Ice cream verbals

In March of this year, the High Court allowed the appeals of Thomas Campbell and Joseph Steele in respect of their convictions in 1984 for murdering six members of a Glasgow family (2004 SCCR 220). The court also quashed Campbell’s conviction for assault to the danger of life, which arose from a shotgun attack on an ice cream van. The appeal of a third man, Thomas Gray, convicted of attempting to murder the occupants of the van, was rejected. All three appeals followed lengthy investigations by the Scottish Criminal Cases Review Commission, which in 2001 decided to refer the cases to the court for determination. For Campbell and Steele, it was the third time their appeals had been heard by the High Court, although the first in which the grounds which led to their acquittal had been argued.

The Commission’s references were based, in the main, on reports it had obtained from a forensic psychologist and a forensic linguist which, in its view, cast doubt on police evidence as to certain incriminatory remarks allegedly made by Campbell and Steele following their arrest. Evidence from one of the experts, Professor Brian Clifford, was led at the appeals, and formed the principal basis of the court’s decision to quash the convictions.

The inevitable media publicity surrounding the cases has arguably obscured some of the real issues raised by the decision. These include important questions regarding the manner in which police officers record admissions by suspects, and the approach taken to giving evidence on such matters. It is with these issues, and similar ones in cases south of the border, that this article is principally concerned. Although the following sections focus on the psychological evidence so far as this related to Campbell, similar principles apply to the court’s reasoning in Steele’s case.

The backdrop to the convictions was the so-called “ice cream wars” in Glasgow in the 1980s, a term used to describe the intimidation and violence employed by rival vendors to gain control of the most lucrative “runs” in the city. A number of Campbell’s relatives owned and operated vans in competition with, among others, a van driven by Andrew “Fat Boy” Doyle. It was Doyle’s van that was the target of the shotgun attack, and it was he who, with five members of his family, died in a fire at their top floor flat, an incident which formed the basis of the murder charge.

The evidence at trial

The Crown’s position at trial was that Campbell planned both the shooting and the murders. Evidence was led from a socius, William Love, who claimed that after the shooting Campbell had thanked him for his role and offered to square up with him once things had quietened down. Campbell’s fingerprint was also found on the interior of the car used to convey the gunman to the locus.

Love also gave evidence in respect of the murder. He claimed that one night he had overheard Campbell, Steele, Thomas Gray and others talking about setting fire to Doyle’s door in order to give him a fright. Evidence was also led of a map said to have been found by police following Campbell’s arrest, on which the location of Doyle’s home was marked with a cross.

More critical to both charges, however, was evidence of a highly incriminating remark allegedly made by Campbell following his arrest. Evidence was led in this connection from four police officers who attended at Campbell’s home in execution of a petition warrant which concerned the shooting, but not the murder. The remark was as follows: “I only wanted the van windaes shot up. The fire at Fat Boy’s was only meant to be a frightener which went too far.”

All four officers testified to noting the remark at the time it was uttered. Aside from certain very minor differences, each officer’s account of the remark in their evidence and respective statements was identical. Two of the officers claimed in evidence that there had been no comparison or collaboration between them and their colleagues as to what they had noted. The remaining officers adopted a similar position at interview with the Commission.

Campbell, supported by his partner, denied in evidence that he had made the remark, and claimed that none of the officers had noted anything in his presence. Their evidence, he said, had been fabricated as part of a police conspiracy to secure his conviction.

Professor Clifford’s studies

Brian Clifford is a professor of cognitive psychology, with an international reputation in the field of “eyewitness memory”. He was asked by the Commission to examine the officers’ evidence and provide an opinion based on empirical studies. Specifically, he was instructed to consider the likelihood that all four officers were able to note Campbell’s alleged remark in such similar terms in the absence of any comparison or collaboration between them.

Professor Clifford conducted two studies, the first of which involved 57 participants drawn from a wide range of occupations. For the purpose of the experiment, the group was divided into two: one, the Informed Group (IG), was provided with relevant background information; the other, the Non-Informed Group (NIG), was not. The IG was intended to simulate, as far as possible, the conditions under which the officers themselves might have noted the alleged remark. This was considered by Professor Clifford to be an important allowance given that prior knowledge and awareness of this kind is recognised as an aid to recall ability. Both groups were presented with a recording of Campbell’s alleged remark, spoken by a Glaswegian male, and instructed to note this verbatim. Participants were scored both for verbatim correctness and semantic completeness (their ability to remember the basic meaning conveyed).

