Why Humans Believe Strange Things

Humans can build particle accelerators.

We can also become convinced that:

• lucky numbers control events,
• hidden groups explain every coincidence,
• anecdote defeats statistics.

Why?

Because the same mind that discovers patterns can also over-detect them.

“Strange” Is Contextual

A belief may seem strange because it is:

• rare,
• culturally unfamiliar.

That alone does not make it false.

The important question is:

How was the belief formed and maintained?

The Pattern-Seeking Brain

Humans are exceptional pattern detectors.

This ability supports:

• language,
• science,
• prediction.

But pattern detection has false positives.

We see structure in noise.

False Positive vs False Negative

Imagine hearing a rustle in ancient grass.

Two possible errors:

False positive

You think predator, but there is none.

False negative

You think nothing, but there is a predator.

The costs are asymmetric.

Error Management

If false negatives are extremely costly, evolution may favor systems biased toward false positives.

Better to occasionally imagine a predator than ignore a real one.

This is one possible route to over-detection.

Hyperactive Agency Detection

Humans readily infer agents behind events.

A branch moves.

Was it:

wind

or:

someone?

Agency detection is socially and biologically useful.

It can also generate over-attribution.

Intentionality Everywhere

We say:

“My computer hates me.”

“The universe is sending a message.”

This reflects our tendency to interpret events through agency and intention.

The metaphor can become literal belief.

Patternicity

Michael Shermer popularized patternicity:

the tendency to find meaningful patterns in meaningless noise.

The underlying cognitive tendency is real even if terminology varies.

Apophenia

Apophenia refers to perceiving meaningful connections between unrelated things.

It can occur mildly in ordinary cognition.

Extreme forms may be clinically relevant.

Pareidolia

Pareidolia is a perceptual example.

We see:

• faces in clouds,
• figures on walls.

The visual system is tuned to detect significant forms.

Superstition

Superstitions often arise when an action happens before a desired outcome.

Example:

wear lucky socks → team wins.

The mind infers causation from coincidence.

Skinner’s Pigeons

B. F. Skinner described pigeons developing repeated behaviors when food delivery occurred independently of their actions.

The interpretation became a famous analogy for superstition.

Modern analysis of the experiment is more nuanced.

Intermittent Reinforcement

When rewards are unpredictable, behaviors can become highly persistent.

Gambling exploits this.

Random success can strengthen false causal beliefs.

Post Hoc Reasoning

The fallacy:

after this, therefore because of this

mistakes temporal sequence for causation.

Humans naturally search for causes.

Coincidence is psychologically unsatisfying.

Anecdotes

A personal story is vivid.

Statistics are abstract.

People often overweight anecdotes.

This is partly due to cognitive availability.

Availability Heuristic

Events easily recalled feel more common.

After seeing news about plane crashes, flying may feel unusually dangerous.

Salience distorts probability judgment.

Representativeness

We judge probability by similarity to a stereotype.

A detailed story may feel more plausible even when mathematically less likely.

This produces errors such as the conjunction fallacy.

Conjunction Fallacy

If:

A and B

is more specific than:

A,

then:

[ P(A \cap B) \leq P(A) ]

Yet people sometimes judge a vivid conjunction as more likely.

Narrative coherence can overpower probability.

Confirmation Bias

Once a belief forms, we notice confirming evidence more readily.

We may:

• search selectively,
• remember selectively,
• interpret ambiguity favorably.

This creates self-reinforcement.

Disconfirmation Is Hard

Contradictory evidence may be:

• ignored,
• reinterpreted.

Beliefs can become resistant precisely because every challenge is absorbed.

Belief Immunization

A theory may say:

“Anyone denying this is part of the conspiracy.”

Now counterevidence becomes evidence.

The belief is insulated from falsification.

Conspiracy Thinking

Real conspiracies exist.

The problem is not believing that people sometimes coordinate secretly.

It is using an explanatory style where:

every anomaly

is evidence of hidden intentional control.

Conspiracy Features

Common features include:

• hyper-agency,
• distrust,
• anomaly hunting,
• self-sealing explanations.

These can make a theory difficult to disconfirm.

Proportionality Bias

People often expect big events to have big causes.

