The genesis of a witness statement

The genesis of a witness statement. Frances Rock, The University of Birmingham f.e.rock@bham.ac.uk Abstract This paper steps back from the plethora of learned articles on simulated interviews with witnesses, the success of interview techniques and cognitive loads on interviewees and interviewers, it reports a detailed examination of the way witness statements are taken, from the first verbal account given by a witness to the final written statement penned by their interviewer. The paper examines a statement-taking session and the resulting statement. It presents examples to illustrate which aspects of the witness’s account are changed during the statement-taking session and how. That is, in what ways, and through what processes does the original version provided by the witness change through the subsequent renderings during the statement-taking session and in the final statement text. There are several reasons for doing this. This enables us firstly to understand a little more about what a witness statement is, secondly to understand how witness statements become what they are and thirdly, in what ways witness statements might not be what they at first appear to be. Keywords: Witness statements, versions, transformations, statement-taking, police language 1 Introduction It is often assumed by lay people, and asserted by police, that witness statements are verbatim accounts of what witnesses have said. Heaton-Armstrong (1995:138) describes this as a “fiction” and field observation suggests that much more goes on during witness statement-taking besides, and instead of, dictation. Labov and Fanshel propose, “we may define an interview as a speech event in which one person, A, extracts information from another person, B, which was contained in B’s biography” (1977:30). This certainly happens in witness interviews; however, it is by no means the only thing that happens. We need to extend the definition for the peculiar case of witness ‘interviews’, in order to take account of the fact that they are not simply about extracting information, but also about working with that information, transforming it, in order to jointly author a written text. The witness interview is therefore, more than, or different from, Labov and Fanshel’s interview; as a speech event, it consists of multiple tasks - telling, listening, writing, formulating, analysing - and has multiple goals - the extraction, communication and use of emotional and factual information. The situation is further complicated by the fact that these tasks may be collapsed, the witness may simultaneously give information and make authorial decisions, whilst the interviewer may simultaneously extract and record information and may, in recording, be doing several things some of which may be much less neutral than the verbal process ‘record’ suggests. Applied linguists who examine legal texts have, for some time, been keen to discover what characterises statements from detainees or witnesses made to and recorded by police officers. This issue is typically raised when a detainee or witness disputes the statement that they have supposedly made, claiming either ‘I didn’t say that’ or ‘I said that but not in that way’ (Coulthard, 2000:419). Linguistic investigations have responded to these claims variously, for example, by scrutinising disputed statements for features which appear uncharacteristic of the linguistic repertoire of the detainee or witness and would therefore not be expected to occur in the statement (Eades, 1993) or for features which are apparently typical of someone other than the witness, including items of policespeak (Fox, 1993) such as postposed then (Coulthard, 1993, 88-9) or features which may be characteristic of an investigating officer (Shuy, 1998:98-100). Less commonly, statements have been analysed for features which seem generically atypical. This paper addresses this issue by offering a preliminary examination of what a witness statement is. It views the statement as being the product of a statement-taking session or interview; as the sea mills rocks into stones and stones into sand, so the session mills text proposed during its duration, into the final statement text. The resulting statement is deeply intertextual, it does not just parts of previous text(s), but rather is moulded through those previous text(s). In this paper I will focus on one case-study, tracing themes raised during a statement-taking session and analysing their ultimate realisation in the statement, to see what the approach might yield. This study is part of a larger project on witness statement-taking which draws on over 20 statements taken by different interviewers over a 2-year period. This research is motivated by the notion that the study of drafting and revising within statement-taking sessions illuminates statements, the written product of the previous verbal interaction which reveals traces of the statement’s origin. 2 Contextualising the data Witness interviewing does not occur in a vacuum. This study does not claim that the interview is an isolated activity which will, through examination, reveal something only about language. On the contrary, witness interviews as a group of related speech activities are influenced by legal procedure and police training and, by examining any interview and the resulting statement, we are also examining these influences and their appropriation within statement-taking sessions and beyond. Statement-taking is an undeniably complex task. Scholarly attention to interview technique highlights this and has focussed attention on training, leading to various initiatives (Fisher and Geisleman 1989, Clifford and George, 1996:232, Shepherd, 1988). In England and Wales, an instructive written guide and week-long course, aimed at improving officers’ general interviewing skills have been introduced. The role of the interviewer is presented, in these materials, as centring on verbal skills, particularly elicitation and attention to interactional structure (1998:25-71). Writing skills, such as transcribing a jointly authored account or recording speech, are neglected. These writing tasks may be unfamiliar even to experienced officers and must be undertaken during the, already complex, interview situation, making statement-taking not only a complex task but a complex literacy event (Barton, 1994:26). The speaking-writing imbalance is being addressed in training for senior investigative officers. Here, the importance of the witness’s contribution is stressed, “verbatim accounts” and use of witness’ “own words where possible” are encouraged. Such training might mark a transition from viewing statement-taking sessions as simply a means to an end to viewing them as important in their own right. Analysis of the relationship between statement-taking sessions and statements and also of the role of writing could interact with such training initiatives. Later in the criminal justice system, the decision to see witness statements as a certain type of text, a complete, true record of sensory input to a particular individual, (the witness), at a specific time or times (during an event defined as a crime or other related events), dictates the way statements are taken-up and used within that process. This stipulates what the texts which emerge from statement-taking sessions are, and how exactly they will be taken to construct a representation of lived reality, how they will, in turn, constitute the institutional