The monkey picks up a potato-sized rock in his tiny hands, raises it above his head and smashes it down with all his might on another stone embedded in the ground. As the creature enthusiastically bashes away, over and over, flakes fly off the rock he is wielding. They are sharp enough to cut meat or plant material. The monkey does not pay much attention to the flakes, save to place one on the embedded rock and attempt to smash it, too. But he has unintentionally produced artifacts that look for all the world like stone tools found at some human archaeological sites.

The monkey is a wild capuchin in northeastern Brazil’s Serra da Capivara National Park, where these animals have long been known to use rocks for a wide range of activities, from cracking open nuts and digging for roots to catching the attention of potential mates. Other nonhuman primates, including West African chimpanzees, also use rocks as tools in the wild. But the Serra da Capivara capuchins are the only ones that scientists have seen banging rocks together to break them—an activity previously thought to be exclusive to members of the human family. Humans do it to create sharp-edged tools for cutting things. The capuchins, in contrast, never use the flakes they make. Exactly why the monkeys want to break the rocks is unclear, but they often pause from smashing to lick the surface of the embedded stone, perhaps in pursuit of mineral dust generated by the impact.

Now a new study has examined the capuchin-produced stone flakes and compared them to human-made artifacts, and it turns out that the chips meet criteria used to distinguish human tools from naturally broken rocks. The findings, published in the October 20 Nature, could fuel debate over controversial archaeological sites in Brazil that are said to have some of the earliest evidence of humans in the New World. The discovery also raises questions about what differentiates humans from other primates, and how our lineage started fashioning tools from stone.

The capuchins bash quartize rocks together, apparently to pulverize them so they can ingest the mineral powder. They do not use the sharp-edged flakes that come off the rocks while bashing. Credit: T.PROFFITT, M. HASLAM, Primate Archaeology Group (University of Oxford)

Tomos Proffitt of the University of Oxford and a group of colleagues carried out the new study. They observed the capuchins selecting stones to use as hammers from a rock outcrop, a conglomerate of sandstone and rounded chunks or “cobbles” of quartzite, and watched as the animals struck the hammer stones against cobbles still embedded in the conglomerate. The researchers later retrieved the fragmented stones, and collected other such artifacts found on the surface and in excavations within the surrounding area—just as they would if they were excavating a human archaeological site. They then analyzed this collection of 111 capuchin artifacts, examining their shapes and sizes as well as the nature of the scars left behind by all the bashing. They were even able to match the fractured flakes to the stones from which they originated.

Remarkably, the team found that the capuchin artifacts exhibit distinctive scoop-shaped or “conchoidal” flaking and sharp cutting edges, and that the monkeys often removed multiple flakes from a single rock—all hallmarks of human-made stone tools. (The authors note that stone fragments produced during chimpanzee nut-cracking, in contrast, lack most of the diagnostic criteria, as do flakes produced by captive bonobos that have been taught to knap.)

Experts have previously linked such characteristics to the emergence of humanlike hands and coordination, and to shifts in human cognition. But the fact that monkeys produced rocks with these same traits demands a different evolutionary explanation. And if modern-day monkeys modify rocks in this way it is possible that extinct monkeys and apes did too, leaving behind archaeological assemblages of their own. Archaeologists thus need to refine the criteria they use to identify stone tools intentionally produced by members of the human family, Proffitt and his colleagues argue.

“It’s clever that they studied monkey tools the same way we study [human] tools,” says archaeologist Sonia Harmand of Stony Brook University, who was not involved in the new research. “Many people are going to be disturbed that these tools can be made by capuchins,” she adds, noting that video footage shot by Proffitt’s team provides solid proof. According to Harmand the monkey artifacts would not look out of place at East African sites containing tools made by human ancestors in one of the earliest technological traditions, called the Oldowan, which dates back to 2.6 million years at the site of Gona in Ethiopia. The capuchin flakes resemble the simplest examples of Oldowan technology. But other Oldowan stone tools exhibit considerably more sophistication and planning, she says. The monkey artifacts also diverge from the oldest known stone tools in the world, 3.3-million-year-old implements that Harmand and her team excavated from the site of Lomekwi in Kenya. The Lomekwi tools are far larger, and are made of basalt and phonolite—rocks that are denser than the quartz and quartzite rocks the capuchins use.

