Born to Carry
Travel back three and a half million years.
The world looks almost nothing like the one you know. The East African Rift Valley is splitting apart at a rate of a few millimetres per year, reshaping the continent’s interior with volcanic upwellings and collapsing fault blocks. The forests that once carpeted the equatorial interior are thinning, fragmenting, retreating before a slow but inexorable expansion of grassland. Mosaic habitat stretches to the horizon: patches of riverine forest, open savanna, seasonal wetlands, rocky kopjes shimmering in equatorial heat. The sky is vast. The distances between water, food, and shelter are growing longer with every thousand years that pass.
Among the trees and at their margins, something is happening that has never happened before in the history of life on Earth. A primate is walking. Not in the hunched, knuckle-assisted shuffle of a chimpanzee moving cautiously across open ground, but fully upright—spine extended, head balanced over pelvis, arms swinging free. The gait is not graceful by modern standards. The hips are wide, the legs slightly bowed, the stride shorter than yours or mine. But the posture is unmistakable. This creature is walking the way you walk.
It is not running.
It is not fleeing a predator. It is not pursuing prey. It is walking from one place to another at a speed that modern exercise physiologists would classify as Zone 2—sustainable, aerobic, hour-capable. And in its hands, cradled against its body or balanced on one arm while the other stabilises its gait, it is carrying something.
The Laetoli footprints, preserved in volcanic ash in what is now Tanzania and dated to approximately 3.6 million years ago, record the gaits of australopithecines moving across a landscape at walking pace. The foot anatomy is clear in the impressions: a non-divergent hallux—the big toe aligned with the others rather than splayed outward for grasping, as in chimpanzees—and a modest arch, not the full plantar vault of modern humans but sufficient to indicate a walking gait rather than a climbing or brachiation-adapted foot. These creatures were committed walkers. They were upright before they were runners. They were striding across Africa before the Achilles tendons, nuchal ligaments, and enlarged gluteus maximi of Homo erectus would appear in the fossil record, a full million years later, to enable something faster.
The first chapter of the human story is not running. It is walking with a load.
In 2012, a research team led by Susana Carvalho at the University of Cambridge published a paper in Current Biology that should have been a landmark in exercise physiology, evolutionary biology, and the popular understanding of why humans move the way we do. It was not. Outside the small world of evolutionary anthropology, it passed largely unnoticed. But the finding it reported may be the most important single data point for understanding the human body as an exercise machine.
Carvalho and her colleagues had been working with chimpanzees in Bossou, Guinea—a well-studied community of wild chimps living in a fragmented forest patch surrounded by farmland. The Bossou chimps are particularly interesting for evolutionary questions because they use tools, they navigate between habitat patches across open ground, and they have been observed for decades by researchers who know individual animals and their behavioural quirks. The team was not studying exercise. They were studying the conditions under which chimpanzees choose to walk on two legs rather than four.
The answer they found was this: when chimpanzees carry objects of perceived value—food, tools, resources worth transporting from one place to another—they preferentially adopt bipedal posture. The more valuable the object, the more consistently they stand upright. A chimpanzee moving on four limbs will often transition to bipedalism the moment it picks up something worth carrying. The load is the trigger. The upright posture is the response.
The load is the cause. The upright primate is the consequence.
Bipedalism—the defining anatomical and behavioural feature of the entire hominin lineage, the adaptation that produced every downstream consequence of human evolution including tool use, language, endurance locomotion, and the peculiar skull-on-a-stick body plan you carry around every day—may have been selected in an ancestral population not because standing upright made them faster or more visible or better thermoregulated, but because it freed their hands to carry things.
The selective pressure, in Carvalho’s argument, was the adaptive value of transport. A primate that could carry more food from a patch to a safe location, carry water across a dry-season landscape, carry an infant and simultaneously gather, carry a stone tool from its knapping site to a butchery site kilometres away—that primate had a meaningful fitness advantage over one that had to choose between locomotion and carrying. The two-legged posture was not the goal. Carrying was the goal. Bipedalism was the solution.
This is not Carvalho’s argument alone. It is a position that has been building across evolutionary anthropology for two decades. The fossil record has never cleanly supported the predator-evasion hypothesis for bipedalism: australopithecines were still capable tree climbers, their speed on the ground was modest by primate standards, and their habitats offered cover. The thermoregulatory hypothesis—that standing upright reduces solar heat load—is compelling but partial, a downstream benefit rather than a primary driver. Carvalho’s carrying hypothesis sits within a broader framework articulated by researchers including Crompton, Thorpe, and Aiello: that the liberation of the forelimbs from locomotion, and the resulting capacity for resource transport, was the fundamental selective pressure that made the hominin lineage what it became.
