Dixon, David --- "Interrogating myths. A comparative study of practices, research, and regulation" [2010] UNSWLRS 40

	Home \| Databases \| WorldLII \| Search \| Feedback University of New South Wales Faculty of Law Research Series

Secondly, the paper contrasts the ways in which jurisdictions (principally in the US, the UK and Australia) have responded to concerns about practices in the police interrogation of suspects. Since the mid-1990s, a stark contrast developed between the strategy of ‘investigative interviewing’ in England & Wales (and increasingly in Australia and elsewhere) and the methods taught to US police offers via the ‘Reid Technique’ and similar training programs. It will be argued that this contrast must be understood in the context of differing responses to miscarriages of justice and investigative failures (caused, at least in part, by inefficient interrogation techniques) and the knowledge which investigations of these miscarriages produced.

This paper is concerned with the questioning of suspects about criminal offences, which is widely recognised to be a standard, indeed often essential, tool of criminal investigation. The first part of the paper challenges an image of interrogation which, it will be argued, is misconceived and misleading. It is concerned with the common tendency to understand interrogation through a series of myths and ‘common sense’ understandings which are unsupported by research evidence and which lead to inappropriate policy responses. In particular, they divert attention and understanding away from the mundane reality of interrogation as carried out routinely by investigators. One of the myths to be considered is that safeguards for suspects’ rights make successful criminal investigation impossible. In challenging this, the paper will comment briefly on contrasting modes of regulation of police practices in interrogating suspects, focusing on the benefits and limits of electronic recording.
Secondly, the paper contrasts the ways in which jurisdictions (principally in the US, the UK and Australia) have responded to concerns about practices in the police interrogation of suspects. Since the mid-1990s, a stark contrast developed between the strategy of ‘investigative interviewing’ in England & Wales (and increasingly in Australia and elsewhere) and the methods taught to US police offers via the ‘Reid Technique’ and similar training programs. It will be argued that this contrast must be understood in the context of differing responses to miscarriages of justice and investigative failures (caused, at least in part, by inefficient interrogation techniques) and the knowledge which investigations of these miscarriages produced.

Perhaps the most significant achievement of the researchers who established policing as an area of serious inquiry in the 1960s and 1970s was to emphasise the mundane, everyday reality of most policing practice. Peace-keeping, order maintenance and service provision, not law enforcement, were what patrol officers did (Reiner 2000; Dixon 1997: 11). These studies fundamentally changed our understanding of policing, challenging the dominant image of police practices and encouraging the development of new policing styles (such as community policing) which responded to public demand for policing services. While some of these researchers studied detective work (Ericson 1981), it attracted less attention than patrol policing.
There is now a considerable body of literature on police interrogation. A series of empirical studies in England and Wales which was prompted by inquiries into malpractice and miscarriages of justice has provided a new basis for understanding and policy. This paper draws on my research in England (Dixon 1997) and particularly Australia, presenting some data from a study of electronically recorded interrogations (Dixon 2007). In both jurisdictions, the emphasis was on everyday policing. In particular, the Australian research drew a random sample of all interrogations across the state of New South Wales for a year, consequently providing significant insight into everyday practices. However, there continues to be a mass of legal and psychological literature particularly in the United States which relies on a misunderstanding of police interrogation which is as damaging as the discredited idea that street-level police work is primarily concerned with law enforcement. If the image of police interrogation which continues to inform much debate has a connection with reality, it is with controversial cases involving serious crimes. In part, my argument is the simple point that most criminal investigation is not about murder or other very serious crime and that we need to understand the everyday reality. However, I will also argue that the place of interrogation in the investigation of very serious crime needs to be understood differently: interrogation seeks not just a bare admission, but a series of statements covering intention and other matters which the prosecution will have to prove (Innes 2003).

... the popular public perception (is) that police interviews are lengthy, gladiatorial conflicts, where police officers employ a repertoire of skilful psychological tactics to undermine a determined suspect's resistance to confess (Pearse & Gudjonsson 1996: 63-4)

The interrogation of suspects is a practice over-laden with myths. These appear to stem from three sources, which flow into each other.

Myths grow in darkness, in this case, the secrecy of the investigative process, particularly inside police stations. Until quite recently, police interview rooms have been inaccessible. As Baldwin suggests,

Because of the secrecy that has always surrounded police-suspect interviews and the traditional reluctance of police officers to allow outsiders access to the interview room, debates on the crucial question of interview procedures have had to be conducted in something of an information vacuum, and in consequence they have tended to be dominated by views from the police service (1993: 334).

Ironically, as will be confirmed below, many of these are 'misconceived or erroneous' (Baldwin 1993: 334). While some researchers have been able to observe interrogation by being present in the interview room (eg Bottomley et al 1991; Dixon 1997; Leo 2008), doing so has obvious limitations. The introduction of electronic, particularly audio-visual, recording of interrogations has transformed research possibilities (athough, as will be noted below, the limitations of recording must be acknowledged). Observation of randomly-chosen interviews, uncontaminated by the researcher’s presence, is now possible.
While Baldwin is correct that police views dominate, it should be noted that others complement them. When those who have been subjected to police questioning have had the opportunity to speak out, it has usually been to complain about malpractice, so emphasizing conflict and drama.
Another significant location for discussion of police questioning has been academic psychology. While some psychologists have made major contributions to understanding (and reforming) police interrogation (Gudjonsson 2003), others have tended to concentrate on dramatic, atypical cases, particularly those involving murder. For a discipline which makes so much of its claims to being a science, there has been a surprising tendency to offer unsubstantiated generalizations. For example, Shuy claims that 'many, if not most, persons from whom (law enforcement agencies) try to elicit confessions are unwilling to reveal all they have done' (1998: 2). It will be shown below that the reality is very different, at least in Australia and the UK.
Secondly, the growth of these myths was closely linked to the police attempt to construct a self-image of crime-fighting professionalism in the twentieth century (Manning 1997). Interrogation must be understood as the definitive skill of the thief-taking detective who, in mid-century, achieved iconic status within western police forces. This was the distinctive product of a much broader development in policing which had emerged in the later part of the nineteenth century – the self-promotion of police as specialists, indeed ‘professionals’, in crime-fighting by means of law enforcement. Amongst police, the detective became pre-eminent, his (usually) status recognized by allowances and higher pay and (ironically in an occupation which so prides its distinctive, disciplined character) freedom from wearing uniform. As the monopoly which the public police won for themselves came to be challenged between the 1960s and 1990s, it became increasingly clear that this development had not been inevitable, that policing is not satisfactorily defined by reference to law enforcement (Bittner 1990), and that, ironically, the police had chosen a mission which (because of the inevitably limited capacity of police to control crime) was to be quixotic (Manning 1977: 16-17). However, claims of police responsibility for declining crime rates in the 1990s led to the renaissance of a model of crime-fighting professionalism (Dixon 2005).
In this distinctive feature of police culture, the ability to interrogate and to get confessions was the key skill of the detective. However, this heroic construction was an ideal, not a representation of reality. As will be shown below, the role of officers who question suspects is quite different from the myth. It is from this background that there emerges the 'elaborate mythology (which) surrounds the subject of police interviewing ... (I)t is extraordinary how often one encounters the confident expression by police officers (and others) of views that one would imagine would be confounded by daily experience' (Baldwin 1993: 331). An example came in a seminar given by the author at the NSW Police Academy (now College) in which a trainer vehemently claimed, on the basis of extensive operational experience, that most suspects refuse to answer questions in interviews. As will be shown below, this is very far from the truth. The issue was not just this officer’s wish to present a particular view of policing practice, but a cultural belief which is a product and prominent part of broader attitudes towards criminal justice. Such myths are significant: police refusal to accept empirical evidence about use of the right to silence drove a successful campaign to restrict the right to silence in England and Wales (Dixon 1997: ch.6).
Thirdly, fictional representations of police work have had a significant influence both on public perceptions and on police self-perceptions. In a multitude of books, films and television series, the hero has been the detective who solves the case. Clues lead to the suspect, then the interrogation provides the classic dramatic finale in which a confession is obtained by the skilled detective. Crime, investigation, interrogation, confession provide an often irresistible structure for script writers feeding the public appetite for TV cop shows. Occasionally, the detective is anti-heroic: from Law and Order (the British 1970s version, not the Dick Wolf productions) to the Australian Blue Murder, to the American The Shield , detectives have been depicted as being morally indistinguishable from their prey. However, more often, the detective has faced the moral dilemma of achieving good ends by bad means: the ‘Dirty Harry’ problem which has spawned its own literature (Kleinig 1996). Much more commonly than either of these, the detective uses his (usually) professional skill – the accumulation of evidence, moral pressure, persuasion, deception - to get the suspect to confess.
Ever since police became the subject of popular drama, police departments have sought to project an image of police as professional crime-fighters. Most well-known (and most influential) was the use of the popular media to construct an image of the Los Angeles Police Department (Domanick 1994). The impact of such images does not merely fall on the public: it may also shape a police department’s self-image, setting an ideal for police officers as well as for members of the public. Most police recruits’ knowledge of policing comes from fiction rather than personal experience. Any tendency to treat the role of media as trivial must surely be dismissed in the wake of the apparently significant influence of 24 on interrogation practices by some US investigators post 9/11.
Unfortunately for researchers watching large numbers of audio-visual records, the image of reality is much more mundane than TV drama. Just as Ericson found that ‘the bulk of the patrol officer’s time was spent doing nothing other than consuming the petrochemical energy required to run the automobile and the psychic energy required to deal with the boredom of it all’ (1982: 206), so it sometimes seemed in my research that a typical interview is an undramatic, non-contentious affair, in which the leading officer shuffles through fact sheets and witness statements without looking at the suspect who sits looking bored at the end of the table, while the second officer stares at the recording machine and cleans out his ears. There is no drama, no conflict, and little tension. As elsewhere in the job, ‘the mundane reality of everyday policing ... is often boring, messy, petty, trivial and venal’ (Reiner 2000: 89). [2]
The following sub-sections examine several key myths about police interrogation

While they disagree about much else, a fundamentally shared assumption of US interrogation manuals (such as Inbau et al. 2001) and of their academic critics (Leo 2008) is that suspects will resist interrogation: it is this resistance which requires the techniques described in such manuals and which is criticised by such academics. However, as Baldwin suggests, the expectation that interviews will be ‘complex or difficult encounters with subjects who are likely to prove awkward or aggressive' is not matched by the reality observed in English (and Australian) police stations. The predominant impression left by the interviews in my research is of police officers carrying out routinized, bureaucratic functions – not engaging in battles of wits with determined adversaries. Vigorous challenges to suspects’ accounts are surprisingly rare (Dixon 2007: ch 4 & 5). Primarily, this is due to the structural forces which determine suspects’ compliance and the legal function of the interview: these will be discussed below. However, it is also necessary to appreciate the social reality of policing as an occupation which is often misunderstood by commentators who depict officers as relentlessly committed to crime control. On the contrary, it is necessary to see policing as a job performed by workers who have the substantial yet limited commitments and interest in their work of most workers. This is missed in many accounts of police culture which present an over-socialized picture of officers (Dixon 1997: 161).
As Baldwin suggests, most interviews ‘involve relatively simple and straightforward interchanges with reasonably compliant suspects' who were 'such co-operative individuals that they should have presented no serious difficulties to a moderately competent interviewer' (Baldwin 1993: 331, 332). There are many structural pressures which lead suspects to cooperate:

the coercive social environment in which interviews take place, the physical isolation of suspects, the pressures that might have been exerted before the recorder was turned on, the possibility of caution rather than prosecution, the existence of the discount in sentencing for those who plead guilty, or advice received from lawyers (Baldwin 1993: 332).

