The greatest advances sometimes come not out of the ground but out of a drawer. In 2015, two Australian researchers combing the collections of a South African site came upon a few slivers of bone, one filed in a box of primate remains, the other at the bottom of a bag of unsorted material. These fragments, excavated years earlier and catalogued as animal bone, had lain there for nearly thirty years. Pieced together, they formed the base of a human skull. And not just any skull: a piece of an adult of a species thought absent from this corner of the world at so remote a date.3
On 25 September 2026, the team of Jesse Martin, of La Trobe University in Australia, published in the journal Annals of Human Biology the description of eighteen homininHomininMember of the subtribe Hominina, comprising the human lineage (Homo, Australopithecus, Paranthropus…) but excluding orangutans and gibbons. The term progressively replaces "hominid" in its narrow sense.→ fossils from the site of Drimolen, including this fragment named DNH 127.1 Assigned, cautiously, to Homo erectusHomo erectusAncient human species, the first hominin to leave AfricaAfricaThe cradle of humankind: the continent where the first hominins appeared, then Homo sapiens around 300,000 years ago, before the expansion to the rest of the world.→ and the maker of the earliest AcheuleanAcheuleanLower Palaeolithic technical culture characterised by hand axes, present across three continents.→ handaxes.→, it is dated to about two million years. It thereby becomes the oldest known adult fossil of this species, roughly 200,000 years before the famous skulls of Dmanisi, in Georgia. A bone the size of a palm is sometimes enough to shift the lines of a story thought settled, the story of the origins of our own genus.
Drimolen, the densest concentration of ancestors in the world
The site of Drimolen Main Quarry lies about forty kilometres north of Johannesburg, at the heart of the Cradle of HumankindCradle of HumankindA karstic area northwest of Johannesburg, inscribed by UNESCO, among the richest deposits of human fossils in the world.→, a region inscribed on the UNESCO World Heritage list for the exceptional richness of its fossil deposits.8 Discovered in 1992, it has since yielded 119 hominin specimens, that is, humans and their extinct relatives. It is, to date, the greatest concentration of early human remains ever brought to light at a single site. Nearby, other famous caves of the same karstic massif have produced Little Foot, an australopith skeleton 3.67 million years old, and the small-brained Homo naledi.
This abundance has a flip side. Excavating is one thing; describing each piece scientifically is another, far slower matter. Many of the fossils recovered from Drimolen between 1992 and 2019, most of them collected by the team of palaeontologist Andre Keyser between 1994 and 1999, had never been formally published. They remained, in effect, unknown to science. It is this backlog that the 2026 study began to clear, by taking up a batch of eighteen skull and jaw fragments left in the stores.
Two of these pieces, as noted, had been filed among the animal remains. Their re-examination by Jesse Martin and his colleague A. B. Leece, in 2015, showed that they did indeed belong to hominins. The episode says a great deal about contemporary palaeoanthropologyPalaeoanthropologyThe science that studies human evolution from the fossil remains of hominins (bones, teeth, footprints) and their context, to reconstruct our biological origins.→: the essential treasures do not necessarily lie beneath the pick, but sometimes in the mislabelled boxes of collections already made. The authors draw from it a methodological plea. One must, they write, publish all hominin fossils after excavation, not just those that are well preserved or relatively complete. The broken shards, the unprepossessing fragments, deserve attention too, for they fill the gaps of a desperately incomplete fossil record.
One must still understand how human bones come to be trapped in such a quarry. Drimolen is an ancient dolomite cave, now largely collapsed and exposed by former lime workings. Over the millennia, sediments, debris and carcasses accumulated there, cemented by limestone into a hard rock called breccia. It is in this breccia, cut into pillars by the old miners, that the fossils lie. This genesis explains both their richness and the difficulty of dating them: one must reconstruct the succession of infills, cross-reference the age of calcitecalciteCrystalline form of calcium carbonate that makes up cave concretions (stalactites, stalagmites, flowstone).→ flowstones, the reversals of the Earth's magnetic field recorded in the sediments and the decay of uranium, in order to reach a reliable bracket. The precision of 2.04 to 1.95 million years is the fruit of this patient work of geochronology.
