Unknown, Possible, Probable, Established: How Evidence Changes a Claim

A surprising number of arguments collapse because two sentences are treated as if they mean the same thing.

"This could have happened."

"This probably happened."

They do not mean the same thing.

Possibility opens a door.

Evidence tells us whether we should walk through it.

Why binary thinking fails

Historical questions are often presented as true or false.

Built by local workers or lost technology.

Natural formation or artificial structure.

Known culture or unknown civilization.

Ordinary aircraft or extraordinary object.

Real evidence usually arrives in degrees.

A claim can be possible without being likely.

It can be plausible without being well supported.

It can be strongly supported while some details remain uncertain.

A better model therefore uses levels of confidence.

Level 1: Unknown

Unknown means the evidence does not currently support an adequate answer.

It is not an explanation.

It is a status.

Examples:

• the exact meaning of a symbol is unknown;
• the precise route of a transported stone is unknown;
• the identity of an undocumented person is unknown.

Unknown is intellectually honest.

It should not be experienced as a vacuum that must immediately be filled.

Level 2: Possible

A claim is possible when it has not been ruled out by important known constraints.

That is a low threshold.

A lost wooden structure may be possible because wood decays.

A coastal settlement may be possible in a now-submerged landscape.

A trade contact may be possible if the distance could be crossed.

Possibility answers one question:

Could this happen?

It does not answer:

Did this happen?

Possibility is cheap

Many mutually incompatible explanations can be possible at the same time.

An ancient object of unknown function could be:

• a tool;
• a ritual object;
• a decorative item;
• a component of something larger;
• reused material;
• misunderstood because its context is incomplete.

The fact that one option can be imagined does not distinguish it from the others.

To move upward, the claim needs evidence that fits it better than alternatives.

Level 3: Plausible

Plausible means the claim fits important background knowledge and does not require too many unsupported additions.

A reconstruction becomes more plausible when:

• the materials were available;
• the technique existed;
• the chronology fits;
• the social organization was capable of it;
• similar evidence appears elsewhere;
• the explanation matches the site context.

Plausibility is stronger than possibility.

It is still not proof.

Level 4: Supported or probable

Support appears when observations favor one explanation over serious alternatives.

Imagine two hypotheses predict different material patterns.

Researchers look for those patterns.

The evidence repeatedly matches one prediction and not the other.

Confidence should move.

This is the logic behind hypothesis testing in many sciences.

Archaeology can also use formal probability and Bayesian methods to express how evidence changes confidence.

The exact number is not always available.

The logic still matters.

Evidence should update the claim.

Level 5: Strongly supported or established

Established does not mean philosophically impossible to revise.

It means the core claim is supported strongly enough that ordinary work can treat it as reliable.

Examples of established historical conclusions can still contain open details.

A monument can be securely dated to a broad period while its exact ritual function remains debated.

A material can be securely sourced to a region while the exact transport route remains unknown.

A settlement can be clearly human-made while the identity of every builder remains unknown.

Confidence can differ inside the same story.

One claim, several confidence levels

This is why responsible explanations sound more complicated than dramatic ones.

Consider a hypothetical monument.

Established:

The stone came from a quarry 30 kilometers away.

Supported:

It was moved using organized human labor.

Plausible:

Part of the route used sledges on a prepared surface.

Possible:

A river was used during one stage.

Unknown:

The exact day-by-day transport sequence.

Nothing is gained by forcing all five statements into one confidence level.

What counts as evidence?

Evidence is not simply information related to a claim.

Good evidence changes the balance between competing explanations.

If two theories predict exactly the same observation, that observation may not help much in choosing between them.

An impressive stone is evidence that the stone exists.

It is not automatically evidence for a specific lifting technology.

A myth about a flood is evidence that a flood story existed in that tradition.

It is not automatically evidence for one global geological event.

The question is always:

What does this observation discriminate between?

Independent lines are stronger

Suppose a migration is proposed.

Pottery similarity supports it weakly.

Ancient DNA supports population movement.

Isotopes show that some individuals grew up elsewhere.

New settlement patterns appear in the same period.

Chronology aligns.

Each line has limitations.

Together, they create a stronger explanation because different methods converge.

The anomaly problem

An anomaly is something that does not fit expectations.

Anomalies matter.

They are often where discovery begins.

But an anomaly does not identify its own cause.

A date that conflicts with the accepted chronology could mean:

• the chronology is wrong;
• the sample moved;
• contamination occurred;
• the material was reused;
• the laboratory result is unusual;
• the context was misunderstood;
• something genuinely unexpected happened.

The anomaly justifies investigation.

It does not automatically justify the largest explanation.

"Nobody has disproved it"

This sentence sounds stronger than it is.

A claim does not receive positive evidence merely because a decisive disproof is difficult.

There may be thousands of possible claims that cannot be individually disproved.

The burden is to show why one deserves confidence.

That means making predictions and finding evidence.

Absence of evidence

The phrase "absence of evidence is not evidence of absence" is important.

It prevents overconfidence when preservation or search is poor.

But it is not a universal shield.

Suppose a claim predicts a massive industry operating for centuries.

That industry should produce mines, waste, tools, transport, settlements, environmental residues, and repeated artifacts.

If well-studied contexts repeatedly lack those signatures, the absence becomes relevant.

Now consider a short-lived wooden camp on a coastline that was later submerged.

The expected preservation is much worse.

Its absence carries less weight.

Missing evidence only becomes informative relative to what should have survived and where people have actually looked.

The moving-goalpost test

Watch what happens when evidence fails to appear.

A testable claim becomes weaker.

A protected claim becomes more flexible.

The civilization left no tools because the technology was non-material.

It left no settlements because the people were nomadic.

It left no biological signature because the group was tiny.

It influenced the whole world despite being tiny and leaving no material signature.

Each new explanation may be possible by itself.

Together they can make the claim impossible to test.

The more a claim changes only to avoid evidence, the less explanatory work it is doing.

What would change your mind?

This question works in every direction.

Ask it of the alternative claim.

Ask it of the mainstream explanation.

Ask it of yourself.

A healthy answer names observations that could move confidence.

A protected belief says no possible observation would matter.

Confidence should move

Changing your mind is often described as inconsistency.

In evidence-based reasoning, it is a feature.

The goal is not to choose the correct position once and defend it forever.

The goal is to maintain a confidence level that responds to the best available evidence.

New evidence can move a claim upward.

It can move it downward.

It can split one claim into several more precise claims.

The DarkBrain evidence ladder

Use this whenever a mystery begins to feel certain.

Unknown

We do not currently have an adequate answer.

Possible

The claim is not ruled out.

Plausible

The claim fits important known constraints.

Supported / Probable

Relevant evidence favors it over important alternatives.

Strongly Supported / Established

Independent evidence converges and the core claim remains robust under serious testing.

Then ask one final question:

What specific evidence moved the claim from one level to the next?

If there is no answer, the confidence may have moved without the evidence.