A puzzling mix of artifacts raises questions about Homo sapiens' travels to China

More than half a century ago, Chinese researchers uncovered thousands of pieces of an ancient cultural puzzle.

Their summertime excavation about halfway up a hill overlooking northern China’s Shiyu River unearthed sharp-edged flakes that had been rapidly pounded off small rocks, a common Stone Age practice in the region. Yet the same sediment also contained more complicated types of stone implements.

Another unexpected discovery, part of a round piece of graphite with a hole in its center, resembled a large button. A chiseled bone, possibly a tool, also turned up, along with the bones of horses, gazelles and other animals.

To top it off, the investigators found a piece of bone that they identified as a Homo sapiens braincase.

The unusual mishmash of artifacts left the Chinese scientists unable to say precisely what had happened at the Shiyu site, where temperatures stay frigid for much of the year, and how long ago toolmakers hung out there.

That puzzle received little scientific attention until the Shiyu site and its surviving array of stones and bones received fresh scientific scrutiny 50 years after the original excavation.

A new report based on that project portrays last century’s finds at Shiyu as the oldest evidence of H. sapiens in northeast Asia. Shiyu artifacts include rectangular stone implements, called blades by archaeologists, and other elements of what’s known as Initial Upper Paleolithic culture, which has previously been linked with H. sapiens, the scientists report January 18 in Nature Ecology & Evolution.

Around 45,000 years ago, hunting groups that had followed animal herds through Siberia and Mongolia turned south and reached a river valley where Shiyu is located, say archaeologist Shi-Xia Yang of the Chinese Academy of Sciences in Beijing and colleagues. From Shiyu, human groups with roots in Africa rapidly forged east to Korea and Japan, researchers suspect.

Shiyu’s unusual artifact array reflects a blending of Stone Age cultures, they contend. H. sapiens newcomers adapted to new surroundings and new neighbors by creating a hybrid toolkit. They combined toolmaking practices carried across northern Eurasia with tried-and-true implements made by local Homo groups, possibly Neandertals or Denisovans.

Shiyu’s mix of stone tools and other artifacts “represents an exceptionally rare opportunity to identify ancient cultural hybridization in Asia,” says archaeologist Evgeny Rybin of the Russian Academy of Sciences’ Institute of Archaeology and Ethnography, Siberian Branch, in Novosibirsk. Rybin does not belong to Yang’s team.

The new report still leaves big gaps in the Shiyu puzzle. It’s possible that Neandertals or Denisovans were the makers of all the Shiyu artifacts, not just the simpler ones. If that was the case, no cultural blending with H. sapiens foreigners occurred. Or shifts in Shiyu sediment layers over time mixed artifacts from occupations at different times by local groups and H. sapiens, muddying the who-did-what-when picture that researchers are now trying to piece together.

Middle East connection

Much of the argument for H. sapiens’ cultural mingling at Shiyu rests on the presence of what Yang’s team regards as Initial Upper Paleolithic tools.

A 1988 publication based on discoveries at an Israeli site called Boker Tachtit introduced the Initial Upper Paleolithic, or IUP for short. Many archaeologists have since treated IUP artifacts as signs of a cultural transition that occurred as H. sapiens groups trekked from Africa into Eurasia between around 60,000 and 30,000 years ago. That period, sandwiched between two ice ages, featured several shifts from cold, dry conditions to a warm, wet climate that would have aided long-distance travel.

Tools and ornaments unearthed at the Israeli site, which date to around 50,000 years ago, suggested that IUP traditions emerged alongside a much older, Middle Paleolithic way of life. Discoverers of IUP artifacts viewed them as H. sapiens’ first steps into Upper Paleolithic cultural practices, which lasted in various parts of the world until about 12,000 years ago.

New-fangled IUP stone blades and triangular points appeared near sharp-edged stone flakes that had been pounded off rocks, known as cores, with prepared striking surfaces. Flakes and cores have a Middle Paleolithic pedigree, dating to as early as 300,000 years ago at sites in Europe, the Middle East and parts of Africa. Blades and points gained favor starting around 50,000 years ago.

Boker Tachtit investigators also saw signs of an increased interest in items with symbolic meanings among the remnants of IUP culture. Perforated seashells found at the Israeli site, once strung from necklaces, reflected novel IUP social or ritual behaviors, researchers suspected.