Some experts wonder whether the capuchins’ flakes could spark doubts that members of the human lineage made the oldest stone tools. Although researchers have attributed the tools to human ancestors, the sites lack diagnostic fossils to establish the connection. “We have no clue” who created the material at Lomekwi and Gona, asserts archaeologist Wil Roebroeks of Leiden University.

But in a commentary accompanying the team’s Nature paper, Hélène Roche of Paris-Nanterre University states that the capuchin findings should not raise suspicions about who produced the early stone tool assemblages found in Africa. Archaeologists have studied hundreds of those assemblages. Many of them are accompanied by cutmarked bones that show how tools were used, and fossils that indicate human ancestors made them, among other contextual clues. The monkey artifacts could, however, necessitate reanalysis of enigmatic modified stones found at the archaeological sites of Pedra Furada—which are near the capuchins’ home in Serra da Capivara National Park—Roche told Scientific American.

Archaeologists have been arguing over the Pedra Furada sites for decades. Some contend that they demonstrate a human presence in Brazil more than 20,000 years ago, long before the Clovis hunters who were once thought to be the first humans to colonize the Americas, around 13,000 years ago; others, including James Adovasio of Florida Atlantic University, warrant that the “tools” there are just rocks that shattered accidentally when they eroded out of a cliff face and fell to the ground below. “The Pedra Furada stuff is not even up to capuchin standards,” Adovasio quips.

Andrew Hemmings, who is also based at Florida Atlantic, thinks ruling out the gravity explanation should be the first order of business at Pedra Furada. But he suspects that distinguishing between naturally broken stones and capuchin-made artifacts may be impossible. “This research highlights exactly how large the gray area spanning natural stone breakage and intentional [human] and/or other animal behavior truly is,” he says.

For his part, Eric Boëda of the University of Paris in Nanterre, who leads the excavations at Pedra Furada, says he is not worried that the monkeys produce stone flakes. He insists that the artifacts at Pedra Furada are far more complex and diverse than the “mere splinters” the capuchins make, and that his stones show traces of actually having been used on meat and plants.

The Pedra Furada controversy aside, the capuchin findings add to a growing list of discoveries that have been steadily eroding the line between humans and other primates. “They help illuminate capabilities of our primate brethren that we thought only we and our immediate ancestors had,” Adovasio says. “They make us rethink how special we are.”

Yet in other ways the capuchin handiwork throws the divide between nonhuman primates and ourselves into higher relief. Researchers agree that the key difference between the capuchin-made artifacts and human-made ones is that the latter were produced intentionally, with a purpose in mind. For the capuchins, sharp-edged flakes appear to be disposable byproducts of their quest for quartz dust. For early humans, they almost certainly aided survival by facilitating access to food.

Although the capuchin discovery demonstrates that nonhuman species can accidentally produce fragments of rock that look just like human-crafted cutting tools, that does not mean the human-made tools are not special, Harmand cautions. Even if human ancestors started creating flakes by mistake like the capuchins do, there was something that made them realize they could put them to use and even make new tools to suit their purposes. Moreover, human technology evolved from the comparatively simple tools seen at Lomekwi and at Oldowan sites to handaxes with carefully shaped cutting edges a million years later, and eventually to the elaborate machinery we have today. Why didn’t technology evolve to the same degree in chimps and monkeys, Harmand asks. Why did humans alone take it to such an extreme?

Definitive answers to those questions may prove elusive. In the meantime, Proffitt is eager to dig deeper into the capuchins’ rock-smashing activities. “We really need to understand why the capuchins are so interested in producing and ingesting quartz dust, as this is a very unique tool use behavior amongst primates,” he says. It is possible they are eating the pulverized quartz as a means of self-medication, either compensating for mineral deficiencies caused by intestinal parasites or damaging those parasites with abrasive dust particles, Proffitt speculates.

He is also eager to determine how long capuchins have been using rocks this way. Other evidence demonstrates that they have been using the cobbles to crack open nuts for at least 700 years. And chimpanzee stone tools from the Ivory Coast of West Africa date back to 4,300 years ago. Beyond that “we have no evidence of what ancient monkeys or great apes were doing,” Harmand observes. Which leaves plenty of room for more surprises in the future.