In New South Wales, the guilty plea discount has a statutory basis: courts must take account of a defendant’s guilty plea, and of when it was made. If the penalty is not reduced, the court must record the reason for its decision.[3] Suspects who have previously experienced the criminal justice process will be aware of this: in some NSW courts, a notice is pinned to the door of the court informing defendants that a guilty plea will be rewarded. Further, the isolation of suspects is greater than in England: as will be noted below, lawyers are almost entirely absent from the pre-charge process in Ausralian police stations.
While many suspects confess, the reality of what suspects say often does not fall neatly into denial or confession. Some suspects admit to one offence, but deny that which police are primarily investigating. Notably, suspects routinely admit to possessing, but deny supplying, illegal drugs. The pressure on such suspects to cooperate in order to avoid supply charges was frankly illustrated by an interviewer’s comment in a case in our sample[4]

You cooperated and everything. That will be it. Mate, if we wanted to be arseholes, we’d give you – you’d be charged with other things. (013)

While such suspects are understandably keen not to be charged with supplying illegal drugs, many of them apparently regarded possession charges as products of an anachronistic prohibition. Guilt was seen as a technical matter: there was nothing morally wrong or socially deviant about their activity. It was, according to the suspect in 046 ‘a lot of hassle over ten bucks of pot, isn’t it? I’m a silly boy aren’t I?’.
Other suspects make statements about their conduct which amount to admissions (notably of assaults or of complicity in offences), but which they appear not to recognize as such, seeing their actions as either justified or innocuous. In still others, the situation is even less clear, with suspects and police alike unsure about what the offence is alleged to be: officers were (perhaps understandably) unsure about how to classify the man who, in the backyard of a suburban home, fired an arrow from a powerful sporting bow straight up to see how high it would go. His neighbour was less concerned about the ascent than the descent, which ended in his garden.
We attempted to replicate Baldwin’s allocation of suspects to categories of cooperative/submissive, remorseful/tearful, cocky/self-assured and awkward/difficult to interview (1993: 332). In brief, we found that these categories overlay a more general condition: whatever their mode of expressing it, most suspects displayed a resigned acceptance of their position. Typical was the suspect in 018, described in fieldwork notes as ‘cooperative, resigned, pissed off, faintly amused’ at the heroin possession charge which he faced (Dixon 2007: ch 3).
If resistance is rare, this is not surprising given the pressures noted above. An example of an awkward suspect was 075 who was questioned about an assault and about illegal drugs found in his home. He refused to answer some questions, answered others obstructively (eg saying he did not know how to spell his own name and refusing to agree that the interviewers’ standard statement about how many people in the room was correct), rejected the scales used to weigh the drugs, and alleged that police had assaulted him during the arrest and planted some of the drugs. What was striking and is relevant in the present context is not these allegations per se, but the interviewing officers’ apparent surprise and upset at their making. An uncooperative, disruptive antagonistic suspect was, apparently, not part of these officers’ normal experience.
Refusal to cooperate in the form of exercising the right to silence, despite claims to contrary, is rare in both England and Australia (see p.12 below and Dixon 1997: ch6; 2007: 92-100). This was illustrated by the tension caused by a suspect’s attempt to refuse to answer questions in 253: the interviewers were clearly not used to dealing with this response. Such rarity should not be surprising given the difficulty of remaining silent and the costs of doing so. This finding undermines a myth which police officers have often encouraged. In England and Wales, police complaints about suspects who exercised their right to silence were a key tactic in the campaigns for extended police powers which began in the 1960s. The extensive English academic research on the right to silence (notably highly reputable research carried out for the Royal Commission on Criminal Justice) shows that the extent of its use is exaggerated and that, when suspects are silent, there may be reasons other than an obstructive, lawyer-encouraged attempt by a guilty suspect to stand on rights (Dixon 1997: ch.6).
It may be due to an officer’s questioning style: ‘One of the primary causes of such resistance (silence, non co-operation, hostility, lying evasion, denial of knowledge and culpability) is, in fact, the interviewer's behaviour' (Milne & Bull 1999: 71). Unskilful as well as aggressive questioning can cause the problem for police: non-response may be encouraged by the common tactic of reading a witness statement and asking ‘what can you tell me about that?’, to which the answer ‘nothing’ or ‘no comment’ may be counted as silence, but which may, in context, be a genuine and appropriate answer. For example, in 231, the suspect denied having been present when a car was stolen, but the interviewer continued to read the victim’s account of the theft. As he had already said he was not there, the suspect could do little but say ‘nothing’ when the interviewer asked repeatedly ‘what can you tell me about that?’
In other circumstances, silence may be due to a suspect’s general attitude to police. For example, in 006, the suspect had admitted possessing and using marijuana, but then smiled, looked at the ceiling, and ignored a series of questions designed to produce more specific admissions of use. Silence may not be motivated by an attempt to avoid charge, but rather by individual and cultural commitments to non-cooperation with police (Dixon 1997: 256-8).
On the other hand, silence may be less obstructive than answering in some cases: for example in 035 a juvenile suspected of stealing a Playstation answered all questions, but did so in a way that treated the interview as a farce. He gave three inconsistent accounts of how he obtained it. When asked what he intended to do with it, he replied facetiously ‘Take it to the police station’. The interviewer’s response was to abandon his attempt to conduct an investigative interview, and he simply went through the motions of putting accusations to the suspect.
Such cases illustrate one of the many problems in counting instances when the ‘right to silence’ is exercised. Much commentary on this assumes that silence is a common-sense, unproblematic category, which it is not. The wide variation in estimates of silence by academic researchers and police officers in England stemmed, in part, from this problem: does ‘silence’ just mean not speaking or saying ‘no comment’, or does it include ‘non-answers’ such as that quoted in the previous paragraph? How is selective answering to be treated? These issues have been discussed in detail elsewhere (Leng 1993), their relevance for present purposes is to undercut the idea that ‘common sense’ is a good guide in assessing police interrogation.
A factor which runs counter to the image of the tension-filled encounter is that some suspects appear to get some satisfaction, even enjoyment from the interview. This may be expressed in pedantic attempts to insist that certain details should be recorded precisely as they describe. For example, in 205, a suspect whose life had spiraled out of control through gambling and a shambolic conspiracy to defraud, seemed to take pride in the precision with which his confession was made. (Sometimes, as shown in the following subsection, such precision is an attempt to resist police control of the interview.) Such suspects, particularly some drug users, appear to get something from being taken seriously, being the focus of attention, even if the outcome is to be a conviction.[5]
Some suspects deal with their situation by tactics which, incidentally, defuse tension. A notable example was 263, an elderly man suspected of bank fraud who described himself as ‘presently unemployed’, but with ‘strong prospects’ of architectural work. He mentioned several times in the interview that he was an architect. Setting up fraudulent accounts in fictitious names was justified (if not explained) as part of his efforts to improve his job prospects: he admitted that, ‘This is an unorthodox way of approaching it’. What is relevant here was the way in which he treated the interview as if it were a professional transaction, remaining business-like and aloof. In this way, the suspect distanced himself from the reality of his predicament. His choice of role (‘professional person’, not ‘criminal suspect’) determined the nature of his interaction with the interviewers.

A crucial part of standard guides to police interviewing (notably and most influentially Inbau et al. 2001) is that officers must dominate the suspect. In consequence, detailed guidance is provided about how police should specify the suspect’s physical position, and so on. Clearly, suspects are controlled while in custody, and such control has both physical and psychological dimensions, even in mundane matters such as access to food and toilet facilities. Officers appear to appreciate this. For example, in 041, there was a subtle negotiation of power between interviewer and suspect as the former responded to the latter’s challenging manner

He told the suspect that ‘it is required’ that he give his name. After the suspect did spell his surname, the constable then asked the suspect to spell it again, saying he could not be heard. He asked the suspect to lean forward, to keep in the picture of the video. He asked him to take his hand away from his mouth when he spoke. And he asked him to stop swaying in his chair. In each instance, the request was explained as facilitating the interview (stopping the chair noise, allowing the suspect to be heard clearly, etc), but at the same time these requests could be interpreted as the interviewer attempting to control a cocky suspect.

In addition, as discussed elsewhere (Dixon 2007: ch 4), interviewing officers exercise close control over the contact between the suspect and third parties.
However, the effective domination of suspects by police should not be overstated. Our samples included several categories of cases in which police control was disrupted. First, a suspect may not be overawed by police for the simple reason that he or she is innocent and confident that their innocence will be established. For example, in 074 a juvenile suspect had been arrested at a school where there had been a burglary. He claimed that he had been playing football with friends in the grounds and that he had climbed onto the roof of a walkway merely to retrieve the ball. He was relaxed throughout, joking with police and the independent person. When asked if any promise, threat or inducement had been offered, he answered 'I wish'. The police cliché – suspects who have nothing to hide have nothing to fear – seemed to underlie this suspect's approach. Similarly, in 278a, a juvenile (who had been arrested for possessing coins which the police suspected had been stolen) coolly corrected the interviewer twice:

Q: Do you agree when I asked you if you intended to sell these to the shop owner you said, ‘No’?

A: You asked me if I asked the shop owner if he wanted to buy these and I said no.

He was not intimidated, and confidently and intelligently stood his ground in dealing with the accusation against him without appearing cocky.
Irrespective of guilt or innocence, some suspects are socially and intellectually more skilled than their interviewers. For example, in 230, the officers were clearly not prepared for the interview, and the suspect coolly rejected allegations in witness statements which police read out, asking for comment. Somewhat similarly in 001, the suspect successfully deflected police suspicion that he had intended to sell the (considerable quantity of) illegal drugs found in his possession. His self-assured, confident manner contrasted markedly with that of the gauche interviewers. At one point, an officer asking a question seemed simply to run out of words:

A: 'I agree that what I had in my hand was illegal. I have no excuse for it. I'm not above the law. I do regret it. It's just a habit.'