Mystery toolmakers

Discoveries in Turkey, southeastern Europe and western and central Asia have since been grouped under the IUP umbrella. Researchers generally attribute IUP artifacts to H. sapiens, although many sites — including Boker Tachtit — have yielded no fossils of their potential hominid toolmakers.

A cave site in southeastern Europe represents one exception. H. sapiens fossils found there, along with IUP artifacts, date to between about 46,000 and 44,000 years ago (SN: 5/11/20).

A 21st century revolution in ancient DNA analysis further ramped up uncertainty about who made IUP items. Investigations now indicate that H. sapiens at least occasionally interbred with Neandertals and Denisovans during IUP times. Any of those populations, or their hybrid offspring, might have made IUP items (SN: 8/22/18).

Only one other Chinese site, located about 500 kilometers west of Shiyu at the edge of the Mongolian Plateau, includes IUP stone blades. Those discoveries date to between 42,000 and 41,000 years ago. Prior to the new Shiyu report, fossil and ancient DNA evidence indicated that H. sapiens reached northeastern China’s Xiamabei site by around 40,000 years ago (SN: 4/4/07).

Shiyu’s unusual array of finds fits a scenario in which H. sapiens — already known to have arrived in southeastern Asia between around 120,000 and 60,000 years ago — took a separate route into northeastern Asia before mixing IUP-style blademaking with simpler tool practices of a native population, possibly Denisovans, says archaeologist and study coauthor Michael Petraglia of Griffith University in Brisbane, Australia. Previous excavations have indicated that whoever already lived in the Shiyu region made tools by striking sharp flakes off small, locally abundant rocks with handheld stones.

Regional variations in IUP tools, often influenced by the quality and size of available rocks, “show that once IUP populations spread [across Asia], they adapted to local circumstances, altering their behaviors and culture,” Petraglia says.

An unusual mix

Piecing together the Shiyu puzzle required taking a new, thorough look at the site and its previously excavated artifacts.

Shiyu’s original excavators briefly described their finds in a 1972 Chinese-language report. They had no way to generate reliable age estimates for what they had uncovered.

Aside from the challenge of establishing dates for that material, Yang’s team faced the sad reality that many Shiyu stones and bones had been lost over the years.

The 1963 dig had unearthed more than 15,000 stone artifacts, thousands of animal bones, that black disc with a hole carved in its center, the possible bone tool and the piece of a braincase. A biological anthropologist on the team assigned that fossil to H. sapiens.

A portion of the Shiyu finds, including 750 stone artifacts, 152 animal bones, the black disc and the bone implement were taken to the Institute of Vertebrate Paleontology and Paleoanthropology in Beijing, where Yang now works. The remaining finds were held in small scientific facilities near Shiyu. Somewhere along the way, most of that material — including the proposed H. sapiens fossil — went missing.

Yang’s team visited Shiyu, a well-known site in archaeological circles, in 2013. The scientists generated dates for sediment layers at the site, including a roughly one-meter-thick layer where stone artifacts and animal bones had been excavated 50 years earlier. Age estimates relied on measures of the approximate time since sediment had last been exposed to sunlight and radiocarbon dates for newly unearthed animal remains.

An analysis of surviving Shiyu finds conducted by the researchers identified a portion characterized by an IUP transitional mix of rectangular stone blades and Middle Paleolithic-style implements hammered off prepared chunks of rock. But many Shiyu stone artifacts consisted merely of flakes struck off small, round rocks. That technique dates to as early as about 2.1 million years ago at northern Chinese hominid sites, the scientists say, long before the evolutionary origin of H. sapiens around 300,000 years ago.

Taking the now-lost braincase fragment into account, they suspect that well-traveled H. sapiens combined their own brand of IUP tools with a simple, practical form of toolmaking that was common among locals.

That initial toolmaking exchange may have heralded others. Yang and colleagues have reported that stone tools and other artifacts excavated at the roughly 40,000-year-old Xiamabei site in northern China show signs of cultural give-and-take between H. sapiens and an unidentified Indigenous population (SN: 3/10/22).

But at Shiyu, cultural blending with locals did not make homebodies out of mobile H. sapiens. For instance, ancient Shiyu people made four tools out of obsidian obtained — possibly via a trade network — from sources 800 to 1,000 kilometers away, Yang and colleagues find. And analyses of butchered animal bones from the Chinese site indicate frequent consumption of wild horses, which hunters must have tracked across vast expanses.

