In the course of an investigative interview, police investigators regularly present evidence in the form of witness statements or technical evidence, such as telephone traffic logs. Previous research has shown that doing so is associated with resistance and confrontational responses from the suspect, leading to decreased cooperation in the interview (Kelly et al. 2025). However, very little is known about how investigators present evidence and how conflict is generated. Using a conversation analytic methodology, this study analyzes the presentation of evidence in a corpus of 80+ video recordings of naturally occurring police investigative interviews in Norway. The focus is on attributions and presuppositions of blame. The analysis shows that the investigator may present the evidence as more or less misaligned with the suspect’s previous account. The suspect’s response, in turn, may treat the presentation as more or less accusatory and disaffiliative by either continuing to collaborate on information exchange or by entering a conflictual or defensive mode by for instance rejecting the legitimacy of the presented evidence. The study thus contributes to uncovering how conversational cooperation may be maintained or challenged in a high-stakes situation of conflicting descriptions of reality.

Reference
Kelly, C. E., Parker, M., Meehan, N., & McClary, M. (2025). Evidence presentation in suspect interviews: A review of the literature. The Police Journal, 98(1), 112-128.

The information-gathering purpose of a police interview is often tied to the type of possible crime committed. Understanding the terms used to describe the crime thus matters for the information provided by the suspect, as certain kinds of crimes are assessed on the basis of specific details. The police interviewer may ask a suspect about their understanding of a legal term, or a term that in the context of a crime may have special meaning or implications. For instance, a police interviewer can ask “when I say the word violence, what do you consider part of that word?”.

Employing Conversation Analysis, my presentation will focus on this type of understanding checks used for legal or crime-relevant terminology in Norwegian police interviews of suspects. I investigate how the sequences play out in the interest of information-gathering and how they arise, what kind of question type they get treated as, and how they reveal the goals of the information gathering. It turns out that these are common in cases of rape and sexual violence, such as how suspects understand terms like sex and violence. Uncomfortable topics and the level of detail sought by the police seem to be a barrier for further elaboration. For suspects who speak Norwegian as a second language, the understanding checks may treat them and their competence in a certain way, beyond the relevant information gathering by requesting specific details.

Police interviews are conducted in phases that differ in terms of procedural constraints and interactional contributions. At the beginning of the interview, suspects receive information about rights, obligations, and provisions, while at the end, they are asked several procedural questions. These formal stages may pose difficulties in understanding, especially when the suspect is unfamiliar with legal procedures and/or legalese. To secure understanding, investigators may use various pre-emptive practices, such as reformulation or simplification (Svennevig et al., 2023). However, such practices are not always employed, nor are they applied uniformly. Moreover, even when used, they do not guarantee that the suspect will understand the information. What remains largely unknown is how investigators handle understanding problems post hoc – that is, how they respond when suspects misunderstand or do not understand what has been procedurally communicated to them during the initial or final stage of the interview.
This study examines such cases by focusing on the investigators’ methods for clarifying legal procedures and formalities in response to suspects’ misunderstandings or non-understandings. Using Conversation Analysis as a method on a corpus of 98 recordings of police interviews with first- and second-language suspects in Norway, the study identifies how understanding problems concerning formalities and procedures are manifested by suspects and how they are subsequently addressed by investigators. More specifically, the analysis focuses on the design and content of investigators’ responses and the extent to which they engage with and succeed in providing sufficient explanations.

Sexual violence in England and Wales is highly prevalent, grossly under reported and rarely results in conviction. The latest figures show that of the 67,125 recorded rape offences (ONS, 2022) there were just 2,537 rape prosecutions (CPS, 2022) leading to 1409 convictions (CPS, 2022). The investigative police interview, enshrined in written guidance, is central in the route to obtaining justice. Recently, updates to interviewing guidance documents (Ministry of Justice, 2022) and the roll out of ‘Operation Bluestone Soteria’ in England and Wales has placed focus on police interviewers taking a ‘trauma informed approach’ to investigative interviewing. Additionally, best practice suggests shifting focus to the behaviour of the accused rather than the complainant in relation to the assault.

The study asks, what does a trauma informed approach look like in practice in the unfolding interaction and what are the interactional impacts of the new guidance, ten years on? To answer this, two data sets of video recordings (n=39) collected ten years apart of investigative police interviews with ‘vulnerable’ adult and child witnesses of rape and serious sexual offence (RASSO) are combined. The study employs conversation analysis to examine what a trauma informed approach looks like in practice. Phenomena of interest include where progressivity is stalled due to distress, and sequences begin with questions such as “So what was going through your head. At that ti:me?” Data are in British English.

The study investigates verbal repair practices on written text (e.g., Balaman, 2021) in the context of criminal and civil case mediation. In mediation, the parties affected by a crime or dispute discuss their case with the help of two volunteer mediators. If they reach resolution, a mediation agreement is written at the end of the meeting. The agreement is between the clients, and legally binds them, while the practical writing work is done by mediators. The writing of the agreement can be organized (Komter, 2006, Mortensen, 2013, Mondada & Svinhufvud, 2016) in two main alternative ways: A) verbal discussion about the agreement is followed by one mediator individually writing it, or B) verbal discussion is intertwined with collaborative writing (on an individual or a projected screen). In both settings, the clients are at some point asked to read the text and request changes if needed.

In this study, I examine how clients initiate repair on the text when reading it from a paper print-out (setting A), or when listening to the mediator read aloud from an individual screen or when observing ongoing writing on a projected screen (setting B). It is discussed how these different constellations afford requesting changes, and, relatedly, how the repair initiations display clients’ low or high entitlement to repair (Bolden, 2011) textual formulations in the agreement. The study contributes to research on collaborative writing, and it has potential implications for mediation practice.

The data are video-recorded mediation meetings in Finnish. The method is multimodal conversation analysis.