If I have misquoted, to borrow the words of the late Groucho Marx: ‘Quote me as saying I was misquoted’.
As a lawyer and part-time academic, I am a huge fan of learning from multi-faceted disciplines such as psychology, sociology, and criminology, to be considered alongside the Law of England and Wales. Indeed, I believe it to be essential if one is to obtain a deeper understanding of a given subject by looking at the same problem from different angles and perceptions. To ignore the wealth of information from other disciplines is to hold a very constricted and narrow view. I very much hope to encourage fellow lawyers to look beyond case-law and Acts of Parliament particularly on the issues of suggestibility, compliance, acquiescence, and confabulation, when considering what factors led to a confession within the context of the reliability on it from a vulnerable witness.
Over the years, evidence taken from vulnerable witnesses (‘VWs’) interviewed, has led to cases collapsing in Court, such as ‘The Tottenham Three’ in 1991, when one of the confessions at police interview was challenged and deemed a false confession, unsafe, and therefore overturned. There were no special measures put in place when interviewing and taking the confession of Engin Redhip who was of low intelligence. During the interview, there had also been an abnormal level of suggestibility. Grounds, A, (1995), p95 In order to ensure a fair and just legal system and treatment of a witness, special care must be taken, and safeguards put in place during the interviewing process to ensure that evidence taken from those who are vulnerable, can be relied upon in Court, or else the evidence becomes unsafe and unreliable. Rozenburg (1992).
‘No legal system should ignore the risk of false confessions…’ The aim is…’…to reduce confessions and prevent miscarriages of justice’.Gudjonsson GH, (2003), Loc 137.
1. Who are vulnerable witnesses (‘VWs’)?
In England and Wales, (hereinafter referred to as ‘England’, for brevity), Zander, M (2013) p348-p352, makes reference to those who are VWs, as defined by the Police and Criminal Evidence Act 1984, (‘PACE’), Code of Practice C, as the young, Zander, M (2013), p348, a suspect who is mentally disordered or otherwise mentally vulnerable, Zander, M (2013), p351, the deaf, Zander, M (2013), p351, those who cannot speak or understand English, Zander, M (2013), p352, and the blind, Zander, M (2013), p352. The Home Office (1992), also makes reference to people who are mentally ill and/or have learning disabilities, as people who are VWs. Gudjonsson, GH, (undated), p16 goes beyond the statutory definition in PACE by considering anyone with psychological vulnerabilities, and learning disabilities, including those who are vulnerable to sexuality, specific fears, and low esteem.The Home Office (2002/2008a), allows special measures to be available when vulnerable and intimidated witnesses are interviewed by the Police.
2. What special measures are in place to protect vulnerable witnesses in the interview process?
The Police and Criminal Evidence Act 1984, Code of Practice C, (‘Code C’), relating to Detention, Treatment and Questioning, and the Notes for Guidance, (‘the Notes’), sets out various measures in place when dealing with VWs. Zander, M (2013), p348-352.
For the young, an Appropriate Adult (‘AA’), must be informed and asked to come to the Police station. An AA can be a parent/guardian/Care Authority, or a voluntary organisation, a social worker, or a member of a youth offending team, or another responsible adult, but not a police officer. PACE, Code C, para. 3.13; Note 3c. Further, ‘a juvenile…must not be interviewed regarding their involvement or suspected involvement in a criminal offence or offences or asked to provide or sign a written statement under caution of record or interview in the absence of the appropriate adult unless paragraph 11.1 or paragraph 11.18 to 11.20 apply’. An AA is also needed for someone who is mentally disordered or mentally vulnerable PACE, Code C, para 3.15. In relation to someone who is deaf, that person cannot be interviewed without the appropriate interpreter. PACE, Code C, para 13.5. Someone who is blind must have someone to assist, to read and understand documents, although it is not an AA as required for those who are young, juvenile, mentally disordered, or mentally vulnerable. Code C, para 3.20. For those who cannot speak or understand English, they must have an interpreter. This also applies to a parent or a guardian or a juvenile being interviewed by the Police. PACE, Code C, para 13.2A.