A major assassination feels too important to result from one ordinary individual.

This can motivate elaborate explanations.

Randomness Is Emotionally Difficult

Large consequences can arise from:

• small causes,
• accidents.

Humans often resist this.

Meaningful causes feel more satisfying.

Control

Belief in hidden patterns can restore a sense of control.

If events are governed by:

secret rules,

perhaps they can be predicted.

Randomness feels threatening.

Uncertainty Reduction

Strange beliefs often offer:

clear explanations

where reality provides ambiguity.

Certainty is psychologically attractive.

Existential Threat

During:

• disease,
• war,
• economic crisis,

uncertainty rises.

Conspiratorial and magical explanations may become more appealing.

Terror Management

Awareness of mortality can shape beliefs and group attachment.

Worldviews provide psychological security.

This is one route linking existential anxiety to belief defense.

Identity-Protective Cognition

People process evidence in ways that protect:

• group identity,
• social belonging.

A factual correction may feel like a social threat.

Tribal Belief

Some beliefs function as membership signals.

The question becomes not only:

Is this true?

but:

What does believing this say about who I am?

Costly Beliefs

A belief can persist even when it creates practical costs if it provides:

• belonging,
• moral identity.

Social benefits can outweigh evidential weaknesses.

Sacred Values

Some values are treated as non-negotiable.

Evidence may be interpreted through them rather than allowed to revise them.

This is not unique to any one culture or ideology.

Social Proof

If many people around us believe something, it becomes more plausible.

This is usually adaptive.

Other people’s beliefs are often useful information.

But crowds can also be wrong.

Information Cascades

An information cascade occurs when people follow earlier decisions instead of relying on private evidence.

A false belief can spread rapidly.

Example

Person A chooses restaurant X.

Person B assumes A knows something and follows.

Person C sees two people choosing X.

Soon a crowd forms.

The initial signal may have been arbitrary.

Repetition Effect

Repeated claims feel more familiar.

Familiarity can increase perceived truth.

This is called the illusory truth effect.

Familiarity Is Not Evidence

A statement repeated ten times does not gain ten units of factual support.

But psychologically it may feel safer.

This is why propaganda uses repetition.

Memory Source Confusion

People may remember:

a claim

but forget:

where they heard it.

Weak sources can become detached from their warning labels.

Misinformation Effect

Later information can alter memory of earlier events.

Memory is reconstructive.

Belief can be built on distorted recollection.

Motivated Reasoning

We scrutinize unwanted conclusions more aggressively.

Desired conclusions receive easier acceptance.

This creates asymmetric skepticism.

“Can I Believe It?” vs “Must I Believe It?”

For agreeable claims, we ask:

Can I believe this?

For disagreeable claims:

Must I believe this?

The evidential threshold shifts.

Intelligence Does Not Immunize

High intelligence does not guarantee unbiased belief.

It can increase ability to generate sophisticated defenses for preferred conclusions.

Reasoning can become a lawyer rather than a judge.

Education Helps, but Not Automatically

Scientific literacy improves many judgments.

But domain knowledge can also be used selectively.

Epistemic virtue requires habits, not intelligence alone.

Overconfidence

People often know less than they think.

The Dunning–Kruger effect is frequently oversimplified, but miscalibration of self-assessment is real.

We should avoid turning one effect into a universal insult.

Explanatory Depth Illusion

People often think they understand mechanisms better than they do.

Ask someone to explain:

how a zipper works

in detail.

Confidence often falls.

This is the illusion of explanatory depth.

Asking “How?” Helps

Requesting a mechanism can expose shallow belief.

This is useful for evaluating:

• political claims,
• technical claims.

Mechanistic detail constrains stories.

Story Bias

Humans prefer coherent narratives.

A good story has:

• cause,
• motive,
• resolution.

Reality is often messier.

Narrative fluency can masquerade as evidence.

Agent-Centered Stories

Stories about intentional actors are especially memorable.

“Someone planned this” is cognitively easier than:

complex interacting systems produced it.

This can overproduce agency explanations.

Essentialism

Humans often assume categories have hidden essences.

We think:

members are similar because of an underlying nature.

This can support stereotypes and pseudoscientific classification.