DNH 127, three slivers behind the ear
The study's star fossil is modest in size. DNH 127 brings together three pieces of bone from the right side and the back of the skull, which fit together just behind the ear. It is essentially a temporal boneTemporal boneA bone of the base and side of the skull that houses the inner ear and bears anatomical traits specific to each species.→, that complex piece at the base of the skull which houses the inner ear and bears the marks of the jaw's articulation. In an adult, its relief and sutures preserve a fine anatomical signature, specific to each species. It is in these details that the attribution is decided.1
Several clues rule out the other major species present at Drimolen, Paranthropus robustus. The bony bump behind the ear, the mastoid process, is only moderately swollen, whereas that of Paranthropus is massive. The squamosal suture, the seam where the temporal bone meets the parietal bone on the side of the skull, is almost straight and shows only a slight overlap of the two bones. A straight seam of this kind is characteristic of Homo erectus. Taken together, these traits are, according to the authors, incompatible with an attribution to Paranthropus. The fragment most closely resembles, they say, the skull D2700 from Dmanisi, an emblematic Georgian Homo erectus.1
The fossil's provenance reinforces its dating. DNH 127 comes from a block detached from the same pillar of sediment that yielded DNH 7, a nearly complete Paranthropus skull. This ties it to deposits dated, by a combination of methods (uranium-lead, electron spin resonance, palaeomagnetism), to between 2.04 and 1.95 million years ago.4 We are therefore around two million years, a threshold that, as we shall see, is anything but trivial.
The authors themselves remain measured. They assign DNH 127 to Homo erectus, they write, provisionally, because only part of the skull has survived. There is still no well-preserved and relatively complete adult Homo erectus skull from Drimolen or the neighbouring caves. Specialists even disagree about which species to assign some other early Homo fossils from the region: more than one form of Homo may have lived there two million years ago. As one astute reader puts it, with only the back of the skull, identification remains necessarily cautious. But what is at stake is considerable.
The choice of the temporal bone is no accident of preservation: it is one of the most information-rich pieces of the entire skeleton. Beyond the mastoid and the sutures, it houses the inner ear, whose semicircular canals bear on balance and posture, and the glenoid cavity, where the jaw articulated. Each of these features varies from one species to another and resists fossilisation well, for the bone there is dense. Reconstructing DNH 127 nonetheless demanded meticulous work: matching three scattered fragments, checking that their breaks coincide, making sure they come from a single individual. It is at this price that a cluster of slivers becomes an object of study again, carrying a species diagnosis. The authors stress the point: even incomplete, such a fragment yields more than an isolated tooth, long favoured in the counts.
Why two million years changes everything
To understand the stir caused by such a fragment, one must recall who Homo erectus is. It is the first hominin species that researchers broadly agree belongs to our own genus, Homo, and that is likely our direct ancestor. With a larger brain and a body of already human proportions, it is the first to have spread well beyond Africa, from Georgia to Indonesia and as far as China. To know where and when it appeared is to touch the starting point of our own history.2
Until now, the oldest adult Homo erectus fossils came from DmanisiDmanisiA site in Georgia yielding Homo erectus skulls about 1.8 million years old, among the oldest known adults outside Africa.→, in Georgia, and were dated to about 1.8 million years ago. These humans were among the first to leave Africa. With DNH 127, the marker moves back by about 200,000 years and shifts several thousand kilometres, toward the southern tip of the African continent. This is not the first alert from Drimolen: as early as 2020, the team had described DNH 134, the braincase of a child aged two to four, dated to about 2.04 million years and recognised as the oldest widely accepted fossil of the species.4
But DNH 134 posed a problem. In a very young individual, the proportions of the skull are not yet those of the adult, and its attribution to Homo erectus had fed debate. It was unknown what an adult of this early population looked like. It is precisely this void that DNH 127 fills. By providing an independent point of comparison, from an adult individual, it turns a fragile hypothesis, hanging on a child, into a far more solid proposition. Two fossils, one juvenile, one adult, now point to the same conclusion: Homo erectus was present in southern Africa two million years ago.