order. Consider how witness statements are used within the criminal justice system. The written statement is seen to ‘stand in’ for the witness, to speak when they are not present throughout investigations (Milne and Shaw, 1999:128) and even at trial (Heaton-Armstrong and Wolchover, 1992:162). By affording the statement such primacy the criminal justice system comments on the witness, the statement and the relationship between both and the legal world. As Smith observes, “the appearance of meaning as a text, that is, in permanent material form, detaches meaning from the lived processes of its transitory construction, made and remade at each moment of people’s talk”. (1990:210–211). Statements must, in detaching meaning, do so wisely, so that when they are used they represent the witness by accurately speaking on their behalf. The true power of the written text becomes fully apparent when we consider the way the criminal justice system comes to prefer the information contained in the statement over the information provided by the witness, when at trial a statement furnishes “a narrative from which, months later, witnesses can refresh their memories… [and] affords a text against which their consistency can be checked” (Heaton-Armstrong and Wolchover. 1992:162). Leaving aside the circularity of this manner of checking witnesses’ consistency, this gives us a sense of the sway that the written text comes to hold. The statement is not used simply to support, confirm or even explore the witness’ presentation of events, rather it becomes an authoritative text against which even the witness, the original source of the information contained in the statement, is assessed. For example, Heaton-Armstrong, a practising barrister, describes the “typical courtroom scenario” in which a witness is asked whether they signed a declaration that their statement was true and correct and is then asked “to explain the inconsistency between oral evidence and statement”. In such instances the reliability of the witness and the accuracy of the statement are put into a strange relationship. In the midst of this, enthusiasm for the written document is unlikely to be tempered by consideration of the possibility that the statement does not neatly stand for the witness’ words, that it does not neatly encapsulate the context of its production for all time. The text becomes extremely powerful, a phenomenon not exclusive to legal texts: The vesting of meaning in … permanent or semi-permanent forms is routine and commonplace, and has transformed our relation to language, meaning and each other. Texts speak in the absence of speakers; meaning is detached from local contexts of interpretation. … In the distinctive formation of social organisation mediated by texts, their capacity to transcend the essentially transitory character of social processes and to remain uniform across separate and diverse local settings is key to their peculiar force. Smith (1990:210–211) Witness statements are taken within specific contexts and using specific processes of production. For example, when statements are taken, certain techniques may be used, such as cognitive interviewing and certain social contexts pertain, such as pre-ordained power relationships. When statements are used, however these contexts and processes of production are rarely mentioned, for example juries are not typically urged to consider what questions might have been asked by a police officer during the taking of a statement. Disquiet about, and abuse of, statement-taking and statement-use led the Anglo-Welsh legal system to introduce audio- and video-recording of interviews with suspects and child witnesses/victims between 1986 and 1991 (Baldwin and Bedward, 1991:671, Police and Criminal Evidence Act, 1984). This has proved to be a valuable and successful safeguard for both police and suspects (Heaton-Armstrong and Wolchover, 1992:160), yet, even so statements with witnesses other than children are still rarely recorded. Written information, filtered from witnesses is given supremacy above spoken information directly from witnesses themselves. However, procedure may change in due course as concern about the standards of witness statements has led to experimental audio-recording initiatives throughout Britain (Milne and Shaw, 1999:130). In addition practical justifications for audio- or video-recording witness interviews have also emerged. An inquiry into shortcomings in the investigation of the racist murder of the London teenager, Stephen Lawrence, in April 1993, highlighted the possibility that interviewers may miss important facts which did not appear important at the time of initial interviews, but turned-out later to be crucial (Macphearson, Cook, Sentamu, and Stone, 1999: para 13.14). Critics of the current paper-and-pen system, in which interviewers usually scribe for interviewees, have called for universal audio-recording of witness interviews. This interest in, and attention to, the recording of statement-taking sessions is, slowly, shifting investigative focus away from the statement as an authoritative text and onto the interview as an interaction which may include details, and details about details, omitted from the statement itself. Indeed, police forces have become so concerned that statements are not taken in sufficient detail, that some forces now encourage the audio-recording of all statements (National Crime Faculty, 1998:160), especially in more serious cases. This has prompted the development of a special class of witness, the significant witness who is always to be recorded and interviewed by a trained officer. 3 Data overview The texts used here were collected from an English police force, in 2000. We will look at one case study using data from two texts: an audio-recording of a single statement-taking session in which a witness describes his understanding of events surrounding a suspicious death; a hand-written statement, composed during the statement-taking session. The investigation of this death had treated it as a possible murder, so, each witness to this crime was interviewed as a significant witness. The analysis sets show what would have been lost without audio-recording. I myself did not make the audio-recording of this statement as I did not attend this interview due to confidentiality. The audio-recording I have analysed is a copy of the police recording, the official record, with personal details removed. The police recordings are otherwise un-broken, made on previously sealed tapes which include a constant audible counter to demonstrate that they are complete and uninterrupted. The events described in the session and in the resulting statement took place over two days during which the witness, victim and friends met at various locations, walking from place to place. Key participants are described in figure 1 in order to orient the reader to the characters who will be mentioned subsequently. All names, locations and dates have been anonymised throughout. Figure 2 also provides orientation, by labelling two meetings described by the witness, and by summarising the location of each, the participants involved and the events which took place at them. I have used “meeting” to conveniently and consistently label the events (it is not a word used by the interviewee or interviewer), however the gatherings were informal and certainly not business meetings, their content and timing were not, apparently, planned. Also, the participants did not meet at them, mostly