Reconstruction
Kanjera South sits on the Homa Peninsula in southwestern Kenya, and its deposits, dated to roughly two million years ago, produced three large and well-preserved bone assemblages. They pre-date every previously published archaeofauna of comparable size, which is what makes the site decisive rather than merely early. Isolated evidence of butchery already existed from older sites in Ethiopia and Kenya, but isolated evidence cannot distinguish a lucky carcass from a way of life. Kanjera can, because the same pattern of hominin activity runs consistently through a stratified sequence covering hundreds of thousands of years. That consistency is the finding: the earliest archaeological evidence of sustained involvement with fleshed animal remains, or persistent carnivory. Two details give the diet its shape. First, the hominins acquired and processed numerous relatively complete small ungulate carcasses. Completeness matters more than size here. A scavenger arriving late to a kill finds a skeleton already stripped by carnivores and birds, so a substantially intact small antelope implies early access, either by hunting it or by reaching it before competitors did. Alongside these, there was at least occasional access to the fleshed remains of larger animals in the wildebeest size range. Second, and more revealing, the site records the treatment of heads. Antelope skulls were not simply broken where the animals died. They were transported some distance to the site and only then broken open, so that the brain inside could be eaten. This is the earliest archaeological evidence of that kind of resource transport in the human lineage, and it says two things at once. It says these hominins recognised value that is invisible from outside a bone, since a skull gives no external sign of what it contains. And it says they were willing to carry food away from where they found it to a place where processing was easier or safer, which is a different cognitive proposition from eating on the spot. The nutritional logic is fat. Lean meat is seasonally abundant on an African savannah but fat is chronically scarce, and an animal running out of fat reserves cannot metabolise unlimited protein safely. Brain tissue and marrow are the two concentrated fat sources a carcass offers, both sealed inside bone, both reachable only with a hammerstone. A hominin population that systematically opened skulls and limb shafts was not competing with big cats for muscle; it was harvesting what the cats left behind and could not reach. Kanjera South is where that strategy stops looking opportunistic and starts looking like a subsistence pattern maintained across many thousands of generations.
Historical context
Around two million years ago, early Homo shared eastern Africa with Paranthropus boisei, and Homo erectus was appearing. Fire was not in controlled use. The Homa Peninsula carried grassland and wooded patches beside a large lake, grazed by antelope and wildebeest and patrolled by sabre-toothed cats, hyaenas and crocodiles. Oldowan tools were in use across eastern Africa. The first hominin departure from the continent, recorded at Dmanisi in Georgia, lay roughly 150,000 years ahead.
Evidence
Zooarchaeological
StrongThree large, well-preserved assemblages with cut-marked and percussion-damaged bone, larger than any previously published archaeofauna of this age, allowing patterns rather than single events to be read.
Food identification
StrongSmall ungulates dominate and are identified taxonomically; wildebeest-sized animals appear occasionally. Brain and marrow consumption is evidenced by skull and limb-bone breakage patterns.
Dating
ModerateThe stratified sequence is dated to approximately two million years by combined stratigraphic, palaeomagnetic and faunal methods.
Preparation method
StrongOldowan tools occur with the fauna, and skull and long-bone breakage indicates percussive opening with hammerstones rather than carnivore damage.
Historical interpretation
ModeratePersistent carnivory is inferred from consistency across the stratified sequence. Whether the relatively complete carcasses were hunted or obtained by early scavenging is not resolved by the assemblages.
Sources
- 1.Joseph V. Ferraro, Thomas W. Plummer, Briana L. Pobiner, et al. (2013). Earliest Archaeological Evidence of Persistent Hominin Carnivory. PLoS ONE 8(4), e62174. doi:10.1371/journal.pone.0062174Scientific literature
- 2.Smithsonian Institution Human Origins Program (2025). Kanjera, Kenya. Smithsonian National Museum of Natural History. humanorigins.si.edu/research/east-african-researModern synthesis
Limitations
- moderatehistorical interpretation
Relatively complete small carcasses imply early access but do not demonstrate hunting. Some coverage of this site describes it as the earliest evidence of hunting; the published claim is persistent carnivory, which is a weaker and better-supported statement.
- minorarchaeological attribution
The toolmakers are not identified by fossils at the site. Early Homo is the conventional attribution, but Paranthropus boisei was present in eastern Africa at this date.