This chronological step back has knock-on consequences. It weakens, for instance, the idea that Australopithecus sediba, a slightly younger South African hominin, could have been the ancestor of the genus Homo: it is hard to be the forebear of a lineage already present elsewhere. It also runs counter to scenarios that placed the origin of Homo erectus in Asia. And above all, it reopens an old question, long settled in favour of a single region of the world.
To fix the appearance of Homo erectus is also to date a cluster of innovations that made it an actor of a new kind. Its body, with long legs and a narrow pelvis, heralds endurance walking and running, built to cover vast territories. Its cranial capacity, intermediate between that of the australopiths and our own, accompanies a more elaborate stone industry, the late OldowanOldowanThe oldest known stone-tool industry (c. 3.3–1.7 Ma), characterised by flaked pebbles (choppers) and basic flakes. Named after Olduvai Gorge (Tanzania).→ then the Acheulean with its symmetrical handaxes. It is again to Homo erectus, or to very close forms, that the oldest serious evidence of fire control and of a more carnivorous diet is attributed. To know that adults of this species already trod southern Africa two million years ago is to push back by as much the horizon of these transformations, and to question the pace at which they fell into place.
The cradle displaced, East versus South and the West African blind spot
For decades, the consensus seemed secure: Homo erectus was born in East Africa, in the great rift that runs from Ethiopia to Tanzania, where famous sites and spectacular fossils abound. The Drimolen discoveries unsettle this certainty. As Stephanie Baker, of the University of Johannesburg, one of the co-authors, puts it, until this find we always assumed Homo erectus originated from eastern Africa.6
The team's argument is twofold. On the one hand, southern Africa now yields the oldest adult and juvenile fossils of the species. On the other, Jesse Martin advances a more provocative reading of the East African data. According to him, several of the supposedly very old erectus fossils of the East rest on dates more fragile than is claimed, and could be markedly younger, on the order of 1.6 million years, the uncertainty reaching barely beyond 1.9 million years. If this reinterpretation were confirmed, southern Africa would no longer merely rival the East: it would precede it.
Martin takes the reasoning one step further. What if Homo erectus was born neither in the East nor in the South, but somewhere else again? He suggests that the species may have emerged in West Africa or Central Africa, regions very little explored from a palaeoanthropological standpoint, before reaching the south of the continent. The hypothesis remains speculative, for want of fossils, and that is precisely the problem. Our image of human evolution is shaped by the geography of research as much as by that of our ancestors. One finds mainly where one looks, and one looks mainly where one has already found. The East African riftEast African RiftA great fracture in the crust running north to south through East Africa, lined with lakes and volcanoes, the setting of much of human evolution.→ and the South African karstKarstA limestone landscape shaped by rock dissolution, rich in caves and passages; its sediments can preserve bone and DNA over long timespans.→ have concentrated excavation campaigns for a century; the forests and basins of the West and Centre, less favourable to preservation and prospecting, remain blind spots.
One must keep a cool head. Nothing today proves that Homo erectus appeared in the west: it is an invitation to search, not a conclusion. Likewise, the attribution of DNH 127 remains provisional, and a single fragment does not close a debate several generations of researchers old. But the strength of this case is to recall a truth too often forgotten: the map of discoveries is not the map of the past. It is the reflection of our own choices, our budgets, our fieldwork habits.
This debate is not new, and its history sheds light on the present. As early as 1924, the discovery in South Africa of the TaungTaungA site in South Africa where the child's skull, the type specimen of Australopithecus africanus, was discovered in 1924.→ Child, an australopith, had made the continent the candidate for the title of cradle of humankind, against the proponents of an Asian or European origin. Then the spotlight shifted toward East Africa, carried by prestigious sites such as Olduvai or the Afar and by fossils as famous as Lucy. For a time, it was even theorised that the great rift had, by its geography, favoured the emergence of hominins. Southern Africa, for its part, suffered from deposits harder to date, in cave breccia rather than in well-calibrated volcanic strata. The Drimolen discoveries, backed by a now robust geochronology, restore to the South a place that the enthusiasm for the East had made it lose.
