The Straight Answer on When Animals Began

Animals first appear as unmistakable fossils about 574 million years ago, in rocks of the Ediacaran period, and then diversify spectacularly in the Cambrian explosion that begins 539 million years ago. Yet the genes of living animals tell an older story. Molecular clocks, which estimate the age of evolutionary splits from DNA differences, repeatedly place the origin of animals between 800 and 700 million years ago, deep in the Tonian period, long before the first fossil that anyone can confidently call an animal. A 2026 study in the journal Science Advances, led from the University of Oxford, argued that the fossil record’s silence before 574 million years ago has been misread, and that the molecular estimate deserves more weight than paleontologists have traditionally given it.

The thesis of this account is a single sentence: animals almost certainly originated tens to hundreds of millions of years before the first unmistakable animal fossils, because the fossil record before 574 million years ago measures the chemistry of preservation far more than it measures the history of life. Everything that follows defends that sentence, and one long section near the two-thirds mark argues against it as honestly as possible.

The question matters beyond curiosity. The origin of animals calibrates the entire tree of life: every molecular date for every later split, from the first fish to the first primates, hangs on where the animal root is fixed. It also sets the template for how science handles the most common situation in deep-time research, which is two excellent kinds of evidence disagreeing with each other. The fossils say one date. The genes say another. The chemistry of ancient oceans and atmospheres says a third, quieter thing. Learning how specialists weigh these against one another is the real payoff, because the same weighing happens in every field that studies a past it cannot visit.

What follows walks the full path. It defines the animal, dates the rocks, opens the great fossil deposits one by one, explains the molecular clock and why its calibration moved in 2026, weighs oxygen and ice and ecology as candidate triggers, stages the famous fights among paleontologists, and then turns around and gives the best case against the thesis. It closes with a reader’s toolkit for the next headline, because there will always be a next headline.

What Does the Word Animal Actually Mean to a Biologist

What Does the Word Animal Actually Mean to a Biologist?

To a biologist, an animal is a multicellular organism that eats other organisms rather than making its own food, and whose cells lack rigid walls. Sponges, jellyfish, worms, insects and humans all qualify. Fungi and plants do not, though fungi are animals’ closest relatives among the great kingdoms. The definition matters because the fossil hunt depends on it.

The everyday word is fuzzier than the technical one, and the fuzziness causes real confusion in this story. When a headline says scientists found the oldest animal, it means the oldest organism meeting the biological definition, not the oldest thing that looks like an animal to a casual eye. Many Ediacaran fossils look like animals and may not be animals at all, while some genuine early animals may have looked like nothing familiar.

Biologists divide animals into roughly thirty phyla, the great body-plan groups: sponges, cnidarians, arthropods, molluscs, chordates and the rest. Nearly all of them appear in the fossil record within the Cambrian period, which is the central strangeness this account must explain. Either the body plans evolved with astonishing speed, or they evolved slowly in bodies too soft to fossilize and only became preservable when skeletons and shells appeared. The whole debate is, at bottom, an argument about which of those two stories the evidence supports.

The definition also sets the search image. An animal fossil from 600 million years ago would not look like a trilobite. It would look like a smear, a tube, a disc, or a microscopic ball of cells. Paleontologists hunting the first animals are looking for the barely visible, in rocks that preserve almost nothing. That the search has found anything at all is remarkable, and that it has not found more is not, by itself, evidence of absence.

Classification itself has a long institutional history, and the modern scheme of phyla took shape through centuries of comparative anatomy. The way animals are grouped into phyla, classes and orders reflects both the history of discovery and the logic of shared body plans, a subject surveyed in depth in guides to zoological classification.

The Two Dates Every Reader Should Hold at Once

Every serious discussion of animal origins carries two dates that refuse to agree, and the reader who holds both at once already understands more than most summaries provide. The first date is 574 million years ago, the age of the oldest fossils that specialists widely accept as animals. The second date is a range, 800 to 700 million years ago, the origin window that molecular clocks keep returning. Between them lies a gap of roughly 130 to 230 million years, a span longer than the entire age of mammals, during which animals either existed without leaving accepted fossils or did not yet exist.

The gap is the subject. It is not a gap in knowledge that one more expedition will close, though expeditions keep trying. It is a structural disagreement between two ways of knowing. Fossils record the organisms that died in the right place at the right time with the right chemistry. Genes record the accumulated differences between living organisms, converted into time by a rate. Each method has failure modes the other does not share, and each is blind to the other’s evidence.

It helps to name the three questions that are often merged into one. First, when did the animal lineage split from its closest non-animal relatives? Second, when did the first organism recognizable as an animal live? Third, when did animals become large, diverse and ecologically important enough to leave an unmistakable record? The molecular clocks answer the first question. The fossil record answers the third. The second question, the one most readers mean, sits in the gap, and neither method reaches it cleanly.

This account will keep the three questions visibly apart. Much of the public confusion comes from writers who quote a molecular date as though it were a fossil date, or dismiss a molecular date because no fossil confirms it. Both moves mistake one question for another.

How Rocks Get Their Dates

How Do Geologists Date a Rock That Is Half a Billion Years Old?

Geologists date the rock around the fossil, using the steady decay of radioactive elements as a clock. In volcanic ash layers, minerals called zircons trap uranium when they crystallize, and the uranium decays to lead at a known rate. Measuring the ratio gives the age of the eruption, and any fossil between two dated ash layers inherits an age bracket.

The method that does most of the heavy lifting in deep time is uranium-lead dating of zircon crystals. Zircons are nearly indestructible, they lock in uranium and exclude lead at formation, and the two uranium isotopes decay to two lead isotopes at different rates, which lets geochronologists cross-check every measurement internally. A zircon from a volcanic ash bed in the Doushantuo Formation of China, which hosts the Weng’an Biota discussed below, can be dated to within a few million years at an age near 600 million years. That precision, better than one percent, is what makes the whole chronology of early animal life possible.

Sedimentary rocks, where fossils actually sit, cannot be dated this way because their grains are older than the rock. The trick is the ash layer. Volcanoes erupt, ash settles across a landscape or seafloor, and the ash becomes a thin, datable sheet inside the sedimentary stack. Find ash above the fossil and ash below it, date both, and the fossil’s age is bracketed. The famous Ediacaran fossil beds of Newfoundland are dated by ash layers in exactly this manner, and the technique underwrites nearly every number in this account.

There is a second, cruder clock: biostratigraphy, the ordering of fossils themselves. Once a distinctive fossil is dated somewhere by radiometric means, its appearance elsewhere marks the same interval. This is how field geologists correlate rocks across continents without a laboratory. It cannot give an absolute date on its own, and it assumes the fossil’s range is the same everywhere, which is not always true. But as a relative ordering tool it is indispensable, and it is the reason the Cambrian period can be recognized from Siberia to Morocco.

Error bars belong on every number. A date quoted as 540 million years ago carries an uncertainty of a few million years in the best cases and tens of millions where the ash layers are sparse. The arguments in this story turn on differences of tens to hundreds of millions of years, so the uncertainties rarely decide them, but they set the resolution limit. Anyone claiming to distinguish 600 from 590 million years ago on fossil evidence alone is claiming more than the methods allow.

The Fossil Record as a Biased Witness

What Is a Lagerstatte?

A Lagerstatte, from the German for a storage place, is a deposit of exceptional preservation where soft tissues survive as films, impressions or mineral replicas. The Burgess Shale of Canada and the Chengjiang biota of China are the famous examples. Without such deposits, whole chapters of animal history would be invisible, because most animals are mostly soft.

The fossil record is not a census. It is a biased witness with strong preferences, and every argument about early animals must discount for the bias before weighing the testimony. Hard parts preserve: shells, bones, teeth. Soft parts usually do not: muscles, guts, nerves, embryos. An animal with a shell is hundreds of times more likely to fossilize than a worm of the same size in the same mud. The Cambrian explosion is, among other things, the moment when hard parts become common, which means it is partly a record of a change in preservability rather than a change in life itself.

The bias has a chemistry. Exceptional preservation needs the right sediment, the right water chemistry, and burial fast enough to beat decay and scavengers. Fine muds, low oxygen at the seafloor, early mineralization: these are the conditions that turn a soft animal into a fossil. They are rare in any period and rarer in old rocks, which have had half a billion years to be cooked, squeezed and dissolved by geology. The older the rock, the more metamorphosis and erosion have destroyed, so the record thins with age for purely geological reasons.

Paleontologists have a name for the deposits that beat the odds, and the name is worth knowing because the whole early-animal story is told through a handful of them. Each is a window, and each window shows only the organisms that lived in that place under those conditions. The Weng’an Biota shows microscopic life in phosphorite. The Ediacara assemblages show seafloor impressions. The Burgess Shale shows a Cambrian marine community in exquisite detail. None shows the whole world, and the temptation to treat a window as a panorama is the standing fallacy of the field.

There is also a subtler bias: the search image. For a century, paleontologists looked for shells and skeletons because those were the fossils they knew how to find. Microscopic embryos, chemical biomarkers and trace fossils required new techniques and new questions. Each technical advance has pushed the apparent origin of animals further back, which is itself evidence that the record’s silence is at least partly a silence of method rather than of life.

The Weng’an Biota: A 590-Million-Year-Old Microscope Slide

High in the hills of Guizhou Province in southern China, rocks of the Doushantuo Formation preserve something no other deposit of its age preserves: entire microscopic organisms, cell by cell, replaced by phosphate with their three-dimensional structure intact. This is the Weng’an Biota, named for the county where it was found, roughly 590 million years old, and it is the deposit at the center of the 2026 argument about animal origins.