Similar cases were noted by Moston and Engelberg, who reported that 'Interviewers would appear unsure of what question to ask and often gave the impression of being more nervous than the suspect they were questioning' (1993: 224). The suspect in 001 emphasized his attempt to define the nature of the interaction by thanking the officers at the end of the interview, shaking hands with the adopting officer, and telling him that the interviewers ‘did their job 100%’.
In the preceding subsection, examples were given of suspects attempting to distance themselves from reality, thereby turning the interview into a non-confrontational encounter. Doing so could undermine police domination of the encounter. In other cases, the resistance was more specific. For example in 003, the suspect confessed to numerous armed robberies. The interview consisted of a recurrent pattern of questions and answers about each incident. A witness statement reported the time of one offence as being 9.55am. It became clear that this was a typographical error: it should have read 9.55pm. While the interviewers were mildly exasperated by this apparent pedantry, the suspect refused to accept responsibility for the alleged offence, insisting that he had not robbed the premises in the morning. Despite his predicament, he wanted some control, however nebulous, over the situation. Similarly, in 038 a suspect refused to agree to a formulaic, insignificant ‘Do you agree...?- (DYA) question about the place of arrest because he did not know the name of the street. In 092, a suspect who gave a patently ridiculous explanation of possessing drugs[6] nevertheless argued with the officers over the precise dimensions of the foil.
The most common group of suspects to be considered in the context of non-domination of interviews were those who were ‘repeat customers’ for whom the process of detention and questioning held no surprises. They had been through it frequently enough that they were not overawed or intimidated. Drug users were prominent among this group. Occasionally, drug-affected suspects went to sleep during an interview. A memorable case involved a heroin addict on rohypnol, a self-described ‘crim’ (038) who provided a comical account of being ‘taken for a gig’ by a woman who got him to break into a house in which she had supposedly left some property and being chased down the street by the irate resident, who was wearing a towel and carrying a baseball bat. Through this story, the sadness of a life involving repeated contact with police was all too evident. One more arrest and interrogation held no surprises or fears for this defendant.
Other suspects expressed their lack of concern about their situation in different ways, as in 244 where the somewhat intoxicated suspect laughed about an alleged assault, referring to the victim as ‘the dickhead upstairs’. Such suspects’ coolness appeared to stem from previous experience, as in 240 where a convict was interviewed in gaol about another offence. He was polite, but dismissive of and unperturbed by the interviewers’ allegations. In 057, the suspect expressed his disrespect for his interviewers by stating his occupation as ‘thrill seeker’, putting his foot on the interview table, identifying himself in a CCTV picture as a ‘good looking bloke, eh’, and telling the interviewer that he found the interview boring.
In 041, the suspect’s disrespect for police even was more overt:

The suspect sat pulling at his finger knuckles, one at a time, and made each one crack. After he completed this for both hands, the partner announced aloud that the suspect had been cracking his knuckles. This was done for the benefit of the transcriber. The suspect responded:

How come police officers are so straight? I did nothing wrong. You don’t even drink, do you, and stuff like that. But when you drink you can’t go out and have fun. You can’t go out running in the streets.

By contrast, some suspects demonstrated their experience in dealing with police by seeking to disrupt accusations against them by a display of geniality and relaxed affability. For example, in 021 stolen property had been found hidden in the roof cavity of the suspect’s house. He claimed that he had bought the items from young people who were going from door to door, but that he had no idea at the time of purchase that they were stolen, and that he had only hidden them when he heard that the vendors had been arrested. On its face, this was a highly unlikely story. However, the suspect maintained a relaxed air, which the officers’ rather stiff, stilted, repetitive questioning did not disrupt. He actively cooperated by locating the serial number on one item when the interviewing officer was unable to do so, and even expressed sympathy for victims of break and enters and condemned the perpetrators.
As these examples suggest, the police interview is a much more complex and contested social field than is usually assumed.

The drama of fictional interviews builds up to the moment when the suspect, worn down or outwitted by the interviewing detective, cracks and confesses. Such events are very rare. In both my Australian samples[7], only 2% of interviews involved a suspect shifting from denial to full confession in the course of an interview. Many suspects - 24% in sample 1 and 22% in sample 2 - confess from the start. Others (50% in sample 1, 16% in sample 2) make partial admissions: typically, these cases feature suspects who admit to possessing and using illegal drugs, but deny selling them. Many suspects (12% in sample 1, 53% in sample 2) denied committing any offence throughout the interview. Very few – 4% in sample one and 3% in sample 2 – refused to answer questions.[8]
In this crucial respect, our data confirm the findings of other researchers: Baldwin (1993: 333) found only 3 of 600 cases in which the interviewer's persuasive skill induced confession to offences of any seriousness (and only 6 other shifts from denial to confession). Pearse and Gudjonsson (1996: 72) report that only 2 of 161 suspects shifted from denial to confession: they conclude that 'suspects enter a police interview having already decided whether to admit or deny the allegations against them – regardless of police interviewing techniques' (Pearse & Gudjonsson 1996: 73).
In some cases, the expected confession never arrives:

As one watches this (record) there is an expectation of a confession. The highly emotive suspect (who at the beginning of the interview has his neck bent, his face looking down at the interview table and his hands over his face and then over his head) gives the impression of a man about to tell all. The suspect appears to cry during the interview (tears seem to be wiped away off his cheek) and his voice quivers, encouraging the impression that the confession is forthcoming. But by the end of the interview the confession has not arrived. After or in spite of such emoting, the suspect makes no more admissions by the end of the interview than he did at the beginning. At first and at the end, he said he was in possession of marijuana and cannabis plants, but he did not sell drugs. He said he was in possession of a Sharp TV, but he did not know it had been stolen when he bought it from an acquaintance. (108)

The suspect was aware of the consequences of confessing to anything other than possession. Meanwhile, the interviewing officers did little other than show (and ask him to account for) exhibits suggesting involvement in supply – bags and scales – and ask if he agrees they are suspicious. ‘Only to a police officer’, he replies.
Even when a suspect does shift from denial to confession, this is not necessarily due to police interviewing skill. For example, in 005, the suspect denied having made a fraudulent application for a credit card. However, when he was shown the document, he immediately acknowledged having signed and submitted it. His 'resistance' was perfunctory, and the police effort in overcoming it was minimal, involving the pre-interview collection of evidence and its presentation, rather than interview tactics. This example illustrates the general conclusion in English research that the crucial determinant of confession is the weight of evidence against the suspect, rather than how he or she is questioned (Moston & Engelberg 1993).
In 130, the suspect was questioned about an explosion in a toilet block at his school. He initially denied involvement, but when he was confronted with evidence from his mother that he had exploded a home-made device at their house, he admitted the offence at school. It was the mother’s statement, not the interviewing tactics that produced this admission. In 113, the suspect evasively said that he was ‘not sure’ why he had taken the fly screen from a house window. However, this resistance was perfunctory: he had admitted to attempted burglary in a preparatory interview, and confirmed this when the interviewer read his record of the exchange from his notebook, asking the suspect DYA questions. Similarly, 123 was counted as a ‘denial to confession’ interview. However, the suspect ‘cracked’ when confronted with an admission that he had made earlier:

At the beginning of the interview, the suspect stated he had no idea the rolls of material that he purchased may have been stolen. He appeared calm and at ease with his responses. In the middle of the interview the interviewer then asked a challenging question based on one earlier admission by the suspect.

Q: When I spoke to you earlier do you agree I asked you if you knew it was stolen and you said, ’I had a sneaking suspicion it may have been’)

The suspect was visibly shaken as he answered.[9]
In 205, appearances were deceptive. Initially, this seemed to be a clear example of a suspect cracking, turning from lies and denials which were challenged by police in an initial interview, to full confession in a subsequent interview. However, by the end of the interview, it had become clear that the suspect had decided to confess (and implicate others in the offence) not because of anything the police had done, but because his co-offenders had ‘shafted’ him in disposing of the stolen goods.
It would be wrong to assume that not getting the suspect to shift from denial to confession amounts to failure. Officers frequently point out that an implausible denial is almost as good as a confession and certainly better than admissions accompanied by factual confusion, justification, or excuse. In the case of the suspect found with stolen property in his kitchen ceiling (021: see above), the denials were thoroughly unconvincing. Similarly, in 064 a man was interviewed about the theft of a remover’s trolley from a truck which he had hired. He consistently denied stealing it. However, police had evidence that he sold an identical trolley (same colour, same faulty wheel) to a second-hand shop. His claims that this was a different trolley, one that he had found it on a rubbish tip, were not merely thoroughly unconvincing: they seemed clearly to indicate guilt.
Another example is 276a, in which police searching the suspect’s vehicle found: more than one kilo of cannabis, 1.3 grams of speed; a black bag that contained small, re-sealable plastic bags and a small box of rubber bands; four syringes; two sets of scales; other drug paraphernalia; and a police radio scanner. The suspect

claimed that when in Adelaide a male asked him casually if he wanted some marijuana, apparently as this male ‘wanted to get rid of it’. He did not know this person who then, according to the suspect, gave him $3,000 worth of marijuana ‘on credit’. The suspect said that as he neither knew the man’s name nor his address, he would return to Adelaide in the future and look for this same man (whose name he did not know) and he would pay him the money owed. Similarly, the suspect said also that the large quantity of cannabis obtained was purely for personal use. The small re-sealable bags could be used for keeping fruit purchased en route, during the drive. The two sets of scales were stored in the suspect’s vehicle to ensure that when he did buy marijuana (for personal use) he obtained the correct amount.

The unconvincing nature of these claims simply harmed the suspect’s case.
In these interviews, police unsuccessfully sought admissions, but did not seem unhappy with the outcome. By contrast, in 249, the interviewer made no attempt to ‘crack’ a man denying sexual assault offences. Instead, he merely gathered the suspect’s account, but did so in a hostile, detached way which did not produce a confession, but which both jarred the suspect, and could have left the audience (prosecutor, judge, jury) in no doubt about the detective’s view of his guilt.[10]

Police interviewers are almost always presented in fiction as competent and effective. If they have any failing, it is that they use morally dubious tactics in order to pressure suspects into confessing. However, such failings are excused, because the audience is assured (indeed has often already been shown) that the suspect did indeed commit the crime. This belief in competence and efficiency is reflected in police self-perceptions. As noted above, interviewing skill has traditionally been regarded as the mark of the good detective: 'few officers see themselves as being poor interviewers, and a potent police culture exists which perhaps inhibits them from doing so' (Baldwin 1992:28)
Baldwin’s research in England pricked this bubble of aura and ego. He found not only that interviewing skills were poor, but that there was a notable gap between self-perception and reality: sometimes the poorest interviewers think they are skilled and it is ‘the officers who are most in need of training or guidance who feel that they need it least' (Baldwin 1992:28). More recently, one senior English officer, exasperated at hearing his officers interviewing burglary suspects, suggested that some of them were ‘so inept at interviewing suspects that viewers of television dramas such as Inspector Morse would have a better chance of solving crimes’.[11]
The interviewing quality in sample 1 was generally not good, although not as bad as that reported by Baldwin. Most common were basic problems of interview technique: officers stare at the recorder machine and shuffle papers rather than interact with suspect. The fact that a suspect provides a confession does not necessarily reflect on the interviewer’s skill. Indeed, in some interviews the opposite seemed more appropriate: for example, in 123, the interviewer was unprepared, shuffled through documents without looking at the suspect, and asked badly phrased, awkward questions. Nonetheless, the suspect was so cooperative and submissive that he confessed anyway. Another problem concerns legal adequacy. While, as will be argued below, a key function of interviews is to provide evidence for court, essential elements of offences were sometimes not covered: for example 131 concerned an alleged assault, but the interviewer did nothing to clarify whether there was an intention to injure or whether the incident was, as the suspect claimed, mere horseplay.
The introduction of investigative interview training drawing on the English PEACE model has improved standards in NSW (Dixon 2007: ch 5). From the perspective adopted in such training, effective interviewing does not require bravado, acting, deception, psychological games, or violence. Much more useful are the simple interpersonal skills which are effective in interviews in other social settings - looking at the person you are talking to; being polite and, within obvious constraints, friendly; indicating that you are listening; asking questions which follow up cues from suspect. While the product may not be a dramatic confession, it may well be useful information in a legally packaged form. This packaging is discussed further below.

