Evolution of Life

Deep time is usually delivered as a list of numbers with no indication of which ones are solid. Here the strength of a claim and the strength of its evidence can be held apart and compared, transition by transition: the end-Permian boundary is known to a few tens of thousands of years while the origin of life is not dated at all, only bracketed, and the difference is not a difference in importance but in what happened to survive. The record also keeps moving. The base of the Cambrian has carried three different numerical ages while the outcrop that defines it has not moved a centimetre, and a claim about the oldest life on Earth was overturned by a team walking the same exposure sideways. The habit that separates those cases — asking what a number was measured on before asking what it means — outlasts every date in the sequence.

• Radiometric datingprocessThe Journey's actual protagonist. Every prominence decision below follows from putting the instrument, rather than the organisms, at the centre of a subject that is normally told as a procession of life.
• Global Boundary Stratotype Section and PointconceptCarries the distinction the whole Journey rests on — that a boundary is a physical point and its age is a separate, revisable measurement of that point. Without it, a redated boundary looks like the science changing its mind about what happened.
• Konservat-Lagerstätte (exceptional preservation)conceptThe single correction that changes how a reader holds every other date here. It earns CORE despite being an idea about mud rather than about life.
• Mass extinctionconceptThe Journey takes a category the reader arrives holding and shows it to be a convention drawn from a biased marine record. That demolition needs the concept present as an object rather than implied by five events.
• Zircon as a geochronometermaterial classThe lattice accident that makes four significant figures possible at all. It carries the first chapter and then quietly underwrites every number after it.
• Molecular clock datingprocessThe dissenting instrument. It appears in three separate chapters disagreeing with the rocks, and the Journey needs it present to keep the fossil record from reading as the only witness.
• Mass-independent fractionation of sulfur isotopesgeological and soil phenomenonThe cleanest illustration of the lens: an event established not by finding something but by a signature ceasing and never returning.
• Meishan GSSP, Changxing, ZhejiangsiteWhere a stratigraphic interval was converted into a duration, which is the move that makes the largest extinction in the record arguable as a rate problem.

Showing 8 of 17 featured records. Atlas does not choose which of the rest matter.

• The method that most often contradicts the fossil record has no absolute scale of its own and must be calibrated against dated fossil horizons. The dissent is therefore not independent: it borrows its units from the record it is used to correct, and inherits that record's uncertainties on the way in.

  Molecular clock dating · Radiometric dating

• Two of the five extinctions coincide with flood basalt provinces, and in both the argued killing agent is not the lava but what the magma passed through on its way up — coal, evaporites and carbonates already sitting in the crust. Put beside each other, the two say something neither says alone: the lethal ingredient was in the sedimentary basin, not in the volcano.

  Siberian Traps large igneous province · Central Atlantic Magmatic Province

• Follow the chain the graph already holds and it closes on itself. Microbial mats produced the oxygen; the oxygen made the eukaryotic merger possible; the eukaryotic cell is the precondition for complex bodies; and the most economical explanation for the sharp decline of those same mats is that animals had learned to eat them. The most abundant fossil evidence of the first three billion years thins out because of what it made possible.

  Stromatolites · Endosymbiosis and the eukaryotic cell · The evolution of multicellularity

• Deep time is legible because of two unrelated rarities. One is a mineral that accepts uranium and refuses lead, so that a crystal keeps an honest account of its own age. The other is a conjunction of rapid burial, no oxygen and no scavengers, so that a soft body leaves a picture. Neither was arranged for anyone's benefit and neither is evenly distributed — which means the parts of the record that look empty and the parts that look precise are both, in large measure, maps of where these two accidents happened to occur.

  Zircon as a geochronometer · Konservat-Lagerstätte (exceptional preservation)

• Are the Nuvvuagittuq filaments the remains of organisms, and how old is the rock they sit in?

  Two disputes run at once and neither depends on the other. The filaments are either mineralised bacteria or abiotic growths in a hydrothermal system that is known to produce filaments; the host rocks are either about 3.77 or about 4.28 billion years old depending on which reading of the isotopic systematics is accepted. Either dispute alone would leave the claim unresolved. The corpus holds both open.

• Are the Francevillian structures organisms at all?