What Weng’an preserves is extraordinary. Algae, cyanobacteria, and enigmatic multicellular balls of cells appear in thin sections as if flash-frozen. Some of the cell clusters look like animal embryos caught at the cleavage stage, dividing from one cell to two to four to many. For a decade after their description, these were celebrated as the oldest animal embryos known, a direct glimpse of animal reproduction 590 million years ago.

Then the interpretation began to fray, as interpretations of ambiguous fossils usually do. Critics noted that giant sulfur bacteria and certain algae also divide in ways that mimic embryos, and that none of the Weng’an structures shows the diagnostic features of a true animal embryo, such as differentiated cell layers. The embryo claim, once a headline, retreated to a maybe. What remains undisputed is the deposit’s quality: if animals had been present in that environment in preservable form, Weng’an is exactly the kind of place that would have caught them, and no definitive animal has been found there.

That absence became an argument. A line of reasoning developed in the 2010s held that Weng’an sets an upper bound: because this exceptional deposit captured everything else alive at the time and no animals, animals must have originated after Weng’an, after 590 million years ago. The absence was promoted from a gap to a date. Molecular clock studies adopted Weng’an as a maximum calibration, a hard ceiling on the origin, and the clocks duly returned younger estimates. The 2026 Oxford-led study attacked exactly this move, and the attack is worth following closely because it shows how a single assumption can steer a whole field.

The Kheseen Biota and the Argument From a Younger Window

The weapon the 2026 study used against the Weng’an ceiling came from Mongolia. The Kheseen Biota, preserved in rocks more than 40 million years younger than Weng’an, is another phosphate deposit of exceptional microscopic preservation, studied in detail only in the years before the paper. It captures its contemporary microbial world in the same fine-grained fidelity. And it, too, contains no identifiable animal fossils.

The logic is a clean piece of reasoning, and it deserves to be stated plainly. During the time of the Kheseen deposit, animals are known to have existed, because animal fossils are documented from other places in rocks of the same age. Yet Kheseen, with preservation as good as Weng’an’s, preserved none. Therefore exceptional preservation does not guarantee the capture of animals even when animals exist. Therefore the absence of animals at Weng’an cannot prove animals did not yet exist. The ceiling collapses, not because anyone found an older animal, but because the argument from absence was shown to be unreliable.

With the ceiling removed, the study’s authors recalibrated their molecular clock using older geological constraints instead: fossil deposits from Norway, Australia and Arizona dated between 850 and 730 million years old, in the Tonian period. The estimated origin of animals shifted backward by roughly 200 million years, landing in the range of 800 to 700 million years ago. The paper appeared in Science Advances in October 2026, with coverage emphasizing that the origin of animals may lie deep in the Neoproterozoic Era, before some of the most extreme ice ages in Earth’s history.

Treat this as the science module demands: a peer-reviewed paper is a data point, and this one is a single study whose central move is methodological rather than evidential. It found no new animal. It argued that an old argument was unsound and recomputed the clock without it. Whether the recomputation is right depends on the clock model, the new calibrations, and the rate assumptions, all of which the complication section will interrogate. The study’s value here is as a permanent fact about how the debate moved, not as a verdict.

The Ediacara Biota: Strange Life on the Seafloor

Before the Cambrian, the seafloor belonged to the Ediacarans. Named for the Ediacara Hills of South Australia where Reginald Sprigg first recognized them in 1946, these are impressions of soft-bodied organisms in sandstone, ranging from small discs to meter-long fronds, dated roughly 575 to 539 million years ago. They are the oldest large, complex fossils known, and nobody fully agrees on what they were.

The classic Ediacarans look like nothing alive. Dickinsonia is a ribbed oval that may have crawled and fed, leaving feeding traces. Spriggina resembles a segmented head with no clear body. The frond-like rangeomorphs, studied in Newfoundland’s Mistaken Point assemblage by researchers including Guy Narbonne, stood upright on the deep seafloor like ferns, absorbing nutrients directly through their fractal surfaces. Adolf Seilacher, the German paleontologist who did more than anyone to systematize them, proposed they belonged to an entirely separate kingdom of life, the Vendobionta, quilted air-mattress organisms with no modern descendants.

Seilacher’s proposal was bold and probably partly wrong, but it forced the right question: are these failed experiments, or the first animals? The current working view splits the difference. Some Ediacarans, like Kimberella, a small bilaterian that scratched feeding traces into microbial mats, are plausibly stem-group molluscs, genuine early animals. Others, like the rangeomorphs, may be animals of an unfamiliar grade or something else entirely. Mikhail Fedonkin’s work on the White Sea assemblages of Russia added crucial detail, showing Ediacarans with bilateral symmetry, directed movement and complex behavior, traits that sit comfortably inside the animal kingdom.

What the Ediacarans establish is that large, mobile, complex multicellular life existed at least 575 million years ago. Whether every Ediacaran was an animal is secondary to the chronological point: the machinery of complex multicellularity was running before the Cambrian began. If the first animals were small and soft, as the molecular dates suggest, the Ediacarans show that their descendants could already be large and ecologically sophisticated within a few tens of millions of years. The fuse, in other words, was already lit.

The Cambrian Explosion, Phase by Phase

How Fast Was the Cambrian Explosion, Really?

Fast by geological standards, slow by human ones. The main burst of animal diversification spans roughly 20 million years, from 539 to about 521 million years ago. That is an eyeblink against the age of Earth, but it exceeds the time since the dinosaurs vanished, and it gave evolution room for extensive work.

The Cambrian explosion is the most famous episode in the history of life, and the name is half the problem. It suggests a detonation, a sudden appearance from nothing. What the rocks actually show is a rapid but structured diversification, in phases, over some 20 million years, of body plans that then persist for half a billion years. Understanding the phases dissolves much of the mystery, because each phase has its own mechanism.

Phase one, the Fortunian age from 539 to 529 million years ago, is the age of small shelly fossils: tiny tubes, caps, spines and sclerites, the disarticulated armor of animals whose full bodies are rarely preserved. These were the first widespread biomineralizers, animals that learned to precipitate calcium carbonate and calcium phosphate into hard parts. Phase two, beginning around 529 million years ago, brings the first trilobites and the great Chengjiang biota of Yunnan Province, China, dated to about 518 million years ago, which preserves soft-bodied arthropods, worms, and the earliest known vertebrate relatives in stunning detail.

Phase three is the Burgess Shale interval, around 508 million years ago, when Charles Walcott’s quarries in the Canadian Rockies captured a mid-Cambrian marine community with a diversity of body plans that startled everyone who saw it. By the end of the Cambrian, nearly every animal phylum that will ever exist has appeared in the record. The pattern is real: a geologically rapid filling of ecological and anatomical space.

The question is what the pattern means. The artifact reading says it means rapid evolution. The preservational reading says it means rapid skeletonization: the animals were already there, diversifying quietly as soft-bodied forms, and the explosion records the moment hard parts made them visible. The molecular dates favor the second reading. The trace fossil record, which shows increasingly complex burrows through the late Ediacaran, also favors it. The explosion was real as an ecological event, but as an evolutionary origin story it may be largely an artifact of what fossilizes.

Windows Onto the Explosion: Burgess Shale and Chengjiang

Charles Doolittle Walcott was the most powerful geologist in America when, in 1909, he found the first slabs of the Burgess Shale on a mountainside in British Columbia. Walcott, then secretary of the Smithsonian Institution, had been told of the fossiliferous shales by workmen on a trail, and his quarries over the following years yielded tens of thousands of specimens of soft-bodied Cambrian animals: the spiny Hallucigenia, the five-eyed Opabinia, the great predator Anomalocaris, alongside early arthropods, worms and chordates. The deposit, about 508 million years old, preserved what ordinary fossilization destroys.

Walcott’s interpretation was conservative to a fault. He shoehorned the strange animals into known groups, minimizing their weirdness, and his authority held the field for half a century. It took the restudy by Harry Whittington, Derek Briggs and Simon Conway Morris at Cambridge in the 1970s and 1980s to show how many Burgess animals fit no known phylum. Stephen Jay Gould’s Wonderful Life, published in 1989, turned their findings into the most famous argument in modern paleontology: that the Cambrian was a riot of anatomical experimentation, most of it pruned by contingency, and that replaying life’s tape would give a different world.

Conway Morris, who had done much of the restudying, dissented from Gould’s conclusion in The Crucible of Creation in 1998, arguing that convergence, the repeated evolution of similar solutions, shows evolution is channeled rather than chancy. The Burgess animals look strange, he argued, but their strangeness has been exaggerated, and many represent stem groups of familiar phyla rather than failed kingdoms. The dispute between Gould’s contingency and Conway Morris’s convergence remains the liveliest philosophical argument in paleontology, and it matters here because it sets the terms for reading the explosion: lottery or unfolding.

Chengjiang, discovered in Yunnan in 1984 and dated to about 518 million years ago, is older than the Burgess Shale and even more diverse, with early fish-like chordates such as Haikouichthys among its treasures. Together the two deposits show that by the mid-Cambrian the oceans held complex food webs with predators, scavengers, filter feeders and grazers. Whatever the deep origin date, the ecological revolution was complete by 510 million years ago. The engines of modern marine life were running.

What Darwin Knew and What Troubled Him

Charles Darwin understood the Cambrian problem better than most of his successors give him credit for, and he stated it with unusual candor in On the Origin of Species in 1859. If species evolve gradually, he wrote, the rocks below the Cambrian should be full of transitional forms, and they are not. The abrupt appearance of complex animals was, in his own assessment, a serious objection to his theory, explicable at the time only by the imperfection of the geological record.

