Donald Phipps v The Director of Public Prosecutions and the Attorney General of Jamaica (Jamaica)

On the 15th April 2005 at around 4 am, the burnt bodies of Rodney Farquharson (“Rodney”) and Daten Williams (also known as “Scotch Brite”) were found in a burning heap of tyres on an open piece of land near Matthews Lane in West Kingston. Forensic evidence showed that they had died from multiple gunshot wounds before being burnt. The appellant, Donald Phipps (known as “Zeeks”), a businessman from Kingston parish, was arrested shortly afterwards and was charged with the two murders, jointly with Garfield Williams.

They were tried before Marsh J and a jury, beginning on 6th March 2006. On the 23rd March 2006, at the end of the prosecution’s case, Garfield Williams was discharged upon the prosecution’s concession that there was no case to answer. The trial continued against the appellant alone. On the 12th April 2006, he was found guilty of two counts of murder and sentenced to life imprisonment, with a minimum term of 30 years. His appeal to the Court of Appeal was heard between 11th and 14th January 2010. On 30th July 2010 the Court of Appeal dismissed the appeal. On the 26th September 2011, he was granted final leave to appeal to her Majesty in Council.

The facts of the case, the relevant evidence, and the submissions of the parties, are set out in meticulous detail in the judgment of the Court of Appeal given by Morrison JA. This makes it possible for the Board, without disrespect either to that court, or to counsel appearing before us, to deal with the issues relatively shortly.

By the end of the trial, the essential facts of these murders, and their timing (around 3 am on the morning of 15th April) were not seriously in doubt. The principal issue was whether the appellant had anything to do with them. On this, the jury were faced with a stark conflict between, on the one hand, three witnesses for the prosecution, based on their alleged telephone conversations with the appellant at or about the time of the murders; and, on the other, an unsworn statement of the appellant himself, supported by the sworn evidence of a witness, David Foster, who claimed to have been not only an eye-witness, but also a victim of the same shootings, but said that the appellant had not been involved.

For the purposes of this judgment it is unnecessary to give more than a summary of this evidence. The three prosecution witnesses were Kelroy Rashford (“Kelroy”), O’Neil Patrick (“Joe”), and Oliver Clue (“Clue”). Kelroy was a friend of Rodney. He had known the appellant since about 1991, and had often heard his voice, but not previously by telephone. He had also heard a CD from Zeeks’ “Cool Tuesday” parties, on which his voice could be heard. Joe had worked for Zeeks for two weeks in 2004 and heard him speak, but not by telephone. Clue had known Zeeks for over ten years through their membership of the PNP party. He had also spoken to him in person and by telephone.

The transcript is not always easy to follow, and there are some discrepancies in the three accounts, but the general sequence is reasonably clear. The conversations started with a call from Kelroy to Rodney on his cellphone around midnight, which led to a conversation with Zeeks. This was followed by Zeeks calling back on the same phone. Kelroy put his phone on speaker so that others with him, including Joe, were able to hear. They then went round to Clue’s house, where again there were calls to and from Zeeks on Rodney’s phone which all present could hear. The general tenor of the conversations was that Zeeks was holding Rodney, and that he wanted Kelroy to come over with another friend called Tim, for reasons which were not explained save that Rodney was “in trouble”. On the last exchange, Zeeks told Clue: “you love Rodney, this is the last time you are going to hear his voice”. The following morning Clue made a call to Zeeks to ask what had happened to Rodney, but was told that Zeeks had not seen him for two weeks. Clue described Zeeks on this occasion as “very calm and humble”, by contrast with his “very aggressive” manner the previous night.

For the defence, the appellant made an unsworn statement from the dock in which he denied any involvement in the murders. Rodney was a good friend. He did not know Scotch Brite. He said that the witnesses who implicated him were lying. He said nothing about whether or not they were known to him, or about the alleged telephone calls.

