There was nowhere to go but everywhere, so just keep on rolling under the stars.

—Jack Kerouac, On the Road, 1957

In 1983, archaeologists were excavating at the medieval site of Dmanisi in the country of Georgia, which was then part of the Soviet Union. The archaeological team was finding coins and other medieval artifacts when they came across a tooth. Thinking it was probably from an animal that had been eaten by traders who stopped at Dmanisi as they traveled the Silk Road, they brought the tooth to Abesalom Vekua, a trained paleontologist. This, he determined, was no cow or pig. The tooth belonged to a rhinoceros.

What was a rhinoceros doing in a medieval grain pit in this mountainous section of Southwest Asia? Vekua and his colleague Leo Gabunia decided to investigate the source of their out-of-place rhinoceros.

One clue was that the tooth wasn’t modern. It was from Dicerorhinus etruscus, a species that went extinct in the Pleistocene. The following year, they dug at Dmanisi and discovered stone tools similar to the simple Oldowan ones Mary and Louis Leakey had found at Olduvai Gorge in Tanzania. The mystery of the rhinoceros tooth began to make sense. It turns out that the Dmanisi citadel was built on Pleistocene sediments. The archaeologists digging through the dirt for medieval relics had penetrated through to a much older layer, a distant time when Dicerorhinus roamed the landscape.

It was also a time when hominins were not supposed to have expanded their ranges beyond the borders of Africa. But those stone tools indicated that they had.

Vekua and Gabunia kept digging, and in 1991, they found a hominin jaw. A decade later, they unearthed two skulls from sediments resting on a 1.8-million-year-old lava bed. The skulls had large faces, but their brains were about half the size of human brains today. They were identified as coming from an early version of Homo erectus, a species that had been known to science since Eugène Dubois’s discoveries in the late nineteenth century. In the two decades since, three more skulls and two partial skeletons have been unearthed from this amazing site. The Dmanisi hominins are the oldest ever discovered outside the African continent.

However, evidence from the other end of the Silk Road, at Shangchen in central China, indicates that hominins were on the move even earlier.

In 2018, Zhaoyu Zhu of the Guangzhou Institute of Geochemistry at the Chinese Academy of Sciences announced the discovery of simple stone tools made by ancient human hands 2.1 million years ago. About the time Australopithecus sediba was hyperpronating around South Africa, another branch of the human family tree had pushed almost 9,000 miles to the east. No bones have been found yet, so we don’t know who made these stone tools, but most paleoanthropologists assume early Homo erectus—or perhaps an even earlier representative of our genus.

At first glance, the spread of ancient humans around the world seems sudden.

At first glance, the spread of ancient humans around the world seems sudden. Hominins had been in eastern and southern Africa for millions of years, but now, seemingly in a flash, they were in China. However, this was not as rapid as it appears. If early members of the genus Homo migrated east at just one mile per decade starting around 2.2 million years ago, they could have reached China 2.1 million years ago, in plenty of time to leave their stone tools at Shangchen.

The discoveries at Dmanisi and Shangchen reveal that soon after Homo evolved in Africa about 2.5 million years ago, their territories expanded, spreading north and east into Eurasia. There was no Welcome to Asia sign greeting them. They didn’t know they were moving into regions that would amaze and puzzle their descendants 2 million years later. But this does raise some questions.

Why did hominins become explorers at this time? And how were they able to move into territories that had been uninhabited by their ancestor Australopithecus?

The clues can be found in the skeleton of a boy.

In 2007, I traveled to Nairobi, Kenya, a densely populated city 6,000 feet above sea level. For two weeks in August, the weather was surprisingly cold and cloudy. It didn’t rain, but the air was heavy and still. The streets were lined with vendors selling fresh fruit and nuts. Goats roamed, eating roadside litter. Smoldering piles of trash added to the unpleasant smell of diesel fuel. The day I arrived in Nairobi, a head cold materialized and pressed hard against my sinuses for a week.

Nairobi is a city of over 3 million people, though that number rises to over 6 million if the surrounding population is counted. This includes the largest slum in Africa—Kibera, where nearly 1 million people live on an average income of less than $1 a day. A few miles north of Kibera, atop Museum Hill in the Westlands district, sits the Nairobi National Museum, where some of the most prized fossils ever unearthed are housed in a vault the size of a small coffee shop.