Teleological Reasoning

People often explain natural things by purpose.

“Mountains exist so animals have homes.”

Children commonly use such reasoning.

Adults can retain versions of it.

Naturalness Bias

“Natural” often feels:

good, safe.

But nature contains:

• toxins,
• pathogens.

Naturalness is not an evidence-based safety criterion.

Appeal to Antiquity

An old belief is not automatically true.

Longevity can indicate cultural importance.

It does not establish factual accuracy.

Appeal to Novelty

Likewise, newness does not imply truth.

“Ancient wisdom” and “new breakthrough” can both be rhetorical devices.

Pseudoscience

Pseudoscientific systems often imitate scientific language while avoiding scientific risk.

Common features include:

• vague predictions,
• selective success stories,
• ad hoc excuses.

Science Is Not Bias-Free

Scientists are human.

They also face:

• confirmation bias,
• incentives.

Science works better because institutions create corrective mechanisms.

Institutional Error Correction

Science uses:

• peer criticism,
• replication,
• open methods,
• statistics.

These do not eliminate bias.

They make bias more detectable.

Belief Communities

Online communities can create epistemic bubbles.

Algorithms may repeatedly expose users to similar claims.

Repetition and social agreement reinforce confidence.

Echo Chambers

An echo chamber does more than repeat beliefs.

It can systematically discredit outside sources.

This makes correction harder.

Epistemic Trust

Humans must trust others.

The challenge is not:

trust nobody.

It is:

calibrate trust.

Radical distrust can make people more vulnerable to fringe sources.

Charismatic Authority

A confident speaker can appear knowledgeable.

Confidence cues often influence trust.

But confidence and accuracy are imperfectly correlated.

Precision Theater

Specific numbers can create a false sense of rigor.

“83.7%” sounds scientific.

Without methodology, precision is cosmetic.

Extraordinary Experience

People sometimes report:

• visions,
• coincidences,
• unusual perceptions.

The experience itself may be sincere.

The explanation remains open.

Experience vs Interpretation

A person can accurately report:

“I saw a light.”

It does not follow:

“It was an alien spacecraft.”

Separate observation from inference.

Sleep Paralysis

Sleep paralysis can involve:

• inability to move,
• sensed presence,
• vivid imagery.

Historically, cultures interpreted it through supernatural frameworks.

A physiological mechanism can explain much of the pattern.

Culture Shapes Interpretation

Similar experiences may be labeled:

• spirit encounter,
• alien abduction,
• neurological event

depending on cultural context.

Experience and explanation are distinct.

Belief Is Socially Scaffolded

No one builds a worldview alone.

We inherit:

• language,
• categories,
• traditions.

This is necessary for cognition.

It also transmits errors.

Why Strange Beliefs Persist

They may provide:

• explanation,
• identity,
• community,
• control,
• emotional relief.

Truth is only one function beliefs can serve.

Epistemic Defense

A good defense against error includes:

• seek disconfirmation,
• compare alternatives,
• ask for mechanisms,
• track source quality.

No method guarantees immunity.

Ask What Would Change Your Mind

This is one of the best epistemic tests.

If the answer is:

nothing,

the belief may be insulated from evidence.

Precommitment

Before seeing results, state:

what evidence would support or weaken your view.

This reduces motivated reinterpretation.

Base Rates

Always ask:

How common is this phenomenon before considering the dramatic details?

Ignoring base rates generates many false beliefs.

Independent Verification

Seek evidence from methods that do not share the same failure mode.

Independent convergence is stronger than repetition.

Slow Thinking

Some errors decrease when people stop and calculate.

But reflective reasoning can also rationalize.

The goal is disciplined procedure.

The Philosophical Lesson

Humans believe strange things not because humans are simply irrational.

We use cognitive systems optimized for:

• fast pattern detection,
• social belonging,
• causal interpretation.

Those systems are powerful.

They also produce predictable errors.

The same mind that discovers patterns must learn when a pattern is only noise.

The Next Question

One of these tendencies deserves closer attention.

Why do we see:

• faces in clouds,
• signals in randomness,
• meaningful coincidence in noise?

The next essay examines:

Patternicity and Pareidolia.