Darwin’s explanation was the record’s incompleteness, and on that narrow point he has been vindicated beyond anything he could have imagined. The Precambrian fossil record he called “unknown” turned out to hold stromatolites, microfossils, the Ediacara biota and the Doushantuo embryos. Each discovery pushed recognized complexity further back. The pattern of Darwin’s objection being answered by new windows of preservation has repeated so often that absence in old rocks is treated as weak evidence by default.

But Darwin also made a prediction that has fared less well. He expected the transitional forms to be found, and in a sense they never were, because the transition he imagined, from simple to complex by small steps in preservable bodies, is not what happened. The molecular evidence suggests the deepest splits among animal lineages happened long before any of the lineages became fossilizable. The transitions are genetic and developmental, written in DNA and embryos, not in stone. Darwin’s mechanism was right; his expectation of what the rocks would show was wrong.

The history of this debate is itself a case study in how science revises received pictures when new evidence arrives, a pattern examined in the broader history of scientific revolutions. Each generation of paleontologists inherited a confident story about the Cambrian and then watched new deposits complicate it. The lesson is not that the science is unreliable but that its reliability comes from the revision, not from the confidence.

The Molecular Clock: How Genes Keep Time

The molecular clock is the second great witness, and it works on a principle of elegant simplicity. DNA accumulates mutations over time. If two species split from a common ancestor, each lineage gathers its own mutations independently, and the genetic distance between them grows with the time since the split. Count the differences, divide by the rate, and the quotient is the age of the split. This is the entire idea, and everything difficult follows from the two words “the rate.”

Rates are not constant. They vary between lineages, between genes, and over time, for reasons ranging from generation length to DNA repair efficiency to population size. Early molecular clocks assumed a strict rate, like a metronome, and produced dates that specialists regard as crude. Modern clocks are “relaxed”: they model rate variation explicitly, allowing different branches of the tree to tick at different speeds, and they estimate the rates and the dates jointly in a Bayesian framework. The sophistication is real, but so is the cost: every additional parameter is another assumption, and the dates that emerge are only as good as the assumptions that shaped them.

The clock needs calibration, and calibration is where fossils re-enter. A genetic distance is a number of mutations, not a number of years. To convert, the analyst anchors certain splits to fossil dates: the split between birds and crocodiles, say, must be at least as old as the oldest fossil on either branch. These anchors are the calibration points, and the choice of anchors dominates the result. Move a maximum bound from 590 to 800 million years ago, as the 2026 study did, and the whole deep tree slides older by a comparable amount.

The chemistry underlying the clock is the biochemistry of inheritance itself: the proteins and nucleic acids whose sequences are compared, built from the same molecular machinery in every living thing. The handedness of those molecules, the fact that life uses left-handed amino acids almost exclusively, is one of the deep regularities that makes sequence comparison meaningful across billions of years of divergence, a subject explored in the study of molecular handedness.

Calibration: The Anchors That Set the Clock

Calibration is the least glamorous and most consequential part of molecular dating, and it is where the 2026 dispute lives. A clock analysis takes dozens of fossil anchors: minimum ages from the oldest fossil on a branch, and maximum ages from deposits where the branch should have been found but was not. The maximum ages are the dangerous ones, because they convert absence into a number, and the Weng’an maximum was the most influential absence in the animal tree.

The pre-2026 consensus used Weng’an as a hard maximum for the animal crown group: no older than about 590 million years. Analysts had choices about how hard to make it, a “hard bound” that the model cannot cross versus a “soft bound” it can cross with penalty, and different teams made different choices, which is one reason published estimates for the animal origin sprawled across hundreds of millions of years even before 2026. The literature already contained nearly every date between 650 and 850 million years ago, defended by someone.

The 2026 study replaced the Weng’an maximum with older Tonian constraints, deposits in Norway, Australia and Arizona dated 850 to 730 million years old that bound certain deeper splits. The Tonian is the period from 1,000 to 720 million years ago, the deep Precambrian world before the great glaciations. Using these older anchors, the animal origin estimate moved to 800 to 700 million years ago. Note what did not change: the genetic data, the tree topology, the rate model. Only the anchors moved, and the date followed them. This is the sense in which molecular dates are fossil dates wearing a mathematical disguise.

The independent commentator Karma Nanglu, an evolutionary paleobiologist at the University of California, Riverside, framed the situation for the press as a “black box”: the game, in his account, is how much the true origin predates the first fossil discovery, and what happened inside the evolutionary interval between the actual origin and the first fossil window. That framing is the honest one. The clock does not see animals. It sees the shadow they cast on the genomes of their descendants, measured through a lens ground by fossil assumptions.

Why the Clocks and the Rocks Disagree

The disagreement between molecules and fossils is not a paradox to be solved by picking a winner. It is the expected outcome of two methods with different blind spots, and the interesting question is what each blind spot hides. Fossils are blind to soft, small, rare organisms in the wrong chemistry. Clocks are blind to rate changes they have not modeled and to calibrations that encode the very fossil assumptions they claim to escape. When the two disagree by 200 million years, both are telling the truth about what they measure, and neither measures the origin directly.

Consider what a molecular date of 750 million years ago actually claims. It claims that the genetic differences between, say, sponges and all other animals are consistent with a split at that date, given the modeled rates and the chosen calibrations. It does not claim that a 750-million-year-old sponge existed, or that one could be found, or that the first animal looked like a sponge. The date is a property of the model, and the model is a property of its assumptions. Change the maximum bound, as the 2026 study did, and the date moves. Change the rate prior, and it moves again. The published spread of estimates is not noise around a true value; it is a map of the assumption space.

Consider symmetrically what a fossil date of 574 million years ago claims. It claims that the oldest organism specialists agree is an animal comes from rocks of that age. It does not claim that no animal lived earlier, and every paleontologist working on the problem will say so when asked directly. The fossil date is a minimum: animals are at least this old. The clock date is an estimate of the actual split. A minimum and an estimate are different kinds of number, and the habit of subtracting one from the other to get “the gap” smuggles in the assumption that they measure the same thing. They do not.

There is a third consideration that both sides sometimes underplay: the stem lineage. The molecular clock dates the split between the animal lineage and its sister lineage, but the first members of the animal stem lineage need not have been animals in any recognizable sense. They were single-celled or colonial organisms diverging genetically while still looking, to any observer, like their protist cousins. The “origin of animals” as a genetic event and the “origin of animals” as a morphological event can be separated by tens of millions of years, and no method distinguishes them cleanly. Part of the famous gap may be nothing more than this definitional shadow.

Oxygen: The Great Enabler

Every account of animal origins eventually reaches oxygen, because animals are, without exception, oxygen-burning machines of a particular kind. Even the simplest sponges need dissolved oxygen to run their metabolism at multicellular scale, and the large, active animals of the Cambrian needed a lot of it. The history of oxygen in the oceans is therefore a hard constraint on the history of animals: they cannot predate the oxygenation that made their lifestyle possible, whatever the clocks say.

Earth’s oxygen story has two great chapters. The first is the Great Oxidation Event, about 2,400 million years ago, when photosynthetic cyanobacteria first flooded the atmosphere with oxygen, poisoning the anaerobic world and rusting the oceans. The second, murkier chapter is the Neoproterozoic Oxygenation Event, a further rise in oxygen sometime between 800 and 540 million years ago, documented in the chemistry of ancient sediments. It is this second rise that matters for animals, and its timing is contested in ways that mirror the fossil-molecule debate.

The evidence for Neoproterozoic oxygenation comes from proxies: the ratios of iron species in shales, the isotopes of molybdenum and uranium, the sulfur chemistry of pyrite. Each proxy is an indirect reading, and different proxies have pointed to different timings, from as early as 800 million years ago to as late as the Cambrian itself. The geochemist Timothy Lyons and his collaborators have argued for a protracted, stepwise oxygenation with the deep ocean remaining largely anoxic until near the Cambrian, while others read the same rocks as evidence for earlier, higher oxygen. The error bars are wide and the stakes are high, because the oxygen date sets the earliest possible date for animal life.

What is the mechanism linking oxygen to animals? Size and activity. Diffusion supplies oxygen to cells, and diffusion operates only over short distances, so a large or active animal needs either a thin body, a low metabolism, or an oxygen-rich environment. The first animals were probably small and thin, able to live on the modest oxygen of the Tonian oceans. The Cambrian explosion’s large predators needed more, and the oxygenation of the late Ediacaran and early Cambrian may have lifted the ceiling that had kept animals small. Oxygen did not cause animals, on this view; it permitted them to become conspicuous, which is a different and more defensible claim.

The cycling of oxygen, carbon and nutrients through living systems is the engine of the planet’s habitability, and the deep history of that cycling connects the animal story to the broader science of ecosystems and the environment. An animal is not just a consumer of oxygen but a participant in its cycle: burrowing animals irrigate sediments, filter feeders clear the water column, and the whole benthic ecosystem engineers its own chemistry. The oxygenation that permitted animals was then amplified by animals, a feedback loop that makes simple cause-and-effect stories inadequate.

Snowball Earth: Ice, Then Animals

Twice in the Cryogenian period, between 720 and 635 million years ago, ice may have covered the entire planet, from poles to equator. These are the Sturtian and Marinoan glaciations, the so-called Snowball Earth episodes, documented in glacial deposits found on every continent, including in rocks that formed at tropical latitudes. If the reconstructions are right, Earth spent tens of millions of years as a frozen ball, with life persisting in refugia: meltwater ponds, volcanic hot spots, thin equatorial seas under the ice.