The youngest, and perhaps the most potent, myth is the belief that deception by a suspect can be readily identified from ‘body language’ and that doing so is core detective work. According to Moston and Engelberg, in England, a 'dangerous mythology pervades the police service which suggests that the ability to detect deception is an essential component of interviewing competency' (1993: 226-7). (This, in turn, relies on a myth discussed earlier that many suspects seek to deceive their interviewers). This myth provides a notable example of how expert knowledge (in this case, from psychology) has been transformed into popular understanding. One of the most disturbing encounters early in our research was with a judge who confidently claimed (on the basis of a session at a weekend professional training course) to be able to assess the veracity of witnesses by observing whether they glanced to left or right.
Professional training courses, schlock psychology texts, and soft-news media and fictions (notably the regrettable Lie to Me TV series), have all played their part in popularizing the belief that body language can be read in order to detect deception. As in drug courier profiling, criteria are broad and flexible. According to members of the FBI’s Behavioral Analysis Program, deception may be indicated both when people avoid eye contact and when they seek it: ‘frequent liars ... increase eye contact because they learned that investigators often gauge veracity by strong eye contact’ (Navarro & Schafer 2001: 10). In New South Wales, the potential for using interpretation of the suspect’s behaviour in evidence is acknowledged in the adaptation of the caution, which now warns a suspect that ‘whatever you say or do will be recorded and may be given in evidence’ (NSW Police 1992: 19).
This is not the place for an extensive review of the literature on this topic. It is sufficient for present purposes to point to clear research evidence which establishes that, whatever a highly trained psychologist may be able to do in detecting deception, an ordinary police officer (or indeed prosecutor, jury or judge) cannot do so accurately, and that standard interview training does not increase the capacity to correctly identify deception.

Consistently, research shows that individuals achieve accuracy around, or worse than, chance, in detecting deception and are only marginally better with truthful responses .. It would seem that professional investigators differ from lay-people only in being more confident in their performance (Mortimer & Shepherd 1999: 302, see also Vrij 1999).

Research has consistently demonstrated that interviewers cannot detect deception through non-verbal cues. Indeed those cues which interviewers ‘read’ as indicators of lying often are a result of their own behaviour (eg being too close to the interviewee) and are instead signs of anxiety with the situation ... (M)ore experienced officers are more confident that they can spot deception but they are, in fact, no more accurate (Milne & Bull 1999:64: see Memon et al 1998)

... the detection of deception is an unreliable process ... training almost universally fails to enhance competence in any such ability (Moston & Engelberg 1993: 226)

Police commitment to detecting deception is indicated by an article by the FBI agents mentioned above who acknowledge that even experienced investigators’ success at detecting deception is no better than chance and that ‘no particular nonverbal or verbal cue evinces deception’, but then proceed to provide the familiar list of verbal and nonverbal cues which supposedly indicate deception (Navarro & Scafer 2001: 9).
Moston and Stephenson warn that unjustified confidence about detecting deception may lead to miscarriages of justice. If officers become convinced of guilt, their claims that they have detected deception may induce a false confession by pressurizing the suspect into accepting the interviewers’ account of what happened (1993b: 107). How this could happen is illustrated by Navarro & Scafer’s claim that deception is indicated by ‘stalling tactics’ such as asking ‘Where did you hear that? Where’s this information coming from? Could you be more specific? or How dare you ask me something like that?’ (2001: 12) These could be perfectly appropriate responses to an inaccurate allegation, yet police are encouraged to treat them as signs of guilt.
As Vrij indicates, '(t)he main problem is that there does not exist typical non-verbal or verbal behaviour which is associated with deception. That is, not all liars show the same behaviour or say the same things’ (Vrij 1999: 325). Conversely, behaviour associated with deception (such as avoiding eye contact and fidgeting) may have very different meanings according to the social or cultural group of the suspect. Most significantly, some Aboriginal people and some people of Asian background may regard eye contact with a person in authority such as a police officer as uncomfortable, disrespectful and inappropriate. Police officers need to be strongly warned that amateurish attempts at psychological assessment may lead their investigation badly astray.
Fortunately, this danger has already been recognized in New South Wales, and the identification of deception is not part of formal interview training. However, there are other pressures in play. As suggested by the anecdote above, there has been considerable interest in the detection of deception from the judiciary. This appears to have driven a good deal of the pressure for the showing of ‘ERISPs’ (electronic records of interview with suspected persons) in court, the improvement in picture quality, and, most significantly, the introduction of video technology which provides an image alternating between a general picture of the interview room and those present to a close-up of the suspect. Indeed, the alternating image system was described by the commander of the electronic recording unit as the ‘Justice Hunt solution’, after the judge who had lobbied for this new technology.
Alternating images have both advantages and disadvantages. The most obvious advantage is that for the first time the viewer can see a large, clear image of the suspect’s face during the interview. After years of (at times frustrating) attempts to make out how the suspect looks (Are his/her eyes closed? Is he/she falling asleep during some questions? How serious an injury is that mark on the forehead which is a blur from the distance? Is she visibly alcohol affected?), it is good to be offered such a large clear image of his or her face. The size and clarity of this image of the suspect greatly reduces what previously may have remained in the realm of guess work.
Among the disadvantages are that other persons present in the interview are only seen briefly, if at all. The video records the whole interview table only for some 20 seconds every three minutes, before reverting to the face of the suspect. For most of the time, the interviewers are not on screen. Often in alternating image videos, the adopting officer enters, performs his/her role and leaves without appearing on screen. Moreover, if the ERISP is to be used to ‘legalize’ exhibits (see below), interviewers will have to put them in the suspect’s face to ensure that they are seen. Similarly, we may not see the notebooks from which DYA questions are asked to legalize pre-ERISP statements. If ERISP is to be used as a mechanism of supervision and accountability of interviewing officers, something is lost by focusing on the suspect.
While both prosecution and defence may gain some advantage from the close-ups showing the suspect’s condition, there are grounds for concern about potential interpretations of these images by both prosecutors and judges. 073 illustrated the potential problem. In close-up, the suspect in this ERISP appears somewhat shifty as he moved his eyes from side to side. However, the brief wider focus showed that these eye movements were a normal mode of interaction with two interviewers who were both attempting to maintain eye contact with him, as a mutual interaction. Seeing him reacting to questions rather than seeing him as one of three people exchanging questions and answers invited incomplete or inaccurate interpretation.
In particular, there is the danger of encouraging uninformed interpretations of body language. It is notable that more than half of both judges and prosecutors in our questionnaire study[12] agreed or strongly agreed that a suspect’s demeanour during interview indicates whether he/she is telling the truth. Given the finding of social psychological research on the identification of deception, it is some relief that police were much less likely to agree.

Table 1: ‘A suspect’s demeanour during the interview indicates whether he/she is telling the truth’

	Police N=123 %	Prosecutors N=71 %	Defence N=77 %	Judges N=49 %
Agree/Strongly agree	28	56	26	57
Disagree/strongly disagree	32	10	38	20
Neutral	36	28	35	16
Don’t know	3	6	1	2
No response	1	0	0	4

Social psychological research has demonstrated convincingly that focusing the video camera on the suspect alone influences the viewer’s perception. The viewer is more likely to believe that a confession is voluntary if the camera focuses on the suspect than if it focuses either on the interviewing officer or on both together. So far as possible, the ERISP camera should provide an objective account: returning to a system in which both police and suspect are shown equally would reduce the tendency to bias (Lassiter et al. 1986, 1992). Less instrumentally, the focus on the suspect indicates a shift in the purpose of ERISP from being a device designed to prevent disputes over admissibility of evidence to being an investigative tool in own right (by recording ‘evidence’ of deception).

When police officers are asked to explain their purpose in interviewing suspects, they typically refer to the process as ‘a search for the truth’. This is attractive rhetoric, but is problematic as a description of what occurs in interrogations. First, it must be acknowledged that an adversary system of justice deals in contested versions of reality, not in absolute truths. Sometimes in complex and controversial cases, ‘what really happened’ is inaccessible. We may have to accept that we do not know what happened, let alone why. None the less, trying to find out what happened is appropriate. However, in dominant US interrogation models, confession is more valued than truth. As will be noted below, a widely approved interrogation tactic is to suggest to the suspect a version of events which makes confession easier by minimising guilt. As Kassin et al summarise:

Minimization tactics are designed to provide the suspect with moral justification and face-saving excuses for having committed the crime in question. Using this approach, the interrogator offers sympathy and understanding; normalizes and minimizes the crime, often suggesting that he or she would have behaved similarly; and offers the suspect a choice of alternative explanations – for example suggesting to the suspect that the murder was spontaneous, provoked, peer-pressured, or accidental rather than the work of a cold-blooded killer. (2009:120)

Kassin et al’s major concern about this tactic is, properly, the potential for false confessions and miscarriages of justice. Here, my concern is more limited: whatever such interrogation tactics do produce, the truth (if this means an accurate account of what happened and why) is not the objective.
Secondly, ‘searching for the truth’ has too often been the gloss on a method of interrogation which consists of the interrogator seeking the suspect’s confirmation of (confession to) a case theory (a ‘truth’) to which the interrogator is already committed. Take, for example, these exchanges in the interrogation of George Heron about the murder of Nikki Allen:

The interviewing officers’ commitment to the ‘truth’ that Heron was a murderer produced a confession so unreliable that the trial judge excluded it. It is now widely acknowledged in England and Wales that ‘case theories’ can be dangerous if they make officers unwilling or unable to consider any other ‘truth’.
At a prosaic level, the rhetoric of ‘search for the truth’ has problematic aspects which go the heart of the real role of interrogation in everyday criminal investigation. Confirming findings of English researchers, our Australian data suggest that most interviews are better seen as a process of information gathering and packaging. If finding truth means getting a suspect to shift from denial to confession, it rarely occurs. More typically, the interviewers obtain the suspect’s account and, while they may check it against other information available to them, there is little in the way of aggressive or even active challenging (Dixon 2007: ch 5). This restriction is, as discussed below, a response to concerns about instances in which ‘searching for the truth’ led police into responsibility for grave miscarriages of justice, not least because what was really involved was not a search for an unknown truth, but a search for confirmation of a ‘truth’ to which police were already committed. As Ashworth suggests, employing the rhetoric of truth is too readily linked to a consequentialist logic that ‘the end may sometimes justify the means’ (Ashworth 1998: 108).
If the interview is not a search for truth, nor is it the recording of an objective account. It is should be trite to say that any account is a selective construction of reality processed through memory:

memory for a complex event is largely constructive. What a person encodes is not recorded in memory ready to be played back like a video-recording. Instead the event is reconstructed using the information the person has encoded about the event and also by using information that the person has about the world in general (Milne & Bull 1999: 17).