None of the participants came close to achieving full verbatim recall. The majority in the IG achieved only 40 per cent recall; in the NIG the figure was 30 per cent. Analysing the performance of four-person groups within the sample, Professor Clifford found that key segments of the remark were missed by all participants. Significantly, no two participants within any such group achieved identical verbatim recall.

The results suggested that verbatim recall of a 24 word remark, such as that allegedly uttered by Campbell, is impossible. Professor Clifford concluded that in the absence of some form of collaboration it was “very improbable” that all four officers who noted the alleged remark could have achieved the level of recall claimed in evidence.

Professor Clifford’s second study was carried out in Scotland. Its purpose was to address the possibility that the results of the first study might be explained by a lack of familiarity with the accent in which the remark was spoken. The second study consisted of 74 participants, including 14 police officers, all of whom were given the same level of information given to the IG in the first study.

The results of the second study mirrored those of the first. Significantly, in terms of verbatim recollection, the police officers performed no better than the other participants. In the light of both studies, Professor Clifford observed that the recall ability displayed by Campbell’s arresting officers must be seen as “truly remarkable: so remarkable, in fact, as to be doubtful”. The probability that any random four officers could recall a sentence of 24 words in identical terms without collaboration was, in his view, “infinitesimal”.

The High Court’s decision

The Crown led no evidence in rebuttal of Professor Clifford’s findings, but sought to cast doubt on the significance of the new evidence, and to criticise aspects of the methodology employed in the studies.

The court began by expressing general doubt as to whether an experienced criminal such as Campbell was likely to have made such an admission in the first place. Assuming that he had, the court found it remarkable that he had not been asked to authenticate this by signing or initialling one or more of the officers’ notebooks. Most remarkable of all for the court was that this crucial evidence, which had constituted a breakthrough in the police inquiry, was not reported by the senior officer to his superior until the following day. The court emphasised, however, that these were merely its own reactions to the evidence, and acknowledged that all such criticisms were before the jury, which had nevertheless convicted Campbell unanimously.

The court considered Professor Clifford’s reasoning and conclusions to be logical and cogent. Although there was some substance to the criticisms concerning methodology, there was nothing in the Crown’s submissions which persuaded the court that it should not accept the key findings. In the court’s view, the contention that the police had fabricated evidence of the remark was crucial to Campbell’s defence at trial, but was something for which there had previously been no independent support. Had the new evidence been heard at trial, it would not have been possible for the judge to direct the jury in terms which emphasised the stark choice which had to be made between the police evidence and that of the accused. Rather, the findings would have constituted independent testimony from a reputable source, capable of undermining crucial evidence on which Campbell’s conviction depended. In the court’s view, any jury hearing the new evidence would have assessed the police evidence in an entirely different light. In these circumstances, verdicts returned in its absence must be regarded as miscarriages of justice.

Similar challenges elsewhere

Campbell is not the first case in which challenges to the veracity of police confession evidence have been made by way of expert evidence, though it is arguably the first in Scotland. (For English challenges to the reliability of confession evidence on the basis of developments in psychological understanding of “suggestibility” of accused persons, see e.g. R v O’Brien, Hall and Sherwood [2000] EWCA Crim 3; R v Raghip, The Times, 9 December 1991; and R v Pendleton [2001] UKHL 66. There are no reported examples of such evidence being led in Scotland, but see Blagojevic v HMA 1995 SCCR 570). Indeed, at an early stage in the appeal, the court appeared to express some doubt as to whether evidence of the kind eventually given by Professor Clifford was admissible. The Crown took up the issue at a subsequent procedural hearing, in which it argued that Professor Clifford’s findings were inadmissible as opinion evidence on the credibility and reliability of the police witnesses. In the event, the court rejected the submission, observing that the findings constituted relevant and admissible evidence of fact (not opinion) that could have a significant bearing on the issue of credibility.

As the court did not issue a written opinion on this question, it is not possible to examine its reasoning further. Other recent decisions, however, suggest that this is an area worthy of clarification (compare, for example, the approaches taken in HMA v Grimmond, 2001 SCCR 708; AJE v HMA, 2002 SCCR 341; and McBrearty v HMA, 13 April 2004).