3. How does an AA know what to do and what to say in an interview with a vulnerable witness, and is the person who attends as an AA suitable?
The consequences of an AA being present, but unable to fulfil their role is that the supposed special measure in place by having them attend, in order to protect a VW is thwarted. An AA’s role is to ensure that the VW does not provide an unreliable statement Gudjonsson, GH (2003) loc 1193. The use of an AA resulted from an inquiry as to the treatment of three juveniles, and how to prevent such treatment of vulnerable suspects in the future. Fisher, H (1977).
Gudjonsson set about analysing the measurement of ‘oppressive Police Interviewing Tactics in Britain’, in 20 serious crime investigations, and all aspects of the police interview process, including the role of the AA. Gudjonsson, GH (2003), loc 1295. In the 20 cases analysed, an AA was present in six of the interviews. Two of those AAs sat silently, at times when they ought to have been proactive. Three of the AAs encouraged a confession of the VW, even when it was not necessarily appropriate to do so, and only once did an AA do their job. Gudjonsson, GH (2003), loc 1362.
4. When does the interview process start, where, and how is what is said recorded as evidence?
Theoretically, upon a caution having been given in England, anything said may be used as evidence. Any formal interviewing is supposed to take place in compliance with PACE, preferably inside of a Police station where rights are repeated, including a right to notify someone of your arrest (S56 PACE), and a right to seek legal advice (S58 PACE). Police interviews are recorded, and on many occasions, there is a camera recording the suspect and interviewers.
The situation is significantly different in America, where there is an initial conversation/informal interview following Miranda rights being read to a suspect, which may not be at a Police station, and which is not recorded either by camera or sound. Zimbardo (1967), as cited by Gudjonsson, GH (2003), loc 257 considered that American Police consider training is ‘psychologically sophisticated and coercive’.He considers that this in itself might lead to false confessions. Zimbardo stresses the importance of interviews being carried out in a Police station in the interests of justice or else the system could be ‘open to public scrutiny’. Zimbardo (1967), as cited by Gudjonsson (2003), loc 464.
Moston and Stephenson (1993), as cited by Gudjonsson, GH (2003), loc 506
Consider that ‘without proper records of conversations there is no way of deterring the tactics used by the police and how they may have influenced the voluntariness and reliability of the subsequent confession’.
In America, the Reid technique is employed by way of ‘theme development’ Gudjonsson, GH (2003), loc 311 in the interview prior to the interrogation, where a Police officer seeks to find out about a suspect; what makes them tick emotionally; their strengths, their weaknesses, in an informal non-accusatory way to lull the suspect into a false sense of security to get the suspect to open up and be more suggestible to persuasion.
In State v Christoff (1997) Fla Cir Ct, the court considered the negative consequences of admissions by way of theme development was very dangerous and could result in false confessions.
Theme development includes: accepting moral excuses Gudjonsson, GH (2003), loc 330, normalizing criminal behavior Gudjonsson, GH (2003), loc 341, attempting to reduce a suspect’s feelings of guilt by minimizing moral seriousness Gudjonsson, GH (2003), loc 348 condemning others as a way of sympathizing with a suspect Gudjonsson, GH (2003), loc 355 using flattery as a way of manipulating a suspect Gudjonsson, GH (2003), loc 355 pointing out that a suspect’s involvement may have been exaggerated Gudjonsson, GH (2003), loc 355 making a suspect believe that it is not in his interests to continue with criminal activities.
5. What techniques are administered by Police in interviews?
There is generally a social interaction, a procedure for questioning, evidence provided which is then commented upon, and a response Gudjonsson, GH (2003), loc 5488. In England, all interviews are subject to PACE, the Codes, and the Notes to Guidance. Current tactics employed in England are based on the PEACE model, in conjunction with the Ministry of Justice’s, Achieving Best Evidence. That is: Planning and preparation, engage and explain, account clarification and challenge, closure, and evaluation. College of Policing, (2013/08)
Prior to PACE, however, England was influenced by the development of interviews and interrogations from America. Justification of employing deceptive, persuasive, and manipulative techniques are rationalised in Police training manuals. ‘…if the truth is to be revealed…persuasive interrogation techniques are essential to police work…’.Gudjonsson, GH (2003), loc 227.