The timing is electric for the animal origins debate. The molecular dates, 800 to 700 million years ago, place the animal origin just before or during the glaciations. The fossil dates place the diversification after them. This has suggested to many researchers that the Snowball episodes are not a coincidence but a cause: that the extreme environmental stress of global glaciation, followed by the hothouse aftermath of deglaciation, drove the evolution of complexity. Joseph Kirschvink, who named Snowball Earth in 1992, proposed that the freeze-thaw cycles fragmented populations and selected for the very traits, multicellularity, mobility, sensory systems, that define animals.

The mechanism deserves a skeptical look, because correlation across 700 million years is cheap and causation is expensive. What would ice actually select for? Cold tolerance, certainly. But the leap from cold tolerance to nervous systems is long, and no model derives the second from the first without several unexamined steps. The refugium argument cuts both ways: if life survived in isolated pockets, those pockets were small, and small populations evolve fast, but they also go extinct fast. Snowball Earth as a cradle of animals is a hypothesis with a mechanism still under construction.

What the glaciations establish firmly is the environmental context: the world into which animals emerged was one of violent climate swings, with continents clustered near the equator and sea level seesawing by hundreds of meters. Reconstructing those ancient continental positions, ocean currents and climate belts is the work of physical geography and climatology, and the deep-time maps it produces are the stage on which the animal drama played. Whatever triggered animal origins, it happened on a planet that had just thawed from the deepest freeze in its history, and that coincidence will keep drawing explanations until a better one arrives.

The Ecological Hypothesis: Predation Changes Everything

Douglas Erwin of the Smithsonian Institution and James Valentine of the University of California, Berkeley, offered in The Cambrian Explosion, published in 2013, the most influential ecological account of the event: that the trigger was not oxygen or ice but animals themselves. Once the first predators evolved, they created an evolutionary arms race. Prey needed armor, burrows, speed and senses; predators needed better weapons to defeat them. The feedback loop drove the rapid elaboration of body plans, hard parts, and behaviors that the fossil record captures as an explosion.

The hypothesis has a precise mechanism, which is its strength. Predation is one of the few selective pressures strong enough to drive rapid morphological change, and the Cambrian record shows exactly the expected signatures: the first drill holes in shells, the first sophisticated eyes, the first burrows deep enough to escape. The small shelly fossils of the Fortunian look like the opening moves of an arms race, mineralized defenses appearing across unrelated lineages at once. Ecology, on this view, is the engine, and the explosion is what an arms race looks like in stone.

The hypothesis also reframes the origin question. If the explosion was ecological, then the long fuse before it, the 200-million-year gap of the molecular dates, is the time when animals existed as small, soft, ecologically marginal organisms, diversifying genetically while remaining invisible. The explosion marks not their origin but their ecological takeover: the moment animals became important enough to reshape the planet. This dissolves the apparent conflict between the clocks and the rocks by assigning each its own event.

Critics note that the arms race needs a starting gun. Something had to produce the first predator, and invoking ecology to explain the explosion risks circularity: animals caused the explosion of animals. Erwin and Valentine answer that the first bilaterians, worm-like crawlers of the late Ediacaran, were enough: once directed movement and a through-gut existed, predation was inevitable, and the race ran itself. Whether that first step needed oxygen, or ice, or just time, remains the open question at the bottom of every account.

The Genetic Hypothesis: The Developmental Toolkit

A different school locates the trigger inside the genome. Animals share a toolkit of developmental genes, the Hox genes that pattern the body axis, the signaling pathways with names like Wnt and BMP, the gene regulatory networks that turn a fertilized egg into a structured body. These toolkits are ancient, shared across phyla, and their assembly may have been the rate-limiting step. On this view, the Cambrian explosion happened when the full toolkit was finally complete, and the diversification was the working-out of its combinatorial possibilities.

The evidence for the toolkit’s antiquity is striking. Sponges, the simplest living animals, possess many of the signaling pathways once thought to define complex bilaterians. The cnidarians, jellyfish and their kin, have Hox-like genes patterning their simple bodies. This means the genetic machinery predates the elaborate bodies they build, which inverts the intuitive order: the genes came first, the morphologies later. The developmental biologist Eric Davidson called this the kernels of gene regulatory networks, conserved cores that were assembled once and then elaborated for half a billion years.

The toolkit hypothesis makes a testable prediction about the gap: the deep genetic splits dated by the clocks correspond to the assembly of the toolkit in small, simple organisms, while the Cambrian morphologies correspond to its deployment. It also explains the odd shape of the explosion, so many body plans so fast, as the exploration of a newly complete design space. Once the regulatory logic existed, evolution could ring the changes rapidly, the way a new instrument lets a musician play new music.

The complication, as always, is the direction of causation. Toolkits do not assemble themselves in anticipation of future use; they are cobbled together by selection for immediate function. The genes that pattern a limb once did something humbler in a simpler body. This means the toolkit’s assembly was itself driven by ecology and environment, pushing the explanation back a level rather than resolving it. Genes, oxygen, ice and predation are not competing hypotheses so much as different chapters of one story, and the honest accounts keep all of them in play.

Sponges First? The Simplest Living Animals

If animals originated 750 million years ago, the first ones probably resembled sponges: simple, sessile filter feeders with no nerves, no muscles, no gut, just pores, channels and choanocyte cells beating their flagella to draw water through. Sponges are the least animal-like animals, and for that reason the best model of the beginning. The question of whether they were actually first is where the fossil, molecular and chemical evidence converge most tantalizingly.

The chemical evidence is the biomarker 24-isopropylcholestane, a molecule found in ancient rocks that is produced by certain living sponges and demosponges in particular. Geochemist Gordon Love and colleagues reported it in rocks of Oman and Siberia dating to around 650 to 540 million years ago, arguing that it records sponges, and therefore animals, tens of millions of years before the Cambrian. The claim is exactly the kind the science module’s evidence standard was written for: a surrogate endpoint. The molecule is not a sponge. It is a chemical made by living sponges, found in old rocks, attributed to sponges in the past.

The attribution has been contested on precisely those grounds. Other organisms might make the molecule, or it might form by geological processes from other precursors. Roger Summons of MIT, a pioneer of molecular fossils, has urged caution about over-interpreting single biomarkers, while defending the broader program of reading ancient chemistry for biological signals. The debate is unresolved in the way that matters: the biomarker is consistent with sponges at 650 million years ago, but consistency is not identification, and the field treats it as supporting evidence rather than proof.

The fossil evidence for early sponges is thinner than the textbooks imply. Claims of Precambrian sponges surface regularly and retreat regularly, because sponge spicules, the tiny skeletal needles, are easily confused with inorganic mineral growths, and soft sponges leave almost nothing. The oldest widely accepted sponge fossils are Cambrian. If sponges are the first animals and they appear in the record only in the Cambrian, then either the first 200 million years of sponge history left no accepted trace, which returns us to the preservation problem, or sponges are not first, which opens the next fight.

Comb Jellies vs Sponges: The Fight Over the First Branch

The most heated argument in animal phylogeny concerns which living group branched off first: the sponges, or the comb jellies, the ctenophores. It sounds like a taxonomic parlor game, but it decides the nature of the first animal. If sponges are first, the ancestor was simple and nerves and muscles evolved once, later. If comb jellies are first, the ancestor may have been more complex, and sponges simplified secondarily, which would mean simplicity is derived rather than primitive, and the whole story of animal origins reads differently.

The ctenophore-first case rests on phylogenomic analyses, trees built from hundreds of genes across many species, published in a series of papers from 2008 onward by teams including Casey Dunn and later workers. These analyses repeatedly placed ctenophores, not sponges, as the sister group to all other animals. The result was shocking because comb jellies have nerves, muscles and a through-gut with an anus, complex traits that would then have to have evolved before sponges lost them, or evolved twice independently.

The sponge-first counterattack came on two fronts. Morphologists and developmental biologists argued that the genomic result was an artifact of long-branch attraction, a known failure mode of tree-building where fast-evolving lineages cluster together spuriously, and ctenophores evolve fast. Then in 2024 and 2025, new analyses with better models and broader sampling swung the result back toward sponges first, though the fight continues and the honest summary is that the deepest branch of the animal tree remains contested.

Why does this matter for dating? Because the ctenophore-first hypothesis, if true, compresses the story: complex traits evolve early, and the long fuse looks shorter. The sponge-first hypothesis stretches it: simplicity first, complexity later, and the molecular dates have room to be right. The argument also illustrates the science module’s warning about single-study framing. Each new phylogenomic paper was reported as settling the question, and each was followed by another paper unsettling it. The body of evidence moves slowly; the headlines move fast.

What the Embryos Say

The Doushantuo Formation’s phosphorite does not only preserve the Weng’an microfossils. It preserves structures interpreted as animal embryos, balls of cells dividing in the pattern of early animal development, some showing what look like differentiated outer and inner layers. If they are embryos, they are direct evidence of animal reproduction 600 million years ago, and they would push the accepted fossil date back from 574 toward 600 million years, narrowing the gap from below.

The embryo interpretation has been through the full cycle of enthusiasm and doubt. Shuhai Xiao and colleagues described the structures in detail in the early 2000s, and the images are genuinely suggestive: spheres within spheres, cells in geometric division patterns. But the critics, led by researchers working on modern analogs, showed that encysting algae, giant bacteria like Thiomargarita, and sulfur-cycling microbes produce structures that mimic embryos stage for stage. The mimicry is so good that morphology alone may never settle it.

This is the taphonomic trap in miniature: exceptional preservation creates exceptional ambiguity. The better the fossil, the more alternative interpretations it can support, because fine detail multiplies the features that must be explained. The Weng’an embryo debate has settled, provisionally, into agnosticism. Some of the structures may be animal embryos. None can be shown to be. The deposit that was supposed to close the question became the clearest illustration of why the question is hard.