The interview is an interactional sequence which produces accounts of action, intention and motive. Crucially, prompts from interviewers also shape the reconstructed account: experience has shown the vital need for interviewing officers not to prompt suspects by leading questions, providing information or, more generally, excessive reliance on case theories. However, it is necessary not to emphasize the link between construction and police malpractice. This is a significant weakness of McConville et al’s influential research monograph The Case for the Prosecution (1991) which treats case construction as a tool of police malpractice rather than a necessary and inevitable (if open to abuse) part of the information work done by criminal investigators.
Police interviewing of suspects involves a very particular form of account construction, which I will refer to as ‘legalization’ which involves the interviewer organizing and ‘framing’ the suspect’s account according to legal criteria (cf Dixon 1997; Innes 2003: ch 3).
A considerable academic literature has shown that the professional policing model which became hegemonic in the mid-twentieth century misrepresented the reality of criminal investigation (Reiner 2000: 118-21). It showed that most such work is not detection, but the transformation of an incident into a case and an individual into a defendant by the collection, categorization and presentation of evidence.[15] As Ericson's account explains, police officers ‘make crime’ by translating the original material (hunch, suspect, physical evidence, witnesses) into legal categories, (1981b: 7-20). Interrogation is an especially important site for the 'legalization' of accounts. Because this point is often misunderstood, it perhaps bears repeating that my argument is not that case construction indicates police malpractice, but rather that it is a necessary and inevitable part of the processing of information which criminal investigation entails.
As Ericson (1981) and Manning (1997; Manning & Hawkins 1989) suggest, the legal processing of a case is information and communication work. The rules of substantive and procedural law provide the language and framework of the accounts which must be given of suspects’ and police officers’ actions. To Ericson, detectives are 'accountants' who 'explain, justify, and legitimate their actions with respect to what they think are the appropriate rules, and these rules are used in conjunction with their accounts to orient their actions' (1981: 16-17; see also id. 1982: 13-14; Manning 1997; Van Maanen & Pentland 1993). An important function of this accounting is to confirm the action's legitimacy and justification: law provides the pigeonholes into which the actions of suspect and police officer alike can be slotted (Bittner 1990: 194, 197, 246). However, legalization is not synonymous with legitimation: both more and less than is involved here. All describing and accounting for action involves active interpretive work. Incidents are reconstructed as stories, in which legal definitions and requirements provide cues, plots and character development (Manning 1992; cf. Bennett & Feldman 1981: 4-5; Shearing & Ericson 1991).
In subsequent research on legal decision-making, Manning has adapted Goffman's concept of framing (1992; cf. Manning and Hawkins 1990; Innes2003: 54-66). Policing activities are fitted into legal frames as cases are 'legalized'. Interrogation provides a good example: there is a significant (but, as yet, inadequately traced) link between the trend in areas of modern substantive criminal law towards requiring proof of subjective intention and police use of interrogation in order to obtain confessions (Sanders 1987; Fletcher 1976: 525-7;). A problem with the 'framing' metaphor is that it can suggest the placing and presentation of given material within boundaries which provide definition without affecting substance. Law does more than this, shaping the matter to which it is applied. This limitation makes it more attractive to think in terms of law in policing as a discourse, allowing for a more active role for participants and for the interpretive shifts in constructing cases which have been suggested above.
The criteria for legalization in police interviews with suspects are of two main kinds. The first is the legal requirements for proving a case. The second is the construction of an accepted, recorded account of events and police-suspect interaction before the ERISP. John Baldwin identifies the former process well, contrasting it with the concept of the ‘search for truth’ which was discussed above. Interviews with suspects are:

conducted within a particular context, namely the context of proof. Although it is the current fashion in official police circles to identify the purpose of questioning as being a ‘search for truth’ or as some neutral collection of information from suspects, it has to be remembered that running through all police interviewing is the expectation that, for any offence, there are a number of clearly defined features – or points to prove – which will need to be addressed .... (T)he idea that police interviewing is, or is becoming, a neutral or objective search for truth cannot be sustained, because any interviewing inevitably involves exploring with a suspect the details of allegations within a framework of the points that might at a later date need to be proved (Baldwin 1993 327).

A critical perspective on this is provided by McConville, Sanders and Leng, who argue that police use ‘legal-closure’ and ‘imperfect syllogistic’ questions as an interviewing tactic, for example getting suspects to unwittingly adopt a description of their behaviour which fits a legal offence. The apparent purpose of such questions is

to invite the suspect to provide information but in reality (they) force information into a legally significant category in the hope that the suspect will 'adopt' it. This may involve introducing some matter not previously mentioned or it may reshape what has been said so that it now 'fits' into an appropriate legal category (1991: 70).

Most importantly, suspects have to be led to speak an account of their actions which satisfies legal requirements of mens rea, so that, for example, reference is made to stealing, rather than simply to taking, or to acting recklessly, rather than accidentally (McConville et al. 1991: 70-1).
As noted above, this concentration on legalization as a manipulative tactic draws attention away from the unproblematic, necessary practice of legalizing accounts as part of the process of case construction. An example from our cases is 129, in which a man suspected of sexual assault was interviewed three weeks after initial questioning:

This ERISP appeared to be tying up loose ends in preparation for the suspect’s court appearance the following week. The interviewers reviewed some of the major claims in the suspect’s version of events (notably his claim that sex was consensual). The interviewer informed the suspect that the day after the alleged assault the complainant attended a hospital where evidence was collected by a doctor. The interviewer informed the suspect that he had arranged for a police doctor to take a sample of blood from the suspect, if he agreed. The interviewer used the ERISP also to formally inform the suspect that two micro-cassette tapes and another larger cassette obtained from the suspect’s answering machine had been returned from police custody to the suspect’s sister, at the request of the suspect’s solicitor. The ERISP also was used to clarify further a possibly misunderstood question (from the first interview) in relation to the date of an AVO (Apprehended Violence Order) violation.

Nor should attempted legalization be seen as inevitably providing benefit to police. As Baldwin suggests, '(o)fficers can be ... preoccupied with establishing relevant ‘points to prove’, albeit tackling the question mechanically and inexpertly, almost regardless of suspects' responses' (Baldwin 1993: 340). A poorly constructed question such as ‘At any time prior to or after sale of that camera did you form the belief that the goods may have been stolen?’ invites the negative response which it received in 017.
The need to fit accounts into legal frames 'makes for a somewhat staged and artificial encounter' (Baldwin 1993: 351). Suspects who are asked, for example whether they had permission to take goods which police allege they stole and whether they know that it was wrong to take the goods often regard their questioners with bemusement or even disdain: yet these questions cover elements of the theft offence. Similar problems occurred in some interviews about drug offences: for example in 148, the interviewer asked the following questions:

Did you know... that you were smoking cannabis?
How did you know you were smoking cannabis?
Can you tell me how you knew it was cannabis?
Does cannabis give you certain effects?

The suspect, clearly unaware of their purpose, was bemused. Another suspect expressed his opinion of the interviewer’s questions in 276:

A: Silly questions, man, fair dinkum. I have just told you I’ve been smoking it. It’s cannabis, all right?

The suspect then turned to the interviewer’s partner to share his amusement.

Q: Can you tell me why you have all these drugs?
A (laughing): I’m a drug addict.

The need to legalize causes some interaction to be stilted. For example, in 130, the interviewer repeatedly specified the location of the alleged offence (‘in the toilet block at X High School’). Doing so ensured that there could be no misunderstanding and that each Q&A could be used in evidence independently. However, such formalism disrupts any attempt to conduct a natural conversation of the kind likely to produce useful information.
The function of the police interview as a legalization process means that assessing police interviewing according to generalized standards of interviewing is often inappropriate. (Baldwin 1993: 327). Unfortunately, some psychologists often ignore the legal context and functions of police interviewing: for example, Shuy (1998: 9) simply refers his readers to criminal procedure textbooks. It is indicative that more than half of the (English) National Crime Faculty’s guide to investigative interviewing is concerned with legal matters (NCF 1998: 73-153).
Appreciating the significance of legalization is vital for an assessment of interviewing practice:

If interviewing techniques were to be assessed in terms of the police claim that they are geared to an objective search for the truth, then they would emerge as thoroughly deficient. But this is to miss the point of interrogations, a central feature of which is concerned with future rather than past events. Interrogations are conducted with an eye to the possibility of any subsequent trial. In that sense, their importance is concerned as much with what may be claimed later in court as with the circumstances of what happened in the original incident ...(T)he central purpose of the interview is to make any contested trial unlikely (Baldwin 1993: 351).

Such ‘legalization’ in police interviews with suspects comes in two forms. First, closed questions are used to ensure that the legal requirements to prosecute successfully are satisfied. In particular, this means proof of intention is supplied when this is a component of an offence. Suspects will also be asked closed questions which are intended to elicit a well-ordered, comprehensive account of what happened. The answers may also provide a basis for checking against other information, or other statements made by the suspect. Closed questions will often be used when checking or challenging the suspect’s account which the interviewer elicits at the beginning of the interview: this is key component of the PEACE model.
Secondly, police use interviews in order to construct an accepted, recorded account of events and police-suspect interaction (including interviews) before the electronically recorded interrogation. As Baldwin stresses, the priority is to 'limit, close-down, or pre-empt the future options available to the suspect' (1993: 351). A key to doing this is to use the interview as an opportunity not just to ask the suspect about the alleged offence, but also about his or her treatment by police. The aim is to lock the suspect into an account which will prevent him or her from subsequently challenging police accounts of or raising complaints about police activities.
In this sense, interviewing practice corresponds closely with the concern to use electronic recording as a means of preventing disputes in court and allegations of police misconduct. Large sections of the interviews in our samples were concerned, not with the alleged offence, but with the interaction between police and suspect before the ERISP began. Legalization involves the suspect being asked, in a series of questions beginning with the phrase ‘Do you agree..?’, to confirm the police account of matters such as initial contact, search, seizure of arrest, transport to station, weighing and packaging of illegal drugs. In one extreme example (246), a suspect was asked 96 ‘Do you agree?’ questions in 15 minutes. Most importantly and potentially problematically, ‘DYA’ is used to incorporate into the electronic record an account of questions and answers in earlier interviews.
DYA questions in ERISPs are used extensively to produce a record of what happened when police arrested a suspect and/or searched his or her person, vehicle of premises. If the suspect agrees with the police account, he or she will find it difficult subsequently to dispute the police account or to make allegations of wrongdoing against police. As might be expected, DYAs are particularly useful in dealing with alleged drug offences.
One Sample I ERISP suggested that the interviewer was so task-oriented in having the note book adopted that the suspect’s attempt to add a comment was overruled.

The notebook interview is read, including admissions, ‘Yeah, I did it.’ When the suspect interjects during the reading of the notebook interview, the interviewer said, ‘I will just read this and then I’ll ask you about that.’ He proceeded to read, and the suspect’s objection to the account to which he was asked to agree was ignored. (158).

