In the north of the Indonesian island of Sumatra lies a lake of tranquil beauty today, about a hundred kilometres long, filling the floor of a gigantic depression. Nothing in this peaceful landscape hints that it is the scar of one of the greatest cataclysms our planet has known since the appearance of humankind. About 74,000 years ago, the Toba volcano exploded with an almost inconceivable force, emptying its reservoirs of magma, darkening the sky, and scattering its ash over thousands of kilometres. For decades, a seductive hypothesis held that this eruption had plunged the Earth into a volcanic winter so harsh that the whole of humanity found itself on the brink of extinction, reduced to a few thousand individuals. Then stone tools, unearthed in a valley of central India, came to crack that narrative. This feature retraces the history of the Toba super-eruption, that of the catastrophe hypothesis, and that of the Dhaba site which, patiently, overturned its logic.
A mountain that collapses on itself
Toba is no ordinary volcano. It is what geologists call a calderaCalderaA large circular depression formed by the collapse of the roof of a magma chamber emptied during a major eruption. The Toba caldera, now filled by a lake, measures about 100 km by 30 km.→, a vast basin born from the collapse of the ground after the sudden emptying of a magma chamber. The Toba caldera measures about one hundred kilometres long by thirty wide, making it the largest known Quaternary caldera in the world. The eruption that carved it, named the Youngest Toba Tuff by specialists, is dated by the argon-argon method to around 73,900 to 73,700 years before present, a figure commonly rounded to 74,000 years.4
The orders of magnitude are dizzying. The volcanologists William Rose and Craig Chesner estimated that the eruption ejected at least 2,800 cubic kilometres of material in dense-rock equivalent, including about 2,000 cubic kilometres of ignimbriteignimbriteCompact volcanic rock formed from pyroclastic flows, harder than tuff, forming the protective caps of the fairy chimneys.→, those incandescent ash flows that race down the slopes, and nearly 800 cubic kilometres of ash hurled into the atmosphere. Expressed as a volume of deposits, the total approaches 5,000 to 6,000 cubic kilometres. For comparison, the 1991 eruption of Pinatubo, one of the most striking of the twentieth century, produced about a thousand times less material.4
With such a tally, Toba is assigned a volcanic explosivity index of 8, the maximum value on the scale, reserved for the rare super-eruptions. It is the largest known explosive eruption of the whole Quaternary, that is, of the roughly 2.6 million years that encompass the history of the genus Homo. One detail is worth stressing: a study of the zirconZirconA highly durable, uranium-bearing mineral whose decay allows precise dating of the grain that holds it.→ crystals contained in the lavas suggests that this paroxysm was not preceded by a massive influx of fresh magma. Deep magmatic activity had even notably declined well before the catastrophe, a reminder that the greatest eruptions do not always announce themselves with spectacular signals.
The Youngest Toba Tuff, a planetary signature
What makes Toba such a precious marker for scientists is that its ash settled over a considerable portion of the globe. The plume drifted westward, carried by the high-altitude winds, and dusted the Indian Ocean, the Arabian Peninsula, the Indian subcontinent, and as far as East 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.→. In many places this grey layer forms a stratigraphic landmark of remarkable sharpness, a moment frozen in the rock that is found from one site to another.
In India, ash deposits several metres thick, sometimes more than two, have been identified, witnesses to the scale of the fallout thousands of kilometres from the source. Even in the sediments of Lake Malawi, in East Africa, researchers have spotted shards of volcanic glass attributable to Toba. These microscopic fragments, invisible to the naked eye, have a name: cryptotephraTephraA generic term for all the solid fragments (ash, lapilli, pumice, blocks) ejected into the air by a volcanic eruption. Tephra layers serve as precise chronological markers (tephrochronology) across vast regions.→, particles of tephraTephraA generic term for all the solid fragments (ash, lapilli, pumice, blocks) ejected into the air by a volcanic eruption. Tephra layers serve as precise chronological markers (tephrochronology) across vast regions.→ so scattered that they must be hunted down under the microscope. Their study has become a discipline in its own right, for it allows very precise correlation of distant sites by tying them to one and the same instantaneous event.