they arrived at and departed from the meetings collectively or at least with one or two other participants. The witness described further meetings but they are not discussed here. 4 Structure and composition of the session Shepherd points out that “interviews often appear to meander in terms of talking turn, structure and topic progression” (1999:139). This was certainly true of the session I examined; on first listening, it appeared a topical jumble. On closer examination, however, something relatively systematic emerged. Heaton-Armstrong ruminates that perhaps “statement recording interviews are preceded by a preliminary chat, or several chats, during which a police officer may endeavour to marshall [sic] a disorganised narrative into more coherent form” (1995:138). The data here includes these very chats and supports Heaton-Armstrong’s hunch in that, unsurprisingly to police officers perhaps, the statement-taking session divides into four extended formulations, tellings or versions, in each the interviewer and witness return to the crime narrative and re-work it. Each reworking is easily identifiable as each is surrounded and demarcated by orientational talk, for example a request for an account precedes the 1st version and the following, from the interviewer, appeared between versions 3 and 4: yeh (.) OK (.) we’re now gonna I’ve just we’ve just gone through one tape (.) where I’ve recorded down everything you’ve told me … in the form of notes hhh I’m gonna now (.) write that out into statement form (.) and at the end I’ll invite you to sign it; The four re-workings each have some key similarities: Each re-working consists of a narrative of the incident with more or fewer interruptions and wider or narrower scope (e.g. some versions narrate events up to the day of statement-taking and others only those on the day of the murder); Each re-working proceeds rather differently, in some the witness talks most and the interviewer rarely interrupts whilst in others the interviewer dominates; Finally, each re-working is functionally different from the others, in some the witness is encouraged to talk whilst others centre on joint text production and result in a written text (notes or the final statement). In the interview devoted to collecting this statement this statement this ‘versioning’ was realised in the structure outlined in figure 3: Therefore each version uses written texts differently, in some the written text helps to structure the verbal interaction and in others the creation of the written text almost becomes a participant, taking turns in talk Komter (forthcoming). Note that versions presented by the witness in the statement-taking session are almost certainly not the only versions the witness has produced or will produce (Heaton-Armstrong and Wolchover, 1992:165). It would be extremely interesting to follow a witness through various tellings of the crime, to friends, family, work colleagues and, if they consulted police more than once, versions to the police. Although I describe the first version of the crime presented on the tape as ‘version 1’, it is clear, even from the content of the statement that this is not the first time this witness has discussed the crime, even if it is the first time he has related events to this particular interlocutor. 5 Tracing themes I will now explore the progress of a number of themes through the statement-taking session into the statement. The main reason for examining this progress is that it enables us to discover whether the witness’s words and ideas have been incorporated into, or rejected from, the statement and if so to what extent, how and, possibly, why. Textual analysis itself cannot address the question of whether and to what extent any witness statement should incorporate the witness’s words and ideas. That is, nonetheless a vital question, addressed towards the end of this paper. I have chosen to trace themes connected to temporal reference, because the witness does not refer to time or duration in his initial, spontaneous version at all, yet preliminary statistical analysis suggests that 6% of information in the final statement references time. We will look at 6 excerpts from the final statement which are reproduced below (figure 4), in the order in which they occurred in the statement. Whenever the statement or part of the statement is reproduced, errors and omissions are as in the original except capitalisations, which have been corrected to ease reading. Temporal references which will be discussed are underlined: Of these 6 references, 4 are preceded by the uncertainty marker “about” (numbers i, ii, iv and v) and 1, number iii, could actually be described as a piece of anti-information because it tells us about something that the witness does not know rather than something the witness knows. “at first” which follows temporal reference iii is also a reference to time but it is used here to describe the sequence of events, not their time or duration, although it clearly implicitly connects with temporal reference iii. We will attend to four questions which will help us to investigate why temporal information has been included in the statement in the way in which it has: How are the timings themselves arrived at, for example why 3pm not 3.30pm? The most obvious reason would be that the witness suggested these times during the session. Why are the 4 approximations, mentioned above as being flagged by “about”, approximations. If the witness was sufficiently aware of time to refer to time why did he not do so accurately? How are the points at which and about which temporal reference is provided, arrived at. For example, when discussing the fight, why is the duration of “pushing” supplied but not the duration of “messing around”? Why is the anti-information, in reference iii, included? Why was the witness unable to say what time he arrived at location 2, despite having estimated the time that he left location 1 and saying that he went straight there? Also, how do the interviewer and interviewee arrive at the rather clumsy formulation in iii? We will begin by looking at the witness’ first version, in the hope that by finding out how the witness referred to time in his first account we might be able to answer some of these questions. Version 1, excerpt 1 started having a laugh and that (.) having some more drink you know (.) and everything just s- started turning around (.) Version 1, excerpt 2 then they (.) went in the house (.) started having some drink and th- (.) then um he looked round and saw saw the ur st- the man and Kigs messing about (.) pushing each other (.) and then (.) the man the ur the man must have got fed up of it It is immediately obvious that, although the witness is describing the events which are presented in the final statement, this first account is managed rather differently. As I mentioned, there is no reference to time here at all, in fact, the witness barely mentions arriving at the locations. If the reference to time which appears in the final statement is not initially made when the witness narrates the incident as he saw it, where does it come from? The answer lies in the intervening versions, 2 and 3. We will therefore examine the origins of the 6 references to time, one by one. 5.1 Time 1 – Time of arrival at Jerry’s and the duration of time spent there To recap, this appeared in the final statement as: On the 5th [month name] 2000 at about 3pm I was a friends house called Jerry It occurred prior to that, during the statement-taking session, in the following ways: From version 2: 1 I you was drinking spirits? 