It remained unknown whether the suspect intended to qualify his admission in some way or how important the substance of the comment might have been
One suspect attempted to correct information recorded in a note book interview:

The interviewer asked the suspect if he agreed that his earlier answers to police questions had been recorded in the officer’s note book. The suspect agreed, but immediately added that he had two problems with what had been recorded. Firstly, the constable had measured the length of the skid mark with his foot. In his notebook, he wrote down 27 metres, but in his report he stated 25 metres. Secondly, in his report, the constable wrote that the other car had pushed the suspect’s car five metres, whereas the suspect remembered it being two metres.

The interviewer listened to the suspect’s comments on these discrepancies, but did not make follow-up questions. The suspect raised the issue of the skid marks another few times during the interview (033).

As this suggests, legalization is not the sole province of the interviewers. The suspect in 121 apparently appreciated this. When asked at the end of the interview whether he had anything else to say, he responded:

I want you two police officers to state that I haven’t said anything off the record... that could be put in court evidence to make me look bad ... that I haven’t said anything else other than what’s on this tape.

Legalization may encourage participants to keep some information off the record. In 003, the suspect began to explain his relationship with a co-offender by saying that they met in gaol: the interviewer interrupted ‘Don’t say anything about that’.
In summary, this subsection has shown that the ‘search for truth’ is both epistemologically and practically a poor description of what occurs in many police interrogations.

It is generally believed that ‘an innocent person will not falsely confess to police unless he or she is physically tortured or mentally ill’ (Leo 2001: 37). As Leo says,

This myth is ... completely false. The social science research literature has amply demonstrated that contemporary methods of psychological interrogation can, and sometimes do, cause cognitively and intellectually normal individuals to give false confessions to serious crimes of which they are entirely innocent (Leo 2001: 37; see also Gudjonsson 2003: 9).

In fact, it is clear that no special skill is required to produce a false confession: persistent questioning, misrepresentation of other evidence, and the unintentional ‘leaking’ of information to suspects can produce unreliable confessions, particularly (but not necessarily) if the suspect is vulnerable through, inter alia, youth, mental illness, or intellectual disability. The short-term benefit – relief from the questioning – may seem worthwhile, even if its long-term cost may be enormous. Confident belief that people will act rationally and in their best interests is inappropriate and potentially very misleading in the context of police interrogation.
Developments in DNA science have fundamentally changed the situation, providing something closer to positive proof of innocence. DNA analysis may prove not only that the convicted person was not guilty, but may identify the real offender. The results of the Innocence Project have raised major concerns about the integrity of the US criminal justice process: one quarter of those exonerated by DNA had confessed to the crime of which they were wrongly convicted (Scheck et al. 2000).
In some crucial cases, it has been definitively shown that suspects confessed to crimes which they did not commit. There is space here to comment on just two. First, the Central Park jogger case, in which a woman was viciously raped and beaten in Central Park. For many, the case defined and symbolized the dangerousness of New York in the early 1990s. According to the confessional accounts which convicted 5 young men (four of which were video-recorded), the attack was an incident of racially motivated ‘wilding’. However, we now know that the five had not attacked the woman: her attacker was a man acting alone who eventually came forward, his confessions confirmed by DNA testing of the semen he left on the woman’s clothing. The five who were convicted had confessed: but these confessions were preceded by the police showing one of the suspects photographs of the victim and taking them to the crime scene, were at times self-contradictory, and drew on information leaked by interrogators (eg about the weapon used). The suspects were (perhaps unintentionally) offered an inducement to cooperate, and were detained and questioned for 14 to 30 hours (Kassin 2002, 2003).
In the Cardiff 3 case, three men were convicted of the brutal murder of a young woman, Lynette White who was a sex worker (Sekar 1997). According to the prosecution’s lurid account, five men and two women had been present at the murder. One man confessed: the women gave witness statements. But again, it eventually emerged that the killer was not one of those charged or convicted, but one man (a client) acting alone. There had been no witnesses. Again, the killer was traced by use of DNA testing. The confessions and the witness statements were false. These cases are important because they provide unchallengeable evidence that people can confess to appalling crimes of which they are not guilty. This does not involve an uncritical faith in science: the limitations of DNA analysis have to be acknowledged. Nonetheless, the myth that innocent people do not confess to crimes that they have not committed has been exposed by the establishment not just that people were not guilty, but that they were innocent.

The image of criminal justice as balancing crime control and due process structures continues to structure public understanding and debate. Police have often encouraged the belief that suspects rights and police effectiveness are at opposite ends of a see-saw: if one goes up, the other must go down. Nowhere has such rhetoric been more prevalent than in discussions of interrogation. A notable example is the resistance by some US police to electronic recording of interrogation and the interrogation trainers’ insistence that effective interrogation requires the isolation of the suspect (Inbau et al 2001).
It is mark of the parochialism of so many criminal justice policy-makers and commentators that such claims continue to be made. The experience of England and Wales since the mid-1980s should be well enough known to require cursory description. The Police and Criminal Evidence Act 1984 introduced wide-ranging protections for suspects, including time-limited detention before charge, contemporaneous (now electronic) recording, and access for vulnerable suspects to ‘appropriate adults’ and for all, crucially, access to publicly-funded and organised legal advice schemes. Criminal investigation is regulated by a complex network of legal and administrative rules. The secrecy of police stations was challenged: officers became used to seeing solicitors, social workers, family members and researchers in custody areas. And, despite the fears of many, the sky remained in the heavens, suspects continued to confess, and police effectiveness was not reduced. On the contrary, the general response of police was retrospectively to welcome the new criminal process. Such changes of heart were even more marked when Australian jurisdictions introduced electronic recording: after years of dogged resistance and obstruction, police came to realise the great benefits that it brought them (Dixon 2007: ch 1).
When, under pressure to respond to demonstrated miscarriages of justice, it is accepted that reform is needed, a characteristic response in Australia and the United States has been to over-rely on one regulatory tool - electronic recording. In the United States, calls for comprehensive recording have become almost de rigeur in recent papers. A disappointing characteristic of almost all such discussion is a parochial failure to refer to the long experience of Australian jurisdictions which have been using audio-visual recording for almost two decades, not just in field trials, research experiments or selected cases, but routinely for questioning about all indictable offences. The Australian experience provides important guidance for other jurisdictions considering the routine, compulsory audio-visual recording of police questioning of suspects.
The key findings of my empirical research (2007) are that the concerns usually expressed about electronic recording have not been substantiated. The interrogator’s job has not become impossible: suspects continue to make confessions and admissions. As seems typical, police opinion has shifted from antagonism and scepticism to acceptance and enthusiasm. The removal of the taint of verballing has benefited the criminal justice process, both ideologically by increasing public confidence and instrumentally by reducing the court time spent on challenges to confessional evidence. However, there are problems. Audio-visual recording may be highly problematic if it is used, as appears typical in the US, to record only a repeat of a confession made during previously unrecorded questioning. While audio-visual recording makes simple verballing impossible, inadequate legal regulation of the process means that recorded interviews are often preceded by unrecorded questioning which may undermine the reliability of any subsequent recorded confession. Secondly, there is a worrying tendency amongst some criminal justice professionals (notably judges) to believe that they can use the recorded image to read the suspect’s body language. The general problems of reading deception from body language were noted above. They are compounded by the influence of the visual image. Thirdly, electronic recording is regarded as a panacea which makes other regulation superfluous. Notably in Australia, this has meant that while some of the structure and terminology of PACE have been imported from England and Wales, there seems at best limited commitment to import the spirit of PACE. Just one example will be given here – the almost complete absence of solicitors from Australian police stations. While Australian jurisdictions provide various rights of access, these remain hypocritically empty rights because no funding is provided for legal advice schemes.

In a criminal justice world increasingly characterised by policy transfer and convergence, interrogation bucks the trend, with notable differences between dominant and approved models in the US and the UK (with Australia, New Zealand and some European countries following the latter). Not long ago, the situation was more normal/expected, with US criminal justice having a significant influence on the UK (Walkley 1987). The first objective of this section of the paper is to examine some of the conditions of this policy divergence

The dominant US model requires little description to anyone who has watched fictional TV ‘police procedurals’ such as Law & Order: once a suspect has been identified through preliminary investigation and questioning, accusatory interrogation techniques are used to draw out a confession. The suspect resists, but the expertise of the police detective leads him or her to crack and to confess, providing neat dramatic closure.
A key source of these fictions is the training manuals used by US police. There are a number, but the most successful and influential is Inbau & Reid (Inbau et al 2001), ‘the most influential approach and the predominant model of interrogation used by law enforcement professionals in the USA’ (Borum et al 2009: 117) which has developed through a number of editions into the ‘Reid nine step technique’. The history of this manual is almost a history of modern interrogation in the USA. As Richard Leo has shown, its origins were in the reaction against the violence in treatment of suspects which became a scandal in the 1930s. Police sought alternatives to violence and found it in psychology: indeed, the history of interrogation provides a great casestudy of the influence of psychology in American life. Psychological pressure is therefore at the heart of police interrogation in the USA (Leo 2008).
In the Reid Technique, a suspect is only interrogated after initial informal interviewing makes the investigator confident of his or her guilt. This is a crucial stage, because thereafter the investigator’s aim is to obtain a confession confirming the original suspicion. To this end, psychological pressure is put on the suspect building up anxiety and facilitating confession. The suspect is helped to confess by, for example, the investigator suggesting accounts of what occurred which minimise the suspect’s culpability, immorality or social abnormality. Interrogators may lie to suspects, typically about the evidence available against him or her. Like most other US interrogation styles, the Reid Technique puts considerable emphasis on the ability of interrogators to read the suspect’s body language.
Whatever else the Inbau & Reid model may be, it has to be understood as a commercial product. In the vast literature which has accumulated on privatization of policing in recent decades, remarkably little attention has been paid to the role of corporate providers of training and other services (including interrogation). The Inbau & Reid model is a product which has been sold in the form of both manuals and training courses. Others compete in the criminal justice market place, but Reid & Co retains a dominant position. Such companies increasingly go beyond training, providing contracted interrogation services either in an advisory or operational capacity. Given these commercial realities, it is little wonder that those associated with Reid & Co fiercely defend their brand (eg Buckley 2000).
Also like other US training manuals, Reid & Co claim that their technique is validated by long operational experience and success. Its critics argue that there has been no scientific evaluation. Indeed, Borum concludes his review by stating

‘Almost no empirical studies in the social and behavioral sciences directly address the effectiveness of interrogation in general practice, or of specific techniques in generating accurate and useful information from otherwise uncooperative suspects’ (2006: 18).
Borum et al are scathing about the ‘lack of science’ underpinning ‘nearly half a century of US doctrine’ (2009: 123). One area in which there has been very extensive research is the detection of deception: as noted above, the general conclusion is that the ‘lie-signs’ of which so much is made are (a) culturally specific (so, most obviously, gaze aversion indicates respect for authority rather than deception in some cultures) and (b) while highly trained psychologists may be able to identify deception at a higher rate, most police, prosecutors, judges and jurors will do so at a rate little better than chance.
There is no doubt that such techniques produce confessions: this is why they have been so consistently popular in US law enforcement. The key problem is that the reliability of such confessions cannot be guaranteed: as will be shown below, the concern over miscarriages of justice have made this a major issue. As noted above, the step from interview to interrogation is vital and, however much Inbau et al defend their model, it is its Achilles heel. It is now almost a cliché to point out that many investigations have gone astray (and some led to miscarriages of justice) by investigators deciding too early that their ‘case theory’ is correct, then interrogating to have it confirmed, and ignoring inconveniently contrary evidence or alternative lines of inquiry along the way.