Think about what this means for the history of human physical capability. The story we have told ourselves, in popular science books and half-remembered undergraduate lectures and the marketing copy of running shoe companies, is a story of flight and chase. We evolved to run from lions. We evolved to hunt by persistence across the savanna. The human body is a running machine, and its highest expression is the marathon, the ultramarathon, the endless miles of pavement.
That story is not false. But it is incomplete to the point of fundamental misdirection. The full story begins not with running, but with carrying. And the chapter on carrying is deeper, older, and more continuous than the chapter on running that follows it.
Dennis Bramble and Daniel Lieberman’s 2004 paper in Nature, “Endurance Running and the Evolution of Homo,” is justly famous. It identified twenty-six derived anatomical features of the human body—features that distinguish us from the great apes and appear in the fossil record at or after the emergence of Homo erectus approximately two million years ago—that are specifically adapted for sustained running over long distances. The enlarged gluteus maximus. The long, springy Achilles tendon. The shortened toes. The narrow waist and decoupled shoulder girdle. The nuchal ligament connecting skull to thoracic spine, preventing the head from pitching forward with each stride. The expanded joint surfaces of the knee and ankle. The arched foot that functions as a spring and shock absorber simultaneously.
The paper proposed the endurance running hypothesis: that Homo evolved the capacity for sustained running as a means of persistence hunting—pursuing prey animals over long distances in the heat of the day until those animals collapsed from hyperthermia. Humans, with their eccrine sweat glands, hairless skin, and upright posture that minimises solar heat gain while maximising evaporative cooling, can thermoregulate across long distances in warm environments better than any other large mammal. A horse overheats. A wildebeest overheats. A human keeps running.
This is real. The anatomy is real. The persistence hunting tradition is real—it has been documented ethnographically in the Kalahari San, the Tarahumara, and Aboriginal Australians. And Bramble and Lieberman were careful to note that these running adaptations appear on top of a pre-existing platform of bipedal walking anatomy: we did not become runners instead of walkers. We became runners in addition to walkers, with the walking baseline intact and the running capability added to it approximately two million years later.
But the endurance running hypothesis, in the years since its publication, has become something of a cultural myth that has distorted how we think about the human body in motion. The popular interpretation—that humans are fundamentally runners, that running is our ancestral mode, that the barefoot-natural-running movement is returning us to our evolutionary essence—overshoots the evidence in ways that matter. Twenty-six anatomical features shared with running are not the same as twenty-six anatomical features selected for running. Many of those features are better explained by load carrying than by running, or by both simultaneously. The nuchal ligament, for example, stabilises the head under dynamic loading—which includes running, but also includes walking with a load on the torso that shifts the centre of mass forward. The enlarged gluteus maximus fires most powerfully not in running but in activities requiring trunk stabilisation under load: carrying, climbing, and the loaded hip extension that keeps you upright when twenty kilograms are pulling your upper body forward.
Bramble and Lieberman gave us the running hypothesis. What they may have given us inadvertently was also the carrying hypothesis with better marketing.
The evidence has continued to accumulate in the two decades since their paper appeared. And the picture it assembles is increasingly clear: Homo sapiens is not a running specialist. It is a locomotion generalist with a unique and defining specialisation in loaded walking over long distances.
In 2007, Cara Wall-Scheffler and her colleagues published a study in the Journal of Human Evolution that took a more direct approach to the question of load carriage efficiency. Rather than inferring carrying capability from anatomy, they measured it. They recruited participants to carry loads using different methods—in the hands, on the head, on the hips, and on the torso—and measured the metabolic cost of each. The question was simple: which carrying method does the human body handle most efficiently?
The answer was equally simple: the back. Torso-centred loading, the backpack position, is metabolically less costly than any other method of carrying the same load over the same distance. Not marginally less costly—significantly less costly. When a load is carried close to the body’s centre of mass, distributed across the posterior chain and transferred through a wide contact area to the pelvis, the metabolic penalty of carrying is minimised. When a load is carried at arm’s length, or asymmetrically, or at the extremities, the metabolic cost rises substantially.
Wall-Scheffler’s finding is deceptively obvious and profoundly important. The rucksack is not a modern invention. It is not a military technology that civilians have borrowed. It is a direct expression of the most efficient load-carrying configuration available to a bipedal primate with a posterior chain optimised for exactly this purpose. Every culture across history and prehistory that required its members to carry loads over distances independently arrived at the same solution: a pack, carried on the back, with the weight riding as close to the spine as possible.
The Neolithic farmer carrying grain. The Inca porters who built Machu Picchu. The Sherpa guides of the Himalaya. The Hadza forager returning from a hunt. The Neanderthal moving camp. The Homo erectus carrying a stone biface from an outcrop to a home base. All of them arrived at the backpack because the human posterior chain—the architecture that connects calcaneus to Achilles to calf to hamstring to gluteus to erector spinae to trapezius—is a load-bearing system, and it is most efficient when the load is placed directly over its axis.