2 W yeah 3 I (2.0) ok and at- what time did you go there? 4 5 W that was about phhh not quite sure on the time but it was quite early (.) when I went to the house 6 I quite early (.) how [ is- how early? 7 W [ say about (.) say about phhh probably after 3 8 I after 3pm yeh? 9 W yeah The interviewer (I) and witness (W) begin discussing timing in line 3 when the interviewer initiates the topic by requesting information, saying “what time did you go there”. Immediately, the witness marks uncertainty, hedging with the words marked in bold “about” and “quite early”. In line 6, the interviewer notes the uncertain formulation by repeating it and then specifically asks for clarification. The witness apparently provides that clarification by proposing a time for his arrival at Jerry’s but in fact his estimate is vague in that he prefaces it with “probably”. Furthermore, the witness fails to suggest a time when he arrived but merely says that it was “after 3” leaving the actual time vague. In lines 7, 8 and 9 the interviewer takes up the most detailed formulation which has been provided, complete with uncertainty markers, and it appears that the witness and interviewer have agreed on a time. Later in the same version, however, the pair return to this issue as if it had not been resolved: From version 2: 1 2 I ok when you say hh you got you at at 2 o’clock you were with Jez and you left Jez’s after about (.) 3 W I’m not really [sure on the times 4 I [say about 2 hours Here, the notion that the witness was with Jez, at his house at 2pm is introduced. We must note that the witness has not mentioned 2pm at all, certainly not in connection with Jerry, it is the interviewer who introduces 2pm here (line 1). However, the witness does not appear to notice either of these things. When he speaks in line 3, he does not comment on the change from 3pm, his proposal, to 2pm, proposed by the interviewer, and instead takes the reference to time as an opportunity to highlight his uncertainty about “the times” in general. The interviewer interrupts the witness’ generalised protestation and continues to pursue the task he began in line 1, proposing a version of events which was not introduced by the witness. The next reference to arrival at Jerry’s and the duration of the visit occurs in version 3. Here, the interviewer proposes to re-cap all of the witness’ contributions. He begins, in line 2, with reference to 3pm, the time originally proposed, earlier: From version 3: I 1 2 just- just to run through (.) everything that you’ve told me (.) alright (.) bout 3 o’clock on Wednesday is that right W 3 yeh yeh I 4 (3.4) you were with a friend called Jez who lives off X Road Interestingly, this is how the interviewer begins the ‘meat’ of the statement. He foregrounds temporal reference by presenting it before details about participants or location, information which the witness selected when he began his version. The interviewer selects “bout 3 o’clock” as the time when the witness was with Jez he does not mention 2pm and, at this stage, does not specify where the witness was when he was with Jerry. 5.2 Time 2 – Leaving Jerry’s Next we will examine the discussion which led to the following being included in the statement: At about 6pm or earlier we left Jerry house leaving Jerry behind. The first excerpt, which relates to this, is from version 2: 1 I so you were all there (.) at what time did you leave 2 3 W we left about (.) we was there for about (.) say about 3 hours I think I’m not sure (.) bout 3 hours 2 hours 4 I 'til about 6 o’clock do you think 5 W yeah around 6 o’[ clock yeh 6 I [(( would that be alright)) 7 W yeh In line 1, the interviewer asks the witness what time he left Jerry’s house. The witness does not answer by providing that information, perhaps he cannot remember what time he plumped for, earlier in the session, as his time of arrival there. In any case, he draws instead, on a reflexive strategy which may refresh his memory. He tries to reconstruct events by remembering how long he was at the house. This is interesting in its own right and also because it is a strategy used by the interviewer as we will see later and encouraged by cognitive interview training (Fisher and Geiselman, 1992). The interviewer then returns to the original question (line 4) by converting the duration provided by the witness into time, based on the 3pm arrival time discussed in the previous section, Time 1. Again, the interviewee exhibits uncertainty and the two negotiate a mutually satisfactory conclusion, with the interviewer asking “do you think” (line 4) and “would that be alright” (line 6). They return to this later in the same version: From version 2: I 1 2 you were with Jez Terry Kevin and another guy that you can’t remember his name = W 3 = yeh I 4 (4.5) ok (.) so you all left about 6 o’clock and you went to this guy’s house W 5 6 yeh (.) I’m not sure if it was about 6 it’s probably ups- earlier about more earlier than that (.)[not I’m not sure of the time] we ((went through ‘til)) 7 [earlier than that you think ] I 8 (.) ok (.) how many of you went Here, the interviewer re-presents information discussed in the previous excerpt. He contextualises it amongst details of participants and locations. Again, the witness responds (lines 5 and 6) by expressing uncertainty and discomfort with the decision which they made earlier, the inclusion of 6pm. The interviewer’s response to these qualms is fascinating. He interrupts the interviewee (line 7), perhaps supportively, but then moves on to a different topic without fully resolving this (line 8). Here, we do not see any attempt to negotiate an agreed time or to help the witness to remember the time. It appears in the next excerpt, however, that the interviewer has noted the witness’ concern that the estimate of 6pm might be incorrect. This excerpt is from version 3, where the interviewer takes up the idea of “6 o’clock or earlier”: From version 3: 1 I you were all there the mood was great yeh 2 W ye[h 3 I [and you left there (.) about 6 o’clock or earlier 4 W Yeh 5 I and then you went to the man’s house This is the formulation which appears in the statement, but in view of the way it was discovered we might want to question its usefulness as evidence and a potential investigative foothold. 