Outside the museum stand statues of Louis Leakey and a large orange dinosaur. I skirted past the public exhibits and through a courtyard to the research collections, where I met Fredrick Manthi, a Kenyan paleontologist who often goes by his middle name, Kyalo.

Manthi’s father worked on Mary Leakey’s expeditions in the 1970s, and young Kyalo caught the hominin bug at an early age. After earning his Ph.D. at the University of Cape Town, he returned to Kenya to lead the paleontology and paleoanthropology division at the museum and thus oversees all prehistory research in Kenya. Three years after I met him, he would discover a beautiful 1.5-million-year-old Homo erectus skull near the village of Ileret on the east side of Lake Turkana.

I gave Manthi the list of fossils I wanted to study, ranging from the 20-million-year-old foot bones of the ancient ape Proconsul to a fossil femur from an archaic Homo sapiens. I figured that on my first day he would bring me a tray of fragmentary foot fossils, the kind of bones only a handful of people in the world cared about. I was still a student, after all, and was groggy from the cold medicine.

Instead, Manthi reemerged from behind the thick steel door of the vault with a wooden tray containing the Nariokotome Homo erectus skeleton. It was like giving the curator of the Louvre a list of Renaissance paintings to study and being handed the Mona Lisa. My arms went weak and my hands trembled. Manthi may have noticed my open jaw as I ogled the skeleton. Instead of handing me the tray, he walked them to where I would be working and carefully rested the precious fossils on the counter.

I love fossils. I travel far to see them, eager to take measurements, photographs, and 3D scans of these fragile fragments of our past. But for the first few minutes of every visit with a new fossil, my calipers, camera, and scanner remain idle. I just sit, alone, with the remains of my ancestors. I appreciate the color, texture, and curve of every piece. I wonder not only about the species but also about the individual whose death and preservation allows us to understand our own place in the story of life. I let myself be moved. I let myself be emotional. This ritual started in August 2007, when I sat, alone, in the Nairobi National Museum with the Nariokotome skeleton.

Then I got to work.

The Nariokotome fossil was discovered in 1984 by Kamoya Kimeu, arguably the most prolific hominin discoverer in history. He was part of the Leakey family’s famous “hominid gang,” whose discoveries in Eastern Africa in the 1960s and 1970s opened the floodgates for paleoanthropological research in Tanzania, Kenya, and eventually Ethiopia. Kimeu’s discoveries are so important that two fossil species—a Miocene ape, Kamoyapithecus, and the early Pleistocene monkey Cercopithecoides kimeui—have been named after him.

In their book The Wisdom of the Bones, paleoanthropologists Alan Walker and Pat Shipman described Kimeu’s approach to fossil hunting as “walk, and walk, and walk, and look while you are doing it.”

On August 22, 1984, he was doing just that on the west side of Lake Turkana. Along the bank of the dried up Nariokotome River, he spotted a tiny fragment of skull bone, camouflaged the same dark color as the surrounding sediment.

“Lord knows how he saw it,” Walker and Shipman wrote.

Kimeu called Richard Leakey and Alan Walker in Nairobi, and the project leaders arrived the next day. For the next five years, the team moved 1,500 cubic meters of dirt. Hidden within it was most of the skeleton of a juvenile Homo erectus who died 1.49 million years ago.

The Nariokotome Boy, as he came to be known, is one of the most complete and important skeletons ever discovered. He reveals the kind of body it took for early Homo to expand its range beyond the borders of Africa.

His brain, which had reached full size, was only two-thirds the volume of a modern human’s. The unerupted wisdom teeth and the unfused growth plates in his arms and legs tell us that he was young when he died—just 9 years old according to a detailed study of his teeth. But his leg bones also indicate that he was already over five feet tall and weighed close to 100 pounds. That is a big kid. My son was almost a foot shorter and 40 pounds lighter at that age.

Scientists calculate that the Nariokotome child would likely have been close to six feet tall had he survived into adulthood. The boy’s large size at such a young age also indicates that his species hadn’t evolved the adolescent growth spurt we have today. Why not? Northwestern University anthropologist Chris Kuzawa discovered a trade-off in energy allocation between the brain and body in children. Kids’ brains use so much energy during their preteen years that the growth of their bodies slows down. In the teen years, bodies play catch-up and rapidly grow in height—the growth spurt. Because the brains of Homo erectus were only two-thirds the size of ours, they probably could still divide energy between the brain and a growing body.

