Evidence Has a Destination

Evidence Does Not Point Everywhere at Once

A fact can be real.

Important.

Emotionally loaded.

Extremely inconvenient.

And still be irrelevant to the particular question you are asking.

That is the part brains sometimes skip.

Something happens, the system lights up, and suddenly one piece of information starts voting on every nearby proposition.

Slow down.

Ask:

Evidence of what?

Not:

Is this fact true?

Not yet:

Is this fact important?

First:

What claim is this fact supposed to bear on?

Name the Proposition

Suppose someone becomes extremely angry when accused of lying.

That reaction is real.

Now imagine several possible claims:

“He is angry.”

“He feels accused.”

“He believes the accusation is unfair.”

“He is guilty of lying.”

“He is innocent of lying.”

The anger is extremely good evidence for some of those propositions.

He is, in fact, angry.

Excellent work.

The anger may also tell you something about how he experienced the accusation.

But does anger itself establish that he lied?

No.

Does it establish that he did not lie?

Also no.

The same observation can have very different evidentiary value depending on the proposition.

So before asking:

“What does this evidence mean?”

ask:

What exactly am I trying to determine?

A True Fact Can Point at the Wrong Question

This is one of the easiest mistakes to make because the fact itself can be completely true.

Suppose:

“He was nervous during the interview.”

True.

Now what are we evaluating?

Was he nervous?

Yes. Strong evidence.

Was the interview important to him?

Maybe. The nervousness could bear on that.

Was he lying?

Not automatically.

Was he incompetent?

Also not automatically.

Was he hiding something?

Maybe there are reasons to investigate further.

But the nervousness does not get to wander around attaching itself to every explanation available.

A true observation does not become universal evidence.

Ask the Question Before Grading the Evidence

People often say:

“That is strong evidence.”

Strong evidence of what?

Or:

“That proves nothing.”

Nothing about what?

Evidence only makes sense relative to a question.

Imagine timestamped footage showing Jordan across town during a meeting.

Relative to:

“Was Jordan physically present at the meeting?”

Very relevant.

Relative to:

“Does Jordan generally care about the organization?”

Much less clear.

Relative to:

“Did Jordan know what was discussed afterward?”

The footage alone may tell you almost nothing.

Same fact.

Different proposition.

Different evidentiary job.

The Destination Changes the Arrow

Take this fact:

“Ten thousand people agree.”

That sounds impressive.

But again:

Agree about what?

And why?

Suppose 10,000 independent, knowledgeable observers examine the same question and reach the same conclusion.

That may legitimately bear on whether the conclusion is correct.

Now suppose 10,000 people repeat one viral claim that all came from the same original source.

You still have 10,000 people saying it.

You do not necessarily have 10,000 independent reasons to believe it.

If the proposition is:

“Is this claim socially widespread?”

Excellent evidence.

If the proposition is:

“Is this claim independently verified by 10,000 sources?”

Absolutely not.

The destination matters.

Your Brain Likes Nearby Questions

Here is where things get messy.

Human questions tend to live close together.

Did she leave?

Why did she leave?

Was I responsible?

Was she fair?

Does she still care about me?

Does this mean I am unlovable?

What should I do next?

Those are not one question.

They are a small village of questions standing dangerously close together.

A piece of evidence may bear strongly on one and barely touch another.

“She ended the relationship.”

That bears strongly on:

“Did the relationship end?”

Yes.

It does not automatically settle:

“Why did it end?”

“Was I entirely responsible?”

“Was she entirely responsible?”

“Will every future relationship end this way?”

“Am I fundamentally unlovable?”

The fact is real.

The emotional consequences may be enormous.

The pointer still has a destination.

Do not redirect it halfway across town.

Separate What Happened From What It Means

This becomes especially important when something hurts.

Pain makes neighboring propositions feel welded together.

You fail at something.

Fact:

This attempt failed.

Possible nearby claims:

I prepared badly.

I lack a particular skill.

The task was unusually difficult.

I am incapable.

I will always fail at this.

I am a failure.

Those claims are not interchangeable.

Evidence for the first does not automatically become evidence for the last.

You may eventually gather evidence that supports some of them.

Fine.

Inspect that evidence.

But make each claim earn its own support.

Do Not Let a Fact Vote Twice

Another useful question is:

Has this fact already done its job?

Suppose your partner says:

“I need some space tonight.”

That fact may bear on:

“They want space tonight.”

Good.

Then your brain may quietly reuse it as evidence for:

“They are pulling away.”

Then again for:

“They are going to leave.”

Then again for:

“I always get abandoned.”

One observation.

Four elections.

No.

If you want to move from one proposition to the next, you need reasoning that actually connects them.

You do not get four votes because the first fact was emotionally expensive.

Write the Question Down If You Have To

When the problem is loaded, make the proposition painfully explicit.

Instead of:

“What does all this mean?”

try:

“Did he say this?”

“Did this happen?”

“Was it intentional?”

“Does this behavior make recurrence more likely?”

“Is this person reliable about this particular kind of commitment?”

“Is my interpretation supported?”

“Do I need more information?”

One question at a time.

Now look at the evidence.

Which pieces actually bear on that proposition?

Which pieces point somewhere else?

Which pieces tell you more about your reasoning conditions than the outside claim?

The room gets much quieter once every fact has an assigned job.

Evidence of What?

This is the habit.

Whenever a fact feels important, pause before letting it move the whole model.

Ask:

Evidence of what?

Name the proposition.

Then ask whether the information actually bears on it.

A fact can be true without answering your question.

A fact can be important without settling your question.

A fact can strongly support one proposition and be neutral toward another.

Keep the arrow attached to its destination.

Then let it vote.