  Unresolved, and the stakes cut both ways. Accepting them moves macroscopic life back by more than a billion years and requires either an enormous ghost lineage or an origin that failed and left nothing. Rejecting them requires an abiotic mechanism that makes regular lobed centimetre-scale pyrite bodies with radial internal fabric, which is not obviously the easier ask.

• Why do molecular estimates and the fossil record still disagree about when the major animal lineages diverged?

  The two answers differ by tens of millions of years for several groups. Relaxed-clock models that let substitution rates vary along the tree narrowed the gap without closing it, and after three decades nobody has shown convincingly whether the remainder is a real cryptic history of small unpreserved animals or a systematic artefact of rate estimation. The limit here is an absence: no study settles it, so none is cited.

• What actually did the killing at the end of the Permian?

  Warming, anoxia and acidification all have supporting evidence in the marine record and all follow from the same degassing, but their relative weight is unresolved — and the terrestrial extinction may not be synchronous with the marine one. The date is known to a few tens of thousands of years; the mechanism is not known at all in that sense.

• How much of the end-Cretaceous biotic stress belongs to the Deccan Traps rather than the impact?

  The Deccan province was erupting across the same interval. The mainstream synthesis treats the impact as the proximate cause while accepting Deccan volcanism as a background stressor; a substantial minority weights the two differently, and the dating is not yet sharp enough to settle it. This Journey does not resolve it.

• How much oxygen did the Great Oxidation Event actually produce?

  Genuinely unclear. Published estimates for the post-event atmosphere range from a fraction of one per cent to a few per cent of present levels, and there is good evidence the rise was not monotonic — transient pulses before it and a large carbon-isotope excursion after. The singular noun "event" flatters the data, and the Journey keeps the range rather than choosing a figure.

The confident numbers here are confident about crystals. The end-Permian boundary is known to a few tens of thousands of years because volcanic ash happened to fall into a shallow sea at Meishan while the extinction was under way; the origin of life is not known to within eight hundred million years because nothing comparable survives from the Hadean. Neither figure is a statement about how important the event was, and reading them side by side as though they were is the commonest error this material invites. Smaller and more portable than any date here is the discipline the dates enforce: a first appearance in a record is a fact about the record, and whether it is also a fact about the world is a separate argument that has to be made every time. One place has that discipline built into it. The base of the Phanerozoic is not a fauna and not a number but a burrow in ordinary rock, chosen because a global boundary marker has to be findable in ordinary rock — on a headland in Newfoundland called Fortune Head.

This is unfinished in the world, not only in the telling.

• Where they are →Where on a body does this subject sit?
• How these connect →What is related to what, and on what evidence?
• When they happened →What happened, in what order, and what changed? · central to this Journey
• What has no place →What in this subject has no location at all?
• Ask the world another question →What else is in the corpus about this? · Dynamic, and nothing it answers becomes canonical

• Coral Reefsecological relationshipReef systems are built, destroyed and rebuilt by different organisms across this record — microbial mats first, then the stromatoporoid and coral systems that collapsed in the Devonian and again at the end of the Triassic. A reader who has spent eight chapters inferring collapse from ash beds and shale should see the one case where the same class of partnership can be observed directly instead of reconstructed.
• Life at the Polesshared systemThe only one of the Big Five driven by cooling rather than warming worked through ice at a pole: glaciation on Gondwana dropped sea level and drained the shallow seas where most marine diversity lived. Going from here to a living polar biosphere is the difference between seeing ice as a mechanism of loss and seeing it as a habitat with its own dependencies.
• Amazon Rainforestcausal consequenceThe most-discussed reading of the Late Devonian crisis has the spread of the first rooted forests driving weathering, nutrient delivery and ocean anoxia — a biosphere altering the chemistry of the planet it sits on. A reader holding that should go next to a forest whose effect on its own atmosphere is measurable now rather than inferred from black shale.
• Life of a Starshared system · crosses into SpaceThe oldest bound in this Journey is not biological. Life cannot be dated earlier than a surface cool enough to hold liquid water, so the first constraint on the history of life is a planetary and stellar one, and a reader who wants to know where that clock starts has to leave the fossil record entirely.

• Think this through in Studio →When a quantity can only be bracketed rather than measured, what would have to be found for the bracket to close — and is that thing the kind that survives?

Ask the world another question. Dynamic keeps this Journey as its context and never changes what is written above.