More about this species can be gleaned from a fragmentary skeleton of a 1.6-million-year-old adult Homo erectus, a specimen named KNM-ER 1808. It was discovered in 1973 by, of course, Kamoya Kimeu. Its large right femur is the size of a thigh bone in a modern human standing a shade under six feet. People tend to think that modern human sizes were not achieved until recently, but that is wrong. Homo erectus was well within the size range of people today.

I turned back to the tray containing the Nariokotome skeleton and plucked his left femur from its aqua-colored foam bed. The femur is dark gray, with splotches of black and brown. I was struck by its length. He also had a large upper arm bone (humerus) 34 percent longer than Lucy’s. That makes sense because he was a bigger individual than Lucy, but you might expect, then, that his femur would also be 34 percent longer than hers. It wasn’t. It was 54 percent longer.

Homo erectus was not a scaled-up Australopithecus. Legs had gotten longer.

“From ants to elephants, the variable that explains how much energy an animal uses to get from one place to another is leg length,” Herman Pontzer, a professor of anthropology at Duke University, told me. As legs get longer, his extensive research shows, travel generally becomes easier.

Equipped with longer legs, our Homo erectus ancestors could range farther than Lucy’s kind. But that’s not all. Homo erectus, it turns out, had also evolved a foot with a fully modern human arch.

In 2009, a team of researchers from the Nairobi National Museum and George Washington University discovered nearly one hundred fossil footprints near Ileret. They were made along the muddy shores of a lake by twenty Homo erectus individuals 1.5 million years ago. They are the size of human footprints today and feature a prominent arch that put a spring in their step—especially when they ran.

Australopithecus species had an arch, too, but it was low by modern-human standards. In Homo erectus, the arch was fully modern, legs were long, and our ancestors finally had the anatomical equipment to range farther and collect more food.

In ecosystems throughout the world, carnivores have, on average, larger home ranges than herbivores. Plants often grow in clumps, so herbivores do not need to range as far each day to find food. Carnivores, however, have to search far and wide to hunt down a meal. It is no coincidence, then, that Homo erectus fossil sites contain a lot of butchered animal bones acquired both by hunting and scavenging.

Stone tools date back to 3.3 million years ago, and cut marks are found on 3.4-million-year-old antelope bones that predate Homo erectus. It seems, then, that Australopithecus and even early Homo dabbled in carnivory through opportunistic scavenging. But they were not hunters. With Homo erectus, scavenging became more prevalent, and there is even evidence of deliberate, coordinated hunting. Plants remained part of their diet as well, making them omnivores just as we are. Of course, some people today choose not to eat meat, but there is ample evidence that meat and marrow were important resources that helped our lineage survive the Pleistocene.

With long legs, arched feet, and expanded home ranges, Homo erectus pushed beyond the borders of Africa and into Eurasia.

The Homo erectus skeletons at Dmanisi, Georgia, are not as tall as the Nariokotome Boy. They are barely over five feet. But they have long legs and the body proportions of modern humans. The Dmanisi hominins could walk with great efficiency and follow game through the Middle East and modern-day Turkey into the Caucasus. Even earlier migrations had made it all the way to China. It remains unclear whether the hominins followed the mainland across the plateaus of Asia or if they followed the coastline through India and Southeast Asia. Either way, 2.1 million years ago they made it clear across the largest continent on Earth.

Human migration narratives are often presented in an oversimplistic and unidirectional way, but the odds that these territorial expansions happened only once and in only one direction are infinitesimally small. 

Human migration narratives are often presented in an oversimplistic and unidirectional way, but the odds that these territorial expansions happened only once and in only one direction are infinitesimally small. Homo erectus almost certainly moved in and out of Africa in pulses, gradually exploring the edges of their expanding range as they ventured into territories never before inhabited by an upright walking hominin. By at least 1.5 million years ago, Homo erectus had pushed as far southeast as they could walk without getting their feet wet.

During ice ages, at least eight of which happened cyclically during the last million years, enough water was trapped at the poles and in mountain glaciers to make sea levels drop, so it would have been possible to walk from Southeast Asia to Java in the Indonesian Archipelago. But no farther. There, Homo erectus would have encountered a twenty-mile-wide, five-mile-deep oceanic trench that delineates Wallace’s Line, named for the nineteenth-century naturalist Alfred Russel Wallace, codiscoverer with Charles Darwin of natural selection. To the west of Wallace’s Line are plants and animals found in Asia; to the east, the startlingly different plants and animals of Australia. This ecological boundary is nearly impossible to cross without a boat.

