French recordings as evidence

A recording in French is not a document in English. Somebody has to write down what is actually on the tape before anyone argues about what it means — and the order of those two operations is what the objections turn on.

L. 01

The recording is the evidence

The first thing to be clear about is what the transcript is for. It is not the evidence. The recording is the evidence, and a transcript exists to help the trier of fact follow it. Courts enforce that distinction, and they reverse over it.

Chavez

“English-translation transcripts may be admitted only in addition to the underlying foreign-language recordings, not in lieu of them.”

United States v. Chavez, 976 F.3d 1178 (10th Cir. 2020)

In Chavez the Tenth Circuit found reversible error where English transcripts of Spanish recordings went to the jury without the recordings themselves. The practical lesson is the same either way you are litigating: produce a transcript that can sit next to the audio and be checked against it, not one that quietly replaces it.

L. 02

Source first, then translation

The professional standard for this work is transcription-translation, and it has an order. The recording is transcribed verbatim in the language spoken. Only then is that transcript translated, aligned to it, and presented beside it — so that anyone who disputes the translation can point at the line they dispute.

NAJIT 2019

“The TTE must not only be highly proficient in the standard usage of the source and target languages (SL and TL), but also have a strong knowledge of varying social and regional characteristics of those languages.”

NAJIT, Position Paper: General Guidelines and Requirements for Transcription Translation in a Legal Setting for Users and Practitioners (May 2019 revision)

NAJIT’s position paper calls the person doing this a Transcription Translation Expert and prescribes a format presenting the source-language transcription alongside the English translation. That is the artefact we produce.

What you receive
Source, verbatimTranslationLine for line00:14 M1 Tu lui as dit quoi, exactement ?00:14 M1 What exactly did you tell him?00:17 M2 [inaudible] ... rien pantoute.00:17 M2 [inaudible] ... nothing at all.00:21 M1 Arrête de niaiser avec moi, là.00:21 M1 Quit messing around with me.00:24 M2 (rires) OK OK, j’ai compris.00:24 M2 (laughs) OK OK, I get it.00:27 ++ [voix superposées]00:27 ++ [overlapping speech]

The source column is produced first and stands on its own. The translation is tied to it line by line, so a dispute about the English can be argued against a specific line of French.

L. 03

What we work from

Raw list, no qualifiers.

Lawful intercepts and wiretap product · body-worn and in-car camera audio · 911 and emergency-services calls · custodial and investigative interviews · institutional and jail telephone calls · covert and consent recordings · voicemail and answering-machine messages · surveillance video with audio · recorded examinations and arbitration hearings · videoconference recordings · social-media video and messaging audio · dictation and field notes read into a recorder

IncludedOn request
SourceVerbatim transcript in the language of the recording, produced first and delivered whether or not a translation is ordered
TranslationEnglish rendering aligned line for line against the source, never replacing itThird column for the transcriber’s notes
TimingTimestamps at every speaker change and at every marked eventTimestamp on every line
SpeakersSpeakers identified as M1, M2, F1 where unknown; named where identification is established on the recordSpeaker-attribution note setting out the basis for each identification
UncertaintyInaudible, unintelligible, overlapping speech and crosstalk marked where they occur, never guessedConfidence marking on contested passages
AttestationSigned accuracy declaration, named practitioner, covering source and translation separately
DeliverySearchable PDF and editable Word, two-column exhibit formatThree-column format, ASCII, RTF, word index
TurnaroundFive business daysForty-eight hours
L. 04

Two things worth budgeting for

First: translation is not an interpreter’s fee, and the distinction has a cost consequence in U.S. federal litigation.

Taniguchi

“Because the ordinary or common meaning of ‘interpreter’ does not include those who translate writings, we hold that the category ‘compensation of interpreters’ in §1920(6) does not include the cost of document translation.”

Taniguchi v. Kan Pacific Saipan, Ltd., 566 U.S. 560 (2012)

Taniguchi means the cost of translating a document — including a transcript — is not taxable against the losing party under section 1920(6). Budget it as a case expense, not as a recoverable cost. We would rather say that now than have you discover it at the bill of costs.

Second: audio quality sets the ceiling. We do not guess. An inaudible passage is marked inaudible, an uncertain one is marked uncertain, and a recording that cannot support a reliable transcript is reported as such before we invoice for one. If that makes the exhibit weaker, it is the recording that is weak, and you are better off knowing.

Send the recording and tell us the forum. We will tell you what the audio can actually support before we quote it.

Submit a proceeding →