The defence case rested principally on the evidence of David Foster. He had known the appellant for about ten years, as the community leader in the Matthews Lane area. He said that on 14th April 2005 there had been a dispute between Rodney and a man known as ‘Scandal’ about the delivery of a quantity of “ganga” being stored by Foster. As a result Foster arranged for them to meet, at Luke Lane, at “about after midnight”. Rodney was with Scotch Brite and Scandal was with two friends, known as ‘My Lord’ and ‘Lion Heart’. There was a struggle over a bag of ganga. This ended when Lion Heart shot Scotch Brite in the head, and Scandal did the same to Rodney. Foster himself was shot in the face and fell unconscious on the ground. He had no recollection of any cellphones being used during these exchanges.

The next thing he knew was when he found himself in a cart with Rodney and Scotch Brite, being pushed by My Lord and Scandal along Beeston Street and onto Rose Lane, where they were dumped on an open lot. He managed to get away before collapsing unconscious at the intersection of Matthews Lane and Heywood Street. He regained consciousness at Kingston Public Hospital, where he remained for about 2 weeks. During this time he had made a statement to the police, in which he had told them (untruthfully) that he had been shot “out by the Ward Theatre area”. He had made this statement because he had been threatened by My Lord.

Both sides gained some corroborative support from the independent technical evidence. For the prosecution, there was evidence of the use of the relevant cellphones in this area between 2 and 3 am. On the other side, Foster’s account of his own treatment was consistent with DNA evidence linking him to blood stains found by the police at Matthews Lane. The hospital records showed that he had been admitted at 3.13 am, with a history of a gunshot wound to the right midface and an exit wound on the left neck.

The first two grounds raised constitutional issues which have not in the event been pursued. Although these may have been the basis for the grant of leave, the respondents have not taken any point on the validity of the appeal on the remaining issues. As appears from the appellant’s supplemental case, they are:

In the Board’s view, grounds (iii) and (iv) add nothing to the others. Mr Rhodes QC criticised the judge for not having addressed the jury on possible alternative offences, such as kidnap or false imprisonment. It does not appear that those possibilities were ever raised by the defence at trial. Their case was that the appellant was simply not involved at all. In any event, as the Court of Appeal said (para 145), once the jury accepted the prosecution evidence as to the nature and content of the telephone conversations with the appellant on the night of the murders, it pointed to no other possible conclusion than that he had been an active participant in the murders, and to a verdict of murder as charged. Furthermore, Mr Rhodes QC accepted that it was not open to the jury to convict the appellant of those offences, given the terms of the indictment. In those circumstances, it would have been inappropriate to direct the jury in relation to those offences.

The complaints relate to the treatment at trial, first of the appellant himself, and secondly of the witness Foster.

As to the first, the Board agrees with the Court of Appeal that no material prejudice arose from the mere fact that the appellant may have been seen to be in custody, while his co-accused was on bail. The judge had properly reminded the jury of their duty to arrive at their decision according to the evidence, uninfluenced by any prejudice against the appellant or bias in his favour. Morrison JA said:

“In our view, in all the circumstances of this case, these remarks by the judge would have sufficed to focus the minds of the jury, as persons of ordinary courage and firmness, on the business at hand, that is, to consider the evidence carefully and to render a true verdict according to law. While it would obviously have been best if the remand status of the applicant had been dealt with as a matter of routine after the jury had withdrawn, we do not think that to the extent that there may have been occasional departures from this ideal during the course of the long trial, any prejudice to the applicant has been demonstrated to have resulted from any such lapse.” (para 128)

The second point has more substance, particularly in view of the critical importance of Foster’s evidence to the defence case. The background is that Foster was arrested by the police in the course of the trial and before he was due to give evidence, and a copy of his statement of evidence was taken by the police. The circumstances of the arrest were the subject of a voir dire conducted by the judge, on the appellant’s complaint that there had been a breach of his constitutional rights, and abuse of process, such as to require the case to be dismissed. The judge dismissed the complaint, finding no evidence of bad faith on the part of the police. There was no appeal against that decision.

The present issue relates principally to the manner in which Foster was presented to the jury at the trial. He was escorted into court by a police officer who, in full view of the jury, brought him up from the lock-ups into court, holding him by the waist. Thus, it was said, he was “displayed before the jury as a prisoner in the hands of the police”.

