Forensic Transcription Basics

Quite frequently interpreters and translators are asked to transcribe and subsequently translate recorded conversations obtained or produced by law enforcement entities. These could be for example, a post-arrest interview of a defendant, suspect, or witness; a recorded conversation from wiretapped cellphones, calls made to or from prison inmates, etc.

SKILLS REQUIRED

A good ear – attentive listening. Since your source material is a recording, you will have to have a good auditory command of the language spoken on the tapes. If, after listening to the audio, you realize that the language is spoken is of a regional variety that you cannot understand or you are not able to comprehend the type of lexicon used, you have an ethical obligation to reject the assignment. You can do this very professionally and if possible, even refer the client to someone you trust can do a good job.

Ability to sit at your computer for long periods of time. Transcription / translation requires one to listen to the same segment of tape many times until you’ve gotten all the utterances down and then spend more translating and editing and if you have a looming deadline, that means a lot of sitting down.

Editing skills. The devil is in details and in this type of work, you have be willing to painstakingly review your written transcription and then the translation to make sure you’ve captured everything and have cleared up any misspellings or typographical errors you may have made.   Always print the document and review it once again on the hard copy. It’s amazing how many mistakes you can overlook on a screen but will pop out on the hardcopy. A second pair of eyes – ideally, another person should look at the final product before you send it to the client.

Conservation.   A transcription is a verbatim written version of what was said. EVERYTHING must be included – false starts, speech errors, repetitions, etc. Oftentimes, the speakers will use “broken” versions of a language, including false cognates, or will pepper their second language speech with English. This must be conserved. If a speaker uses the wrong word, this MUST be indicated in the transcription AND in the translation. If you can find an equivalent error in the target language, use it. Otherwise, simply indicate that a wrong word was used in the original language.

Include the original speech in the non-English language: Many attorneys will ask you to not bother including what said in the original language. You should include it anyway.   Best practices standards dictate that you should. It’s a matter of ‘transparency’. One should be able to cross-reference what was originally said at any time. This is very important.   No matter how tempting it is, there should always be a column with the original utterances.

Format: We recommend a four-column format as shown below.

  1. The first column is the line number. This is necessary so attorneys can identify a particular comment as they reference it during court proceedings.
  2. The next column is for identifying the speaker.
  3. Then, what was said in the original language.
  4. And finally, the translation of the message.

Labeling the speakers. You are not a voice expert but as a human being with good hearing ability, you will usually be able to identify the different voices in the recording. We recommend you use V1 (voice 1), V2 (voice 2)… or MV1 (male voice one), FV1 (female voice 1). If you are certain who the voices belong to – say it’s a female agent interrogating a male suspect, you may use ‘Agent’ and ‘Mr. so-and-so’ on the speaker column. That’s a decision you will to need to make. If you are uncertain which one of the voices uttered something (this happens often with overlapping speech) just say so.   Example: Transcriber’s note: It is unclear which speaker stated this. Or put ‘V1 or V2’ in the speaker column, as the case may be.   Or add a note on the cover sheet, e.g., Transcriber’s note: It’s possible that some messages may have been attributed to the wrong speaker, particularly in overlapping speech. If you are unsure, make it known. Don’t just guess and leave it as it if were a sure thing.

Legend. We’ve attached an example of a legend that you can use for this type of work. Eventually, if you do this often enough, you will come up with your own legend.

The most important factor, is that your transcription be precise – that it accurately mirrors everything that was said in the recordings. This includes uh-huh, huh-uh, uh, um, etc.   These linguistic indicators can be very important for the attorneys or jurors who will read the transcription as they can demonstrate hesitation, doubt, insecurity, and other cues that may be important to determine credibility and other issues. Don’t clean up the speech or make it sound more eloquent!! If grammatical errors were made in the original language, or the speaker used a lot of low-level terminology, that should be reflected in the translation.

Remember, your transcription and translation will be used in further legal proceedings and may also be contested by opposing parties. Do not be surprised if you are required to testify as to the accuracy of transcription / translation or if you’re asked why you translated something in a certain way and not another.

For a more expansive guide to forensic transcription and translation, we recommend you read the respective chapter in Fundamentals of Court Interpretation, Theory, Policy, and Practice, 2nd edition, written by Dr. Roseann D. Gonzalez, Holly Mikkelson, and Victoria Vazquez.

EXAMPLE:

V1       VOICE 1

V2       VOICE 2

V3       VOICE 3

 

Line# speaker Original Translation
1. V1 No te hagas güey, tú también estuviste allí. No andes ahora con que-, que eres el muy inocente.   [o] [u] Don’t play the fool, dude, you were there too.   Don’t act now like-, like you’re so innocent [o] [u]
2. V2 [o] Miren… [o] Look…
3. V3 [o] Nosotros no vamos a pagar el pato solos. Si tú empiezas a soplar, te vas a arre-, arrepentir. [o] We’re not going to pay the piper by ourselves.   If you start singing, you’ll be sor-, sorry.
4. V2 [o] ¡Yo no soy dedo! Ese chota vino y me quiso sacar información, pero yo no le dije nada – ni pio. [o] I’m not a snitch! That cop came and tried to get information out of me, but I didn’t tell him anything – he didn’t get a peep from me.

TRANSCRIPTION AND TRANSLATION LEGEND

[sound] indicates background or other noises as described
[U] Unintelligible
[O] Overlapping voices
[overlapping – U] overlapping voices, unintelligible
[p] Pause
[lengthy p] lengthy pause
::::: prolongation of sound
ALL CAPS indicates shouted or increased-volume speech
[sil] Silence
[WSW] wrong Spanish word
[NSW] non-Spanish word
[WEW] wrong English word
[NEW] non-English word
[garbled] inexistent word, or word makes no sense in context, or distorted pronunciation; can combine with WSW, NSW, etc.
[v] Vernacular
italics indicate English utterance/s spoken in Spanish context, or vice versa
[1]footnote

 

supplemental information regarding meaning, context or cultural significance
[gen] Error in gender marker
[sic] as stated, despite error
[gen] error in gender marker
[pers] lack of agreement between personal pronoun and verb
I think that– sample of interrupted speech
–that’s not right sample of interrupting speech
/phonetic/ sound or unrecognized word, transcribed phonetically
Bold words pronounced with greater emphasis
[PRO] Mispronunciation

 

Leave a Reply

Your email address will not be published. Required fields are marked *