Two neighbours, two destinies, erectus and Paranthropus
Drimolen is not only a matter of Homo erectus. The site tells above all of the coexistence, over a hundred thousand years, of two radically different forms of humanity. Alongside our ancestor lived Paranthropus robustusParanthropusA genus of fossil hominins known as robust australopithecines, with a powerful chewing apparatus, that lived in Africa about two to one million years ago.→, a robust hominin with a crested skull, endowed with a brain barely larger than a chimpanzee's, but with enormous molars driven by powerful chewing muscles. Two species, one landscape, two opposed evolutionary bets.2
The two lineages took divergent paths. Homo erectus was developing a larger brain and smaller molars, a change that Jesse Martin associates with a growing share of meat in the diet. Paranthropus robustus, by contrast, kept a small brain while acquiring oversized grinding teeth and ever stronger chewing muscles, to process hard and fibrous foods. Martin compares this apparatus to a mortar and pestle: a crushing machine, specialised in tough plant resources.
The most striking contrast lies in the numbers. At Drimolen, Paranthropus robustus far outnumbered Homo erectus, by more than ten to one. In other words, at that time it was our ancestor that looked like the outsider, the long-odds bet, where the robust one seemed the safe value of the moment. We know, with the hindsight of two million years, which of the two lineages flourished and which died out. But nothing, two million years ago, foreshadowed this outcome. A palaeontologist betting on the winner would probably have backed the wrong horse.
An attentive reader will object, rightly, that the number of fossils does not mechanically reflect the number of living individuals: the vagaries of fossilisation and discovery distort the proportions. The authors are aware of this. But the gap is so wide, and the Drimolen series so rich, that it probably reflects a demographic reality, and not merely a preservation bias. On this Early PleistocenePleistoceneThe geological epoch of the great ice ages (c. 2.6 Ma–11,700 BP), spanning most of human prehistory.→ landscape, the robust one was at home; our ancestor was still only a discreet newcomer.
One must still picture the scene. These two species did not necessarily share the same hours or the same corners of the landscape, and nothing proves that a given erectus individual ever crossed paths with a given Paranthropus. The cave, moreover, was not their habitat but a bone trap: carnivores, raptors and simple run-off carried in remains from the surroundings. In South Africa, the big cats, and leopards in particular, are formidable accumulators of bone, dragging their prey into the shelter of trees or into crevices. A number of the Drimolen hominins probably ended up that way, prey rather than predator. This taphonomic reading, the study of how bones are laid down and preserved, is indispensable if one is not to mistake the composition of a deposit for a faithful portrait of a living community.
Microevolution, from Drimolen to SwartkransSwartkransA fossil-bearing cave in the Cradle of Humankind, South Africa, rich in remains of Paranthropus robustus and early tools.→ in 200,000 years
The richness of Drimolen offers a luxury very rare in palaeoanthropology: to follow a species as it changes over time. By comparing the Paranthropus robustus of Drimolen, about two million years old, with those of the neighbouring cave of Swartkrans, dated to 1.8 million years, the researchers were able to trace anatomical transformations over an interval of about 200,000 years. This is what is called microevolutionMicroevolutionSmall anatomical changes within a single lineage over short periods, here captured between Drimolen and Swartkrans.→, those small steps that are usually hard to grasp, for want of fossils numerous enough and close enough in time.2
The finding is clear: over these two hundred millennia, the teeth of Paranthropus robustus grew larger still, and with them the chewing muscles. The species was pushing its grinding specialisation further, probably under the effect of climatic pressures that altered the availability and texture of foods. Hominin fossils are so rare that, most often, the differences between them are enormous and researchers are reduced to conjecture about what is missing. The large number of Drimolen fossils allows, for once, an unprecedented glimpse of the small-scale changes that shaped our relatives.
