Abstract
When police officers interview people with intellectual disabilities who allege sexual assault and rape, they must establish rapport with the interviewee but deal with their distress in a way that does not compromise the interview’s impartiality and its acceptability in court. Inspection of 19 videotaped interviews from an English police force’s records reveals that the officers deal with expressed distress by choosing among three practices: minimal (e.g. okay) or no acknowledgement, acknowledging the expressed emotion as a matter of the complainant’s difficulty in proceeding (e.g. take your time) and rarely (and only if the complainants were apparently unable to resume their talk) explicit reference to their emotion (e.g. it’s obviously upsetting for you). We discuss these practices as ways of managing the conflicting demands of rapport and evidence-gathering.
Introduction
Police investigations of sexual assault and rape inevitably require the alleged victims to revisit their experiences in interview and be subjected to detailed probing. This may be distressing. Our interest in this article is to examine how police interviewers deal with the distress of one particular group of such alleged victims, that is, people with an intellectual disability (ID), whose cognitive powers are likely to be uncomfortably tested by the requirements of a formal police interview.
People with IDs are disproportionately likely to be sexually abused (McEachern, 2012) and sexually assaulted (Petersilia, 2001) compared with people without such disabilities. They are likely in consequence to suffer ‘marked increases in the frequency and severity of emotional, physiological and behavioural symptoms of psychological distress’ according to Rowsell et al. (2013: 257). They are, however, less likely to speak out about the abuse, and the number of victims pursuing alleged offenders through the criminal justice system is but a tiny fraction of those abused (Brown et al., 1995, for example, estimate 6%). For those who do bring a complaint, the legal process can cause them emotional and psychological distress as they re-live the incident. Indeed, research suggests that adult rape victims with learning disabilities or psychiatric problems are over-represented in terms of cases that drop out of the system (Lea et al., 2003).
One element in the complex of reasons for the failure of sexual assault cases involving adults with ID may be their experiences with the criminal justice system at an early stage of the investigation. In this article, we use a sample of videotapes collected from an English police force to examine the way the police handle the complainant’s first recorded interview. This interview is a crucial part of the police’s decision as to whether to proceed with the case. It requires the police interviewer to solicit a description of events, in forensic detail, to assess whether (a) a crime has been committed, (b) there is enough evidence to pursue an investigation and locate the perpetrator and (c) the complainant’s account – specifically its reliability and plausibility – would stand up in court. In doing so, the officer will ask for detailed descriptions of the complainant’s experiences and, if necessary, probe and challenge apparent inconsistencies or vagueness. The intrusive and intimate nature of the information being solicited (involving explicit description of genitalia) would risk causing discomfort to any complainant; where the complainants have difficulty in understanding questions and in expressing themselves – as people with ID do – the risk of frustration and distress may be magnified.
The difficulties that people with ID have in police interviews (mostly as suspects) have been much researched since the pioneering work of Clare and Gudjonsson (1993), and a useful summary is provided in a recent report by the Royal College of Psychiatrists (2014). The report notes that police and other members of the legal profession systematically have failed to acknowledge the tendency of people with ID to acquiesce, their deficiencies in vocabulary, their tendency to report apparent contradictions and their memory limitations. One of the recommendations of the report is that ‘all police officers, in particular custody officers and community support officers, need to have intellectual disability awareness training as part of their induction process. This training should be provided with input from health professionals’ (Royal College of Psychiatrists, 2014: 11).