This worldwide dispersal fed a simple and spectacular idea: if the ash of Toba reached Africa and India, then the human populations of those regions must have suffered the effects of the eruption. From this arose the hypothesis that Toba had not only marked landscapes but also bent the course of human evolution. It still remained to be shown that these effects had been catastrophic, and that is precisely what an American researcher attempted at the end of the 1990s.
The catastrophe hypothesis: Ambrose and the bottleneck
In 1998, the anthropologist Stanley Ambrose of the University of Illinois published in the Journal of Human Evolution an article that would durably mark the discipline: "Late PleistocenePleistoceneThe geological epoch of the great ice ages (c. 2.6 Ma–11,700 BP), spanning most of human prehistory.→ human population bottlenecks, volcanic winter, and differentiation of modern humans".3 In it he proposed an audacious synthesis linking a geological phenomenon, a climatic phenomenon, and a genetic enigma.
The starting point was an observation from population genetics. By comparing the DNADNAThe molecule carrying genetic information, used to reconstruct kinship between species.→ of present-day humans, researchers had found that our species shows a surprisingly low genetic diversity for such a widespread mammal. It is as though humanity descended from a very small number of ancestors who lived in a remote period. This narrowing is called a genetic bottleneckBottleneckA sharp, temporary reduction in a population's size that lastingly impoverishes its genetic diversity.→: an episode in which the population shrinks drastically, leaving only a fraction of the original diversity to survive. Estimates placed this bottleneck between about 50,000 and 100,000 years ago.
Ambrose then proposed a culprit. The Toba eruption, pinned to around 74,000 years ago, falls squarely within this window. He put forward the idea that it had triggered a volcanic winterVolcanic winterA prolonged global cooling of the climateClimateThe long-term average atmospheric conditions of a region; its variations (glaciations, aridifications) shaped migrations, agriculture and the collapse of prehistoric societies.→ caused by the injection of sulfate aerosols and ash into the stratosphere during a large eruption, which reflect sunlight and lower temperatures for several years.→, a brutal cooling caused by the sulfur aerosols injected into the upper atmosphere, which reflect sunlight back into space. This winter would have lasted six to ten years, followed by a phase of prolonged cooling. Plant resources would have collapsed, famine would have decimated human groups, and the world population would have plummeted to only a few thousand individuals, perhaps a thousand breeding pairs. To support his reasoning, Ambrose relied in particular on the analysis of pollen taken from a marine core in the Bay of Bengal, above the Toba ash layer. The hypothesis was elegant: a single event explained both the worldwide volcanic trace and the narrowness of the human gene pool.
Dhaba: an 80,000-year archive in the Son Valley
It was in central India, thousands of kilometres from Sumatra, that this fine construction would meet its most solid counter-fire. In the state of Madhya Pradesh flows the Son River, a tributary of the Ganges, which has carved a wide valley in the plain whose banks preserve thick sediment deposits. It is there, at a site named Dhaba, that an international team brought to light an exceptional archaeological sequence, published in 2020 in the journal Nature Communications under the direction of Chris Clarkson, of the University of Queensland.1
The interest of Dhaba lies in the continuity of its deposits. The sediments piled on the banks of the Son record, without notable interruption, a period of about 80,000 years. Thanks to luminescence datingLuminescence datingA method dating the last heating or light exposure of sediments and minerals by measuring energy trapped in the crystal.→, which measures the energy stored by quartz grains since their last burialBurialThe intentional deposition of a body, sometimes with offerings; a marker of symbolic behaviour.→, the researchers were able to anchor this sequence in time. The human occupation brackets the Toba eruption, marked on site by the presence of cryptotephra, between ages of about 79,600 and 65,200 years. In other words, archaeologists here have a record covering the precise moment of the catastrophe: before, during, and after.