These observations reinforce an earlier proposal by the same team: to recognise the Drimolen population as a distinct subspecies, under the name Paranthropus robustus ukusa. The 2026 study adds to this case a possible jaw trait. On a lower jaw, DNH 5, the joint at the back of the mandible sits higher than the bony point where a chewing muscle attaches. The authors see in this a sign of a perhaps weaker bite in the Drimolen population than in the later one at Swartkrans, even though no formal biomechanical test has yet been carried out to establish it.
Caution remains in order. A very large Paranthropus jaw, DNH 6, comes from rubble left by former lime miners, which makes its age uncertain. If it turned out to be older than 1.95 million years, it would complicate the neat story of a gradual enlargement of teeth and jaws over time. The microevolution read at Drimolen is a slender thread, precious but fragile, that will have to be confirmed by other pieces.
Why do the molars swell in this way? The climatic context offers a lead. The Early Pleistocene sees, in southern Africa, a trend toward aridification and the expansion of open environments, grasslands and savannas at the expense of forests. Plant resources there become harder, more abrasive, more fibrous: tubers, rhizomes, seeds, tough stems. For a lineage that had bet on grinding, the adaptive response runs through an ever more powerful chewing apparatus. The enlargement of Paranthropus teeth between Drimolen and Swartkrans thus reads as the signature of an environment growing harder, in the literal sense. Conversely, Homo erectus seems to have met the same challenge by another route, broadening the diet toward meat and, no doubt, processing food with tools. Two strategies, one and the same changing world.
The collection, eighteen fossils and all that remains to be described
Of the eighteen newly classified specimens, only one was assigned to Homo erectus, the famous DNH 127; fourteen go to Paranthropus robustus; three remain too fragmentary to be tied to a genus.3 Behind these figures hides a gallery of remarkable pieces, some of which already rank among the emblems of African palaeoanthropology.
The most famous is DNH 7, the most complete Paranthropus robustus skull ever discovered, unearthed by Andre Keyser in 1994 and since recognised as that of a female. To it are added DNH 155, of exceptional preservation quality, as well as DNH 134, the child braincase assigned to Homo erectus, and DNH 152. To these skulls are added pieces from the rest of the skeleton, rare and precious: the pelvis DNH 43, dated to about two million years, whose digital reconstruction revealed an anatomy recalling that of the australopiths and Paranthropus, without its attribution being certain, the Paranthropus robustus hypothesis remaining the most reasonable provisionally.5 There are also mandibles such as DNH 8, which enrich the record of the site's jaws.
This profusion made it possible to isolate a mandibular trait that could distinguish the Drimolen population from those of the other South African sites, and attest to gradual changes over time. But the essential message perhaps lies in what has not yet been said: much of the Drimolen material remains undescribed. The eighteen fossils of 2026 are only one batch among many others, and the three-dimensional scans now available will open the way to future comparisons with discoveries across the continent.
Finally, one must salute the lesson in method that this work carries. For thirty years, decisive fragments waited, mislabelled, in the stores of a museum, within reach of science but ignored by it. Their rediscovery is a reminder that the story of human origins plays out not only in the field, under a blazing sun, but also in the obscure patience of collections. At Drimolen, two million years after two humanities crossed paths there, it is a forgotten bone that, resurfacing from a drawer, pushed back in one and the same movement the age of Homo erectus and the frontiers of our ignorance.
What comes next will play out on several fronts. The three-dimensional digitisation of the fossils, circulating between laboratories the world over, now makes it possible to compare DNH 127 finely with D2700 or with the East African specimens, without moving the fragile originals. New excavation campaigns, at Drimolen as at still unexplored sites, could yield the complete adult that is so sorely missing to settle the matter. And if Jesse Martin's call to prospect West and Central Africa were heeded, the centre of gravity of our story of origins could, once again, shift. In the meantime, DNH 127 imposes a double humility: before the antiquity of our ancestors, greater than we thought, and before the extent of what museum stores may still hold, within sight, awaiting a fresh eye.
No comments yet. Be the first.