Wall-Scheffler also documented something that, once you see it, you cannot unsee. When loads are carried asymmetrically—on one hip, in one hand, slung over one shoulder—the metabolic cost rises disproportionately to the load increase. The body spends significant energy compensating for the lateral shift in centre of mass, recruiting trunk musculature to prevent lateral trunk lean, constantly recalibrating balance with every step. Symmetric, torso-centred loading requires almost none of this compensatory work. The body simply carries. The trunk muscles work to maintain extension and resist the forward moment, which they are architecturally suited for, and the lower limbs simply walk, which they are architecturally suited for. The entire system functions together in the pattern for which it was selected.
This is what your body was built for. Not abstractly, not poetically, but mechanically, anatomically, enzymatically, hormonally. The system was selected across millions of years to do this specific thing. When you put a pack on your back and walk, you are not doing something unusual or novel or challenging for your musculoskeletal architecture. You are doing what it was designed to do.
The ethnographic record is unambiguous on this point. Modern hunter-gatherer populations—the closest living analogs to the Pleistocene foragers in whose footsteps we are tracing this argument—carry loads as a central feature of daily life, not as an exceptional exertion. Hadza women in northern Tanzania, studied by anthropologists across multiple decades, routinely carry infants, water containers, and gathered food for distances of several kilometres per day. Hadza men carry game, hunting equipment, and honey over comparable distances. Loads of fifteen to thirty percent of body weight are not unusual. The carrying is not considered exercise. It is not considered a challenge. It is simply the background activity of survival, as unremarkable as breathing.
The Ju/’hoansi San of the Kalahari move camp multiple times per year, carrying everything they own—their entire material culture—on their backs and in their arms. Women carry children and food. Men carry tools and skins. Children carry what they can. Everyone walks. The distances between water sources in the Kalahari can be extraordinary, and the carrying capacity of each individual directly affects the group’s ability to range further and access resources unavailable to less mobile or less load-capable groups.
Among the Aché of Paraguay, ethnographic studies have documented foraging ranges of fifteen to twenty kilometres per day during intensive foraging periods, with meat and gathered food carried back to camp in nets slung from the forehead—a head-strap carry configuration—or distributed among multiple carriers. The Aché are interesting because the head-strap configuration they use is metabolically less efficient than the torso-centred configuration Wall-Scheffler identified as optimal, yet they have used it for generations because it frees the hands and allows simultaneous gathering and carrying. The body adapts. The neck musculature of lifelong head-strap carriers is visibly more developed than in non-carrying populations. The human body is not just built for carrying; it is built to adapt to the specific demands of the carrying it is asked to perform.
What these ethnographic records collectively establish is this: for the vast majority of human evolutionary history, and for the majority of human cultures that have existed since, carrying loads over distance was not a form of exercise. It was not a fitness practice. It was the fundamental physical requirement of survival, as quotidian as digestion. The body that evolved in that selective environment is the body you have inherited. And when you put a pack on your back and walk, you are not imposing something foreign on that body. You are returning it to the baseline condition for which it was designed.
Before moving from the ethnographic record to the anatomy, an honest accounting is required. The three populations above—Hadza, Ju/’hoansi San, and Aché—are not the full picture. They are the populations that happen to have been studied. The full picture is mostly blank.
Consider what is missing. The Quechua and Aymara peoples of the Andes have carried loads at extreme altitude for at least five hundred years of documented history and probably far longer, using tumplines, woven aguayo cloths, and back-frames across terrain that maxes out above four thousand metres. The Inca built Machu Picchu, maintained thousands of kilometres of road network, and supplied armies across the cordillera entirely on the backs of human porters. The physiological consequences of a lifetime of loaded walking at high altitude—by a population with documented genetic adaptations to hypoxia, blunted ventilatory response, and elevated haemoglobin—have never been measured. The single peer-reviewed study of Andean porters is a descriptive health survey of 101 workers on the Inca Trail in Peru, documenting complaints of back pain and inadequate equipment (Bauer, 2003). No biomechanical data. No metabolic measurements. No gait analysis. Nothing that would tell you what this body does when it does what it has been doing for centuries.
Japan has an equally ancient and equally unstudied tradition. Historical porters—nimotsu carriers, mountain bokkashi working the alpine routes, and the sango relay system of the Edo period—moved goods across a mountainous archipelago using tenbin bo shoulder poles and back-frames for centuries. Ukiyo-e woodblock prints from the seventeenth century depict them in fine anatomical detail. Not a single biomechanical study exists. Korea’s jige carriers, who used a distinctive A-frame backpack system, and Vietnam’s don ganh workers, who balance loads on flexible shoulder poles at remarkable efficiency, are in the same position: rich living tradition, zero scientific documentation of what their bodies do.
