Apply the Survival Test

Most arguments about lost civilizations stall because the two sides are answering different questions — one about what is possible, the other about what is demonstrated. A shared checklist makes the disagreement productive, and it works just as well on a claim you brought with you as on the ones catalogued here.

• Falsifiability requires a claim specific enough that some possible observation would contradict it.
• Different materials and depositional environments have radically different preservation probabilities.
• Dating context is not the same as dating construction.
• Residual and intrusive material routinely distort single dates at multi-phase sites.
• Conventional geological and biological processes can produce regular forms, unusual chemistry and apparent alignments.
• The evidentiary weight of absence depends on the proposed population size, duration, geographic extent, material durability and sampling intensity.

• How much sampling is sufficient before absence becomes strong evidence in a given region — this varies by setting and survey history.
• Where the line falls between a productive open question and an unfalsifiable claim.
• Whether some proposals classed as unfalsifiable could be reformulated into testable form.

Bounded unknowns. Nothing here is treated as evidence that the possibility is true.

• The framework is reusable: Atlantis and Lemuria, ancient aviation and sky vehicles, acoustic levitation, floating cities, unknown hominins, worldwide flood traditions, lost metallurgy, sophisticated stoneworking, ancient global navigation, Titans and unicorn traditions.
• Those applications are not written yet. They will be added when there is real material to work with, and existing Arcanum entries are linked rather than duplicated.
• Anyone invoking preservation bias inherits a constraint on the scale of what they propose: the shield and the claim cannot both be maximised.
• Establishing that something could have happened and left no trace is a statement about detection limits, not about the past.

The traces this subject should have left, listed before we ask what we actually find.

• Step two of the test: derive the expected traces before assessing them — physical, geological, biological, chemical, linguistic, cultural, astronomical.
• Durable end of the spectrum: mine voids, ceramics, geochemical anomalies, isotope signatures, domesticated genetics.
• Fragile end: textiles, timber, documents, knowledge, provenance records.

Preservation bias: the processes that erase, degrade or relocate the evidence.

• Erosion, burial, flooding, glaciation, tectonics and sea-level change.
• Corrosion, decomposition and diagenesis.
• Human reuse, previous excavation, museum removal, war and later settlement.
• The interpretive frame itself — evidence can survive and still be unrecognised.

Concrete findings that would strengthen, weaken or change this assessment.

• For any specific claim: name it in advance. A claim that cannot say what would falsify it has left the evidence conversation.
• Objects, sites, publications or datasets a reader can independently check.
• Results that a conventional explanation cannot accommodate, not merely results consistent with the claim.

This is the method the rest of the category applies. It is deliberately reusable: the same ten questions work on Atlantis, on ancient aviation, on acoustic levitation, on floating cities, on unknown hominins, on worldwide flood traditions, on lost metallurgy, on sophisticated stoneworking, on ancient global navigation, on Titans and unicorn traditions. The framework is being published now; individual applications will be added only when there is real material to work with, and existing Arcanum entries are linked rather than duplicated. The ten questions. One: what exactly is the claim, stated precisely enough to be wrong? "Something strange happened in the deep past" cannot be tested. "A metal-working society of at least a hundred thousand people occupied this basin for at least five centuries before 10,000 BCE" can be. Two: if it happened, what traces should it create? Physical, geological, biological, chemical, linguistic, cultural, astronomical. This is the step that almost every popular claim skips, and skipping it is what makes a claim unfalsifiable. Three: which of those traces would survive one thousand, ten thousand, a hundred thousand, a million years? Survival curves differ enormously by material and by environment. Ceramics and mine voids are near the durable end; textiles, wood and documents are near the fragile end; geochemical anomalies outlast almost everything made. Four: which traces would be destroyed or displaced? Erosion, burial, flooding, glaciation, tectonics, sea-level change, corrosion, decomposition, human reuse, previous excavation, museum removal, war, later settlement. A claim about a formerly coastal site on a subsided continental shelf is in a completely different evidentiary position from one about an eroding upland. Five: could surviving evidence be misdated because what was dated is context rather than construction? Charcoal, organics, occupation debris and sediment date events near the structure, not the moment stone was worked. Six: could the site have been excavated, reused or reburied, producing younger intrusive layers or older residual material? Multi-phase sites routinely carry several honest dates at once. Seven: if the evidence survived, would a later civilization recognise it? Function is not self-evident. An unfamiliar institution, an unfamiliar material or an unfamiliar practice can be present in the ground and invisible in the interpretation. Eight: what natural or conventional process could produce the same evidence? Geology makes remarkably regular shapes, unusual chemistry and apparent alignments. A conventional explanation that fits is not a boring explanation; it is the one that must be defeated on evidence. Nine: what evidence actually exists today, as opposed to what is asserted to exist? Track each item to a specific object, site, publication or dataset that a reader can check. Missing artefacts, unnamed institutions and uncited photographs are not evidence. Ten: what discovery would materially strengthen, weaken or falsify the claim? If nothing could, the claim has left the evidential domain, and everyone involved should say so plainly. Two rules govern how the answers are used. First, absence of evidence is not proof of absence — but its weight is not constant. Absence is weak evidence against a small, short-lived, localised, non-durable proposal in a poorly sampled setting. It is strong evidence against a large, long-lived, widespread, materially durable proposal in a well-sampled setting. Anyone invoking preservation bias must therefore accept the corresponding constraint on the scale of what they are proposing; the shield and the claim cannot both be maximised. Second, possibility never becomes probability without evidence. Establishing that something could have happened and left no trace is a statement about detection limits. It changes what we can rule out. It does not raise the likelihood that it occurred, and no accumulation of unfalsifiable possibilities ever adds up to one supported claim. Applied honestly, the framework is not a debunking tool. It frequently strengthens open questions by identifying searchable traces that nobody has looked for, and by separating the defensible core of a claim from the indefensible packaging around it. That is the useful output: not a verdict, but a list of things that could actually be found.