Dhaba is not an isolated site in the region. The Son Valley has long been known to prehistorians, and earlier work had already spotted Toba ash layers there. What changes with Dhaba is the quality of the record and the fineness of the chronological anchoring, which make it possible, for the first time, to follow step by step what the local populations did on either side of the event. The answer lies in the stone tools found at every level of the sequence.
Unchanged tools on either side of the ash
The heart of the demonstration by Clarkson and his colleagues rests on an observation of great sobriety: the lithic industry of Dhaba does not change as one crosses the Toba ash layer. The tools found below the eruption level are the spitting image of those above it. There is no technical break, no abandonment of the site, no cultural collapse, only the quiet continuation of one and the same tradition.
The artefacts belong to the Middle PalaeolithicMiddle PalaeolithicA Palaeolithic period (c. 300,000 to 40,000 years ago) associated mainly with NeanderthalsNeanderthalsA fossil humanity of Eurasia, robust and cold-adapted, extinct around 40,000 years before present.→ and early Homo sapiens, marked by Levallois tools.→, a period characterised by elaborate knappingknappingThe set of operations for fracturing a stone block to extract flakes or blades.→ methods. One recognises in particular Levallois knappingLevallois techniqueA Middle Palaeolithic flintFlintA hard, brittle siliceous rock, knapped by prehistoric people to produce blades, points and sharp tools.→-knapping method: a core is prepared so a flake of predetermined shape can be struck off in one blow; a hallmark of Neanderthal skill.→, a refined technique that consists in methodically preparing a block of stone, the core, so as to detach from it, with one precise blow, a flake of predetermined shape. This predetermination implies advanced mental planning: the knapper knows, before striking, what shape the final flake will take. It is one of the markers of the technical sophistication of the Middle Palaeolithic, and it is present at Dhaba both before and after the rain of ash.
The continuity does not even stop at the eruption. The Dhaba sequence continues well beyond it, recording, around 48,000 years ago, the appearance of microlithicmicrolithicRelating to the production of very small stone tools, often hafted, characteristic of the later periods of the Palaeolithic.→ technologies, that is, the production of very small tools a few centimetres long, often hafted. This late evolution testifies to a gradual, homegrown transformation of technical traditions, and not to a collapse followed by a repopulation. As Chris Clarkson sums it up, the same toolkits are present at Dhaba before and after the super-eruption, indicating that populations survived the so-called catastrophe.1
Jwalapuram, the Petraglia precedent
Dhaba did not spring from nowhere. It extends and consolidates a controversy opened thirteen years earlier by the archaeologist Michael Petraglia, who also works on the prehistoryPrehistoryThe span of human history before the invention of writing, from the Palaeolithic to the Metal Ages, known mainly through material remains.→ of the Indian subcontinent. In 2007, Petraglia and his team published in the journal Science an article with an explicit title: "Middle Paleolithic assemblages from the Indian subcontinent before and after the Toba super-eruption".2
The site in question was Jwalapuram, in southern India, where a particularly thick Toba ash layer crops out, measured in places at more than two and a half metres. By excavating the deposits located immediately below and above this layer, the team collected hundreds of stone tools. And here too, the finding was one of strong resemblance between the assemblages before and after the eruption. For Petraglia, this technical continuity meant that the region's toolmakers had not been annihilated by the catastrophe: they had persisted, their material culture intact.
Petraglia went even further. He noted that the Jwalapuram tools strongly recalled the industries of the African Middle Stone AgeMiddle Stone AgeA long phase of African prehistory (c. 300,000 to 40,000 years ago), the African equivalent of the Middle Palaeolithic, marked by the rise of symbolic behaviour.→, that phase of the Palaeolithic when the first modern humans lived. From this he drew the hypothesis, audacious for the time, that Homo sapiensHomo sapiensThe present-day human species, which emerged in Africa around 300,000 years ago, the only surviving human lineage after the extinction of Neanderthals and Denisovans.→ might have reached India earlier than thought, perhaps even before Toba. The idea was far from unanimous, and it remains debated. But it laid the groundwork for a debate that Dhaba would revive with more robust data: not only had humans survived Toba in India, but they may have been present there as representatives of our own species.