In contrast, the UK model sees questioning as investigative rather than accusatory. This approach has been widely adopted in Australia and New Zealand, and is attracting increasing interest in continental Europe (Williamson et al eds, 2009). The title ‘investigative interviewing’, not interrogation, is deliberate and significant. Officers are warned against assuming the suspect’s guilt and encouraged to investigate with an open mind. Investigative interviewing is structured around the mnemonic PEACE, which means: preparation and planning; engage and explain; account, clarification and challenge; closure; evaluation Rather than directive questioning, the investigator is trained to allow the suspect to provide their own account of events, which can then be checked and challenged if necessary against other evidence. Questioning of suspects is downgraded: it is just one of a series of skills, including questioning witnesses. Deception of suspects and psychological coercion of the kind used in the Reid Technique are unacceptable, and any resulting evidence should be inadmissible. PEACE is more important for its impact in undermining previously accepted strategies than for the specific techniques of information gathering and checking which it incorporates. At its heart is a simple, but crucial shift of emphasis: rather than setting out to gain a confession which confirms a case theory to which the officer is firmly committed, the interrogating officer is encouraged to get the suspect’s account and then to check its authenticity by questioning and testing it against other evidence.
Like the Reid Technique, investigative interviewing has to be understood as a product. While the commercial aspects of it are less significant, it would be naïve to overlook its material dimension: books are sold and academic and practitioner careers have been progressed on the back of investigative interviewing. However, this is a secondary element. Investigative interviewing was championed by a moral and policy entrepreneur in Tom Williamson, an influential senior police officer whose commitment to improving investigative practice had a strong ethical foundation. Investigative interviewing was developed by a network of psychologists and police practitioners. Its influence spread through a network of police trainers and academic psychologists, notably through the International Investigative Interviewing conferences

In order to understand how and why interrogation has diverged so markedly in the UK and the US, the role of miscarriages of justice must be understood. Reform of the criminal justice process in England and Wales was driven by a well-documented series of major miscarriages of justice beginning with Confait and running through a series in which the Guildford 4 and Birmingham 6 were the most publicised. These provided a background of distrust of police and shaken confidence in beliefs about the matchless superiority of English justice.
However, two less well known cases had the decisive influence on interrogation policy by making the police realise that change was needed (and that if they didn’t change themselves, that change would be imposed). First, in the Cardiff 3 case, the Court of Appeal delivered damning criticism of police practices. As had been happened in the Confait case twenty years before, subsequent disclosures caused even more problems for the police: Sekar’s relentless investigative journalism and the development of DNA science proved not just that the Cardiff 3’s conviction had been legally unsustainable, but that it was factually wrong: someone with who they had no connection was eventually convicted. The second case was a miscarriage of a different kind: not a false conviction, but a botched investigation which led to no conviction and the escape from justice of the murderer of a child (which may or may not have been the accused: the way a suspect was interrogated made the line between truth and conjecture indiscernible. George Heron’s ‘confession’ was the product of interrogators who used techniques familiar in the US – identification of a ‘good suspect’, followed by pressure, misrepresentation of witness evidence, and the offering of culpability-reducing ‘themes’. However, the prosecution did not make it past the legal first base of a Crown Court trial: the confession was excluded, and the case collapsed. Heron was released to a life of hiding from the family of his alleged victim’s family. Largely overlooked by lawyers because it did not reach their notice via an appeal court report, how then did a failure by a provincial police force have such an impact? The answer lies in an internal report reviewing the case which included the participation of Tom Williamson, a senior police officer who became the key figure in the development of a new model of interrogation.
Appalling as the miscarriages of justice in England and Wales were, they are overshadowed by what happened in the USA. Researchers and activists associated with the Innocence project and its off-shoots have proved that scores of Americans were sentenced to death for crimes that they did not commit. If they were saved from execution by positive proof of innocence, the conclusion is unavoidable that some innocent people were executed after making false confessions. While one might have expected a liberal-democratic society to treat wrongful incarceration and probably execution of the innocent as a national disgrace. Instead, electronic recording (or, sadly, ‘recordation’) of confessions has been promoted and, it appears, widely accepted as a panacea.

It would be wrong to exaggerate the distance between policing practices in the US and the UK. Most obviously, the fact that investigative interviewing is the approved model in the UK does not mean that it is universally applied. Indeed, the major evaluation by Clarke and Milne reported PEACE’s impact to be quite limited. No doubt, Inbau & Reid is still read by some UK police. More significantly however, there is growing interest in the USA in alternative approaches to interviewing. Notably, perhaps surprisingly, some such interest has come from those responsible for questioning terrorist suspects.
Since the revelations of how terrorist suspects were treated in Guantanamo Bay, Afghanistan, and the American gulag of secret prisons, there has been much debate about the definition, morality and legality of torture. There has been less intense discussion of whether practices such as water-boarding are ‘even effective, necessary or useful’ (Borum et al 2009: 117 ). It is sometimes regarded as unnecessary or distasteful to go beyond the ethical/legal debate except to repeat the accepted wisdom that torture produces unreliable information. Recently, this area has been subjected to valuable expansion and deeper analysis by American officials with operational experience in questioning terrorist suspects. Their conclusion is that the debate about the ethics of torture is made virtually redundant for policy purposes by, first, the appreciation that the ‘ticking bomb hypothetical which structures much of the torture debate is ‘rife with assumptions that run counter to nearly any “real-world” intelligence interrogation’ (Borum et al 2009: 115) and that non-coercive means are more effective in obtaining information from suspects.

‘Some in the US law enforcement community believe that experience with detainees in Guantanamo Bay, Afghanistan, and Iraq – both positive and negative – generally support the effectiveness of “rapport-based” methods and reveal the myriad of problems that are associated with more coercive or aggressive tactics’ (Borum et al 2009: 118 ).
These experiences led to the development of what is hailed as a ‘new era in US intelligence interrogation’. Almost inevitably it seems, this has been christened with a grisly neologism – ‘educing information’[16]. ‘The term educe was chosen because it denotes a drawing out or elicitation of information, specifically information that may be hidden, unexpressed, or latent’ (Borum et al 2009: 119). This approach is much closer to investigative interviewing than to the Reid technique: it is not coincidental that one of the authors, Michael Geddes, has been an active participant in International Investigative Interviewing conferences. EI seeks information rather than a confession not just (as in investigative interviewing) because of the problems in an accusatory approach but also because the primary objective is to gain ‘actionable intelligence’ which can be connected to other information resources (Mackey 2004) rather than evidence against the individual suspect. It will be of great interest to see the extent to which this penetrates the US law enforcement community, not least how Reid & Co respond to the potential challenge to their market position.

How do the models of interrogation outlined above fit with the account of police interrogation which was developed in the first part of the paper by challenging a series of eight myths? As this sort of critique informed the debate in England and Wales, it is not surprising that investigative interviewing’s conception is nearer to what I suggest is reality than the American model, exemplified by the Reid Technique. The myths are put in four groups for the purpose of this concluding comment.

While the American model assumes that properly trained officers will acquire the skill necessary to break down resistant suspects (which includes dominating the suspect through isolation and behavioural control), the investigative interviewing model downgrades confession as the objective of questioning suspects in favour of information gathering. In the latter, training focuses on basic principles of good interaction rather than specialist skills (although such training is also provided in the higher ‘tiers’ of the model: see below).

While this appears to be taken for granted in most American studies, almost the opposite is true among exponents of investigative interviewing (at least as regards everyday policing). This stark contrast between the US and the UK would have to be explained in part by the very specific role which attempts to find a scientific replacement for the third degree played in the history of American interrogation (Leo 2008: ch3) but also more deeply in cultural transatlantic contrasts.

The American model assumes that miscarriages of justice are rare and that if they do occur in the context of interrogation, it is because the suspect was particularly vulnerable through youth or intellectual disability. Leo’s research on miscarriage cases shows that this assumption is quite wrong (2008: ch 6). The continuing problems which American jurisdictions are having in coming to terms with the evidence that many people have been wrongly convicted are surely due at least in part to this fundamental misunderstanding of how cases miscarry. As research into the impact of Miranda and subsequent Supreme Court cases has shown, suspects’ rights remain more significant as rhetoric than as substance (Leo 2008). In England and Wales, miscarriages have become major public scandals in a way that the American (and, despite Chamberlain the Australian) cases have not. This recognition led directly to PACE and accompanying reforms of justice processes in which provision of substantial rights for suspects has been achieved without reducing efficiency (indeed from many perspectives increasing it)

Why do suspects confess? The short answer, of course, is police interrogation, since suspects almost never confess spontaneously but virtually always in response to police pressure. (Leo 2008: 162)

(Confessions) are actively elicited, and often demanded of the suspect, then jointly shaped, typically after sustained psychological pressure (Leo 2008: 119)

This would suggest that there are much greater differences between the American and the English (and Australian) justice processes than previously thought. However, there appears to be no convincing evidence that this is the case: similar pressures for cooperation and non-conflictual settlement characterise both systems. Indeed, bargaining (over pleas, charges and sentences) is even more structured into the American system than its comparators. It is hard, therefore, to avoid the conclusion that the American literature does not tell the full story. Inbau & Reid and the authors of training manuals have no more incentive to de-emphasise the conflictual elements of interrogation than a dietician has to say the way to lose weight is simply to eat less. Richard Leo presents his Police Interrogation and Criminal Justice as ‘a comprehensive study of police interrogation in America’ (2008:5). It is perhaps better read as a study of how non-compliant suspects in serious cases[17] are questioned by detectives. The last point is significant: all Leo’s interrogators were detectives (and all apparently men), so we are left to wonder about the role of uniformed officers (and female detectives) in questioning suspects. Leo’s paints a picture of detectives characterised by ‘one-sided scepticism, conviction psychology, and game approach to interrogation’ who are ‘motivated by a means-end rationality in the interrogation room’ (2008: 23) and for whom interrogation is ‘a strategic, multistage, goal-directed, stress-driven exercise in persuasion and deception’ (2008: 119). This is an example of what I described above as an over-socialised account of police work. The empirical base for his book is observation of ‘more than one hundred interrogations’ inside the Criminal Investigation Division of the Oakland Police Department, as well as sixty video-taped interrogations by two other police departments, followed by ‘several hundred more electronically recorded interrogations by American police departments across the United States’ (2008: 5). Given that electronic recording in the US has, until recently been reserved for very serious cases (and has usually been the recording of a rehearsed confession), it seems likely that Leo’s data do not provide the basis for a ‘comprehensive study’: this would need to be of a sample randomly chosen from all interrogations.[18] Leo’s account of American interrogation may be accurate as regards serious, contested cases, but it is not a convincing account of everyday investigative practices in the cases which make up the bulk of the criminal justice process’s workload. I pick Leo’s book for critical comment not because it is bad: on the contrary it is by a long distance the best book in the field, providing an excellent account of the rise of modern interrogation and of miscarriages of justice. Indeed, it is the only substantial empirical study carried out in the USA. As Leo notes, ‘though interrogation is central to the study of policing, the criminal investigation process, and modern detective work..., criminologists and sociologists have largely ignored it’ (2008: 5). A factor which seems inexplicable is the dearth of field-based, empirical research on US police interrogation. So much talent, so many resources ... so many case-focused law journal articles, and so few sociological or criminological studies. Perhaps if this gaping hole could be filled, a different perspective on American interrogation would emerge.
A final word on the mundane reality of policing: very few police officers will ever interrogate a murder suspect. Most of them, however, will question the kind of suspects who feature above in cases drawn from my research: people who shoplift food from supermarkets, steal trolleys from hired vans, blow up school toilets, or use illegal drugs. This mundane reality needs to be recognised by researchers, practitioners, and policy makers.