Leo, RA (1994), p239 considers that deceptive strategies in America, including concealing identity, are tactics that the Police can and do employ. In England, such tactics are employed very cautiously for fear of being accused of entrapment, which would in itself be an abuse of process of the due administration of justice.
The Reid technique Inbal, F, Reid, J and Buckley, J (2001), p12 uses psychological tactics ‘involving trickery and deceit…in order to secure incriminating information from the guilty’, Gudjonsson, GH (2003), loc 270, ‘by breaking down denials and resistance and increasing a desire to confess’. Gudjonsson, GH (2003), loc 283, 288.
Kassin, S and McNall, K (1991), loc 445 argue that interrogation techniques such as the Reid techniques consist of two main strategies namely ‘maximisation’ exaggerating the strength of evidence, and ‘minimisation’ lulling a suspect into a false sense of security, ‘offering sympathy, face-saving excuses. Such tactics, Kassin and McNall argue are ‘inherently coercive’. Kassin, S and McNall, K (1991), loc 451.
Gudjonsson, GH (2003) loc 345, para 3 argues in strong opposition to the Reid technique that: ‘There is a general reluctance among the authors of police interrogation manuals to accept the possibility that their recommended techniques could in certain instances make a suspect confess to a crime that he or she had not committed’.
From tactics and strategies employed by adoption of the Reid technique amongst others sanctioned by the Police in America, creating environments of pressure by confusion, misinformation, and deception are perfect ingredients for someone who is a VW to be subject to subjectivity, coercion, false confessions and admissions, and ultimately miscarriages of justice.
6. How do the Police know when someone is a vulnerable witness, to employ special measures?
The young, the deaf, the blind, and those who cannot speak or understand English, are generally obvious to define. How do the Police know when someone is mentally disordered or vulnerable? There is no obvious definition or guidance in PACE, other than 11.18 – 11.20, Annex E, Code C,1:
‘If an officer has any suspicion, or is told in good faith, that a person of any age may be mentally disordered, or otherwise mentally vulnerable, or mentally incapable of understanding the significance of questions…that person shall be treated as mentally disordered or otherwise mentally vulnerable’.
In relation to the young, and the mentally disordered/mentally vulnerable, an appropriate adult must be present during the interview at the Police station.
So, if a Police interviewer does not know or is not told, a VW could be subjected to a Police interview without an appropriate adult present. A Police interviewer who is not told, and who does not appreciate that a witness is vulnerable, may not be treated as such. This is disturbing; how can an interviewer know that an interviewee is mentally disordered or vulnerable? Gudjonnson considered that ‘there is an absence of an operational definition for police officers, who are expected to identify the relevant psychiatric conditions without any formal training’. Gudjonsson, GH (1993a), loc 4138. Although a questionnaire has been given in advance to those being interviewed in recent years to enquire as to a suspect’s mental health including whether on medication or having feelings of depression or suicidal thoughts, such a questionnaire evolved following a number of deaths whilst in custody awaiting police interview.
What however, the Police now have in their armory for their own protection to ensure that those who are vulnerable receive special measures including an appropriate adult being present, is disclosure of a suspect’s physical and mental health, allowing Annex E, of the Code, to be effected, and to avoid an interview in which a confession is subsequently considered an abuse of process and unsafe.
Without an AA present, the interview is seemingly conducted in privacy without anyone else there other than the interviewer. That is pressure in itself to be alone in a room with a questioner. Inbal, F (2003), p51 considers that this pressure of being alone is a ‘principal psychological factor’. How much more pressure would a VW feel in such a situation in such a closed social arena. It is a highly stressful situation in which a police interviewer holds all the power and authority.
