Archaeologists have long debated when early humans began hurling stone-tipped spears and darts at large prey. By throwing a spear, instead of thrusting it, humans could hunt buffalo and other dangerous game from a safe distance, with less risk of a goring or mauling. But direct evidence of this hunting technique in early sites has been lacking. A new study of impact marks on the bones of ancient prey shows that such sophisticated killing techniques go back at least 90,000 years ago in Africa and offers a new method of determining how prehistoric hunters made their kills.
Other researchers have used indirect methods to study the use of projectiles, such as analyzing impact fractures on ancient stone points or identifying traces left by hafting on the points. Such evidence suggests that early humans created throwing spears as early as 500,000 years ago in Africa. But that kind of evidence leaves room for doubt and is frequently disputed.
Archaeologist Corey O'Driscoll of South East Archaeology in Canberra became interested in the traces left by hurled spears after reading studies of the wounds that medieval weapons inflicted on humans. In preliminary work, European archaeologists had fired reproductions of Upper Paleolithic points made of antler at the carcasses of oxen and deer, then studied the marks that they left on the bones. But many archaeologists remained unconvinced by the findings, seeing little clear difference between projectile marks and cut marks from butchering. O'Driscoll decided to build on these studies for his undergraduate honors thesis.
He and a colleague knapped flint reproductions of spear and arrow points from the Middle Stone Age in Africa and attached them to wooden shafts. With a group of University of Queensland students, he ran 15 experiments, throwing replica spears and firing replica arrows with bows or a calibrated crossbow at lamb and cow carcasses. After boiling the carcasses or burying them for rapid defleshing by microbes and insects, O'Driscoll found 758 wounds on the bones, which he examined microscopically, and compared to 201 cut marks in an experimentally created reference collection of butchered animal bones.
He found "quite a difference between the butchering marks and projectile impact marks," he says. His study revealed six types of distinctive projectile impact wounds, from drag marks to fracture marks and punctures. O'Driscoll also noted that most projectile impact marks were located on vertebrae or rib bones and that 17% percent of the marks overall—and 50% of the punctures—held microscopic bits of embedded stone from the flint points, due to the high velocity of impact. By contrast, none of the butchering marks contained such stone fragments, another key distinction.
These findings prompted O'Driscoll and the University of Queensland's Jessica Thompson to take a new look at three bone specimens from large unidentified mammals—a rib and two vertebrae—from Pinnacle Point Cave in South Africa. Thompson had earlier detected embedded stone fragments in marks on these bones. Using O'Driscoll's diagnostic criteria, the pair identified projectile impact marks on all three bones. Two dated to between 91,000 and 98,000 years ago—making them the oldest direct evidence of the use of projectile weapons, according to a paper  presented at the Society for American Archaeology meeting in Honolulu in April. (O'Driscoll's thesis will be published by the Australian Archaeological Association in June.) A third bone dated even earlier, between 153,000 and 174,000 years ago.
"This is great work," says Curtis Marean of Arizona State University, Tempe, noting that the projectile impact marks, "have a clear and recognizable morphology."
Archaeologist Tiina Manne at the University of Queensland also finds the identification of projectile impact marks—at least on the two later bones—highly persuasive. "This strongly suggests that projectile technology at Pinnacle Point was in use by at least 90 to 95,000 years ago," she says. But she's less convinced by the evidence on the oldest bone, noting that only a "single grain" of stone from the projectile point was embedded in the bone.
Despite this reservation, Manne says that these "exciting" findings can help researchers recognize projectile impact marks on bone in many times and places. They have "incredibly wide-ranging applicability and the potential to further our understanding of when this technology was adopted elsewhere."