5.3 Time 3 – Arriving at the victim’s house and the duration of time spent there Next, we move to the proposed arrival time at the subsequent location, meeting 2. This is presented in the final statement using the following: We went directly to this mans house which is on X rd almost opposite Y rd. On the left. At this time, I’m not sure what time it was, but I can say that I had not seen this man before The inclusion of “directly” is intriguing. The witness and interviewer have extensively discussed the journey between meetings 1 and 2 and the witness has talked fluently about the route taken and events which occurred en route. Given the witness’ familiarity with this journey and the fact that the group travelled “directly” from meeting 1 to 2, it seems a little strange that a statement from a witness who specified the time he left meeting 1 would not be able even to estimate the time he arrived at meeting 2. In investigating how this section of the statement came about then, we must reflect on possible reasons that a time was provided for leaving meeting 1, despite the uncertainty in the previous 3 excerpts yet time was omitted here. This excerpt is also interesting in that it has an extremely dialogic feel. We could imagine the anti-information “I’m not sure what time it was” as the answer to a question and the “but” and associated clause which follow as a mitigating conversational turn. If these features had been presented in a disputed statement, they would almost certainly have caught the attention of an analyst who was looking for traces of dialogicity which might reveal an attempt to disguise a dialogue as a monologue. The crossing-out of “not” might also have attracted anyone lucky enough to have access to the original hand-written document, and rightly so as this is, in fact, the result of discussion between the interviewer and witness after the interviewer had initially misunderstood the witness’ description of his previous relationship with the man. Investigating the origins of this strange temporal formulation therefore enables us to consider text which appears to be dialogic as perhaps signifying something even more complex than we might imagine. From version 2: 1 I what time do you think you got to the man’s house 2 W ur phhh (.) I’m not sure 3 I not sure 4 5 W it was in between about 3 3 o’clock 3 3 about 3 to 4 o’clock something like that 6 I about 3 to 4 7 W yeh Remember that in the previous excerpt from version 3, the interviewer and witness agreed on “about 6 o’clock or earlier” as being a suitable way to express the likely time that the group left meeting 1 yet here, in version 2, which of course occurred in the session before version 3 the witness proposes a time 2 or 3 hours earlier than that as being the likely time that they arrived at meeting 2. Of course, the group could not have arrived at the second location before they had left the first, so it is intriguing to note this great discrepancy and to consider how different the statement might have been had this been pursued. The other important question which this excerpt raises is the following: if the witness proposed a time for arrival at meeting 2 (3 to 4 o’clock which he says twice in the excerpt above) how is it that this is not incorporated in the statement given that the times he has proposed elsewhere, perhaps with less conviction, were incorporated. This is answered by the next excerpt, from version 3: 1 2 3 4 I yeh (.) everyone was cheerful when we walked down the road and the girls were about 16 years old (.) we went directly to the man’s house (.) you’re not sure of the time (.) but you’d describe the house as being al- a little bit opposite 5 W yeah By the time version 3 is being produced the police officer has perhaps forgotten the time-estimates which were volunteered in version 2. He certainly appears to have decided against including time references in this section of the statement and instead his assertion in line 3 “you’re not sure of the time” enters the statement, with a change of pronoun, becoming “I’m not sure what time it was”, as we saw at the beginning of this section. 5.4 Time 4 – Duration of time spent at the victim’s house before “messing about” The witness presents a complex and decidedly jumbled sequence of events, as having taken place at the murder scene. These include pushing, shouting, arguing and attempted strangulation. In dealing with this account, the interviewer gives some support to Heaton-Armstrong’s inkling that officers marshal disorganised narratives (1995:138) although, in this data, it appears that the marshalling is accomplished collaboratively. During the session, the interviewer and witness delineate a sequence for the events, teasing out their chronology and interrelationship and formulating an order and categorisation of events at meeting 2: Before the fight Phase 1 – Messing about / around Phase 2 – Pushing Phase 3 – Arguing Phase 4 – The change Phase 5 – The attempted strangulation Phase 6 – After the attempted strangulation Leaving the flat A vital part of the negotiation of the murder location event sequence involves discussing the duration of these phases. Duration and timing of events at the murder scene make their way into the statement three times: first, in describing the length of time the group was at the man’s house before the activity which the witness describes as “messing about” began; secondly, specifying a duration for an activity which the witness describes as “pushing” and thirdly, specifying a duration for an argument. We will look first at the beginning of the second meeting, the things that happened before “messing about”. In the final statement, this is referred to as follows: Everything was ok at first the man was asking if I was OK and we were all laughing and joking About ½ hour later Kevin and Terry and the man were messing around When the interviewer first requests information on the time spent at the man’s house, before the fight began, the following exchange ensues: From version 2: 1 W I’m not really sure on the time ((too tough)) 2 I (4.5) would you say it was evening time 3 W it was about evening time yeh 4 I do you was there a TV on was there anything on the [television 5 6 W [nah there weren’t no there weren’t no TV on 7 I there weren’t 8 W No 9 I ok [((inaud, 3 syllables)) 10 W [there was a telly in there though but there weren’t no TV on there 11 I ok but there wasn’t anything to (.) do you wear a watch or anything like that 12 W nah 13 I was there anything on that really could assist you with the time or 14 15 W nah I weren’t just look- I weren’t even look- I weren’t looking round just talking to the girls and that Earlier, we saw the witness use the strategy of trying to invoke the duration of events in order to remember precise times. Here, the officer suggests alternative remembering strategies: using consensual definitions, in this case the definition of “evening time” (lines 2 and 3); using information about surroundings and incidental information, television (lines 4 – 10) and a watch (lines 11 – 12). Finally, the interviewer appeals to the witness to attempt to prompt his own memory (line 13). As exemplified here, “discourse technologies involve simulation: in particular, the simulation for strategic and instrumental purposes of interpersonal meanings and discursive practices” (Fairclough, 1992:215-6). The interviewer simulates concern, presenting himself as caring, helpful and supportive, as someone who would help the witness to remember the facts he is groping for. However, the witness does not seem to buy into this simulation, rather than using the remembering strategies which the officer suggests, he side-steps them, becoming defensively distracted by details of a television at the murder scene and re-stating his own role in proceedings rather than attending to his failure to estimate time. In version 3, the interviewer returns to this unresolved matter: From version 3: 1 I how long were you there for before they started messing around 2 W say about (4.3) good half hour a good half an hour 3 I (3.8) ok they’d started messing around Here, the witness’ response is strikingly different from that in version 2. Rather than resisting the question, the witness immediately proposes an approximate duration (line 2). This second request for information, and the obliging response, occurred some way into the session and whilst it is impossible to say why the witness’ stance changed so dramatically, it is interesting to note the change as it demonstrates the extent to which asking for the same information at different stages of an interaction