Around the time Homo erectus reached Java, hominins were also spreading into western Eurasia. In 2013, Spanish paleoanthropologists announced the discovery of a single hominin tooth and stone tools from a cave in the town of Orce in southeastern Spain. They were deposited in 1.4-million-year-old sediments. A few years earlier, a more complete 1.2-million-year-old lower jaw was found by Eudald Carbonell at a cave called Sima del Elefante, meaning pit of the elephant, in the Atapuerca region of northern Spain. The researchers called the fossil Homo antecessor, which means “pioneer.”

Homo erectus and its cousins had become cosmopolitan apes, ranging from the tip of South Africa to as far west as Spain and as far east as Indonesia. There were no wagons, planes, trains, or automobiles. No domesticated horses to ride. They walked.

During that time, a strange and wonderful thing was happening. Brains were getting bigger. A lot bigger. Two nonmutually exclusive hypotheses explain why this might have been happening, and they both have to do with food.

The first, formulated by anthropologists Leslie Aiello and Peter Wheeler in 1995, is called the expensive-tissue hypothesis. Aiello and Wheeler collected data on organ weights of primates and reported that humans are unusual in having a very large brain (which everyone knew) but an extremely short gut (which everyone did not know). Guts are energetically expensive to maintain, constantly sloughing off old tissues and regenerating new ones. By evolving a shorter gut, these hominins had ostensibly freed up energy that could be reallocated to brain growth. This would not work in a strict herbivore. Plant eaters require long hindguts to digest tough cellulose fibers in plants. Carnivores, in contrast, have short hindguts that absorb nutrients from meat and marrow without meters and meters of intestines. Aiello and Wheeler proposed that as our ancestors consumed more animals, those with shorter guts and larger brains thrived and multiplied. Between 2 million and 1 million years ago, the average hominin brain roughly doubled in volume.

More recently, Richard Wrangham, a human evolutionary biologist at Harvard University, introduced another variable into the equation: fire.

Tantalizing evidence from the east side of Lake Turkana in Kenya and from Swartkrans Cave in South Africa indicates that by 1.5 million years ago, Homo erectus had learned how to control fire. The 1-million-year-old South African cave Wonderwerk has incontrovertible evidence of fire use. With it, our ancestors could cook their food, making it easier to digest. This, Wrangham suggests, provided them with the energy they needed to evolve larger brains. Fire also would have allowed our predecessors to spread into territories previously too cold to inhabit. And it would have freed them from escaping to the trees for safety at night since fire is a predator deterrent. As Homo erectus evolved longer legs, they became better walkers, but climbing became difficult. With fire, these disadvantaged climbers could survive and multiply.

And as our ancestors walked, they began to talk. As the phrase goes, “If you’re going to talk the talk, you’ve got to walk the walk.” It turns out, walking and talking are indeed linked.

In four-legged animals, the muscles of the shoulders, chest, and even abdomen absorb the impact as their front limbs hit the ground. This means that animals moving on all fours have to coordinate their breathing and walking, one breath for each step. That’s why animals can’t gallop and pant at the same time. Such short, quick breaths are not possible when digestive organs slam up against the diaphragm with every stride. Because they can’t pant, most running animals are unable to cool down, so they have to stop and rest in the shade after short sprints. Humans, however, can breathe rapidly as they stride. Unlike many four-legged animals, we also sweat. This allows us to cool off while running. We are slow, but we can go for miles.

It turns out, walking and talking are indeed linked.

But what does this have to do with language?

You can mimic the role of chest and arm muscles in a quadruped by carrying something heavy while you walk. Your chest muscles tense and you draw a breath with each stride. Aside from the occasional grunt, it’s difficult for you to make sounds. That is what it is like for a four-legged animal. But animals that walk on two feet have finer control of their breathing, and that gives them the flexibility to make a great range of sounds.

Geladas, terrestrial monkeys that live in the Ethiopian highlands, sit upright while they graze on seeds. In a seated position, they communicate through a series of complex vocalizations. As bipeds, humans have evolved language, combining sounds produced through fine muscle control of our breathing into a seemingly infinite number of combinations and meanings. Even in our children, the onset of walking and talking are closely linked.