7. Suggestibility, compliance, acquiescence, and confabulation:
Primary suggestibility is a suggestion given to someone who results in an ideomotor response Noizt as cited by Gudjonsson, GH (2003), loc 5349. This is something like hypnotic auto-suggestion, and motor suggestibility, where a person is being told that he or she is being influenced, begins to sway, or unable to move certain parts of their body at the suggestion of the person giving the instructions. Gudjonsson, GH (2003), loc 5367.
There are generally two schools of secondary suggestibility theoretical approaches, which focus on sensory processes and perceptions of individuals, amongst other processes.
The first is the ‘individual differences approach’, which views ‘…suggestibility as being dependent upon the coping strategies people can generate and implement when confronted with the uncertainty and expectations of the interrogative situation’ Gudjonsson, GH (2003), loc 5316. This theory seeks to explain individual differences in interrogative suggestibility with particular emphasis upon application to a police investigation. It is based on three components namely: uncertainty, interpersonal trust, and expectation. Gudjonsson, GH (2003), loc 5591.
The measure is a change or shift in a response to questions, which are first given, and then when challenged with a negative feedback, to repeat answers to the questions, and see what answers were changed. This is known as the Gudjonsson-Clark theoretical model, used to measure levels of suggestibility based on the uncertainty triggered by a subject being told that not all of the answers given were correct.
The second secondary suggestibility theoretical approach is that of the ‘experimental approach’, put forward by Loftus. This approach considers gaining an understanding as to how leading questions might affect the way in which witnesses respond. Gudjonsson, GH (2003), loc 5322. Schooler and Loftus (1993) as cited by Gudjonsson, GH (2003), loc 5322 consider that children and persons with learning disabilities, (some of the categories of vulnerable persons), may be more accepting of misinformation, so as not to argue or challenge figures in authority. Police are clearly such people in authority, and the desire therefore to have the AA is ever more important in such circumstances, so long as they fulfill their role appropriately.
In either of the secondary suggestibility theories, an interviewer who states precisely what is expected of a VW, could directly influence how someone is to respond. Loftus (1981) as cited by Gudjonsson, GH (2003), loc 5616. In either the PEACE model, or the Reid model, there is a possibility for abuse of an influential figure suggesting an answer must be in a certain way, which could influence a decision. Given VWs are more likely to be suggestible, in the absence of an effective AA, VWs are subject to saying what the interviewer wants them to say, which could be misleading, and lead to an inaccurate or forced confession.
Schooler and Loftus consider that negative feedback to a witness reduces confidence in their own memory, and makes them more suggestive to the interviewer’s version of events. Also increased anxiety from negative feedback ‘may decrease the subjects’ ability and/or motivation to adequately scrutinise the content of the interrogator’s questions’. Schooler and Loftus, as cited by Gudjonsson, GH (2003), loc 5627.
When applying negative feedback after the first set of questioning, comparing the same with the same set of questioning and monitoring the shift, empirical findings of Gudjonsson in terms of interrogative suggestibility show:
1. There is a positive correlation between confabulations (the ability to make up facts, to fill gaps), and memory deterioration.
2. Misleading and interrogative questioning increases confabulation Gudjonsson, GH (2003), loc 5813.
3. Depressed people confabulate less because they are less imaginative and creative in that condition Sigurdsson (1994) as cited by Gudjonsson, GH (2003), loc 5819.
4. Hypnotic states (i.e. primary suggestibility), had no influence upon suggestibility Gudjonsson, GH (2003), loc 5838.
5. Those with low self-esteem, feelings of powerlessness, incompetence, high anxiety, an eagerness to please, pride, avoidance of conflict, and an aversion to a confrontation with people, and dominance of authority tend to lead to compliance rather than suggestibility. (i.e. accepting a version of events as true, that they know not to be true) Gudjonsson, GH (2003), loc 5858, 6297.