The main instrument for such training and guidance given to police interviewers in England and Wales is set out in Achieving Best Evidence in Criminal Proceedings: Guidance on Interviewing Victims and Witnesses, and Guidance on Using Special Measures (henceforth ABE) published by the UK Ministry of Justice (2011). These guidelines mandate that the interviewer develop a rapport with the interviewee. This extract from the guidelines gives a clear instruction: 2.230 Some witnesses may be unhappy or feel shame or resentment about being questioned, especially on personal matters. In the rapport phase, and throughout the interview, the interviewer should convey to the witness that they have respect and sympathy for how the witness feels. (UK Ministry of Justice, 2011: 66)
But the guidelines also remind the interviewer that this is an interview in the service of a possible court case: 3.1 The basic goal of an interview with a witness is to obtain an accurate and reliable account in a way which is fair, is in the witness’s interests and is acceptable to the court. (UK Ministry of Justice, 2011: 68)
This overarching mandate requires the interviewer to ensure that they are seen to be impartial so as to collect evidence as free as possible from any implication of influence. It is important to note that were the case to come to court, the tape of the interview would be made available to the counsel defending the alleged perpetrator, who would be free to call for any part of it to be played in court; hence the need for the information solicited from the complainant to be seen to be uncontaminated by the interviewer’s behaviour. There is, then, something of a conflict between the need for rapport and the imperative to obtain full and accurate information without the taint of partiality.
Some institutional ways of dealing with distress weigh the balance towards empathy. Call-takers on a peer-run ‘warm’ helpline, for example, will offer explicit support to callers as a definitional part of the service they offer (Pudlinski, 2005). Other institutional objectives point the other way. When there is a tension between expressing empathy and progressing with the institutional task in hand, the interactional research literature on dealing with clients suggests that many practitioners – for example, medical professionals (Heritage and Lindström, 2012; Ruusuvuori, 2005) – will put non-emotion-oriented tasks first and avoid offering encouraging receipts to expressions of trouble or distress. This may be so even when the troubles being reported are obviously highly upsetting. Hepburn and Potter (2007) find that when NSPCC
1
call-takers hear reports of child neglect and abuse: full scale surprise at the abuse, explicit condemnation of the perpetrator, and overt sympathy for either victim or reporter are unusual. Instead, abuse reports are typically followed by actions that perform institutional tasks – for example, evidence gathering or advising. (p. 100)
Given the evidence-gathering objectives of the police interview, police interviewers are very likely to respond as NSPCC call-takers do. But the further legal obligations they work under (their need to be seen to be impartial and to test the complainant’s accounts against forensic standards of proof) may constrain them still more closely and make the pattern that Hepburn and Potter (2007) identify even more pronounced. What is of interest is how it is brought off while maintaining the ‘respect and sympathy’ for the complainant that the national police guidelines require.
Data and method
Our data come from a police force in England, who provided us with 20 videotapes of, and summary information about, interviews from archived cases involving complainants with IDs. The selection of these cases was made by local personnel, on their judgement that they were representative of such complainants. The interviews were recorded between 2010 and 2013, in one city station, by trained officers. Interviews lasted between 21 minutes and just over two hours. The tapes were visually and aurally anonymised before leaving the police station. We excluded from analysis one tape which was incomplete. Of the 19 cases, 10 dealt with alleged rape and the rest sexual assault. Internal evidence from the tapes showed that two of the complainants were children under 16 (a boy and a girl), 14 were young women, one was a mature woman and two were men (one young and one young-middle aged). Three of the 19 cases led to a court case, two resulting in a guilty verdict. The other 16 led to no further action.
Description of the environment of the interview
Once a case is brought to the attention of the police by the person alleging assault or someone acting on their behalf, basic details are recorded, an officer is assigned to the case and the alleged victim is invited for interview. Before this interview, the officer may or may not investigate the matter further. In the case of complainants suspected of being, or known to be, intellectually disabled, steps may be taken to identify the nature of their disability (e.g. by consultation with educational, medical or social work authorities). At the station from which our cases come, such information is recorded only briefly (e.g. ‘Learning disability, attends special school’).
On the day of the interview, at the station, the complainant (and any accompanying friend or relative, carer, social worker or other professional) is met by the assigned officer who, before the interview proper, engages them in a preliminary conversation. This will typically be short and neutral and would normally serve as an opportunity for the officer to alert the interviewee to the nature of the upcoming discussion as an official one which will be recorded for use in court (should the matter proceed). The complainant’s party is then shown the layout of the interview room and the observation room, and the cameras are located and explained. A companion (e.g. a relative or case worker) may be present in the interview room but take no part in the conversation.