There is a broader lesson that the field has absorbed. The search for the first animal fossil may be misconceived if the first animals were microscopic and soft. What the Doushantuo shows is that even when preservation is perfect, identification can fail, because the organisms themselves lacked the diagnostic features that would identify them. A 700-million-year-old animal might be unrecognizable as an animal even if it were found, which means the fossil record’s silence is overdetermined: geology, chemistry, biology and epistemology all conspire to hide the beginning.

The Complication: The Case Against the Thesis

This account has argued that animals originated long before their first accepted fossils, and it is time to give the opposing case its full strength, because the thesis has real weaknesses and the honest version of this story names them. The complication has three strands, and each is held by serious researchers.

The first strand is the fragility of the clocks. The 2026 study moved the animal origin by 200 million years by changing the calibration anchors, and that very move demonstrates the method’s sensitivity rather than its accuracy. If a defensible recalibration can swing the answer by 200 million years, then the next defensible recalibration can swing it again, and the 800 to 700 million year window is not a measurement but a model output. Critics in the paleontological tradition, including researchers who work on the fossil calibrations themselves, argue that the relaxed-clock models have so many free parameters that they can accommodate almost any date, and that the dates they produce reflect the priors as much as the data. On this view, the gap between clocks and rocks is not evidence that the rocks are incomplete; it is evidence that the clocks are underdetermined.

The second strand is the possibility that the Cambrian explosion was genuinely rapid evolution, not a preservational artifact. The artifact reading assumes a long cryptic history of soft-bodied bilaterians, but that history should have left trace fossils, burrows and trails, and the late Ediacaran trace record, while increasingly complex, does not show the diversity the artifact reading requires. Where are the diverse burrows of the cryptic fauna? Their relative scarcity is a real datum, and it favors a shorter fuse. The developmental toolkit, moreover, may have assembled faster than the gradualist picture assumes, if the key regulatory innovations were few and their effects were large.

The third strand is the most philosophical: perhaps the question is malformed. “When did animals originate” assumes a moment, but evolution produces populations, not moments, and the lineage that became animals diverged from its sister lineage over millions of years of gradually accumulating difference. Asking for the date is like asking for the exact moment a dialect becomes a language. The fossil date and the clock date may both be correct answers to different precisifications of a vague question, and the gap between them may be an artifact of demanding precision the subject cannot supply.

The thesis survives these objections in narrowed form. The clocks are fragile, but they are fragile around an old date, not a young one: no defensible calibration returns an origin younger than the fossils. The trace fossil record is thinner than the artifact reading would like, but the Ediacaran traces that do exist show the right trend, increasing complexity toward the boundary. And the question may be vague, but science routinely precisifies vague questions, and the precisified versions still separate by a hundred million years or more. The honest position is not that the thesis is proven but that the burden of evidence lies with anyone who claims the fossil date is close to the true origin. That burden has not been met.

How the Story of the Story Changed

The interpretation of early animal life has been revised so many times that the revisions are themselves part of the subject, and a brief historiography belongs in any complete account. Each generation inherited a confident picture and then found the deposit, or the method, that broke it.

The first picture belonged to the nineteenth century: the Cambrian was the beginning of life, full stop. Darwin’s contemporaries treated the “Primordial” strata as life’s dawn, and the absence below as absence. The second picture arrived with Walcott’s Burgess Shale and the slow realization, through the twentieth century, that the Precambrian was not empty but poorly searched. The discovery of the Ediacara biota in the 1940s and 1950s, and its global recognition through the 1960s and 1970s, established complex Precambrian life as a fact and moved the question from whether to when.

The third picture was Gould’s: the Cambrian as a carnival of experimentation, contingency ascendant. The fourth was the molecular revolution of the 1990s and 2000s, when the first clock studies pushed animal origins deep into the Precambrian and paleontologists dismissed them as methodological artifacts. The fifth is the current synthesis, in which the clocks are taken seriously as minima-plus-models, the fossil record is read as a biased witness, and the geochemistry sets the boundary conditions. The 2026 recalibration belongs to this fifth picture as a refinement, not a revolution.

What changed at each step was not just the data but the questions. The nineteenth century asked what the rocks contain. The twentieth asked what the rocks can preserve. The twenty-first asks what the models assume. This progression, from evidence to bias to assumption, is the normal maturation of a historical science, and it is why the current debate is more sophisticated than its predecessors even where it remains unresolved. The field has learned to distrust its own confidence, which is the beginning of getting things right.

The pattern of confident stories giving way to better ones under the pressure of new evidence is a recurring feature of the historical sciences. The field has learned to distrust its own confidence, which is the beginning of getting things right.

The Measurement Problem: What Deep Time Hides

Every number in this account carries an uncertainty, and the uncertainties compound in ways that casual summaries hide. The zircon date has an analytical error of a few million years. The correlation of the ash layer to the fossil bed has a stratigraphic uncertainty. The fossil’s identification has a taxonomic uncertainty. The clock’s rate has a model uncertainty. By the time a headline reports “animals originated 750 million years ago,” four layers of maybe have been compressed into one number, and the compression is where misunderstanding lives.

The hardest thing to observe is the thing the whole debate needs most: a small, soft-bodied animal in a Tonian rock. Consider the requirements. The rock must be sedimentary and unmetamorphosed, which eliminates most Tonian strata. It must have been deposited in a setting where soft tissues could preserve, which eliminates most of the rest. It must have been found, recognized and correctly interpreted, which requires a paleontologist to look at the right slab with the right question. Multiply the probabilities and the expected number of such fossils in museum drawers is near zero, which is consistent with what is actually there. The absence is predicted by the search parameters, not just by history.

There is also the problem of what counts as evidence for a negative. To establish that animals did not exist at 700 million years ago, one would need a representative sample of the world’s 700-million-year-old environments showing no animals, with preservation good enough to have caught them. No such sample exists or could exist, because the Tonian rock record is fragmentary and its environments are undersampled. This asymmetry is fundamental: presence can be established by one fossil, absence requires an exhaustive survey that geology cannot provide. The entire “when” question is therefore biased toward older answers by the logic of evidence itself, a point the clock advocates make and the fossil advocates concede in principle while disputing in practice.

The practical consequence is that the origin date will always be given as a range with a one-sided error bar: at least 574 million years from the fossils, probably older by an unknown amount. Anyone who quotes a single year for the origin, from either camp, is overselling. The ranges are the honest unit of deep-time knowledge, and the width of the range is itself information: it measures how much the methods cannot see.

Common Misreadings, Corrected

Why Do Paleontologists Argue About the Same Fossils?

Because a fossil is an observation plus an interpretation, and the interpretation depends on the theoretical context the observer brings. The Weng’an spheres look like embryos to an embryologist and like algae to a microbiologist. Arguments persist because the evidence underdetermines the conclusion, and each round sharpens what would count as decisive.

Five misreadings recur in public discussion of this subject, and each deserves a direct correction.

First, the misreading that the Cambrian explosion created animals from nothing. It did not. It diversified body plans that were already genetically distinct, in oceans that already held complex Ediacaran ecosystems. The explosion is a diversification event, not a creation event, and the distinction matters because it reframes the search: not for a moment of invention but for a long fuse.

Second, the misreading that molecular clocks and fossils are rival claimants to one truth. They are complementary instruments measuring different things, the genetic split and the preservable body, and the interesting science is in relating the two, not in declaring a winner. A reader who understands this will never again be confused by a headline that pits them against each other.

Third, the misreading that “no fossils” means “no life.” The history of this field is a sequence of such inferences being overturned: no Precambrian life, then stromatolites; no Precambrian animals, then Ediacara; no Tonian animals, then the clock dates. Each overturning followed the same pattern, a new window or a new method, and the rational response is to discount absence heavily in old rocks.

Fourth, the misreading that a new study settles the question. Studies move the debate; they do not end it. The 2026 recalibration is important for its methodological argument, not for its headline number, and the number will move again as models improve. Science advances by the accumulation of such moves, not by single decisive experiments, and deep-time science least of all.

Fifth, the misreading that uncertainty means anything goes. It does not. The constraints are real: animals cannot predate the oxygenation that their metabolism requires, they cannot postdate their oldest fossils, and the clock estimates, for all their model dependence, cluster in a range rather than scattering at random. Uncertainty with boundaries is knowledge, and the boundaries here are informative.

The Evidence at a Glance

The table below compresses the argument into the form a reader can scan: each major line of evidence, what it actually measures, the date it supports, and what it cannot show. No single row settles the question. Read together, they show why the field converges on an origin older than the fossils but younger than the wildest clock estimates. A companion reference tool is available for readers who want to trace any of these lines further.

Line of evidence	What it measures	Date it supports	What it cannot show
Oldest accepted animal fossils	First preservable animal bodies	574 million years ago, a minimum	Animals that were small, soft or rare
Ediacara biota	Large complex Precambrian organisms	575 to 539 million years ago	Which were true animals
Cambrian explosion	Diversification of body plans	539 to 521 million years ago	The genetic splits behind the morphologies
Molecular clocks	Genetic distance converted by rate models	800 to 700 million years ago, model dependent	Rate constancy or correct calibration
Weng’an Biota	Microscopic preservation quality	Absence of animals at 590 million years ago	That absence means non-existence
Kheseen Biota	A second exceptional window	Preservation does not guarantee capture	The true first appearance
Oxygen proxies	Ocean and atmosphere chemistry	Rising oxygen 800 to 540 million years ago	The exact threshold animals needed
Biomarkers	Molecules attributed to sponges	Sponge-like chemistry near 650 million years ago	That the molecules came from sponges

The table is worth revisiting after the sections that follow, because its last column is the point. Every method has a characteristic blindness, and the origin of animals sits precisely in the region where the blindnesses overlap. This is not a counsel of despair. It is the reason the field uses multiple methods, and the reason their convergence, where it happens, carries weight.