Baldwin, J (1992) Videotaping Police Interviews with Suspects – an Evaluation Police Research Series Paper #1 (London: Home Office Police Department).

--- (1993) 'Police interview techniques: establishing truth or proof?' British Journal of Criminology 33: 325-52

Bennett, WL & Feldman, MS (1981) Reconstructing Reality in the Courtroom (New Brunswick: Rutgers University Press)

Borum, R (2006) ‘Approaching truth: behaviral science lessons on educing information frm human sources’ in Intelligence Scince Board 2006.

--- Gelles, MG, & Kleinman, SM (2009) ‘Interview and interrogation: a perspective and update from the USA’ in T.Williamson, B.Milne & S.Savage, eds International Developments in Investigative Interviewing (Collumpton: Willan).

Bottomley, AK, Coleman, CA, Dixon, D, Gill, M & Wall, D (1991) The Impact of PACE: Policing in a Northern Force (Hull: Centre for Criminology & Criminal Justice)

Buckley, J (2000) ‘Defending the Reid Technique of interrogation’ paper for Reid Conference for Investigative Training, Nov 6-8 2000, available at http://www.reid.com/educational_info/critictechniquedefend.html (accessed 19 Nov 2009).

Clarke, C & Milne, R (2001) National Evaluation of the PEACE Investigative Interviewing Course, Police Research Award Scheme Report # PRAS/149.

Dixon, D. (1997) Law in Policing: Legal Regulation and Police Practices (Oxford: Clarendon Press).

---- (2005) ‘Why don’t the police stop crime?’ ’ Australian & New Zealand Journal of Criminology, (2005) 38: 9-16.

--- (2006) ‘”A window into the interviewing process”? The audio-visual recording of police interrogation in NSW Australia’ Policing & Society 16: 328-48

--- (with G Travis) (2007) Interrogating Images (Sydney: Institute of Criminology)

Domanick, J (1994) To Protect and Serve: The LAPD’s Century of War in the City of Dreams (New York: Pocket Books).

Drizin, S.A. & Leo, R.A. (2004) ‘The problem of false confessions in the post-DNA world’, North Carolina Law Review

--- & Reich, M.J. (2004) ‘Heeding the lessons of history: the need for mandatory recoding of police interrogations to accurately assess the reliability and voluntariness of confessions’, Drake Law Review 52: 619-46.

Ericson, R. (1981) Making Crime: A Study of Detective Work (Toronto: Butterworths).

Fletcher,GP (1976) ‘The metamorphosis of larceny’ Harvard Law Review 89: 469-530

Griffiths, A & Milne, B (2006) ‘Will it all end in tiers?’ in T.Williamson, ed. Investigative Interviewing (Cullompton: Willan) 167-89

Gudjonsson, GH (2003) The Psychology of Interrogations and Confessions (New York: Wiley)

--- (2007) ‘Investigative interviewing’ in T.Newburn, T Williamson & A.Wright, eds Handbook of Criminal Investigation (Cullompton: Willan) 466-91

Inbau, FE, Reid, JE, Buckley, JP & Jayne, BC (2001) Criminal Interrogation and Confessions 4th edition (Gaithersburg, MD: Aspen).

Intelligence Science Board (2006) Educing Information: Interrogation Science and Art: Foundations for the Future (Washington: National Defense Intelligence College)

Kassin, SM,(2002) ‘False confessions and the jogger case’ New York Times 1 November 2002.

---- (2003) ‘The Central Park jogger case’, paper presented to the Psychology & Law conference, Edinburgh, July 2003.

---, Drizin, SA, Grisso, T, Gudjonsson, GH, Leo, RA & Redlich, AD (2009) ‘Police-induced confessions’ Law & Human Behaviour (Available at SSRN: http://ssrn.com/abstract=1483878)

Kleinig, J (1996) The Ethics of Policing (Cambridge: Cambridge University Press)

Lassiter, GD, Slaw, RD, Briggs, MA & Scanlan, CR (1992) 'The potential for bias in videotaped confessions' Journal of Applied Social Psychology 22 (23) 1838-1851

--- & Irvine, AA (1986) 'Videotaped confessions: the impact of camera point of view on judgments of coercion' Journal of Applied Social Psychology 16: 268-76.

Leng, R. (1993) The Right to Silence in Police Interrogation, Royal Commission on Criminal Justice Research Study # 10 (London: HMSO)

Leo, R (1994) 'Police interrogation and social control' Social & Legal Studies 3: 93-120.

--- (1996) 'Inside the interrogation room' Journal of Criminal Law & Criminology 86: 266-303.

--- (2001) ‘False confessions: causes, consequences and solutions’ in SA Westervelt & JA Humphrey, eds. Wrongly Convicted: Perspectives on Failed Justice (New Bruswick: Rutgers University Press) 36-54

--- (1992) ‘”Big-bang” decisions: notes on a naturalistic model’ in K. Hawkins ed The Uses of Discretion (Oxford: Clarendon Press) 249-85

--- & Hawkins, K (1989) ‘Police decision-making’, in M.Weatheritt ed Police Research (Aldershot: Avebury) 139-56

--- & --- (1990) ‘Legal decisions: a frame analytic perspective’ in S.Riggins ed Beyond Goffman (Berlin: Aldine DeGruyter) 201-33

McConville, M., Sanders, A & Leng, R (1991) The Case for the Prosecution (London: Routledge)

McGurk, BJ, Carr, MJ and McGurk, D (1993) Investigative Interviewing Courses for Police Officers: An Evaluation, Police Research Series Paper # 4 (London: Home Office Police Department).

Mortimer, A & Shepherd, E (1999) 'Frames of mind: schemata guiding cognition and conducting the interviewing of suspected offenders' in Memon & Bull: 1999: 292-315

Moston, S. & Engelberg, T. (1993) 'Police questioning techniques in tape recorded interviews with criminal suspects' Policing and Society 3: 223-37.

---- & Stephenson, G.M. (1993b) 'The changing face of police interrogation' Journal of Community & Applied Social Psychology 3: 101-15.

Navarro, J & Schafer, JR (2001) ‘Detecting deception’ FBI Law Enforcement Bulletin 70 (7) 9-13.

NCF (1998) A Practical guide to Investigative Interviewing 2nd ed (Bramshill: National Crime Faculty)

Pearse, J & Gudjonsson, GH (1996) 'Police interviewing techniques at two South London police stations' Psychology, Crime and Law 3:63-74

Scheck, B., Neufeld, P., & Dwyer, J (2000) Actual Innocence (New York: Doubleday)

Sanders, A (1987) ‘Constructing the case for the prosecution’ Journal of Law & Society 1: 229-53
Sekar, S (1997) Fitted In: the Cardiff 3 and the Lynette White Inquiry (London: The Fitted In Project).

Shearing, CD & Ericson, R (1991) ‘Culture as figurative action’ British Journal of Sociology 42: 481-506.

Shuy, RW (1998) The Language of Confession, Interrogation and Detection (Thousand Oaks: Sage)

Skolnick, J.H. (1993) 'Justice without trial revisited', in D. Weisburd & C. Uchida (eds.) Police Innovation and Control of the Police (New York: Springer-Verlag) 190-205.

Van Maanen, J and Pentland, PT (1993) ‘Cops and auditors’, in SM Sitkin & RJ Bies, eds The Legalistic Organization (Thousand Oaks, Cal.:Sage) 53-90

Williamson, T. (1993) 'From interrogation to investigative interviewing', Journal of Community & Applied Psychology 3: 89-99.

[1] Dean, Faculty of Law, UNSW, Sydney. This is a working draft: please contact the author before citation: d.dixon@unsw.edu.au
[2] An obvious criticism of these comments will be to say that it is naïve and that the real action (including threats, inducements and violence) takes place elsewhere, before the recorder is turned on. My response is in the section on regulation below and, at more length, in Dixon 2007: ch 2.
[3] Crimes (Sentencing Procedure) Act 1999 s.22.
[4] For details of the samples and methodology, see Dixon 2007: ch2.
[5] The desire for a ‘turn in the spotlight’ may be connected to more grotesque examples of public confessions on TV shows like Oprah and Jerry Springer.
[6] He had no idea how the heroin had got into his wallet, and had found the bag of marijuana in grass at the side of the road. Moreover, he had dropped it when his bicycle went over a bump, not because he knew police were following him.
[7] For details, see Dixon 2007 ch2. A strength of these data is that the samples were randomly selected from, respectively, all interrogation in one year and all interrogations by officers who had received investigative interviewing training.
[8] For further details, see Dixon 2007: ch 4. In both samples, a number of records were indecipherable for technical reasons.
[9] Such cases also raise questions about the reliability of such confessions made in ‘informal’ interviews and of the propriety of allowing them and subsequent recorded confessions to be used as evidence (Dixon 2007: 45-54).

[10] It is of course possible that suspects in our sample were subsequently re-interviewed. For example, the interviewee in 206 was suspected of armed robbery, but the interview in our sample was merely a preliminary discussion about his earlier movements. In 225 the suspect was questioned about a fraudulent mortgage application. However, it was clear that the interviewer was interested in this suspect primarily as a lead to more significant players in a series of organized frauds. It is often not clear at the end of the interview whether the suspect is to be charged.

[11] Mike Todd, Greater Manchester Police, quoted, Weekly Telegraph 27 August 2003, p.2.
[12] For details, see Dixon 2007: 27-8 & ch 6.
[13]
[14] R v. Heron, unreported, Leeds Crown Court, 1 Nov. 1993; see Dixon 1997: 172-6; Gudjonsson 2003: 96-106
[15] For a more detailed analysis, see Dixon 1997.
[16] Other examples are the redesignation of interrogators as ‘HUMINT (human intelligence) collectors’ or ‘educing information professionals’ (Borum et al) and the ‘recordation’ of interrogation
[17] Leo qualifies the second quoted statement above with ‘especially to serious crimes’ (2008: 119), which undercuts the claim to generalisability
[18] For an example of how this can be done, see Dixon 2007: ch 2

University of New South Wales Faculty of Law Research Series

Dixon, David --- "Interrogating myths. A comparative study of practices, research, and regulation" [2010] UNSWLRS 40

University of New South Wales Faculty of Law Research Series

Interrogating myths. A comparative study of practices, research, and regulation [2010] UNSWLRS 40 (5 October 2010)