To examine the interviews for the way in which distress is dealt with, we shall use Conversation Analysis (CA) (for an overview, see the collection in Sidnell and Stivers, 2012), which, building on a base established in ordinary conversation, has been used extensively in research on how practitioners deliver their services in a wide variety of environments where clients may present accounts of troubles – recent overviews can be found of primary medical care (Robinson and Heritage, 2014), welfare services (Drew et al., 2014), police and mediation interviews (Stokoe, 2014) and telephone helplines (Hepburn et al., 2014). We shall use CA here to identify the sequential locations in which distress appears and identify the interactional formats with which specific interviewers deal with it, juggling the need for rapport on the one hand and the requirement for dispassionate evidence-gathering on the other.
Analysis
Inspection of the tapes revealed that the interview fell, as mandated by the national guidelines ABE, into four phases: an orientation phase, in which the purpose, organisation and recording of the interview are explained; a free-narrative phase, in which the complainant recounts their experiences; a probing phase (which usually takes up the bulk of the interview), in which the interviewer solicits detail and checks inconsistencies and other snags in the account; and a final closing phase. Distress emerged in the narrative and the probing phase.
Reports of distress
We distinguished between two senses of distress: reports of distress, where the complainants described how they had felt at the time of the incident or how they were currently feeling in general, and distress expressed in the moment (by crying, whimpering, holding their head in the hands, moaning, etc.). Our main analysis is of this latter kind. But it is worth briefly mentioning that when the former kind – reports of distress – appeared in the complainant’s narrative or in response to questions, they were received by the interviewing officer only with minimal, or sometimes no, acknowledgement. Examples of such reports of past distress and the interviewers’ responses are shown in Extracts 1 and 2:
2
Extract 1: INT9 Line 157: Scared
In Extract 1, the complainant’s report that she felt scared (Line 9) is met with a second of silence, then a quiet okay from the interviewer. The interviewer may have been going on to give a more elaborate reply, but the complainant adds, in overlap to the outbreath at the very end of okay, a further element to the story; this, too, is only minimally receipted: Extract 2: INT20 Line 414: Pee off
Note that in Extract 2, the arrows indicate the multiple occasions at which the interviewer could have taken a turn at talk to acknowledge the complainant’s reported distress, but does not do so (and the video, although it does not capture the interviewer’s face, shows no obvious nodding or other head movement). The reports of distress are treated as being part of the ‘narrative world’ being described in the complainant’s account. The emotion described in the story is probably better conceived of as an economical way for the teller to display her contemporaneous stance towards the events (for emotion as stance, see Goodwin et al., 2012). In receiving the account, the interviewer is co-operating by aligning (Lindström and Sorjonen, 2013) with the activity of storytelling, but not affiliating (Lindström and Sorjonen, 2013) with the stance expressed in it.
Extract 3 is a combination of both kinds of reported distress. It shows the interviewer’s non-response both to the complainant’s report of her current feelings of shame and to her report of the fear and upset she felt at the time of the assault: Extract 3: INT13 Shame and upset. Note that the ~ sign in Line 24 indicates ‘wobbly voice’
In Extract 3, the interviewer at Line 18 receipts the complainant’s report of current feelings of shame by a repeat response-confirmation (and deleting the amplifying adjective really). Such repeats are common in institutional questioning such as classroom exchanges (Park, 2014): when delivered without rising intonation, they serve to affirm that the response has been heard. Without further elaboration, it withholds an appreciation of the content – here, of the emotion that the answer expresses. At Lines 21 and 22, the interviewer asks the complainant how she felt at the time. The complainant’s answer has a ‘wobbly’ quality, implying incipient crying (Hepburn, 2004). However, the interviewer responds with silence and an immediate move to a follow-up question, making no reference to the emotions reported. In sum, the reports of distress here, as in Extracts 1 and 2, are treated as part of a dispassionate account-for-the-record and not as warranting acknowledgement of the emotional content as such. In effect, the format of the exchange is like that of the medical encounter where, as Maynard and Heritage (2005) put it, practitioners ‘design their questions so as to display a neutral stance toward the lifestyle matter at issue’ (p. 433) and maintain that neutrality in the face of reports of distressing conditions.