can elicit wildly altered responses. These two excerpts raise the issue of possible shortcomings of the accepted reporting style of witness statements. Witness statements are, as we have seen, recorded in the first person. The interviewer scribes on the witness’ behalf and does not mediate or evaluate their words in a way that is external to the text, in that it is not visible on the text’s surface. The statement-taker mediates and evaluates by deciding what to include, what to exclude and what precise formulations to use on the basis of what the witness has told them, but their evaluation is a part of text-production processes, not a part of the text itself. This means that evaluation is invisible, the interviewer cannot comment on, or acknowledge, it and those who read the statement cannot access it. Remember that in the statement the two sets of exchanges reproduced above, one in which the witness was completely unable to recall the duration of ‘messing about’ and one in which he easily supplied a duration, became: Everything was ok at first the man was asking if I was OK and we were all laughing and joking About ½ hour later Kevin and Terry and the man were messing around How different this summary would have been if it had been accompanied by a note, mentioning its origins, or if it had been narrated in the third person, with some recognition of the dialogue which spawned it. Use of a reporting style which acknowledges the textual processes which are used in producing witness statements would be more honest, it would enable interviewers to step outside their positioning as mouthpiece and act as scribe, more realistically. It would also allow more completeness than the current style, uncertainty and indeed certainty could be noted and evaluated. Here, for example, the witness sounds relatively certain of the duration spent at the man’s house before fighting began. This certainty is manifest in that he uses the formulation “good half hour”, suggesting that he is certain of having spent at least half an hour, and reiterates, “a good half an hour”, perhaps to add emphasis, yet this certainty is not apparent in the statement. Instead, in the statement “½ hour” is prefaced by “about”, “about” has been used to preface many of the times presented in the statement and it is used here too, despite the witnesses relative certainty in this instance. Conventionally, “a good half hour” is taken to mean something like “at least half an hour, possibly longer” yet this certainty is also omitted from the statement. If the officer had been provided with a voice of his own in the statement, rather than simply being lent the voice of the witness, he could have conveyed these finer points, had he felt that they were important. Alternatively and ideally, if the witness had been truly provided with his own voice these things could be clarified. 5.5 Time 5 – Duration of time spent “pushing” Time was also quantified in the final statement in relation to the duration of “pushing”: they began to push each other for about 3 minutes I was looking back and forward while talking to the girls and I heard them begin to argue From version 2: 1 2 W um (.) Kegs was pushing the m- the man and I was pushing Terry and just st- going on like that 3 I (.) how long did that go on for 4 W say about phhh bout 3 minutes 5 I what were you doing while this was going on 6 7 W I was just looking back and forward to see (.) if they ((were)) gonna start fighting (2.5) I was talking to the girls at the same time From version 3: 1 2 3 4 I (.) Kegs was pushing the man and the man was pushing (.) Terry and this went on for about 3 minutes [sniffs] you were looking backwards and forwards and talking to the girls hhh when they began to argue could you hear what they were arguing about = 5 W = nah I weren’t really (.) listening Again, we see expression of some uncertainty when the witness first suggests “3 minutes” as the possible duration of pushing (line 4, version 2). As in some of the previous examples, this time estimate is taken through to the final statement via version 3 with uncertainty maintained using “about”. Interestingly, the witness declares involvement in this phase of fighting (version 2, line 1), lack of interest in the fight (version 3, line 5) and involvement in a different activity, talking (version 2, line 7). Contradictions in these assertions are not investigated. Furthermore if the witness was involved in various activities while the crime was going on, that may have implications for his ability to accurately assess the passing of time, not to mention implications about his involvement in the crime, but this is not noted in the final statement. 5.6 Time 6 – Duration of time spent “arguing” In version 3, the following exchange takes place: 1 I was he still in the chair at this time 2 W (.) yeh 3 I yeh (.) and you say he was proper drunk 4 W Yeh 5 I (.) ok (.) y- you said some was it how long (.) before the man flipped 6 7 W say about phhh (.) not quite sure because they was quite- arguing for quite a long time 8 I (10.2) ok and suddenly the man flips In line 5, the interviewer asks how long the victim and alleged assailant were arguing before ‘phase 4’, the change of mood, occurred and the fight began. At this stage, the witness is unable to supply a time and instead suggests that the argument went on for “quite a long time”. Unsurprisingly, this is not incorporated in the statement. However, in version 4, when the statement-taker is drafting the final text of the statement aloud for the witness they return to discussion of the duration of this phase: From version 4 – Drafting the statement: 1 2 I ok (10.4) so when you say he wa- he continued to argue a bit is that when he was shouting (.) urm (.) you bastards I’ll knock you out 3 W Yeh 4 I (9.2) who was he saying that to 5 W to Kegs 6 I (11.8) he said that to Kevin (2.9) this went on for how long 7 W about not actsh- not actually sure how long it went (.) on for 8 I how long do you think 9 W say about (.) say about 10 minutes 10 I (4.2) this went on for about 10 minutes (.) suddenly the man flip- flipped In line 7, the witness maintains that he is “not actually sure” how long the argument lasted, but the question in line 8, “how long do you think” persuades him to give his opinion “about 10 minutes”. This is incorporated into the final statement rather differently: From the final statement: I was looking back and forward while talking to the girls and I heard them begin to argue I don’t know what about but it was loud, I said don’t start arguing in the mans house and they calmed down, the man continued to argue a bit saying “you bastards I’ll knock you out” he was saying that to Kevin [surname]. This went on for 10 minutes. Suddenly the man flipped In this excerpt, in contrast to some of the earlier examples, the interviewer does not note the witness’ uncertainty about the timing of the activity at all in the final statement. He has removed the hedge “about” despite the fact that the witness used it and that the witness only supplied this duration when asked for his opinion. 