Two unusual objects from Shiyu may represent innovations by H. sapiens as they mixed with a local culture, the researchers suggest. That disc-shaped object made of graphite, with a hole in its center, may have served as a button, possibly for closing a cloak or a bag, they suspect. The bone tool had uncertain uses.

Despite such cultural tweaks, “IUP technologies have commonality across Eurasia and represent a key transition, suggesting the movement of human populations across great distances,” Petraglia says.

Siberian travelers

Shiyu’s IUP crowd did not exist in a geographic vacuum. Stone tool excavations conducted by different teams indicate that IUP cultures, presumably the products of mobile H. sapiens communities, spread through northern Asia around 45,000 years ago, says Rybin.

Growing evidence documents movements of IUP groups through open grasslands of three northern Siberian river valleys, located near Lake Baikal roughly 2,000 kilometers northwest of China’s Shiyu site, Rybin and colleagues report in the December 2023 Archaeological Research in Asia. Northern Siberian IUP sites excavated so far date to between roughly 45,000 and 40,000 years ago.

Stone tools at these sites include IUP-style stone blades and flakes. Siberian makers of IUP tools followed some distinctive practices, such as snapping large, thick blades in two to use as cores for striking off smaller implements. Different types and qualities of rock found across Eurasia influenced variations in the size and shape of IUP implements, Rybin says.

Hominid fossils have not turned up at Siberian IUP sites. But ancient DNA evidence identified a 45,000-year-old leg bone found near a present-day Siberian settlement in 2008 as that of a H. sapiens man with a small genetic inheritance from Neandertals (SN: 10/22/14). No stone tools accompanied that fossil find.

Researchers have not unearthed any hints of IUP populations in northern Siberia encountering culturally distinct groups already living there, in situations akin to the newly proposed scenario at Shiyu, Rybin says.

Dueling scenarios

Hardy H. sapiens travelers possibly merged with Indigenous Homo communities at Shiyu around 45,000 years ago, says archaeologist John Shea of Stony Brook University in New York. But he views other scenarios as equally plausible.

For instance, Neandertals or Denisovans based in northeast Asia may have added stone blades and flakes to their toolmaking repertoire without any input from H. sapiens. These implements could have served as tips of spears or arrows well-suited to hunting animals across grasslands that expanded after around 50,000 years ago.

If that were the case, the now-lost H. sapiens fossil at Shiyu could have been present “because some early human wandered too deep into Neandertal country, got spotted, tracked, killed and eaten,” Shea speculates.

Or perhaps Shiyu’s contrasting types of stone artifacts were made and discarded around the same time by different Homo groups living near one another.

For instance, many Shiyu artifacts, including stone flakes and blades, resemble Middle Eastern and Iranian finds associated at some sites with H. sapiens fossils and at others with Neandertal remains, Shea says.

Using ancient stone tools to determine which hominids hung out at Shiyu 45,000 years ago “is like trying to reconstruct how many cultures contributed to the durable metal and plastic contents of a municipal trash can,” Shea says.

Sedimental journey

Whoever bashed stones at Shiyu, the products of their efforts do not look like classic IUP artifacts, says archaeologist Nicolas Teyssandier. Unlike IUP sites in the Middle East and southwest Asia, excavations at the Chinese site uncovered little debris typically generated during blade production and no triangular stone points, contends Teyssandier, of University Toulouse-Jean Jaurès in France.

“Most of the Shiyu stone artifacts look just like Middle Paleolithic [tools],” he says.

Shiyu artifacts, which were recovered before the adoption of modern excavation techniques, might originally have been deposited in older and younger sediment layers that became mixed over time, he says. If so, Shiyu artifacts could have accumulated during several occupations by Homo populations at different times.

But Petraglia doubts that scenario. Two closely aligned age estimates for different parts of Shiyu’s artifact-bearing layer indicate that this deposit formed rapidly as a geologically undisturbed unit, over perhaps a few hundred years, he says.

It may take just as long for scientists to reach a consensus on who did what at Shiyu 45,000 years ago. As the site’s original excavators would no doubt agree, ancient cultural puzzles come out of the ground far easier than they get reassembled.