How to Read a Claim About Deep Time

The reader who has come this far has earned a toolkit, because the next headline about the origin of animals, or the origin of anything ancient, will arrive soon, and the same five questions will sort the substantial from the noise. They are the working scientist’s questions, translated for the non-specialist.

First, ask what kind of evidence is being reported: a new fossil, a new analysis of old data, a new method, or a new model applied to existing data. The 2026 study was the fourth kind, a recalibration, which is why its importance lies in the argument rather than the number. New fossils change the minimum. New models change the estimate. The two are different news.

Second, ask what would falsify the claim. A claim that cannot be falsified is not a finding. “Animals originated 750 million years ago” is falsified, in principle, by a well-preserved Tonian deposit showing diverse non-animal life and no animals, though in practice such a deposit may never exist. “The Cambrian explosion was purely preservational” is falsified by evidence of genuinely rapid genetic diversification. If the article reporting the claim does not say what would disprove it, the claim is being sold, not stated.

Third, ask about the error bars and whether they are shown. A date without an uncertainty is a rhetorical device, not a measurement. Serious work reports ranges, and the width of the range tells the reader how much is known. Narrow ranges on deep-time dates should raise suspicion, not confidence, because the methods do not support them.

Fourth, ask who disagrees and what their best argument is. A field with no disagreement is either settled or stagnant, and animal origins is neither. The credible reports quote the skeptics, because the skeptics are often right about the weaknesses even when they are wrong about the conclusion. The complication section of this account exists on exactly this principle.

Fifth, ask whether the claim is about the world or about the method. Many deep-time controversies are method controversies wearing world costumes. The Weng’an ceiling was a methodological choice that looked like a fact about the world. Learning to see the costume is the single most valuable skill a reader of science can develop, and it transfers to every field that reconstructs an unobservable past.

Why the Date Matters: Calibrating the Tree of Life

The origin of animals is not just the root of the animal tree; it is the anchor for the molecular clocks that date everything else. Every estimate of when mammals diversified, when birds split from dinosaurs, when the human and chimpanzee lineages parted, depends on calibrations deeper in the tree, and the deepest calibration is the animal origin. Move the root by 100 million years and the branches shift, not by the same amount but by amounts that propagate through the models in complex ways. Getting the root right is therefore infrastructure for the whole of evolutionary dating.

The human story illustrates the dependence. The split between humans and chimpanzees is dated by combining fossil hominids with molecular clocks calibrated, ultimately, against deep nodes that trace back to the animal root. The fossil record of human evolution, with its own celebrated deposits and debates, is the shallow end of the same methodology: the same clocks, the same calibration logic, the same arguments about absence and presence. Work on human evolution and fossil hominids shows the method operating where the record is rich, which throws the deep-time difficulties into sharp relief.

There is also the astrobiological stake. Animals are the only known example of complex multicellular life, and their origin is the sole data point for how long a planet needs to produce such life. If animals took 3.7 billion years from Earth’s formation, or 3.9, the difference matters for estimates of life’s cosmic frequency, because it changes the apparent difficulty of the transition. The origin date feeds, through the Drake-adjacent reasoning of astrobiology, into humanity’s oldest question about whether it is alone. A single data point cannot bear much weight, but it is the only one available, so it bears what it can.

Finally, there is the intrinsic stake, which needs no utilitarian defense. Human beings are animals, and the question of when animals began is the question of when the lineage that led to observers emerged from the microbial world. Every reader of this account is a late product of the events it describes. That is not a scientific argument, but it is why the question will still be asked in five years, and in fifty.

What Is Settled and What Is Open

Settled, to the standard that historical science can settle things: animals are at least 574 million years old, because the fossils say so and the fossils are real. The Cambrian explosion was a genuine ecological and morphological revolution between 539 and 521 million years ago, whatever its deep causes. Oxygen rose in the Neoproterozoic and animals needed it. The molecular clocks, across many studies and methods, keep returning pre-Ediacaran dates, and no one has produced a principled reason to dismiss them all. The burden of evidence lies with the young-origin claim, and it has not been met.

Open, and likely to remain open for years: the exact date, which will probably never be known to better than a 100-million-year window. The trigger, or more likely the combination of triggers, that moved animals from marginal to dominant. The identity of the first animal lineage, sponges or comb jellies, which decides whether simplicity or complexity came first. The interpretation of the Weng’an structures, which may remain ambiguous forever. The oxygen threshold, which geochemistry has not yet pinned down.

The thesis of this account stands in its narrowed form. Animals almost certainly originated before their first accepted fossils, because every line of evidence except the most literal reading of the fossil record points older, and the literal reading has been wrong at every previous stage of this investigation. The fossil record measures preservation. The genes measure divergence. The chemistry measures possibility. None measures the origin directly, and wisdom about the beginning of animal life consists in holding all three measurements at once without collapsing them into one.

There is a coda that the evidence table’s last column already suggested. The search for the first animal may never find a single specimen that ends the debate, because the first animals may have been unrecognizable and the rocks that held them may be gone. What the field will find instead, and has been finding, is a progressively better understanding of the conditions, the mechanisms and the uncertainties. That is how historical sciences end their arguments: not with a fossil on a podium but with a convergence of independent lines, each insufficient alone, persuasive together. The animals emerged in darkness, in every sense, and science is learning to see in the dark. A companion reference tool is available for readers who wish to follow the primary literature from here.

Before Animals: The Choanoflagellate Connection

The closest living relatives of animals are not sponges but choanoflagellates, single-celled organisms that swim with a whip-like flagellum surrounded by a collar of microvilli, filtering bacteria from the water. A sponge’s feeding cell, the choanocyte, is essentially a choanoflagellate embedded in a multicellular body. The resemblance is not a coincidence. It is the visible seam where the animal lineage meets its unicellular ancestry, and it tells us what the first animals were built from.

Choanoflagellates do something remarkable for single cells: some species form colonies, rosettes of cells that coordinate their flagella and share a common extracellular matrix. The biologist Nicole King and her laboratory showed that these colonies use some of the same adhesion and signaling proteins that animals use to hold their bodies together, including cadherins. The implication is that the molecular toolkit for multicellularity predates animals themselves. The parts were lying around in the unicellular world, and animals assembled them.

This reframes the origin question once more. If the genetic components of animal multicellularity existed in choanoflagellate-like ancestors, then the “origin of animals” was not the invention of new genes so much as the regulatory rewiring of old ones: the same proteins, deployed in new patterns, to build bodies instead of colonies. The fossil record cannot capture a rewiring event, because rewiring leaves no skeleton. The transition from colony to organism may have happened in creatures the size of a grain of sand, invisible to every method except the comparative genomics that reads the toolkit’s history in living DNA.

The choanoflagellate connection also sets a floor under the molecular dates. The split between choanoflagellates and animals must predate the animal origin, and clock estimates for that deeper split run older still, into the deep Tonian. Whatever date the animal crown gets, the stem lineage stretches further back, into a world of colonial protists experimenting with the multicellularity that animals would inherit. The first animal was not a breakthrough from nothing. It was a colony that learned to be a body.

The Trace Fossil Record: Behavior Before Bodies

Body fossils are not the only fossils. Trace fossils, the burrows, trails, tracks and borings that organisms leave in sediment, record behavior rather than anatomy, and they have their own story to tell about the approach to the Cambrian. A trail does not need a skeleton. It needs a mover, and movers leave signatures in mud that persist where bodies do not.

The Ediacaran trace record shows a clear trend of increasing behavioral complexity. The earliest traces, from around 565 million years ago, are simple horizontal burrows, probably made by small bilaterians grazing on microbial mats. Kimberella, the slug-like Ediacaran organism, left feeding traces with paired scratch marks, evidence of a muscular foot and directed movement. By the late Ediacaran, traces show meandering, branching and deeper penetration, the signatures of animals exploring the sediment in three dimensions. The trend is unmistakable: behavior was complexifying before the Cambrian began.

The base of the Cambrian period is defined not by a body fossil but by a trace: Treptichnus pedum, a distinctive branching burrow system that appears worldwide at the boundary. The choice is deliberate. Body fossils are facies-dependent, tied to particular environments, while the behavioral innovation that Treptichnus records, systematic sediment probing, spread across environments. The golden spike that marks the boundary, at Fortune Head in Newfoundland, is anchored to this trace, a rare case where behavior, not anatomy, defines geological time.

Trace fossils constrain the artifact hypothesis from the other direction. If a diverse cryptic fauna of soft-bodied bilaterians existed for 100 million years before the Cambrian, it should have left a diverse trace record, and the Ediacaran traces, while increasingly complex, are not as diverse as the hypothesis would predict. This is the strongest empirical point the short-fuse advocates have, and the honest response is that it is genuinely puzzling. Either the cryptic fauna was less diverse than its genetic shadow suggests, or its traces are not recognized for what they are. Both remain possible, and the trace record is where the next decisive evidence is most likely to come from.

The Oldest Claims: Otavia and the Francevillian Enigma

Every few years a claim surfaces of an animal far older than the accepted record, and two deserve discussion because they show how the field evaluates extraordinary claims. The first is Otavia antiqua, described in 2012 by Anthony Brain and colleagues from the Otavi Group of Namibia, and presented as a 760-million-year-old sponge. The structures are millimeter-scale, with pores and internal chambers that the authors interpreted as sponge anatomy, preserved in carbonate rocks far older than any accepted animal.