6 Interference or translation? Several authors comment on potential interviewer interference with information supplied by witnesses during statement-taking (Heaton-Armstrong and Wolchover, 1992:160 and 163, Shepherd and Milne, 1999:131). However, frequently expressed fears that degree of formality and techniques of statement-taking might distort the statement are somewhat mono-dimensional. To introduce further debate about this question, I propose 3 stances on what a witness statement should aim to do and thus how it should be composed, in terms of compilation, organisation and content. Advocates of each stance might claim: I will now expand these basic positions and will then discuss them and possible compromises between them: A. A witness has seen the events, people, locations, items and so on that they have been asked to describe in their statement. They are therefore the only possible source of information on their perspective on those things. A witness statement should encapsulate the ideas that the witness presented in their first narration of the crime events, no additional ideas should be included and none removed. The statement should be recorded in a witness’ own words and would, in some senses be akin to a deposition. B. A witness has seen the events, people locations, items and so on that they have been asked to describe in their statement. They are, however, unaware of the investigative importance of some items of information. The statement should therefore be based on what the witness has said, but should be transformed from an impressionistic and uninformed account into something which informs the investigation and complements other information by attending to the key facts in dispute, for example. This position may sound contentious, because it may appear that advocates believe that the police should ‘doctor’ statements by inserting information or ‘coaching’ or otherwise manipulating the witness to provide information that ‘fits’ with other witness statements, forensic evidence and so on. However, this position is in reality altogether more subtle, and perhaps best explained with an example. Suppose an officer interviews a witness who has seen an armed robbery and knows that the robbers made off in a car. The officer may be particularly keen to include in the statement information about cars that the witness saw in the area shortly before and after the crime and may therefore introduce the topic of cars or may seek to develop the topic if the witness mentions it (as advocated in cognitive interviewing). In this scenario, the interviewer is not ‘putting words into the witness’s mouth’ (i.e. using forces of ‘suggestibility’), although critics would suggest that that could happen, even unintentionally. Neither is the officer necessarily introducing something into the statement that the witness did not want to discuss or did not feel was important. Rather, the witness is afforded the opportunity to include something that may otherwise have been forgotten and in recalling that item to also remember others. C. A witness has seen the events, people locations, items and so on that they have been asked to describe in their statement. They are, however, unaware of generic requirements of statement taking (particularly structure and level of detail). The statement should, therefore, be based on what the witness has said but should be translated from witness-speak into something that the legal process can use. This does not mean that the witness’ own wording in flawed in itself, rather that the particular sorts of textual processes which the statement will subsequently be subject to (summarising, placing into a sequence along with other statements and so on) require that the statement is in a particular form. Police officers are familiar with this ideal form, thanks to training and experience, and they should convert the witness’ formulation accordingly. In the hazy areas between these positions are combinations: Blend 1. A witness has seen the events, people locations, items and so on that they have been asked to describe in their statement, so, all of, and only, the information they have should be included in the statement. The interviewer knows best what information should be included (i.e. what information is typically needed and what information is likely to be salient in this case) and how best to conform to generic requirements in order to make the statement as useful as possible on both scores. Producing a maximally useful statement is a matter of uniting the witness’s information with the interviewer’s formatting and formulating expertise. It is also a matter of both parties responding to the question ‘what is important’ by communicating with one another about what their own expertise tells them should be included. Blend 2. The statement is not intended to represent the witness. In the statement-taking session, all aspects of what the witness saw, heard, smelled, felt, experienced and so on should be discussed. Then, points which the witness and/or interviewer hold to be important should be included in the statement, but the statement should only be seen as a summary of the witness’ understanding, incomplete but offering an overview which can be used as just that, with reference to the full deposition being made if necessary. The area between and around these perspectives is mediated by such considerations as: Who knows what information is important? What lexical items, grammatical and discoursal structures should be used? A clear understanding of what is typically included and excluded in statement-taking sessions and in the resulting statements might help us to answer these questions. At present, jurors and other lay people are given the impression that witness statements are artefacts produced under the conditions of scenario A, above, purely the words of the witness. If one of the other scenarios is more accurate, or a blend of the three scenarios, as is suggested by the analysis above, jurors and lay people should be made aware of that. Furthermore, if a one of the ‘blends’ of positions would provide a more ethical and realistic approach to statement taking than scenario A, then that should be acknowledged and embraced. As Komter (forthcoming) notes, discussing the related issue of taking statements from suspects, interviewers must balance the need to “do justice to the suspect’s version of events” with the production of a document that can “accommodate the demands of its future users in the criminal law process”. There seems to be no reason that this balancing should not be demystified. 7 Concluding themes and thoughts It may be that precise timings are not important in this investigation. However, if timing was genuinely important to this investigation, for example in helping to merge the testimony of several witnesses, or in aiding investigators in locating other potential witnesses who were in the area at the time, the audio-recording of this session would become invaluable. An audio-recording of this interview enables an investigative team to do exactly what I have done here, to trace the development of salient facts through the statement-taking session and to check that incongruities or apparent trivialities, which might turn out to be pivotal, are not overlooked. It may be unreasonable to expect interviewers to ask pertinent questions, note answers, identify contradictions, resolve inconsistencies, construct a text and attend to interpersonal aspects of the interview situation simultaneously, if this is so, it is not at all surprising that some information which witnesses present gets ‘lost’, despite good intentions. Furthermore, it is entirely unreasonable to expect a written text to represent a series of spoken texts in a truly meaningful way. By locating paradoxes which were not explored during interview and noting information which is recorded on tape, but not paper, this