Distress displayed in the moment
Before examining what happens in our data, it is worth noting how an expression of in-the-moment distress (e.g. crying or incipient crying, whimpering, holding the head in the hands, moaning, etc.) might be responded to when the interlocutor is free of institutional constraints. In this example, taken from Hepburn and Potter (2012), two friends are on the phone. Jill is apparently upset by something that happened at the weekend, and we join as Kerry probes the trouble: Extract 4 (from Hepburn and Potter, 2012: 201–202, notation slightly simplified): Note that the # sign in Line 6 indicates ‘creaky voice’
In Line 4, Jill’s ‘wobbly voice’ ‘no’ (especially given the preceding signs, not shown here) allows the fairly confident inference that she is currently feeling some kind of emotional upset. Note Kerry’s response. As Hepburn and Potter say, her use, in the arrowed line, of a ‘creaky’ and prosodically drawn-out change-of-state token ‘oh’ (Heritage, 1984) and a yes/no question projecting Jill’s agreement that she was ‘having a bad time’ makes us hear the response as warmly sympathetic, encouraging Jill to expand on her troubles. This, of course, is in a non-institutional conversation, between friends.
In our police data, it was apparent that the kind of visible distress that Jill was displaying in Example 4 was not common to all the interviews, despite the nature of the events that the complainant was describing (of the 19 cases, distress was only manifested unambiguously in six cases and more ambiguously in two more). However, in all these cases, there was a fairly clear pattern in the way in which the interviewing officer dealt with the complainant’s distress, and both were unlike the sympathetic response shown by Jill’s friend Kerry and much more like the institutional responses documented by Hepburn and Potter (2007) and Ruusuvuori (2007).
What the police interviewer did, roughly speaking, was to acknowledge complainants’ difficulty in continuing, and that only gave way to an explicit reference to their emotional state if the problem persisted. And in cases where distress had already been acknowledged in one of those ways, acknowledgement was withheld if the complainant expressed distress again. The analysis of examples below will flesh out the details:
1. Graded responses if complainants’ distress persist within a spate of talk
Where the complainant’s expressed distress seemed apparently to prevent a return to the action in hand (responding to questions, continuing a narrative), the interviewer’s response began with an acknowledgement of difficulty and escalated to a more empathic response – that is, one that explicitly recognised the emotions that were being displayed – until the current action was reinstated. In Example 4, the complainant has just begun her open narrative account and is quickly overwhelmed by emotion. The interviewer begins receipting this distress at Line 9: Extract 5: INT1 Bus stop. Note that the ~ sign indicates ‘wobbly voice’
Note the interviewer’s graded responses at the arrowed lines: although they acknowledge that there is something amiss, they do not at first explicitly empathise with the complainant’s distress as a felt emotion; that is to say, the first responses acknowledge that the complainant is experiencing something that is impeding the onward trajectory of the interview. As Hepburn and Potter (2007) put it in their account of how NSPCC call-takers use the phrase take your time, the phrase seems to be deployed to acknowledge, and license, a break in talk to be heard as a ‘response to sequences of crying that involve abnormally long delays combined with disrupted or incomplete TCUs [turn-constructional units]’ (p. 98). Only as the emotion continues, however, and after many non-committal acknowledgements does the officer ultimately explicitly recognise the distress with the observation that ‘it’s obviously really upsetting for you’. But note that the use of ‘obviously’ distances the officer from the implication that she was empathetically interpreting what she was seeing – it is cast as a public fact, implicitly requiring no personal investment in understanding. In Extract 6, again we see that the interviewer’s first responses to the complainant’s expressed distress are to treat it as a difficulty in proceeding: Extract 6: INT3 – Pinned me up against the wall
Although the interviewer does acknowledge the complainant’s distress, it is initially in terms of hindering her from answering the question (Lines 8 and 10), and only after continued distress does she explicitly acknowledge the memory being ‘upsetting’ (Line 31). Once again, however, this upset is ‘known’, which, like ‘obviously’ in Extract 3, implies a more abstract appreciation of what the complainant is going through, rather than full emotional empathy.