Resilience on a planetary scale
If the catastrophe hypothesis were correct, it is everywhere in the affected zone that one should read, in the archaeological record, a trace of the collapse. Yet the opposite emerges, including far from India. In southern Africa, on the south coast of South Africa, two nearby sites, Pinnacle Point and Vleesbaai, have yielded shards of Toba volcanic glass. A team led in particular by the volcanologist Eugene Smith published in 2018 in the journal Nature a study showing that human occupation continues there before, during, and after the eruption.5
Better still: at Pinnacle Point, far from showing a decline, the intensity of human activity, measured by the abundance of bones, tools, and hearths, seems to increase in the levels after Toba, with even the rise of an early microlithic tradition. The ash had nonetheless arrived there after a journey of nearly 9,000 kilometres from Indonesia. Far from the vision of a dying humanity, the picture is one of populations that prosper and innovate in the decades following the greatest eruption of the Quaternary.
This resilience is part of a deeper movement: the great dispersal of Homo sapiens out of Africa. Setting out from East Africa, our ancestors spread toward the Near EastNear EastA region of western Asia (Levant, Mesopotamia, Anatolia, Iran), cradle of the Neolithic revolution, agriculture, the first cities and writing.→, Arabia, then South Asia, following a route that passes precisely through the Indian subcontinent. The culmination of this journey is documented at the other end of the world, at Madjedbebe, a rock shelterRock shelterA shallow cavity at the foot of a cliff or under a rocky overhang, offering natural shelter; a favoured site of prehistoric habitation and rock art.→ in northern Australia where human occupation has been dated to about 65,000 years ago by another study by Chris Clarkson, published in Nature in 2017.6 That humans reached Australia barely a few thousand years after Toba argues, once again, for a dispersal that was not broken by the eruption. The date of 65,000 years is certainly debated, with some researchers invoking soil disturbance, but it clearly places the peopling of Australia in the wake, and not in the shadow, of the Toba catastrophe.
Faceless tools: who knapped these stones?
One point deserves all the reader's caution. At Dhaba, as at Jwalapuram, it is stones that speak, not bones. No human fossil contemporary with Toba has been found at these Indian sites. The attribution of the tools to Homo sapiens therefore rests on the comparison of lithic industries alone, and not on the direct identification of their makers.
The authors of the 2020 study argue that the Dhaba industry strongly resembles the assemblages of the African Middle Stone Age, the Arabian assemblages, and the oldest Australian artefacts, which leads them to see it as likely the work of Homo sapiens dispersing eastward.1 But they themselves acknowledge that the timing of our species' arrival in India remains the subject of intense debate. Without a fossil, it remains conceivable that these tools were knapped by other 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.→ populations present in Asia at that time, whether archaicArchaicRefers to an ancient, now-extinct human population or form (Neanderthals, Denisovans, ghost lineages), as opposed to anatomically modern humans.→ lineages or groups whose existence modern genetics is only beginning to glimpse.
It is precisely on this ground that Stanley Ambrose pressed the counter-argument. As soon as the Jwalapuram work was published, he judged the data insufficient, summing up his skepticism in a formula that has remained famous: the only way to prove these assertions would be to find, beneath the ash, human skeletons that look like Africans. In the face of the Dhaba study, his reservation did not weaken. The debate illustrates a fundamental tension of the discipline: how far can one make tools speak when they are mute about the biological identity of those who made them? Technical continuity proves the survival of traditions; it does not by itself prove the identity of the survivors.
The twilight of a catastrophe
Beyond archaeology, it is the whole climatic and genetic edifice of Ambrose's hypothesis that has cracked over the years. On the climate side, the weak point is the estimate of the amount of sulfur injected into the atmosphere, for it is this that determines the magnitude of the cooling. In 2009, models published in the Journal of Geophysical Research still envisaged global temperature drops of 8 to 17 degrees Celsius.7 But more recent work argues that these models rested on sulfur dioxide values overestimated by one to two orders of magnitude, and thus on an exaggerated cooling.