Arcanum framework, synthesised from standard archaeological, geochronological and taphonomic reasoning for reuse across this category

Sources are labelled by type and access. Arcanum preserves claimant material where it documents the claim, while using academic, institutional and primary records to evaluate it.

• Historic England — Radiocarbon Dating and Chronological Modelling(opens in a new tab)

  Open guidance document — no login. The sampling and modelling standards that questions five and six depend on.

  Reference work

• Kidwell & Holland — The Quality of the Fossil Record (Annual Review of Ecology and Systematics, 2002)(opens in a new tab)

  Peer-reviewed foundation for questions three and four: what the record preserves and what it filters out.

  Peer-reviewed

• Schmidt & Frank — The Silurian hypothesis (arXiv preprint 1804.03748)(opens in a new tab)

  Open full text — no login. A worked example of the framework applied rigorously by working scientists.

  Peer-reviewed

• Council for British Archaeology — Research frameworks and evidence standards(opens in a new tab)

  Open professional resource — no login. Context for what counts as a documented find versus an assertion.

  Reference work

Each question searches Arcanum’s own index — nothing is sent to an external search provider. Use “Copy question” to save the full wording anywhere you like.

• What exactly is the claim or hypothesis, stated precisely enough to be wrong? — search the Arcanum index for this questionCopy question
• If it happened, what physical, geological, biological, chemical, linguistic, cultural or astronomical traces should it create? — search the Arcanum index for this questionCopy question
• Which of those traces would plausibly survive 1,000 / 10,000 / 100,000 / 1,000,000+ years? — search the Arcanum index for this questionCopy question
• Which traces would erosion, burial, flooding, glaciation, tectonics, sea-level change, corrosion, decay, human reuse, excavation, museum removal or war destroy or displace? — search the Arcanum index for this questionCopy question
• Could surviving evidence be misdated because the dated material is context — charcoal, organics, occupation debris, sediment — rather than the moment stone was worked? — search the Arcanum index for this questionCopy question
• Could the site have been excavated, reused or reburied, creating younger intrusive layers or older residual material? — search the Arcanum index for this questionCopy question
• If evidence survived, would a later civilization recognise what it was? — search the Arcanum index for this questionCopy question
• What natural or conventional process could produce the same evidence? — search the Arcanum index for this questionCopy question
• What evidence actually exists today, as opposed to evidence that is asserted to exist? — search the Arcanum index for this questionCopy question
• What specific discovery would materially strengthen, weaken or falsify the hypothesis? — search the Arcanum index for this questionCopy question

• How Archaeology Dates Stone Sites
• Life After People: The Preservation Experiment
• The Silurian Hypothesis
• The Museum Paradox: Archaeology After Archaeology
• Giants and the Fossil Record

Notes are stored locally on this device only in this phase.