The claim was received the way such claims always are: with intense scrutiny. Critics argued that the structures could be inorganic mineral growths, bacterial colonies, or artifacts of the rock’s diagenesis, and that the “pores” lacked the organized aquiferous system of a true sponge. The defenders pointed to the size range, the wall structure and the ecological context. More than a decade later the claim remains in limbo: not refuted, not accepted, cited by clock advocates as consistent with old dates and by fossil conservatives as an example of over-interpretation. Its status is the honest status of most oldest-claims: unresolved.

The second is stranger. The Francevillian biota of Gabon, described in 2010 by Abderrazak El Albani and colleagues in the journal Nature, consists of centimeter-scale structured fossils in 2.1-billion-year-old rocks, more than a billion years before the Ediacaran. The authors argued for coordinated multicellular life, possibly even the earliest experiment in large body size. If the structures are biological, and if they represent anything on the road to animals, they would rewrite the timeline completely.

The field’s response has been cautious to the point of cold. The structures are accepted as probably biological, but their affinity is unknown, and most specialists do not connect them to animals. A single enigmatic assemblage, however well documented, cannot carry the weight of a billion-year revision without independent corroboration, and none has arrived. The Francevillian remains what the science module calls a data point awaiting a body of evidence, and the responsible account reports it as exactly that: fascinating, unplaced, and not part of the animal story as the published evidence draws it.

Both cases illustrate the evaluation standard this account has applied throughout. An extraordinary claim needs extraordinary preservation, independent replication, and a mechanism that makes sense. Otavia has the first arguably, the second no, the third maybe. The Francevillian has the first, the second no, the third unknown. Until those ledgers change, the accepted floor stays at 574 million years, and the argued ceiling stays in the Tonian, and the gap between them remains the most productive argument in paleontology.

Frequently Asked Questions

Q: What is the oldest animal fossil ever found?

The oldest fossils widely accepted as animals come from rocks about 574 million years old, in the Ediacaran period. These include small, simple organisms preserved in exceptional deposits, along with the enigmatic Ediacara biota that appears in the same interval. The date is a scientific consensus rather than a single specimen: different researchers defend different candidates, but none commands agreement at an older date. It is important to understand what this number represents. It is the oldest fossil that specialists agree is an animal, not the oldest animal that lived. Every older claim, from 600-million-year-old embryo-like structures to 760-million-year-old sponge-like fossils, remains contested. The 574-million-year figure is therefore a floor. Animals are at least this old. How much older is the central question of the field, and the answer depends on whether one trusts the genetic clocks, which point to 800 to 700 million years ago, or insists on fossil confirmation.

Q: How do molecular clocks estimate evolutionary ages?

A molecular clock converts genetic differences into time. When two lineages split from a common ancestor, each accumulates mutations independently, so the DNA of their living descendants differs by an amount that grows with the time since the split. If the mutation rate is known, dividing the genetic distance by the rate gives the age of the split. In practice the rate is never simply known. It varies between genes, between lineages and over time, so modern analyses use relaxed clock models that estimate rates and dates together in a statistical framework. The critical step is calibration: anchoring certain splits to fossil dates so that genetic distance can be translated into years. The choice of these anchors dominates the result. A clock is therefore not an independent witness in the naive sense. It is a model whose output reflects its fossil inputs, its rate assumptions and its mathematics, and responsible users report all three.

Q: What was the Cambrian explosion?

The Cambrian explosion was the geologically rapid diversification of animal body plans between about 539 and 521 million years ago. Nearly every major animal phylum appears in the fossil record within this interval: arthropods, molluscs, chordates, echinoderms and the rest. The classic windows onto the event are the Chengjiang biota of China, about 518 million years old, and the Burgess Shale of Canada, about 508 million years old, which preserve soft-bodied animals in extraordinary detail. The name is misleading in one respect. It suggests life appearing from nothing, whereas the rocks show a diversification of lineages that were already genetically distinct, in oceans that already held complex Ediacaran ecosystems. Specialists debate whether the explosion records genuinely rapid evolution, driven by ecology and developmental innovation, or largely records the moment animals evolved hard parts and became preservable. Most specialists favor a combination: real ecological revolution, amplified by a preservational bias.

Q: Why do fossils and DNA give different dates for the origin of animals?

Fossils and DNA measure different things. Fossils record organisms that died in conditions favoring preservation, which means they record the appearance of preservable bodies, not the genetic origin of lineages. DNA-based molecular clocks estimate when lineages split genetically, which can long predate the evolution of a fossilizable body. A lineage can diverge genetically while its members remain small, soft and invisible to paleontology. There is also a methodological entanglement: the clocks are calibrated with fossil dates, so they are not fully independent of the record they seem to contradict. When the clocks return 800 to 700 million years ago and the fossils say 574 million years, the gap represents the interval in which animals may have existed as small, soft organisms, plus the uncertainty in both methods. The two dates are a minimum and an estimate, not two measurements of the same quantity, and much confusion comes from treating them as rivals rather than complements.

Q: What is the Weng’an Biota and why does it matter?

The Weng’an Biota is a 590-million-year-old fossil deposit in Guizhou Province, China, that preserves microscopic organisms cell by cell in phosphate, with three-dimensional detail unmatched at its age. Algae, bacteria and enigmatic multicellular spheres appear as if flash-frozen. For years some of the cell clusters were interpreted as animal embryos, which would have made them the oldest animal fossils known, but that interpretation has retreated under scrutiny, since algae and giant bacteria can mimic embryonic division patterns. The deposit matters for a different reason. Because its preservation is so exceptional, some researchers argued that the absence of definitive animals at Weng’an proves animals had not yet evolved, and used it as a maximum age constraint in molecular clock studies. A 2026 study challenged this reasoning by pointing to the Kheseen Biota of Mongolia, an equally well-preserved but younger deposit that also lacks animals even though animals existed at the time. The Weng’an absence, the study argued, reflects preservation bias rather than history.

Q: What did the 2026 Oxford study conclude about animal origins?

The study, published in October 2026 in the journal Science Advances by a team led from the University of Oxford, concluded that animals may have originated up to 200 million years earlier than the fossil record’s traditional ceiling allowed. Its central argument was methodological rather than evidential: it did not report a new animal fossil. It argued that the long-standing practice of using the 590-million-year-old Weng’an Biota as a maximum age for animal origins was unsound, because the absence of animals in that deposit does not demonstrate their non-existence. The authors demonstrated this with the Kheseen Biota of Mongolia, a younger deposit of comparable preservation quality that likewise contains no animal fossils, even though animals are documented elsewhere from the same interval. Removing the Weng’an ceiling and recalibrating the molecular clock with older Tonian geological constraints, dated 850 to 730 million years old, moved the estimated origin to between 800 and 700 million years ago. The paper is best understood as a single peer-reviewed data point that shifted the debate’s assumptions, not as a final verdict.

Q: How do scientists date rocks that are hundreds of millions of years old?

The workhorse method is radiometric dating, which uses the steady decay of radioactive isotopes as a clock. For deep time, uranium-lead dating of zircon crystals is the gold standard. Zircons crystallize in volcanic eruptions, trapping uranium and excluding lead at formation; the uranium then decays to lead at a precisely known rate, and measuring the ratio gives the age of the eruption. Because sedimentary rocks cannot be dated directly, geologists date volcanic ash layers above and below the fossil-bearing beds, bracketing the fossils between two ages. Zircon dating can achieve precision better than one percent at 600 million years. A second method, biostratigraphy, orders rocks by their fossil content: once a distinctive fossil is dated somewhere by radiometric means, its appearance elsewhere marks the same interval. Every deep-time date carries uncertainty, typically a few million years in the best cases, and responsible reporting includes it. Dates quoted without error bars should be treated with skepticism.

Q: What is the Ediacara biota?

The Ediacara biota is an assemblage of large, soft-bodied organisms that lived on the seafloor between about 575 and 539 million years ago, named for the Ediacara Hills of South Australia where they were first recognized in 1946. They include ribbed ovals like Dickinsonia, segmented forms like Spriggina, and fractal fronds called rangeomorphs that stood upright in deep water. They are the oldest large, complex fossils known. Their nature is debated. Some, like Kimberella, which left feeding traces suggesting a muscular foot, are plausibly early molluscs and therefore true animals. Others, particularly the rangeomorphs, may represent an extinct grade of organization with no living descendants. The German paleontologist Adolf Seilacher proposed they belonged to a separate kingdom, the Vendobionta. Whatever their affinities, the Ediacarans prove that large, mobile, complex multicellular life existed tens of millions of years before the Cambrian, which means the biological machinery of complexity was already operating when the explosion began.

Q: Were sponges the first animals?

Sponges are the leading candidate for the earliest-branching animal lineage, but the question remains contested. The traditional view places sponges first because they are the simplest living animals, lacking nerves, muscles and a gut, which suggests the first animal was similarly simple and complexity evolved later. This view was challenged by phylogenomic studies from 2008 onward that placed comb jellies, which possess nerves and muscles, as the sister group to all other animals instead. If comb jellies branched first, simplicity may be a secondary loss rather than the primitive condition. More recent analyses with improved models have swung support back toward sponges first, though the deepest branch of the animal tree is still debated. The fossil evidence is thin either way: the oldest widely accepted sponge fossils are Cambrian, and Precambrian claims remain disputed. Chemical biomarkers attributed to sponges in rocks around 650 million years old offer supporting but not conclusive evidence. The honest answer is that sponges are probably first, with the qualifier doing real work.

Q: What is the difference between the origin of animals and the Cambrian explosion?

