There is a quiet assumption in court reporting that deserves to be examined:

If every misspeak, stutter, and restart is preserved, the transcript must be more accurate.

It sounds reasonable.
It feels safe.
It also isn’t true.

Accuracy is not achieved by capturing every physiological artifact of speech. It is achieved by faithfully conveying what was meant and what occurred. Those are not always the same thing.

Speech Is Noisy by Design

Human speech is not a polished delivery mechanism. It is an improvised cognitive process happening in real time.

People:

• start sentences before they know how they’ll end

• revise mid-thought

• repeat themselves while searching for precision

• stutter under stress or fatigue

• use fillers while organizing ideas

Most of this noise carries no evidentiary value.

Rendering it verbatim does not preserve testimony. It preserves how the nervous system behaved while producing it.

The record is not a neurological scan.

When Over-Verbatim Transcription Distorts Meaning

Including every “uh,” half-word, and restart is often defended as neutrality. In practice, it does the opposite.

A transcript saturated with disfluency:

• makes speakers appear confused or evasive

• exaggerates stress into incompetence

• turns normal thinking into apparent uncertainty

• invites uncharitable interpretations from readers

This is especially dangerous because tone cannot be reliably transcribed. The reader sees disruption but hears none of the vocal nuance that made it intelligible in the room.

The result is manufactured meaning.

That is not fidelity.

The Record Is About Meaning, Not Mechanics

Court transcripts exist to answer substantive questions:

• What was said?

• What was admitted or denied?

• What was understood?

• What occurred interactionally?

They do not exist to document:

• every false start

• every syllabic repetition

• every moment of verbal friction

A witness who says:

“I—I—I didn’t, uh, I didn’t see it.”

is not giving different testimony than one who says:

“I didn’t see it.”

In most cases, preserving the former adds nothing except bias.

Where I Do Not Agree With Blanket Smoothing

This is where the line must be drawn carefully.

Some disfluency is not incidental. It is interactional.

Speech disruption matters when:

• a witness repeatedly avoids answering a direct question

• a sentence stops short of a commitment

• a correction materially changes the facts

• an interruption prevents completion of an answer

• non-answers accumulate in response to pressure

In those moments, smoothing would erase something real.

The ethical obligation is not to remove disfluency wholesale, but to distinguish signal from noise.

A Defensible Ethical Standard

Here is the standard that holds up under scrutiny:

The transcript should preserve disfluency only when it affects meaning, interaction, or the integrity of the exchange.

Everything else is speech mechanics.

That standard:

• protects the record

• protects the speaker

• protects the reader

• and protects the reporter

It also requires judgment.

Why “Strict Verbatim” Is Not a Neutral Position

Strict verbatim is often framed as the safest ethical choice.

In reality, it is a philosophical position that assumes:

• speech should look like speech

• disruption should be visually preserved

• clarity is secondary to literalism

Those assumptions are not neutral. They favor process fidelity over meaning fidelity.

And they quietly shift responsibility away from the reporter and onto the reader.

A Practical Test (That Keeps This Honest)

When deciding whether to preserve a misspeak or stutter, ask:

If I remove this, does the meaning, force, or interactional reality of the answer change?

• If no, smoothing improves fidelity

• If yes, preservation is required

That test is simple, defensible, and repeatable.

Why Fear Drives Over-Inclusion

Many reporters include everything not because it’s necessary, but because it feels safer:

• safer from challenge

• safer from accusation

• safer from responsibility

But ethics is not about maximal inclusion.
It is about faithful representation.

Avoiding judgment does not eliminate risk. It merely hides it.

The Bottom Line

Not every misspeak belongs in the record.

Not every stutter is testimony.

The ethical responsibility of the court reporter is not to transcribe the nervous system, but to preserve meaning and interactional truth.

That requires discernment.
It requires accountability.
It requires courage.

And it requires rejecting the comforting fiction that fidelity means copying everything that happened to come out of someone’s mouth.

Because a transcript that includes everything can still get the truth wrong.

And a transcript that exercises judgment—carefully and honestly—often gets it right.