The natural archives point in the same direction. The Greenland ice cores show, for this period, peaks of sulfate and calcium, but no volcanic ash links this signal directly to Toba: the attribution remains uncertain. Above all, the sediment core from Lake Malawi, which does indeed contain Toba glass shards, indicates a local temperature drop that would not have exceeded 1.5 degrees after the eruption. This is a far cry from the apocalyptic winter of several degrees supposed to last for years.
The genetic pillar has also weakened. Several recent reviews conclude that no solid signal supports a bottleneck precisely pinned around 74,000 years ago. The trough of diversity that is observed could be explained otherwise, for example by an ancient structure of populations, made of successive separations and admixtures, rather than by a single near-extinction. One must, however, guard against tipping into the opposite excess. Researchers such as Martin Williams, a specialist of the Indian subcontinent, recall that Toba did have real and sometimes marked environmental effects at the regional scale, notably on Indian vegetation. Ambrose himself maintains that one cannot prove the survival of populations without human remains. The position that emerges today is therefore not that of a harmless eruption, but that of a real catastrophe, locally severe, which Middle Palaeolithic human societies managed, in many regions, to absorb and overcome.
What Toba teaches us about ourselves
The story of Toba and Dhaba is exemplary of how prehistory advances. A powerful hypothesis, almost novelistic, that of a humanity that came within a hair of annihilation, dominated for nearly twenty years. It had coherence, elegance, and the thrill of our species' fragility on its side. Then facts, slowly accumulated in the mud of an Indian valley, on a South African coast, or in the ice of Greenland, came to put it to the test.
What the file reveals is not that Toba was insignificant, far from it. It was one of the greatest eruptions in the whole history of humankind, an event that transformed entire landscapes and cooled, at least in places, the planetary climate. But the lesson of Dhaba is that human groups, armed with their technical know-how and their capacity for adaptation, came through this episode without breaking the thread of their traditions. The Levallois knappingLevallois techniqueA Middle Palaeolithic flint-knapping method: a core is prepared so a flake of predetermined shape can be struck off in one blow; a hallmark of Neanderthal skill.→ practised before the rain of ash is found unchanged after it. The fabric of human gestures did not give way.
There remains the shadow, honest and fruitful, that comes from the absence of fossils. We know that someone knapped these stones, before and after the catastrophe, but we do not know the exact face of these survivors. Were they already modern humans come from Africa, or other hominins whose memory has faded? The question remains open, and it is perhaps this openness, more than any certainty, that gives the Dhaba site its value. It reminds us that humanity, in its multiple forms, was not a flickering flame ready to go out at the first breath, but a tenacious presence, scattered from one continent to another, and already able to survive the most violent thing the Earth could throw at it.
Mon pote m'avait parlé de ce site. Il avait raison, c'est vraiment bien expliqué.
C'est beau de voir que des gens comme moi peuvent comprendre ces sujets grâce à vous.
Cet exemple sera parfait pour illustrer une notion difficile à transmettre en visite guidée.
La catastrophe du Toba est un cas fascinant en paléoécologie humaine. Ce qui est remarquable, c'est que des populations humaines ont survécu à l'une des plus grandes éruptions volcaniques des deux derniers millions d'années. Cela témoigne d'une capacité d'adaptation et de résilience extraordinaire, probablement liée à la flexibilité comportementale propre à notre espèce.
La super-éruption du Toba il y a 74 000 ans est souvent présentée comme un goulot d'étranglement démographique majeur pour l'espèce humaine. Les analyses génétiques de la diversité actuelle d'Homo sapiens montrent effectivement une perte de diversité compatible avec un effondrement de la population à cette période. Mais des sites africains montrent une continuité d'occupation, nuançant le scénario catastrophiste.