The origin of animals is the evolutionary split of the animal lineage from its closest non-animal relatives, an event dated by molecular clocks to roughly 800 to 700 million years ago. The Cambrian explosion is the diversification of animal body plans in the fossil record, dated to 539 to 521 million years ago. They are separated by at least 150 million years and they are different kinds of events. The origin is a genetic divergence, invisible in stone. The explosion is an ecological and morphological radiation, spectacularly visible. Confusing the two is the most common error in popular writing on the subject. The explosion did not create animals; it diversified them, gave many of them hard parts, and made them ecologically dominant. Understanding the distinction dissolves much of the apparent paradox: there is no contradiction between an old genetic origin and a young fossil diversification, because they were never the same event. The interesting question is what happened in the interval between them.

Q: How much oxygen did the earliest animals need?

All living animals require dissolved oxygen to power their metabolism at multicellular scale, and even the simplest sponges need a minimum concentration to survive. The exact threshold for the first animals is unknown, but it was certainly far below modern atmospheric levels. Early animals were probably small and thin-bodied, which allowed oxygen to reach their cells by diffusion, so they could have lived in the modestly oxygenated oceans of the Tonian period. The large, active predators of the Cambrian needed substantially more, and the oxygenation of the late Ediacaran and early Cambrian oceans may have lifted the ceiling that had kept animals small. Geochemical proxies suggest a stepwise rise in oxygen between 800 and 540 million years ago, though the timing is debated. The key point is that oxygen set a constraint, not a trigger: animals could not predate the oxygenation their bodies required, but oxygen alone did not cause them. The relationship is one of permission rather than causation.

Q: What was Snowball Earth?

Snowball Earth refers to two episodes in the Cryogenian period, the Sturtian and Marinoan glaciations between 720 and 635 million years ago, during which ice may have covered the entire planet from poles to equator. Evidence comes from glacial deposits found on every continent, including rocks that formed at tropical latitudes. If the reconstructions are correct, Earth spent tens of millions of years largely frozen, with life surviving in refugia such as meltwater ponds and volcanic hot spots. The timing is provocative: the glaciations fall squarely between the molecular dates for animal origins and the fossil diversification. Some researchers, following Joseph Kirschvink, who named the hypothesis in 1992, propose that the extreme environmental stress and subsequent thaw drove the evolution of complexity. Others regard the coincidence as just that. What is certain is that animals emerged in the aftermath of the deepest freeze in Earth’s history, on a planet of violent climate swings and shifting continents.

Q: What are trace fossils and what do they reveal about early animals?

Trace fossils are the preserved signs of organismal activity: burrows, trails, tracks, borings and feeding marks, as opposed to body fossils, which preserve the organism itself. They record behavior rather than anatomy, and they can capture animals too soft to fossilize as bodies. The Ediacaran trace record shows increasing behavioral complexity toward the Cambrian boundary: simple horizontal burrows around 565 million years ago, the paired feeding scratches of Kimberella, and eventually deep, branching burrow systems. The base of the Cambrian period is formally defined by the first appearance of the trace fossil Treptichnus pedum, a branching burrow marking a behavioral innovation in sediment probing. Trace fossils also constrain hypotheses about a long hidden history of animals: if diverse soft-bodied bilaterians existed for 100 million years before the Cambrian, they should have left diverse traces, and the relative scarcity of such traces is a genuine puzzle. The trace record is where the next decisive evidence about the fuse before the explosion is most likely to be found.

Q: Who discovered the Burgess Shale?

The Burgess Shale was discovered in 1909 by Charles Doolittle Walcott, then the most powerful geologist in America and secretary of the Smithsonian Institution. Workmen on a mountain trail in British Columbia told him of unusual fossiliferous shales, and Walcott’s quarries over subsequent years yielded tens of thousands of specimens of soft-bodied Cambrian animals, about 508 million years old. The deposit preserved what ordinary fossilization destroys: the spiny Hallucigenia, the five-eyed Opabinia, the predator Anomalocaris, alongside early arthropods, worms and chordates. Walcott interpreted the strange animals conservatively, fitting them into known groups, and his authority held for half a century. In the 1970s and 1980s, Harry Whittington, Derek Briggs and Simon Conway Morris restudied the fauna at Cambridge and showed how many of its members fit no known phylum. Stephen Jay Gould’s Wonderful Life, published in 1989, made their findings the most famous argument in modern paleontology.

Q: What did Darwin say about the fossil record before the Cambrian?

In On the Origin of Species, published in 1859, Charles Darwin identified the abrupt appearance of complex animals in the Cambrian as a serious objection to his theory. If species evolve gradually, the rocks below the Cambrian should contain transitional forms, and in his day they contained almost nothing. Darwin attributed the gap to the imperfection of the geological record, arguing that the Precambrian had simply not been adequately searched. On that narrow point he has been thoroughly vindicated: the Precambrian record turned out to hold stromatolites, microfossils, the Ediacara biota and the Doushantuo embryos, each discovery pushing recognized complexity further back. Where Darwin fared less well was in his expectation that transitional forms would eventually be found in stone. The molecular evidence suggests the deepest animal divergences happened in small, soft-bodied organisms that left no preservable transitions. Darwin’s mechanism was right, but his expectation of what the rocks would show was wrong, and the pattern of his objection being answered by new windows of preservation has repeated ever since.

Q: What are Hox genes and why do they matter for animal evolution?

Hox genes are a family of regulatory genes that pattern the body axis of animals, determining which end grows a head and which grows a tail, and where limbs and segments form. They are shared across the animal kingdom: the same basic toolkit operates in a fruit fly and a human, which is why their discovery in the 1980s revolutionized evolutionary biology. They matter for animal origins because the toolkit is ancient. Sponges and cnidarians, the simplest living animals, possess many of the signaling pathways and regulatory genes once thought to define complex bilaterians. This means the genetic machinery for building bodies predates the elaborate bodies they build: the genes came first, the morphologies later. One influential hypothesis holds that the Cambrian explosion occurred when the full developmental toolkit was finally assembled, and the rapid diversification was the exploration of its combinatorial possibilities. The toolkit’s antiquity supports the molecular dates, since assembling it in small, simple organisms would leave exactly the kind of invisible history the clocks detect.

Q: How do chemical biomarkers provide evidence of ancient sponges?

Biomarkers are organic molecules preserved in ancient rocks that can be attributed to particular organisms. The most discussed animal biomarker is 24-isopropylcholestane, a sterane found in rocks of Oman and Siberia dating to around 650 to 540 million years ago. It is produced by certain living demosponges, and its presence in Precambrian rocks has been argued, notably by Gordon Love and colleagues, to record sponges tens of millions of years before the Cambrian. The method is powerful because chemistry can survive where anatomy cannot, but the attribution is the weak point. The molecule is a surrogate: it is not a sponge but a chemical made by living sponges, found in old rocks and attributed to sponges in the past. Other organisms might produce it, or it might form geologically from other precursors. Specialists treat the biomarker as supporting evidence rather than proof, consistent with sponges at 650 million years ago but not demonstrative. It is one row in the evidence table, not the table itself.

Q: Why are there so few animal fossils from before the Cambrian?

Three factors conspire, and all three are about preservation rather than absence. First, the first animals were probably small and entirely soft-bodied, lacking shells, bones or any hard parts, and soft tissues fossilize only under exceptional chemical conditions. Second, the rocks themselves are the problem: Precambrian strata have endured half a billion years of heat, pressure and erosion, and most have been metamorphosed beyond the point where delicate fossils could survive. Third, the search has been biased: for a century paleontologists looked for shells and skeletons because those were the fossils they knew, and the microscopic, chemical and trace evidence that could reveal early animals required techniques developed from the 1980s onward. Each technical advance has pushed the apparent origin further back, which suggests the silence is partly methodological. The fossil record is a biased witness with strong preferences for the hard, the large and the young, and discounting for that bias is the first step in reading it honestly.

Q: What evidence would settle the debate over when animals originated?

No single discovery would end the debate, but several would narrow it dramatically. The most decisive would be an unmistakable animal body fossil from Tonian rocks, 800 to 700 million years old, ideally from more than one deposit: that would confirm the molecular dates directly. Almost as valuable would be a well-preserved Tonian deposit with excellent microscopic preservation, comparable to Weng’an, that showed diverse non-animal life and no animals, which would genuinely challenge the old molecular dates rather than merely failing to confirm them. On the methodological side, independent convergence of different clock models on the same date range, with different calibration schemes, would strengthen confidence in the genetic estimates. Improved geochemical proxies that pinned down Neoproterozoic oxygen levels would tighten the boundary conditions. What will not settle it is another model applied to the same data, or another press release about a single ambiguous structure. Historical sciences converge through independent lines of evidence, not through decisive experiments, and this debate will end the way such debates always end: gradually, and then all at once.

Q: How should a non-scientist evaluate the next headline about the origin of life?

Five questions sort the substantial from the noise. First, what kind of evidence is reported: a new fossil, a new analysis of old data, a new method, or a new model? New fossils change the minimum age; new models change the estimate, and the two are different news. Second, what would falsify the claim? If nothing could, it is being sold rather than stated. Third, are the uncertainties shown? A deep-time date without an error bar is a rhetorical device, and narrow ranges on ancient dates should raise suspicion. Fourth, who disagrees and what is their best argument? Credible reports quote the skeptics, because the skeptics are often right about the weaknesses. Fifth, is the claim about the world or about the method? Many deep-time controversies are method controversies in disguise, like the debate over whether the Weng’an deposit sets a maximum age. Learning to see the method behind the headline is the most transferable skill in reading science, and it applies far beyond paleontology.