The bamboo shoulder pole deserves a brief detour. Chinese and Vietnamese carriers who use the compliant carrying pole—a flexible bamboo staff with loads suspended from each end—have been studied in small biomechanical investigations. The physics are elegant: the pole oscillates at a natural frequency, and experienced carriers tune their step frequency to match it, creating a coupled oscillator system that reduces the metabolic penalty of the load (Castillo et al., 2014; Huang et al., 2019). Skilled carriers routinely transport loads exceeding their own body mass using this method (Xu et al., 2019). Whether the experienced Japanese or Vietnamese carrier achieves similar efficiency through similar mechanics remains unknown. No one has looked.
There is also an important honest caveat about the “free-ride” phenomenon that has received attention in the African head-loading literature. Maloiy and colleagues’ 1986 paper in Nature reported that experienced East African women could carry loads of up to twenty percent of body weight with no measurable increase in metabolic cost—a finding that seemed to confirm extraordinary load-carrying economy in traditional populations (Maloiy et al., 1986). Subsequent work has complicated this substantially. A 2022 scoping review of forty-five studies found that eighty-two percent of investigations failed to detect the free ride, and that its presence was independent of sex, experience level, load position, and walking speed (Prado-Novoa et al., 2022). The free-ride effect, where it exists, is real but not universal. The carrying advantage of lifelong practice may be genuine while remaining smaller and more conditional than the early literature suggested.
One finding from the head-loading literature warrants particular attention because it bears directly on the argument for torso-centred loading that Wall-Scheffler’s work supports. Habitual head-loading is associated with accelerated cervical spine degeneration. Studies across Nepalese porters, Indian manual labourers, and sub-Saharan African populations document cervical spondylosis and neck pain in lifelong head-loaders, and a 2025 systematic review confirmed a significant association between occupational head loading and cervical degeneration (Ansari et al., 2021; Journal of Clinical and Diagnostic Research, 2025). The neck musculature of the Aché head-strap carrier may be visibly developed, as noted above, but it is being loaded in a configuration the cervical spine was not primarily designed to bear. The backpack argument is not merely about metabolic efficiency. It is also about directing compressive force through the structures best able to absorb it: the posterior chain and pelvis, not the cervical vertebrae.
Let us look more carefully at the anatomy.
The human gluteus maximus is, by any comparative measure, extraordinary. In chimpanzees, our closest living relatives, the gluteus maximus is a relatively small muscle with a modest role in locomotion. In humans, it is the largest muscle in the body by mass, and its expansion during hominin evolution is one of the most distinctive skeletal changes in the fossil record. The question that Bramble and Lieberman correctly identified is: why?
The gluteus maximus does relatively little during level walking. Electromyographic studies show that it is largely inactive during the swing and early stance phases of an unloaded walking gait, activating primarily during late stance to assist push-off. But during running, it fires powerfully to control trunk flexion and stabilise the hip against the forward lean that running imposes. And during loaded walking, it fires continuously, fighting the forward moment that a torso-mounted load creates and stabilising the pelvis against the asymmetric vertical forces of each footfall.
The enlarged human gluteus maximus is not primarily a running adaptation. It is a trunk-stabilisation adaptation for bipedal locomotion under load. Its size makes evolutionary sense not because our ancestors ran from predators but because they walked for hours with weight on their backs. Every step with a loaded pack requires continuous posterior chain activation that unloaded walking does not—and over millions of years of selection pressure, the posterior chain grew to meet that demand.
The Achilles tendon tells a similar story. Longer than in any other primate, elastic enough to store and return significant energy with each step, the human Achilles tendon is presented in most popular accounts as a spring for running. And it is. But it is also a mechanical benefit to loaded walking: the energy storage and return function of the Achilles reduces the metabolic cost of each stride, and this benefit exists at walking speeds as well as running speeds, becoming more consequential as the load increases and the total work per stride rises. A longer, more elastic Achilles is an advantage for any bipedal locomotion, including the hour-long loaded walks that characterised most of hominin foraging.
The nuchal ligament—the thick connective tissue band running from the skull base to the thoracic vertebrae, which stabilises the head in the sagittal plane during dynamic locomotion—is perhaps the most telling of all. In great apes, it is vestigial. In humans, it is a substantial structure that prevents the head from pitching forward with every step or stride. Bramble and Lieberman highlight it as a running adaptation, and the evidence for this is good: during running, the head pitches forward at foot strike and the nuchal ligament acts as a passive restraint. But the nuchal ligament also stabilises the head against the forward moment that a torso load creates. Walk with a heavy pack and notice where your head wants to go: forward, down, in the direction the load is pulling your upper body. The nuchal ligament resists this. It was selected for dynamic stabilisation of the head during bipedal locomotion under any loading condition, not only running.