The judge addressed this point in his summing up. He said

“He said that when he came to court the previous day he came to court being brought by the police and the policeman was behind him holding him in his waist and that he has been taken from the lock up and brought up. Now, Mr. Foreman and your members, this bit of information about how he was brought into court ought not to affect in any way shape or form what you have to decide, and that is to assess the evidence of David Foster fairly and impartially. It should not be used against him in any way at all. (5/2063-4)”

The Court of Appeal accepted that Foster’s treatment by the police was “inadequately explained and wholly unfortunate”, but did not think that the defence was handicapped thereby (para 130). It was clear from the transcript that, notwithstanding what had gone before, the witness had been able to give his evidence in full detail, and the jury would have had “ample time to observe him and to make a careful assessment of his credibility”.

Before the Board, Mr Rhodes submits that the Court of Appeal gave insufficient weight to the central status of Foster’s evidence in the defence case. The effect of what happened was to undermine the credibility of this critically important defence witness in the eyes of the jury before he had even begun his evidence. No direction by the judge could cure “the inevitable and overwhelming prejudice” so caused.

The Board is unable to accept Mr Rhodes’ criticisms. It is of course the judge’s duty to ensure that evidence is presented as fairly and impartially as possible. Where irregularities occur, it is the judge’s responsibility to consider their significance in the context of the trial as a whole, and take appropriate action. In most cases the giving of a suitable warning to the jury will be sufficient to ensure that no material prejudice results. That is what the judge did in this case. Mr Rhodes was unable to refer us to any authority to support the proposition that an irregularity, such as the Court of Appeal acknowledged to have occurred in this case, required the trial to be abandoned. This ground of appeal must be dismissed.

In his summing-up, the judge reviewed the identification evidence in considerable detail. The relevant passages are quoted extensively in the judgment of Morrison JA. It is enough to give some examples. By way of introduction to his review of the evidence, the judge emphasised the importance of familiarity:

“Now, in order for the evidence of a witness to be accepted who said that he recognized an accused person by voice, to be cogent there must be evidence of the degree of familiarity the witness have had with the accused and his voice including the time the witnesses may have had to listen [to] the voice of the accused and the occasion when the recognition of the voice occurred must be such that such words used to make a recognition of that voice is safe to act on.” (4/1884)

He reminded them that, of the three witnesses, only Clue claimed to be familiar with his telephone voice (4/1885). Later, he reminded them of his earlier directions, and of the need for “very special caution” before they could accept that the voice alleged to have been heard was in fact the voice of the accused (5/2046).

Finally, at the end of the summing-up, he was requested by the prosecution to elaborate on the issue of voice identification, by making clear that even if the three witnesses were found to be honest, they could nonetheless all be mistaken. The judge responded by adding to his direction:

“… I omitted to indicate to you that sometimes people can be very convincing although they are mistaken when they say that they identify somebody by their voice on the telephone. And you are going to be very careful in your assessment of the evidence because an honest witness can also be a mistaken witness. The witness may honestly feel that the person they heard on the ‘phone was John Brown, but in fact it turns out to be otherwise.

So you look on the evidence, the circumstances under which the identification of the voice was made. You look at the previous history of that person who heard the particular voice. The person who seeks to identify the person by voice, what opportunity that other person would have had to have heard the voice.

I told you that of the three person who said they heard the accused, only one had given evidence that he had spoken to and heard the accused on a telephone. So please remember that.” (5/2097)

On this issue, the Court of Appeal referred to a number of authorities on identification evidence in general (starting with R v Turnbull [1977] QB 224), and on voice recognition in particular (notably R v Clarence Osbourne [1992] JLR 452). They reviewed the judge’s summing-up in the present case. They described the directions on voice recognition as “full and proper” (para 142), adding:

“He reminded them of all the relevant factors to be taken into account, including potential weaknesses in the evidence, and gave warnings appropriate to the circumstances of the case. At the end of the day, it was entirely a matter for the jury to decide what weight should be given to the various factors and it appears to us that in this regard they had the benefit of as much assistance as could reasonably be expected from the judge.”

