Guest blog: Who uses transcriptions of conversations as formal evidence?

Emma Richardson

Emma Richardson

It’s not just researchers who go out with their taperecorders and bring back data to transcribe: formal transcripts are part and parcel of the work of Parliaments, law courts, the police, and many others. Emma Richardson has been looking into the reach and scope of official recordings, and asks us to compare officials’ practical interests with our more academic ones.

My PhD thesis was bang in the ethnomethodology / CA tradition, and I had a lot of fun working on service encounters (in pubs!) with the wonderful Liz Stokoe  and the DARG group in Loughborough. Since then my horizons have widened, though I’m still deeply wedded to the idea of working with video records.

My current job is with the excellent  Aston Institute for Forensic Linguistics (AIFL), and specifically the Centre for Spoken Interaction in Legal Contexts (SILC). One of SILC’s main jobs is tracking down places where spoken language is recorded, transcribed – and, crucially, later serves an evidentiary purpose.

New to the role, I began to wonder and went online to ask the Twittersphere: “In what contexts is spoken interaction recorded and then transcribed and what interests are being served by doing so?”

 Who’s doing it, why, and how?

Tweeters responded with many settings where spoken interaction is recorded and in some form transcribed. The degree to which the detail of interaction is transcribed varies greatly. Your GP, for example, may write down someting you’ve said, but it’s not likely to be verbatim, and it’s not meant to stand as an official record.  In the House of Commons, though, interaction is transcribed and stored in Hansard, as a “substantially verbatim” report of what is said in Parliament. But speech is ‘tidied up’ to remove repetitions and obvious mistakes, retaining only the meaning of the original speech. However, features such as hesitation and disfluency can be an interactional resource employed by speakers during exchanges such as Prime Minister’s Question Time, for effect (Cribb & Rochford, 2017). Removing this is consequential for assessing how well a speaker performed.

Example 1: Hansard and Jefferson (2004) comparison (from Cribb and Rochford (2017))

Screenshot 2020-03-12 at 11.51.36

Tweeters also suggested that I look at interviewing across a range of contexts. Interviewers seek information from interviewees in a range of contexts; social work, admission to CAMHS services, immigration assessments. One of particular interest to us in SILC is police investigatory interviews, where the goal is to uncover the truth about an alleged event. Historically, police officers captured the verbal suspect account in written statements, which the suspect would then sign to confirm an accurate account. This method gave officers’ primary control over the production of any final written documents (Coulthard, 1996). The written record was often found to be shorter, denser in information, to have included more complex noun phrases and to be more structured than the oral telling (Eades, 1996; Jönsson & Linell, 1991). The practice of ‘verballing’ was also common, where police officers fabricated a suspect’s confession.

Screenshot 2020-03-12 at 14.14.56A number of international high profile cases of miscarriages of justice resulted in a call for audio taping investigating interviews (Dixon & Travis, 2007). Audiotaping, while initially met with resistance, was implemented in 1992 under PACE (Police and Criminal Evidence Act 1984).  At least in England, transcripts of the audio tapes now serve as a record of the interaction between officer and suspect, called a ROTI (Record of Taped Interview). ROTI clerks who work within police forces transcribe these tapes. However, they have been found to do so with individualised practices and no standardised transcription system to speak of (Haworth, 2018). They also ‘edit’ the transcripts to summarise actions in the interview.

Example 2: ROTI clerk editing of spoken talk (from Haworth (2018))

Screenshot 2020-03-12 at 11.53.45

In court the transcripts of investigative interviews are ‘used’ by the prosecution, but the audio is rarely ever played. Instead the transcript is ‘performed’ in what Haworth (2013) refers to as the normalised but ‘bizarre’ custom of the police witness acting as the interviewer and the prosecution lawyer as the defendant interviewee. These parties are free to perform features to the talk, which might, or might not have been present in the oral version. While members of the court have copies of the transcript available to them, it is likely the oral performance will be more memorable.

In the UK, some courtroom interactions are recorded and transcribed (but not all).  Hearings at Crown Court and at civil and family courts are always recorded and transcribed. Tribunal hearings are not always, and hearings at magistrates’ courts are never, recorded and transcribed. The transcripts of court hearings can be used in two ways, first as a record of the proceedings, but they can also be used as evidence in cases (Fraser, 2003). The defence can use what was said at interview and then later in court to show consistency, or by prosecution to highlight inconsistency in a suspect’s case (Haworth, 2010, 2018).  The media also make use of transcription of what was said in a courtroom in their reports, often directly quoting sections of the interview interactions (Johnson, 2013). In a sense, the interviewee is ‘ventriloquized’ (Tannen, 2007).Therefore, the detail and quality of those transcripts is highly consequential.