6. Anger and paranoia where there is an absence of interpersonal trust, lead to lower suggestibility levels. Gudjonsson, GH (2003), loc 5896, 6349 Also, excessive pressure can lead to increased resistance to questioning, extreme disagreement, and a ‘boomerang effect’. Brehm’s (1966) ‘reaction theory’, as cited by Gudjonsson, GH (2003), loc 6335. Blatant misleading information creates suspicion and the suspect becomes less receptive Loftus (1979b) as cited by Gudjonsson, GH (2003), loc 6335.
7. A person may acquiesce rather than be open to suggestibility when they do not understand big words, and phrases; they are unable to read and understand. It is a lack of intellectual and education that lead VW to acquiesce. Gudjonsson, GH (2003), loc 5977, 6001, 6279, 6286. Those with low intelligence and children may well give an answer which is plausible than true or guessed. If unavoided in a VW, this could lead to an unsafe confession.
8. A person with a poorer memory will be more open to suggestibility, and confabulation. Gudjonsson, GH (2003), loc 6097.
9. Memory rapidly deteriorates over a 40 or 50 minutes period of interviewing. Gudjonsson, GH (2003), loc 6116. The elderly are therefore more open to suggestibility, acquiescence, and confabulation.
10. Sleep starvation causes greater levels of suggestibility Blagrove (1996) as cited by Gudjonsson, GH (2003), loc 6182.
11. The less compliant and lower level of suggestibility, the more one is likely to resist. Higher levels of suggestibility and compliance are more likely to result in false confessions. Gudjonsson, GH (2003), loc 6459, Fig. 14.1 mean scores.
12. Compliance is more likely to lead to false confessions, than suggestibility. The Reid model uses systems of compliance and suggestibility by way of psychological manipulation Gudjonsson, GH (2003), loc 6481.
13. Younger children are more likely to make a false confession. They are more compliant and unchallenging of authority. These were the results of Redlich (1999) following accusation of the pressing of the alt key on a keyboard, the consequences of which the subjects were told, would cause the computer to crash and data to be lost. 69% accepted they hit the alt button even though none of them did. Gudjonsson, GH (2003), loc 6493.
There is a tertiary level of suggestibility, ‘involving attitude change resulting from persuasive communication originating from a prestige figure’.Eysenk and Furneaux (1945), p167. Milgram (1974), probably falls into the tertiary level, where he found a positive correlation to a subject obeying an instruction, even though it might be socially abhorrent, as a result of authority, status, and power. (i.e. electrocuting a subject on the orders of someone authoritative). Milgram (1974) considered that within a hierarchal structure, people chose to be compliant with what others were doing, and what would be the social norm to do. Applying this theory to a police interview and suggestibility to do or act or answer in a certain way, Irving and Hilgendorf (1980) as cited by Gudjonsson, GH (2003), loc 5440, considered that a strong and dominant interview may effect compliance with instructions based on obedience to a figure in power.
An interviewer can influence the way in which someone answers questions, by the positive or negative response they, in turn, give to an interviewee’s answer.
Gudjonsson, GH (2003), loc 5550 suggests that a positive response from the interviewer would encourage a witness on a particular train of thought or memory. It might not just be a verbal communication. The reward for data the interviewer wants to hear may be given with suggestions of food, or a toilet break, or something to drink; a seeming act of kindness, or perhaps lulling a witness into a false sense of security. Whereas a negative response could have an opposite effect. It would infer that the response displeased the interviewer and that the witness should adjust or modify their reply. A person who was vulnerable is more likely to shift their response and is more suggestive to the interviewer’s perceived desire to be pleased, and for the interviewer in turn to praise the vulnerable witness. The other response to negative feedback from an interviewer is a heightened response to leading questions. Gudjonsson, GH (2003), loc 5556.
8. Special measures employed for the young:
Under Code C, 11.15 a juvenile ‘must not be interviewed regarding their involvement or suspected involvement in a criminal offence or offences or asked to provide or sign a written statement under caution or record of interview in the absence of the appropriate adult unless para 11.1 or para 11.18 to 11.20 apply’.