The origin of human language remains unknown and controversial. Many factors beyond breathing flexibility help us produce sounds. The base of our skull and the vocal apparatus in the back of our throats form a resonance chamber absent in the apes. Our hyoid, a bone rarely found in the human fossil record, is thick to anchor muscles and ligaments used when we speak. Regions of the brain such as Broca’s and Wernicke’s areas are critical for language production and comprehension. Our inner ear bones are fine-tuned to the frequencies of human voices.

Early on, hand signals may have been as important as the spoken word in the development of language. The relationship between sounds and meaning may have started with onomatopoeias—words that sound like what they mean. Birds chirp. Bees buzz. Hands clap. But onomatopoeia doesn’t work for everything. What does a hunt or a sunrise sound like? So some sounds were needed to symbolically represent meanings. We had to become symbolic apes. Eventually, songs and music also played a role in spreading ideas and preserving memories. These pieces did not emerge all at once, but we can try to reconstruct when the first languages were developing in our ancestors by picking through the fossil record.

Bipedalism likely gave Australopithecus the fine-controlled breathing required to make a larger range of sounds than a chimpanzee can make, and it freed their hands to communicate with gestures. But there is little evidence that they actually talked.

A hyoid bone from the 3.4-million-year-old Dikika Child looks like an ape’s. Fossilized brain impressions and CT scans of the inside of fossil skulls reveal that the folds and fissures of the earliest Australopithecus brains were quite apelike. But in some Australopithecus, there appears to be asymmetry in the Broca region, suggesting that the brain was primed for producing and understanding language. Certainly, this was the case for early Homo brains by around 2 million years ago.

Half-million-year-old fossils from Spain reveal both humanlike hyoid bones and inner ears fine-tuned to detect and process sounds in the vocal bandwidth. And genetic evidence indicates language may have been present by this time. DNA extracted from fossil hominins in Europe and Asia shows that a gene that impacts language, although exactly how it does that is unknown, evolved into its current form by at least 1 million years ago.

All of the key ingredients for language appear to have been in place by half a million years ago, but the first step in this evolutionary sequence was upright walking, which provided the fine control of breathing required to make a large repertoire of sounds.

Homo erectus walked, and as it spread through the world, it also talked.

The coming and going of ice ages through the Pleistocene at times permitted hominins to reach places that would otherwise be inaccessible, only to then trap and isolate them. Homo erectus individuals living on what is now the island of Java, for example, could wander throughout southeastern Asia during a glacial maximum but then were stuck on the island for tens of thousands of years when the seas rose during a warm spell. In Western Europe, glacial periods permitted hominins to even reach England. We know this because they left footprints there.

Around 800,000 years ago, a group of hominins sometimes given the species name Homo heidelbergensis walked along a muddy shore near modern Happisburgh, England, leaving footprints nearly identical to yours and mine. The shoreline is eroding fast, however. Soon after researchers photographed and measured the prints, they were washed away. The Pleistocene population that made the footprints was also fleeting, eventually pushed south and isolated in pockets along the Mediterranean as glaciers advanced from the north.

These climatic pulses led to the intermittent genetic isolation of Pleistocene Homo populations. One of them, initially isolated in pockets in Europe and western Asia, evolved into the Neandertals. Their bones are plentiful and have been known to science since the mid-nineteenth century. Once glaciers retreated, they expanded their range from Portugal to Ukraine. Two dozen complete skulls have been unearthed.

In 2019, scientists from the National Museum of Natural History in Paris announced the discovery of an astounding collection of 257 Neandertal footprints made in the dunes of Normandy, France, 80,000 years ago. A dozen children walked with an adult or two through the wet sand, immortalizing a day in the life of a Pleistocene day care. I imagine the Neandertal kids playing and laughing while the adults scanned the horizon for threats.

At this time, parts of Asia were inhabited by the Denisovans, a group known not just from the anatomy of a handful of scarce fossils, but from DNA extracted from tiny scraps of bone found in Siberian and central China caves.

Homo erectus and its cousins, with their long legs, enlarged brain, and control of fire, had spread throughout Africa, Asia, and Europe. The stage was now set for the final phase of our journey—Homo sapiens.

But recent discoveries have disrupted this narrative, hinting that human evolution and the migration of hominins around the globe were much more complicated, and even more interesting, than any of us imagined.

Except, maybe, J.R.R. Tolkien.

Excerpted from Jeremy DeSilva’s First Steps: How Upright Walking Made Us Human, published by Harper, an imprint of HarperCollins Publishers.