analysis shows that an audio-recording is a totally different sort of investigative tool from a witness statement. This analysis reveals that a tremendous amount of interactional currency is spent in the production of a witness statement. The collapsing of interviewing, note-taking, composing and so on makes the process of obtaining a witness statement and the product, the statement itself, quite unique. Audio-recording offers a way of seeing how the statement is fashioned through multiple versions, whether we are analysts, lawyers, police, observers or legal decision-makers. Close inspection of the verbal interactions through which written statements are crafted shows: If witnesses express degrees of uncertainty or certainty about important facts during statement-taking sessions this may or may not be noted in the resulting statement; Information may be ‘lost’ during statement-taking; Apparently contradictory claims by witnesses may be overlooked by investigators; Both witnesses and interviewers draw on strategies in attempts to aid the witness’ memory; Witness statement-taking is a complex task in that multiple demands are placed on the interviewer who needs to do several things (notably listening, and writing) simultaneously. More research is necessary to discover whether such an undertaking is at all realistic. More attention to writing skills in police interview training may assist; Audio-recordings of statement-taking sessions may enable police officers to conduct more detailed, thorough investigations, where necessary, by providing them with access to the origins of witness’ claims. We may then use these insights to contribute to the technologisation of this activity, to reflect critically on the productive processes at work or possibly to do both of these things. ‘The Guinness Advice’ includes the following: The general rule must be to preserve documents which come into existence during a criminal investigation … notes of interviews with actual or potential witnesses … draft witness statements. The Director of Public Prosecutions (14 August 1992: paragraph 8) This research supports the claim that documents which are produced in the course of an investigation should be collated and used, because it demonstrates that, during statement-taking, despite the best efforts of the statement-taker, information may be summarised ambiguously or lost. Preservation of original documents would not only make the processes of statement-taking more transparent if things go wrong, but would also enable investigations to become more sophisticated through use of draft witness statements to investigate disputed or unclear points and to pursue important insights more fully. A failure to maintain a record of the development of a statement is a failure to acknowledge any degree of infallibility in eyewitnesses. Police officers may problematise a witness’ ability to remember complicated strings of events or high levels of detail relating to events, people and locations which they may have observed whilst under stress: [Uncritical] acceptance of eyewitness evidence by police officers may be misleading and hinder the appropriate investigation of crimes. On the other hand, the police may minimise the disruptive potential of incorrect information if they are critical of eyewitness memory. Kebell and Milne (1998:324) This critical approach to eyewitnesses should extend to include an honest recognition of the processes through which statements are constructed. This would mean that witness statements would be seen, by the police and by other legal specialists (judges, barristers and solicitors) and lay people (jurors) downstream of the witness interview, as functions of their productive processes. If, for example, the processes of statement-taking were acknowledged, they could be explored in court rather than being used as a weapon against the witness who may, as the system stands, find themselves under pressure to conform to the version provided in the statement. Any statement-taking session risks creating an incorrect or incomplete record. The only way to guard against these risks completely is to make, and use audio-recordings of witness statement-taking sessions. It could be that these issues have already been played-out and fully assessed in the arena of the audio-recording of statements with suspects. If this is the case, by analogy, incentives for the audio-recording of witness statements include clarity and justice but disincentives include demands on time and resources. Use of audio-recordings would need to be considered carefully to avoid wasting time and money and to avoid abuses, particularly the creation of an impression of improved practice where improvements have actually not occurred. Statement-taking is in a formative stage and the introduction of audio-recording of some witness statement-taking sessions has been an exciting development. It will be interesting to see whether this trend gathers pace and if so, how these recordings come to be used and how, indeed whether, applied linguists become involved in their use. References Baldwin, J. and Bedward, J. 1991 ‘Summarising tape recordings of police interviews’ Criminal law review 671-679 Barton, D. 1994 Literacy: An introduction to the ecology of written language Blackwell Publishers Ltd: Oxford Clifford, B. and George, R. 1996 ‘A field evaluation of training in three methods of witness/victim investigative interviewing’ In: Psychology, crime and law 2 231-248 Coulthard, R. M. 1993 ‘On beginning the study of forensic texts: corpus, concordance, collocation’ In: Hoey, M. (ed) 1993 Data, description, discourse HarperCollins: London 86-114 Coulthard, R. M. 2000 ‘Suppressed dialogue in a confession statement’ In: Coulthard, R. M., Cotterill, J. and Rock, F. (eds) 2000 Working with Dialogue Niemeyer: Tbingen 417-424 Director of Public Prosecutions 14 August 1992 Advice to chief police officers on disclosure of unused material CPS Headquarters: London Fox, G. 1993 ‘A comparison of ‘policespeak’ and ‘normalspeak’: a preliminary study In: Sinclair, J. M., Hoey, M. and Fox, G. 1993 Techniques in description: spoken and written discourse Routledge: London 183-195 Heaton-Armstrong, A. 1995 ‘Recording and disclosing statements by witnesses – law and practice’ In: Medicine, science and the law 35, 2 136-143 Heaton-Armstrong, A. and Wolchover, D. 1992 ‘Recording witness statements’ In: Criminal law review 160-172 Kebell, M. and Milne, R. 1998 ‘Police officers’ perceptions of eyewitness performance in forensic investigations’ In: The journal of social psychology 138, 3 323-330 Komter, M. 2001 ‘La construction de la preuve dans un interrogatoire depolice’, (‘The construction of evidence in a police investigation’), Droit et Société: Labov, W. and Fanshel, D. 1977 Therapeutic discourse: psychotherapy as conversation Academic Press: New York Macphearson, W., Cook, T., Sentamu, J. and Stone, R. 1999 The Stephen Lawrence Inquiry report The Stationary Office: London Milne, R. and Shaw, G. 1999 ‘Obtaining witness statements: the psychology, best practice and proposals for innovation’ In: Medicine, science and the law, 39, 2 127-137 National Crime Faculty 1998 A practical guide to investigative interviewing National Crime Faculty: Bramshill Shepherd, E. 1999 ‘Non-barking dogs and other odd species’: identifying anomaly in witness testimony’ In: Medicine, science and the law 39, 2 138-145 Smith, D. 1990 Texts, facts and femininity: exploring the relations of ruling Routledge: London