Comment by Mace Ojala

On the 10th of November RUSTlab had the pleasure to welcome in person Dr. Martha Komter, emerita and visiting research fellow from the Netherlands Institute for the Study of Crime and Law Enforcement (NSCR) to share her research of analysing police interrogation transcripts in action.

The presentation was on-point with our lab’s theme this semester, data at work, and brought us into a very specific professional situation, namely routine police interrogations of street crime suspects. Under the Dutch criminal law, police interrogations must be transcribed and transcripts included in the case file, which is then used in the courtroom as material by the judge, as well as the defence and prosecution lawyers to build their respective cases. It therefore matters greatly how exactly what the suspect says is transcribed into text – paying close attention to the practices and implications of transcription should perk up the interest of any social scientists and everyone who deals with interviews in their own research. The presentation certainly spoke to my own experiences of the interesting chore of transcription, as well as the time I was myself interrogated by the police.

How then is this routine work done by the police officers? Dr. Komter walked us through three styles of transcription she had identified, reading them against the backdrop of criminology and trends in law enforcement practice. Traditionally what the suspect says is typed in as a monologue. This style has the effect of removing the interrogator from the transcript and imply that the suspect has given the account out of their own volition. This is far from what happens in interaction of the interrogation, as is obvious from the audio recordings we heard, and the from the extended transcripts Dr. Komter has produced to include things such as what the police actually typed, time it takes for the police officer to do so, and how this typing action maintains the turn (in conversation analytical terms) and control of the interaction with the interrogating police officers. The trending, alternative style is more question and answer based. This style re-stores the interrogator’s voice in the transcript, and gives a more adversarial effect which police officers sometimes find preferable for their transcripts. Dr. Komter labels the third transcription style she identified as the recontextualized monologue. In this style, the suspect repeats the question of the interrogator, and places what the interrogator has said within the suspects own statements. This style highlights the present absence of the the interrogator from the transcript, and enhances the idea that the suspect comprehends and accepts what has been asked of them.

Regardless of style, the work of transcription navigates the transmedia task of turning a dynamic, bodily and highly staged adversarial interaction event into a sequential text to be used in the courtroom as an agreeable dataset. Dr. Komter herself drew from conversation analysis literature, and I am sure many of us in the audience were listening through our science and technology studies sensibilities; even “transcripts in action” in her title suggested a connection to Bruno Latour’s (RIP) classic 1987 book relevant to the work of translation. Indeed, an audience comment in the lively discussion reframed the interrogation situation as coordination of the production of the transcript-object by the suspect, the one or multiple interrogating police officers, and I would add the keyboard, the word processor, the desk, screen and other material actors. Once established and sufficiently stabilized, these transcripts then travel as immutable mobiles from the interrogation interaction to the case file, and to the courtroom where they participate in the doings of institutionalized justice.

We were told that different police stations would cultivate different styles, and indeed a classic ANTy move would be, I figure, to trace the social life and sociology of these translations, how they travel and are reproduced and maintained between police officers, their educators, the suspects, tools of the trade such as word processors, keyboards, sound recorders and MP3 files, victims and witnesses, crime scenes, courtrooms and more. Before long, the ANT anthropologist would encounter a sociologist – perhaps you!

Regarding Dr. Komter’s presentation, it is amazing how illustrative an analysis of a mere fleeting moment no longer than a few seconds can be when detailed at by an attuned and curious conversation analyst. I was reminded of gatecrashing a series of annual symbolic interactionist conferences, some of the absolutely nerdiest conferences I ever went to! Obviously in the asymmetric police interrogation the stakes are high for the suspect (and possible victims and other people in their absence), while routine or banal for the interrogator, a relation in which epistemic and testimonial justice (Fricker 2007) materializes (or doesn’t, or to a varying degree and quality) as the transcript comes to be.

While hopefully less adversarial, a sociologists interview transcript does similar creative datafication work of decontextualization and recontextualization, always-already predicting future uses and setting their conditions of possibility. It seems to me that many pragmatic questions from Dr. Komter’s work mirror directly interview work conducted by social scientists, including many STS scholars; what counts as a satisfying answer, what is typed and what not, how is the turn passed from the interviewer to the interviewee and then seized again for example by act of typing? Further, our late-night discussions after the presentation spiralled to grapple with what is the intended and factual thing the interrogation transcript of an interrogator (or a social scientist, for that matter) represents; is it the interrogation interaction, a projection of suspicion, solidification of the relation of data-extraction or voluntary “handing over” of data, a replay of the question protocol, the subject’s testimony, or how their internal state refracts onto their testimony, the historical flow of events of potentially criminal nature, establishment of truthyness of the case file, the consummerability of interrogation interactions with content and type of other data… or something else entirely?

I was left thinking that transcription and transmediations of interactions into listings of text seems like such an obvious thing to do, but are only obvious given the technological a-priori of search functions, the latin alphabet and its mapping onto ubiquitous and mass-produced typing equipment, achievable speeds of its operation with two human hands, printers, routine word processing operations such as cursor movement or erasure, cheap typeset, wide and successful literacy education, authoritative power of text, (presumed) immutability of bureaucratic documents et cetera. I would like to ask the sociologically inclined readers of this comment: what is your transcription style, and why do you transcribe in the first place? Could you stop yourself from doing it altogether?