The picture that emerges from this anatomical survey is not a running machine. It is a carrying machine that can also run. The posterior chain, the foot arch, the Achilles tendon, the nuchal ligament, the narrow waist, the decoupled shoulder girdle—all of these features serve loaded walking at least as well as they serve running, and several of them make more evolutionary sense in the context of carrying than in the context of pursuit.
The Laetoli footprints were pressed into a layer of volcanic ash approximately 3.6 million years ago, most likely by Australopithecus afarensis, the same species whose most famous representative is the fossil known as Lucy. In 1978, a team led by Mary Leakey uncovered a seventy-metre trail of footprints near the Laetoli site in northern Tanzania. The prints were preserved by a thin layer of cemented ash that hardened before subsequent ashfalls buried and protected them. They remained in the rock for 3.6 million years before human hands brushed them clean.
What the footprints reveal is not dramatic by the standards of popular palaeontology. There are no signs of running, no evidence of a pursuit or a flight, no dramatic action preserved in the substrate. The individuals who made these prints were walking. They were walking at a pace that biomechanical analysis of stride length and foot placement places at roughly 1.0 metres per second—just over three and a half kilometres an hour, a comfortable stroll. The toe anatomy shows no sign of a divergent hallux: the big toe is aligned with the others, as in modern humans, not splayed outward for arboreal grasping. The arch impression is shallow but present.
These creatures were committed bipeds. They walked, apparently at ease, on a substrate that also preserved the prints of the ancient savanna fauna around them—the three-toed horses, the elephants, the small mammals moving in every direction across the ash field. They were not special in that landscape. They were part of it, moving through it on two legs the way everything else moved on four, carrying—we must presume, because they had hands and they had to move things—whatever needed carrying.
The emotional weight of the Laetoli footprints is hard to convey in scientific language. You are looking at the actual footprints of an actual ancestor, pressed into actual mud, 3.6 million years before the present moment. The fossil record is full of bones and teeth and occasionally soft tissue impressions, but footprints are different. Footprints are action. They catch a moment of motion, of a living body moving through a living world. The individuals who made these prints had family relationships, social hierarchies, fears, hungers, and the basic primate drive to survive and reproduce. And they walked the way you walk.
They walked upright long before they ran. They carried long before they ran. The anatomical story of the last three and a half million years is the story of a body that was first shaped by walking and carrying, and then, approximately two million years later, refined by running. The carrying came first. The running was built on top of it.
If you want to understand what the human body was primarily designed to do, look at the Laetoli footprints. Not at a marathon. Not at a sprint track. At those seventy metres of walking prints pressed into ash on the Tanzanian plain, three and a half million years before anyone invented a treadmill.
There is a dimension of the carrying argument that evolutionary biologists rarely discuss, but that may be as consequential as the anatomical evidence: the social implications of carrying.
When a chimpanzee at Bossou picks up a precious oil palm nut and goes bipedal to carry it, the behaviour is fundamentally possessive. The chimp is monopolising a resource, keeping it from competitors. It is a solitary carrying act, and bipedalism is the mechanism that makes monopolisation possible: two free hands holding the object, preventing others from taking it. This is consistent with the evolutionary logic Carvalho proposed. But it is also, implicitly, an asocial carrying behaviour.
The hominin carrying that built our bodies was different in a crucial way. Carrying food from a foraging patch back to a home base—the pattern that drove most of the selective pressure for loaded locomotion—is not a monopolisation behaviour. It is a sharing behaviour. You cannot eat a five-kilogram root tuber in the field while simultaneously fending off competitors and watching for predators. You carry it home, where there is a group, a fire, tools for processing, and the social context that makes eating large amounts of food safely possible. Carrying enabled home-basing. Home-basing enabled food sharing. Food sharing is one of the foundational behaviours of human social organisation.
The anthropologist Richard Wrangham has argued that cooking, which required a home base with a controlled fire, drove the anatomical changes associated with early Homo. Bramble and Lieberman argue that running drove many of those same changes. Both are probably partly correct. But the carrying that enabled home-basing, and therefore cooking, and therefore the social structures that cooking and shared food created, is arguably prior to both. You cannot cook without a home base. You cannot have a home base without the ability to carry food to it from foraging grounds that may be several kilometres away. Carrying is the enabling technology.
This is not merely an academic point. It changes the meaning of what you are doing when you ruck. You are not just performing an efficient exercise. You are activating a behavioural and physiological system whose evolutionary function was fundamentally prosocial: provisioning, transport, home-basing, the material foundation of human community. The capacity for loaded locomotion is not just what shaped your muscles. It is, in part, what made human society possible.