Example 3: Media quoting transcript as if Shipman was speaking at the trial (from Johnson (2013))

Screenshot 2020-03-12 at 12.03.13

What have I learnt?

This preliminary search of the literature reveals that different contexts occasion different levels of transcription. In all contexts, accuracy of what was actually said is important. However, institutional standards of what counts as accurate transcription vary greatly from what we, as scholars of interaction, would deem acceptable, as we’ve seen in Example 1.  Academics such as Bucholtz (2000, 2007, 2009), Jefferson (1996, 2004), Hepburn (2004) and many, many others in the field have written on the importance of capturing the micro detail of transcription. These and others acknowledge that each transcription is in itself an analysis, serving a purpose, conducted by a professional hearer and even amongst ourselves, we dispute a transcription.

Building on some of Haworth’s prior work (Haworth, 2010, 2018), the initial project of SILC involves creating a comprehensive map of current procedure for transcribing police-suspect interviews. There is known to be variation in practice at both local and national level, and we will conduct an analysis of this variation and examine the consequences for the use of interview data as evidence.  The research findings will be used to work alongside practitioners and other data users (such as the courts) to produce standardised transcription conventions and training, thereby making a significant contribution to evidential integrity and accuracy.


Bucholtz, M. (2000). The politics of transcription. Journal of Pragmatics, 32(10), 1439-1465. doi:
Bucholtz, M. (2007). Variation in transcription. Discourse Studies, 9(6), 784-808. doi:10.1177/1461445607082580
Bucholtz, M. (2009). Captured on tape: professional hearing and competing entextualizations in the criminal justice system Text & Talk – An Interdisciplinary Journal of Language, Discourse & Communication Studies (Vol. 29, pp. 503).
Coulthard, M. (1996). The official version: Audience manipulation in police records of interviews with suspects. In C. Caldas-Coulthard & M. Coulthard (Eds.), Texts and Practices: Readings in Critical Discourse Analysis (pp. 166–178). London: Routledge,.
Cribb, M., & Rochford, S. (2017). The Transcription and Representation of Spoken Political Discourse in the UK House of Commons. International Journal of English Linguistics, 8, 1. doi:10.5539/ijel.v8n2p1
Dixon, D., & Travis, G. (2007). Interrogating Images: Audio-visually recorded police questioning of suspects. Sydney, Australia: The Sydney Institute of Criminology.
Eades, D. (1996). Verbatim courtroom transcripts and discourse analysis. In H. Kniffka (Ed.), Recent Developments in Forensic Linguistics (pp. 241–254). Frankfurt: Peter Lang.
Fraser, H. (2003). Issues in transcription: factors affecting the reliability of transcripts as evidence in legal cases. International Journal of Speech, Language and the Law, 10(2). doi:10.1558/sll.2003.10.2.203
Haworth, K. (2010). Police interviews as evidence. In M. Coulthard & A. Johnson (Eds.), Routledge Handbook of Forensic Linguistics. Abingdon: Routledge.
Haworth, K. (2013). Audience design in the police interview: The interactional and judicial consequences of audience orientation. Language in Society, 42(1), 45-69. doi:10.1017/S0047404512000899
Haworth, K. (2018). Tapes, transcripts and trials:The routine contamination of police interview evidence. The International Journal of Evidence & Proof, 22(4), 428-450. doi:10.1177/1365712718798656
Hepburn, A. (2004). Crying: Notes on description, transcription, and interaction. Research on Language and Social Interaction, 37(3), 251-290.
 Jefferson, G. (1996). A case of transcriptional stereotyping. Journal of Pragmatics, 26(2), 159-170. doi:
Jefferson, G. (2004). Glossary of transcript symbols with an introduction. In G. Lerner (Ed.), Conversation Analysis: Studies from the First Generation. Amsterdam: John Benjamins.
Johnson, A. (2013). Embedding police interviews in the prosecution case in the Shipman trial. In C. Heffer, F. Rock, & J. Conley (Eds.), Legal-lay communication. Textual travels in Law (pp. 147-167). Oxford: Oxford University Press.
Jönsson, L., & Linell, P. E. R. (1991). Story generations: From dialogical Interviews to written reports in police interrogations Text – Interdisciplinary Journal for the Study of Discourse (Vol. 11, pp. 419).
Tannen, D. (2007). Talking Voices: Repetition, Dialogue, and Imagery in Conversational Discourse (2 ed.). Cambridge: Cambridge University Press.