2. Distress acknowledged only as difficulty in proceeding
More often than the escalation described above, a minimal, or non-emotion-implicative, acknowledgement would not be escalated if the complainants resumed the activity that they had been pursuing before displaying distress.
In Extract 7, the complainant is being asked to give more detail about what her father did to her as a child, and the question turns to an incident in her bed: Extract 7: INT10 13.54.40 Finger
At Line 4, the complainant abandons her turn, and there are six seconds of silence. She is not audibly crying or visibly moved, nor is her facial expression (somewhat blurred by the anonymisation of the video) distinctly distressed. Nevertheless, the abrupt end of her turn, and the especially graphic nature of what she is describing, seems to suggest to the interviewer that she is indeed upset. She handles this by treating the matter as requiring effort and time – not by allusion to the complainant’s emotional state as such.
3. Distress met with minimal or no acknowledgement
The last practice of dealing with expressed distress that we identified was for the officer to treat it with no mark of empathy: to receipt it only with a minimal, or no, acknowledgement and to allow, or expect, the complainant to resume the description that they had in train.
In Extract 8, the complainant’s marked inbreath and outbreath at Lines 8 and 17 represent what on the video we would gloss as ‘stifled sobbing’, and the ‘wobbly voice’ at Line 10 indicates a display of strong emotion held in check: Extract 8: INT3 Smug look
The interviewer receipts the first of these expressions minimally (okay in Line 9) but the subsequent ones not at all. Over two seconds pass before, in Line 12, she gets ‘back on track’ (Antaki and Jahoda, 2010) by a so-formulation which deletes the complainant’s term ‘rape’ and replaces it with the more neutral ‘put his cock in you’. In dealing with the complainant’s third expression of distress at Line 17, she issues a ‘follow-up’ question (Romaniuk, 2013) in Line 18 without further ado.
One case out of the 19 was notable for the almost complete lack of recognition by the police interviewer of the distress manifested by the complainant. Extract 9 gives a sense of the way in which the complainant expressed his distress (which we can gloss as snuffling, chest heaving and marked voice quality). Only once is this even minimally acknowledged. In Extract 9, at the start of the interview, the officer is instructing the complainant about the importance of telling the truth (this is mandated by the guidelines that the officer works to, but may be exacerbated by the perceived deficits of the interviewee; see Williams (2013: 41ff) on the mistrust of reports from people with ID). At Lines 15–17, the complainant expresses what we might gloss as frustration and despair: Extract 9: INT13 16.11.30 Why me.
The officer’s ‘of course’ in Line 18 receipts the complainant’s uncompleted announcement (disfluently delivered, with much gulping and chest heaving) as if confirming something already known (as from a ‘knowledge-plus’ position, in the terms set out by Heritage, 2012a, 2012b); in this context, it comes across (together with its soft delivery and in its sequential position before the complete termination of the utterance) as reassurance at least that his self-questioning distress is understandable. Thereafter, none of this complainant’s similar displays of emotion over the next 1.5-hour interview receive even such minimal responses.
Discussion
This article sets out to examine how police officers deal with people with IDs who approach them alleging sexual assault and rape. The sample of 19 interviews we obtained included people with Down’s syndrome, autism and unspecified ‘learning disabilities’, and who were sometimes challenged by other difficulties such as deafness or physical disability. Prompted by the finding that such vulnerable victims of sexual crime are likely to suffer marked psychological distress (Rowsell et al., 2013: 257), and conscious that other researchers had found low levels of empathy in environments such as medical consultation (Heritage and Lindström, 2012; Ruusuvuori, 2005, 2007) and – most closely to our situation – in calls to a child-abuse helpline (Hepburn and Potter, 2007), our interest was in seeing how the police officers dealt with the distress that might arise in the interview that they must necessarily put the complainant through.