There is a deeper argument here, one that goes beyond anatomy and fossil evidence into the logic of evolutionary selection itself.
Natural selection does not optimise for performance metrics. It optimises for reproductive success. The question is never “what is the most impressive thing this body can do?” but “what pattern of movement and behaviour, repeated across a lifetime, maximises the probability of surviving long enough to reproduce and raise offspring to reproductive age?” These are different questions with different answers.
Running at high intensity is metabolically expensive, injury-prone, and—crucially—provides only brief windows of sustained performance. No human can sprint for more than a few minutes. No human can run at even moderate intensity for more than a few hours without significant caloric input and risk of thermal injury. Running is a tool. A useful tool, in specific contexts: evading a predator, making a final push to close distance on fleeing prey, responding to an emergency. But it is a tool for exceptional circumstances, not for the continuous background work of survival.
The continuous background work of survival, for a foraging hominin, was movement over long distances at low-to-moderate intensity, carrying resources between locations. An australopithecine or early Homo waking at dawn faced a day that looked, in motion terms, like this: walk to a water source, carry water back or drink and continue; walk to a foraging area several kilometres away, gather, and carry the gathered food back to a home base; walk to a tool source, carry raw materials to a knapping site, transport finished tools to wherever they were needed; carry infants who could not walk the full distance; carry meat from a scavenged or hunted carcass back to a place where it could be processed and consumed safely.
This is not running. This is all-day loaded walking. And it was the dominant mode of physical activity for the entire span of hominin evolution until the invention of the wheel, the domestication of animals for burden-carrying, and the eventual mechanisation of transport.
The endurance running that Bramble and Lieberman describe was real and was selected for. But it was an enhancement to a carrying baseline, not a replacement of it. The twenty-six derived running features of the human body did not erase or diminish the carrying-optimised architecture they were built on top of. They augmented it. The fully modern human body is a carrying machine that can also run with extraordinary efficiency—not a running machine that happens to be able to carry things.
This distinction matters enormously for how you understand your own body’s needs. If humans are runners, then running is the return to baseline, the natural movement, the thing your body is asking for. If humans are carriers, then carrying is the return to baseline. The evidence says carrying. And your body, if you have ever put a well-fitted pack on your back and walked for several hours, knows this in a way that is difficult to articulate but unmistakable in experience. Something settles. Something that is usually tense and braced relaxes into function. The load feels, after the first uncomfortable minutes, like it belongs there.
That sensation is real, though its explanation requires a caveat. Evolutionary narratives—including both the “born to run” hypothesis and the “born to carry” hypothesis presented here—are compelling but ultimately unfalsifiable in the strict scientific sense. We cannot rerun hominin evolution with different selection pressures. The carrying hypothesis is consistent with the fossil evidence and with comparative anatomy, but “consistent with” is not “proven by.” The biomechanical arguments in this chapter and the next are on firmer ground than the evolutionary narrative, because biomechanics can be measured directly. The evolutionary story provides context and motivation. It does not provide proof.
In comparative anatomy, there is a concept called the functional complex: a suite of morphological features that work together to perform a specific function, and that can be understood only in relation to each other rather than in isolation. The human lower limb and posterior chain constitute such a complex, and the function they are optimised for, when you examine the entire suite rather than individual features, is unmistakably loaded bipedal locomotion.
Consider the architecture from the ground up. The plantar arch—that elegant biomechanical spring that chimpanzees lack entirely—stores elastic potential energy during the contact phase of each stride and releases it at push-off, reducing the metabolic cost of bipedal locomotion. It does this in walking as well as running, and the energy storage and return function becomes more important, not less, as the carried load increases, because the total work per stride is higher and any energy recovery reduces the net cost. The foot of Homo sapiens is a carrying foot as much as a running foot.
The Achilles tendon, already discussed, operates on the same principle. The human Achilles is approximately three times longer, relative to the muscle it connects, than the Achilles of a chimpanzee. This makes it a more effective spring. It also makes it more vulnerable to injury under acute high-loading conditions—a trade-off that makes sense if the primary selective environment was sustained moderate loading rather than explosive loading, which is exactly what carrying requires.
The hip joint of Homo sapiens is configured for sustained weight-bearing in a way that the chimpanzee hip is not. The acetabulum faces more inferiorly and laterally in humans than in chimps, orienting the femoral head to bear axial load most efficiently in the upright posture. The femoral neck angle is narrower, bringing the femur more directly under the centre of mass. The iliac blade is shorter and broader, orienting the gluteal musculature more posteriorly, improving its leverage for hip extension and trunk stabilisation. These features improve walking and running efficiency equally, but they improve loaded walking—where trunk stabilisation demands are highest—most of all.