Some support for the approach taken by the court to admissibility in Campbell can be found in R v Meads [1996] Crim LR 519, CA. There, the principal evidence against the appellant consisted of admissions which police officers claimed to have recorded in handwritten notes during various interviews. At appeal, the appellant sought to introduce expert evidence which suggested that in order to be consistent with the police evidence, certain interviews would require to have been conducted at the speed of a horseracing commentary. The Crown argued that the new evidence was inadmissible on the basis that it was opinion evidence founded on an insufficiently organised body of knowledge. In rejecting this submission, the court observed that the evidence was no more opinion evidence than that of a police officer who times a journey in order to test the veracity of an alibi.

The facts in Campbell are similar to those of R v Maynard, Dudley and Others [2002] EWCA Crim 1942, which followed a reference by the (English) Criminal Cases Review Commission. There, three of the appellants had been convicted in 1977 of murder, partly on the basis of alleged admissions during police interviews. According to the evidence, the senior investigating officer in the case had been concerned to ensure that the accuracy of the record was beyond doubt. It was therefore decided that a note-taker would enter a full contemporaneous record of each interview in a red hardback book. At the conclusion of each interview, the note-taker would sign the record, and the interviewing officer would then check it for accuracy. At no time, however, were the appellants asked to check or sign the records.

At trial, each of the appellants denied making the admissions. According to them, such notes as were taken during the interviews had been written on loose sheets of paper. Although the notes in the hardback books contained details of answers they had in fact given, the appellants claimed that certain damaging replies and comments had been added subsequently.

At appeal, the appellants sought to lead evidence from a leading forensic document examiner, who had provided an opinion on whether the interviews could have been contemporaneously recorded. Save for one interview, involving the appellant Dudley, he found that they could. The record of Dudley’s interview consisted of over 11,000 handwritten characters, which police claimed had been produced within a 50 minute period. The expert considered this to be a physical impossibility. In order for the police evidence to be correct the note-taker would require to have achieved a “noting” speed of around 222 characters per minute, whereas the fastest speed the expert had ever encountered was less than 170 characters per minute. In the court’s view, the evidence demonstrated that the police evidence in certain respects was necessarily wrong and unreliable. In these circumstances, Dudley’s convictions were held to be unsafe. The convictions of the other appellants were also quashed on the basis that the integrity of the hardback book system was central to the cases against them.

Motives and circumstances

As will be apparent, there is nothing in the expert evidence led in any of the above cases capable of directly contradicting the police evidence that the admissions in question were made. Rather, the importance of such evidence lies in its ability to demonstrate that the police evidence on certain peripheral matters is unlikely to have been true. Although this is significant in itself, an inference of possible malfeasance arises only when one considers the possible motives behind the falsification of such evidence.

The limited relevance of the expert evidence used in Campbell is, of course, reflective of the general difficulties associated with challenging police evidence of the kind given in that case. By its nature, such evidence frequently derives from circumstances in which a suspect is alone in the presence of two or more officers. Accordingly, in many cases the only realistic means of testing its veracity is to scrutinise the surrounding facts and circumstances of the admission. This is precisely the approach taken in each of the three cases mentioned.

Of course, no-one is denying the propensity among accused persons to retract a genuine admission once it becomes apparent that, without it, there is little evidence against them. However, it is not difficult to imagine cases in which police officers themselves may have an equally strong interest in securing a conviction, particularly where the crime is emotive or high profile, and the pressure for results intense. In cases in which an accused person’s admission is genuine, it is not surprising that police witnesses may be reluctant to admit discussing with colleagues the precise terms of the remark. There is, first of all, the fear that such an admission will lead to allegations of fabrication. Connected with this is the more general concern that collaboration also opens the evidence up to attack on the basis that one cannot be certain whether a particular officer’s account is based on his own observations, as opposed to those of his colleagues. In this sense, even if the officers in Campbell had admitted to such comparison, it is unlikely that this would have been the end of the matter.

Although many of the difficulties in past cases may have been solved by the introduction of audio and video recording of interviews, such innovations clearly do not cover situations in which admissions are allegedly made by a suspect at his home, or en route to a police station. It is not immediately obvious how potential disputes can be avoided in such cases, except perhaps by requesting that a suspect sign the officer’s notebook, a measure which may not always be possible or practicable. On the one hand, it would be wrong to exclude such evidence as a matter of principle, given that in the vast majority of cases there seems to be little cause for complaint; on the other, it is clearly important that evidence of alleged admissions is given honestly and accurately, something which the emerging case law will perhaps re-emphasise.

The views expressed in this article are entirely those of the writer, and do not represent those which may be held by the Scottish Criminal Cases Review Commission.