lake
Victoria, a movement for which evidence is otherwise lacking. Be this
as it may, we are still on firmer ground than with Pama- Nyungan.
The other aspect of the spread of farming to southern
Africa concerns
its adoption by hunter- gatherers already living in the region.
Pastoralism (but not cultivation) was at some point adopted by people
of the same physical type as the Bushmen. Such brown pastoral
populations are known as the Khoekhoe peoples; they are still a
significant presence in the central and western regions of southern
Africa, and they were more extensive in the past. They speak click
languages of a family to which a number of Bushman languages also
belong. Archaeologically they are not well known, but their way of life
seems to go back a couple of thousand years. The origins of their
pastoralism must lie in northern Africa, whence their livestock
derives. The simplest hypothesis would be that the ancestors of these
pastoralists obtained their livestock from the iron-using black
farmers. But there are indications that their pastoralism may have
reached the south by a different route.