The spine deserves particular attention. Human lumbar lordosis—the characteristic inward curve of the lower back—is unique among primates. Chimpanzees have a relatively flat lumbar spine. Humans have a pronounced curve that positions the lumbar vertebrae in a configuration that bears axial load through the cancellous bone of the vertebral bodies rather than through the posterior elements and facet joints. This matters because the vertebral bodies are much better designed to bear compressive load than the facets are. The lordotic curve optimises the spine for bearing weight from above—which is exactly what a torso-mounted load requires. A flat lumbar spine, under a heavy pack, would concentrate stress on exactly the structures least able to bear it. The lordotic spine distributes that stress across the most appropriate structures. It is the spinal architecture of a load-bearing animal.
The thoracic cage is similarly revealing. Human ribs are more cylindrical than the conical rib cage of great apes, creating a torso cross-section that is roughly barrel-shaped rather than funnel-shaped. This configuration provides a stable platform for a dorsal load: a pack resting against a barrel-shaped torso contacts a large surface area and does not tend to slide sideways. A pack on a funnel-shaped torso would be perpetually fighting to maintain its position. The human rib cage is, among other things, a load-bearing platform.
Every level of the anatomy, from the plantar arch to the lumbar curve to the rib cage geometry, tells the same story. This body was built to carry.
There is one more piece of the evolutionary argument that deserves careful attention, because it challenges the most common objection to the carrying hypothesis: if we evolved primarily as carriers, why are we such exceptional runners?
The answer is that these capabilities are not in tension. They are additive. The anatomy that makes loaded walking efficient also makes running possible, with refinements. The anatomical features that specifically enhance running—the nuchal ligament, the semi-circular canal configuration for balance at running speeds, certain specific features of the foot’s energy return geometry—appear in the fossil record approximately two million years ago, at the emergence of Homo erectus, and represent enhancements to a carrying platform rather than replacements of it.
Think of the evolutionary sequence this way. The baseline platform is a bipedal carrying primate, established in the australopithecine lineage three to four million years ago, with the full suite of torso-load-bearing anatomy in place. On top of that platform, beginning around two million years ago and continuing through the emergence of Homo sapiens, a set of running-specific enhancements are added. The resulting creature can both carry and run, and can do both with extraordinary efficiency compared to any other primate. The running is real. The running adaptations are real. But the substrate on which those adaptations are built is a carrying body, not the other way around.
This is why, when you ask people who have recently begun rucking what the activity feels like, the most common answer involves some version of recognition rather than novelty. Not “this is hard” or “this is strange” but “this feels right” or “my body knows how to do this.” That sensation of rightness is not mere sentiment. It is the subjective experience of a body performing the movement pattern for which its deep architecture was optimised. The feeling is species memory in the form of kinesthetic ease.
Running, by contrast, requires significant practice and adaptation before it feels natural. Beginning runners experience injury, discomfort, and a persistent sense that their body is being asked to do something it finds effortful in a particular way—not the good effort of working muscles but the effortful wrongness of tissue under inappropriate stress. This is not because running is bad or unnatural. It is because the running adaptations were added on top of a carrying baseline, and the carrying baseline is the deeper substrate. Running requires learning. Loaded walking, done at appropriate intensity and load progression, comes home.
Consider the injury data alone. Epidemiological studies consistently find that approximately forty to sixty percent of recreational runners sustain an overuse injury in any given twelve-month period. The knee is the most common site—the iliotibial band, the patellofemoral joint, the patellar tendon. The plantar fascia, the Achilles, the tibial stress reaction. The injury landscape of recreational running is so extensive and so consistent that sports medicine clinics in major cities are built substantially around the running patient. This is not the injury profile of a body perfectly adapted to its exercise. This is the injury profile of a body performing a movement that, while within its biological capability, lies at the edge of its tolerance when performed at the volumes that recreational culture demands.
The injury landscape of rucking, at appropriate load progression, is dramatically different. The ground reaction forces are lower. The repetition rate is lower. The tissue loading, while real, falls within the adaptive range that the musculoskeletal system was selected to handle across hours of daily activity. Rucking is not injury-proof—no physical activity is—but it is injury-resistant in a way that running fundamentally is not, because it operates more fully within the biomechanical parameters of the body’s evolutionary design.
This is not an accident. It is a consequence of evolutionary history. The body was not primarily designed to run. It was primarily designed to carry. When you do the thing it was primarily designed to do, it tends to hold together. When you push into the secondary capability, the edges show.