A null finding worth mentioning is that there was nothing in the data to suggest that the interviewees were distressed as a consequence of any cognitive difficulties they experienced in either understanding the interviewer’s questions or in articulating a response to them. This is perhaps testament to the care with which the interviewers approached their task and designed their questions. But most interviewees reported having felt very negative emotions (fear, disgust and pain) at the time of the alleged events, or reported having strong feelings about the events or the alleged perpetrator currently (unclean-ness and hatred), and some displayed visible distress in the telling.
Our first finding is that complainants’ reports of distress were routinely not topicalised, and sometimes indeed not explicitly acknowledged. Extracts 1–3 were examples of cases in which the complainant reported what they felt at the time or what they currently felt in the sense of a general assessment of their stance towards the event or the perpetrator. Neither of these classes of report prompted a response from the interviewer. Our analysis, following the distinction made by Lindström and Sorjonen (2013), is that the officer was aligning with the complainant’s action in storytelling, while withholding affiliation to the stance that the story was expressing. This is consistent with other institutional interviews where clients may report troubles; indeed, Ruusuvuori (2005) reports that the most usual reception of the patients’ troubles tellings was a minimal acknowledgment, silence, and/or a continuation of the task-related activity at hand, such as interviewing the patient (40% of the responses in homeopathy, 55% in general practice consultations). (p. 208)
What, however, about distress expressed outright? Our main finding was that the interviewer had three ways of handling it: it could be acknowledged minimally (with okay) or not at all; it could be treated as a temporary difficulty, either of effort or memory, or without specification; or, more rarely, it could be recognised for its emotional charge, as Hepburn and Potter (2007) found in analysing calls to a children’s helpline. But even in this last receipt, the few occasions on which it occurred showed that the interviewer was careful to acknowledge the emotion, as it were, intellectually, not empathically (e.g. by such receipts as I know it’s really difficult or you’re doin’ ever so well). And in one 1.5-hour-long interview with a man alleging prison rape, the interviewer gave only one, very minimal, response of any kind.
If there is, as Stevanovic and Peräkylä (2014) claim, an ‘emotional order’ in talk, which requires a speaker to display a recognition of the emotional stance being expressed by their interlocutor – the kind of response shown by Kerry to her friend Jill in the telephone conversation in Example 4 – then it is abeyance here. The explanation for this pattern of dealing with distress is probably to be found in the dual, and to some degree conflicting, institutional demands of the interviewing officers’ job. Their guidelines require not only that they establish rapport with the interviewee, but also that their conduct be seen to be neutral and non-leading – to be, in the words of ABE, ‘acceptable to the court’. Although our sample of complainants were people with IDs, and the findings are possibly exaggerated for that reason, it might well be that this is a general pattern in police interviews with anyone alleging assault; certainly, it squares with the critical discourse analysis reported by MacLeod (2010).
Consulting with the police officers with whom we were in contact, we formed the strong impression that for evidence to ‘be acceptable to the court’ effectively meant, among other things (such as its audibility and coherence), being seen to be solicited impartially and unjudgementally. It is here perhaps that the complainant’s identity as having an ID comes most strongly into play, as the officers strive to avoid any hint that they are leading someone who the defence counsel may argue will be likely to be suggestible. Part of the interviewer’s job is to forestall any accusation by the defence counsel that the interviewer had, by being empathic, encouraged a heightened or exaggerated account. This extra contingency, over and above the constraints on practitioners such as medical personnel, health workers and helpline call-takers, makes it very difficult for the police officer to do more than – at best – acknowledge that the complainant is experiencing something that is interfering with the telling of their story; it favours, as it institutionally is obliged to, forensic probity over empathy.
Footnotes
Appendix: Transcription Symbols
Acknowledgements
We are grateful for the help and advice of personnel in the regional police force with whom we collaborated, who must remain anonymous. We would also like to thank Chris Pudlinski and Val Williams for insightful comments on a draft of this article.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.