They made one criticism:

“While it is a fact that, as Mr Phipps submitted, the judge did not tell the jury that it is a notorious fact that mistakes have been made in voice recognition in the past, we think that this omission is more than outweighed in this case by the judge’s repeated emphasis of the need for caution in assessing the evidence of voice identification.”

Mr Guthrie QC, for the respondents, disputes the criticism. The reference to “notorious mistakes” in some of the cases reflects UK experience in the context of visual recognition. He suggests that that there is no equivalent evidence of mistakes in respect of voice identification. The Board has no material on which it could properly comment on the extent of such mistakes in the UK or Jamaica. On any view, however, it agrees with the Court of Appeal that this was not a material defect in the summing-up.

Mr Rhodes challenges the conclusion that the summing-up was adequate. He submits that the judge failed to deal with certain issues. He failed to warn them that even a number of witnesses can be mistaken, and that where the evidence of identification is weak, it is not improved by mere repetition by another witness. More specifically he says the judge should have warned them of the particular risks of “contamination”, innocent or otherwise, between the evidence of the three witnesses. Finally the judge should have commented on the prosecution’s failure to conduct any “controlled voice identification procedure” to test the accuracy of the witnesses. In this connection he refers to Aurelio Pop v R [2003] UKPC 40.

He points to particular weaknesses in the evidence, which should have been more clearly highlighted. For example, Kelroy and Joe were not familiar with the telephone voice of the Appellant. Kelroy's CD, which had not been mentioned in his statement, was not produced for analysis. There were discrepancies in Kelroy’s evidence as to how often he had heard the appellant's voice. Although Clue claimed some familiarity with the appellant's telephone voice, this had not been tested under controlled circumstances. He himself had said that the voice he attributed to the appellant in the calls during the night was different in tone to that of his call to the appellant on the following morning. His evidence as to the content of the calls he witnessed was at times inconsistent with the accounts given by the other witnesses.

In the view of the Board, these criticisms are unfounded. There is no doubt as to the importance of the guidance in Turnbull nor as to its application in principle to identification by voice recognition. In that context more detailed guidance has been given more recently by the English Court of Appeal in R v Flynn and St John [2008] 2 Cr App R 20. However, as has been emphasised on many occasions, and as the Court of Appeal recognised in this case, “no precise form of words need be used so long as the essential elements of the warning are given to the jury”: Shand v The Queen [1996] 1 WLR 67, 72. The judge could hardly have done more to bring to the attention the need for caution in considering this part of the evidence, and the reasons for it, and their duty to consider each of the witnesses individually. As to the CD, it does not seem that any request was made for an opportunity to analyse it, nor is there any indication of what such analysis might have revealed to the advantage of the defence.

It is true that the issue of “contamination” was not in terms addressed by the judge or the Court of Appeal: that is, the possibility that one or more of the witnesses had been influenced by the identification made by the others, rather than forming an independent view. This is perhaps not surprising as it does not seem to have been mentioned as a separate point until the appellant’s supplementary case to the Board, submitted at the end of April 2012. Such a suggestion would have been particularly difficult in the case of the witness Clue, who not only knew the defendant well, but had heard his voice on the telephone. In truth it is merely one aspect of the process of considering the witnesses individually, on which the jury were directed in meticulous detail. Nor was Mr Rhodes able to indicate with any clarity what form of “identification procedure” should have been adopted. Aurelio Pop v R was a very different case. No full Turnbull direction had been given, and the identification evidence was “more than usually open to criticism” being a sighting in a dark street, by a single eyewitness who had given divergent accounts to the police and to the court (para 17).

In conclusion the Board sees no basis for challenging the decisions of the Courts below. Indeed the Court would commend the judge’s thorough and balanced summing-up in a complex case, and the very full and convincing review by the Court of Appeal. It was for the jury to resolve the direct conflict of evidence as to the involvement of the appellant. There was ample material on which they could reach their conclusion.

The Board will humbly advise Her Majesty that this appeal should be dismissed.