The recreational hiking literature, while limited, supports this picture. Long-distance hiking surveys consistently find that the injury landscape of people walking with packs looks nothing like the injury landscape of runners. A 2021 survey of 1,295 Appalachian Trail hikers found that sixty-one percent reported musculoskeletal complaints—but the dominant sites were the knee, ankle, foot, and lower leg (Chrusch & Kavin, 2021). Back injuries, when they appeared at all, were a minor secondary category. A NOLS risk-factor analysis of 1,283 participants found that pack weight, age, sex, and body weight were not statistically significant predictors of musculoskeletal injury during hiking (Hamonko et al., 2011). The body, when allowed to set its own pace across terrain it can choose, does not appear to struggle under recreational loads.
The caveat here is important and should be stated plainly: no longitudinal spinal imaging data exist for recreational backpackers. No one has ever performed serial MRI of people who hike with packs over years or decades and compared them to sedentary controls. Military data clearly show that loads of thirty to sixty percent of body weight accelerate disc degeneration under forced-march conditions—but recreational hikers typically carry ten to twenty-five percent of body weight, at self-selected pace, with full autonomy to stop and rest. These are categorically different loading regimes. Inferring spinal harm to recreational hikers from military data is like inferring tendon damage from marathon runners and applying it to casual joggers. The comparison lacks basis.
What the recreational data do show is that conditioning before you hike matters substantially. The Appalachian Trail survey found that hikers who did no pre-hike training had 2.82 times the odds of musculoskeletal injury compared to those who prepared (Chrusch & Kavin, 2021). Load carriage, at appropriate volumes and with progressive preparation, appears to be self-pacing enough to remain within the musculoskeletal system’s adaptive tolerance. This aligns with what Malville and colleagues found in Nepali porters: lifelong carrying without persistent musculoskeletal problems, attributable to “self-paced, intermittent exercise strategies” (Malville et al., 2001). The body knows how to do this. The body has known for a long time.
Return to the moment at the beginning of this chapter. An australopithecine in the East African Rift Valley, upright and walking, carrying something of value. It is not a romantic image in the conventional sense. There is no drama, no speed, no predator, no chase. Just an animal moving through a landscape, carrying weight, at a pace it can sustain across the hours between sunrise and the evening cooling.
That image contains the entire evolutionary logic of your body. Every pound of muscle in your posterior chain. Every millimetre of Achilles tendon. Every lordotic curve in your lumbar spine. Every mechanical detail of your plantar arch, your gluteus maximus, your barrel-shaped rib cage, your posteriorly-oriented hip musculature. All of it was shaped by selection pressure acting on that image, repeated millions of times across millions of years, in millions of ancestral bodies that either carried well and survived to reproduce or did not.
You are the survivor of that selection. Your body is its product.
When you put a pack on your back, you are not doing something new or unusual or particularly athletic. You are returning to the baseline. You are performing the movement that your musculoskeletal architecture was selected to perform, in the loading pattern that your metabolic systems were evolved to sustain, at the intensity that your cardiovascular machinery was calibrated for across millennia.
The fitness industry will tell you that exercise requires novelty, intensity, variety, and constant escalation. That you need to shock your muscles, confuse your metabolism, push past discomfort, and find the next level. This is marketing. It is marketing built on a fundamental misunderstanding of what the human body actually needs, because it is built on a misunderstanding of what the human body actually is.
What you actually are is a porter primate. The only primate on Earth that can carry meaningful loads on its torso while walking for hours at aerobic intensity without stopping. No other primate can do this. Not the chimpanzee, not the gorilla, not the orangutan. Their anatomy does not support it. Their cardiovascular systems, their posterior chains, their spinal geometry, their foot architecture—none of it is configured for the sustained loaded bipedal locomotion that is the human specialisation.
You inherited that specialisation. It was given to you by three and a half million years of selection acting on your ancestors in the landscape where bipedalism began. It is the deepest physical fact about you. It is more fundamental than your height, your fitness level, your weight, your age, your injury history. It is the thing your body was built to do before it was built to do anything else.
Wall-Scheffler showed that the torso-centred position is the most efficient carrying configuration available to a bipedal primate. Carvalho showed that the desire to carry drove the original shift to bipedalism in the primate lineage. Bramble and Lieberman showed that the running features of the human body were built on top of a walking and carrying baseline that was already millions of years old. The Laetoli footprints showed that the baseline was established before any of the running enhancements appeared.
The complete picture is not complicated. It is, in fact, the simplest possible story: a primate started carrying things upright, and the body that results from millions of years of selection for that task is the body you have. The backpack is not an accessory. It is not a military tool or an outdoor recreation product or a fitness equipment category. It is the external expression of an internal architecture that has been building since before the word “human” had any meaning.
You were born to carry.
The pack belongs on your back.
It has always belonged on your back.
Put it there, and walk.