Note 1D, Guidance, states that it is better that the appropriate adult has experience or training in regard to mentally disordered or mentally vulnerable, rather than a relative lacking such qualifications.
It is often the case, however, that relatives and friends of the family, stand by juveniles in police interviews knowing little or nothing of what they are supposed to be doing, or they are supposed to be assisting the youth being interviewed to ensure their vulnerability or volatility is not being exposed or being exploited.
What then are the circumstances and exceptions in which the Police can interview a juvenile suspect without an AA present? ‘When a Superintendent or above considers that a delay will lead to consequences such as interference or harm to evidence, harm to persons, thee alerting of accomplices, or hindering of recovery of property…only if it would not significantly harm their physical or mental state’.Annex E, Code C, 11.1(a) to (c).
A juvenile can only be interviewed at school in exceptional circumstances, and then only when the principal or his nominee agree. The head teacher should be present Zander, M (2013), p348.
At Note 11 c, Guidance the Police are warned that juveniles (like the mentally disordered or otherwise mentally vulnerable) are ‘particularly prone in certain circumstances to provide information that may be unreliable, misleading, or self-incriminating’. It is important therefore to obtain corroboration of any facts admitted wherever possible.
Can the venue and uniform influence how someone may respond? Bickman (1974) as cited by Gudjonsson, GH (2003), loc 5445 demonstrated that where someone was dressed in a guard uniform requesting pedestrians to give money to someone for a parking metre, 83% of people asked, complied. Whereas, only 46% complied when the person making the request was dressed as a civilian.
A police officer dressed in uniform as opposed to plain-clothes may unwittingly influence a vulnerable witness. Usually, children are not so influenced by plain-clothes police for that reason, whereas this is not the case so much for vulnerable witnesses when interviewed by officers in uniform, and this is a potential failure to safeguard. The non-wearing of a uniform should also be considered as a special measure so as not to unnecessarily influence a vulnerable witness.
a. The usual setting of a police interview, the control and dominance of the police interviewer as to the use of a room, seating arrangements, lighting, heat, food, drink, breaks, refreshments, and toilet breaks, without giving reasonable expectations and explanations in basic terms to a VW is cause for uncertainty and anxiety, which can trigger greater levels of suggestibility, compliance, acquiescence, and confabulation, leading to unsafe confessions.
b. The PEACE model is preferable to the Reid model, avoiding willful deliberate deception, confusion, and false security/trust. However, responses to negative feedback, positive feedback, leading questions, and repeated questions can be very suggestive for a VW, and ought to be either avoided or monitored carefully by an AA;
c. PACE leaves ambiguous the identity, and role of AAs who are supposed to support an VW, and reduce unfair, coerced, and false confessions. There is an absence of application of the medical questionnaire completed before a witness is interviewed, to be considered as part of assessing whether a person being interviewed is vulnerable;
d. Even when an AA is required, they are not generally aware or understanding of their role during the interview process, and most certainly are unaware that negative feedback, leading questions, and repeated questioning, without breaks, over a long period of time, can induce an unreliable confession;
e. The wearing of a Police uniform when interviewing VWs rather than just interviews with children should be avoided in order not to trigger obedience and compliance unwittingly making someone confess what the interviewer wants to hear, rather than what actually happened.
f. Arguably, the elderly and those with notable deterioration of memory should be treated as VWs and special measures applied, so as not to confuse and illicit a false/forced confession.
g. Those being interviewed should be well rested, and not deprived of sleep, prior to a police interview being conducted with a VW or any witness, or else they are open to greater levels of interrogative suggestibility.
This blog post is for information purposes and should not be relied upon as legal advice because it does not consider or take into account your own personal circumstances. If in doubt, seek legal advice.
Professor David Rosen is a solicitor-advocate, partner and head of litigation at Darlingtons Solicitors LLP. He is a strategic legal advisor for Help4LiPs, a member of the Society of Legal Scholars amongst other memberships, and honorary professor of law at Brunel University where he regularly lectures on practical legal skills and procedure, and advocacy amongst other subjects.