PRINGLE v THE QUEEN (JAMAICA), [2003]

“Members of the jury, if you accept the scientific evidence called by the Crown, this indicates that there are probably only about four or five males in [the given area] from whom that semen stain could have come. The defendant is one of them. If that is the position, the decision you have to reach, on all the evidence, is whether you are sure that it was the defendant who left that stain or whether it is possible that it was one of that other small group of men who share the same DNA characteristics.”

The requirement of transparency lies at the heart of the appellant’s complaint that there was no DNA expert for the defence in this case. Reference was made to section 20(6) of the Constitution of Jamaica, which provides that an accused person must be provided with adequate facilities for the preparation of his defence. But Mr Guthrie assured the Board that funds would have been provided if an application had been made for the defence to be provided with the services of a DNA expert. Their Lordships do not find it necessary to dwell on this matter or on the question whether Dr Cruickshank’s use of the HLADQa marker in her laboratory was inappropriate. The important issues in this case relate to the requirements of accuracy and fairness in the presentation to the jury of this evidence.

Dr Cruickshank was asked in her examination in chief to describe the tests which she had performed and to describe her findings and conclusions after subjecting the samples she was given to the two different methods which she used for her analysis. She said that her object was to identify the male fraction in the vaginal swab, and that having done so she found that the two markers which she obtained from the D1S80 marker test were similar to the two markers obtained from the sample of blood taken from the appellant. She said that her conclusion from the results produced by the HLADQa marker test was that the male fraction in the vaginal swab contained the same two markers as those found in the appellant’s blood sample but that Donald Davidson was not represented on the vaginal swab. In response to a further question by the prosecutor she said that the spermatozoa that she found on the swab did not come from Donald Davidson. None of this evidence was controversial and none of it has been criticised in this appeal.

At this point the trial judge put questions to Dr Cruickshank which invited her to express her views on a matter which the prosecutor himself, so far, had been careful to avoid:

“His Lordship: Can you say where is came from?

Witness: The spermatozoa could have come from Michael Pringle.

His Lordship: When you said could have …?

Witness: In science we have 99.999 per cent certainty. So, what I would say, it is with a high degree of certainty.

His Lordship: 99.999 per cent that it came from Michael Pringle?

Witness: Yes, my Lord.”

Towards the end of her cross-examination the following exchange took place between her and the judge:

“Witness: You asked me, I can’t recall the question – what came to my mind – �Why not the degree of certainty with which I state something?’

His Lordship: With which you stated that it was his semen in Kevan Davidson’s vagina?

Witness: Yes.

His Lordship: And you told me it was 99.999 per cent certainty?

Witness: No. I said to you I would have to say it is with a high degree of certainty.

His Lordship: Not 99.999 …

Witness: I said 99.999. In science we say 99.99999. It goes on. But we did not address the probability in this.

His Lordship: Should I qualify this 99.999 now?

Witness: Not in the context of which we spoke. It will still stand.”

When he came to this point in his summing up the judge said that the readings on the D1S80 test and on the HLADQa test on the male fraction in the vaginal swab:

“would indicate that the spermatozoa in the vaginal cavity of the deceased woman came from the accused man, Pringle.”

After referring to the readings which she had obtained on the HLADQa test from the female fraction he said:

“So, it is based upon these results that she comes to the conclusion that the spermatozoa there came from Pringle, that it, that Pringle had sexual intercourse with the deceased.”

This conclusion was fallacious, as Phillips LJ explained in R v Doheny [1997] 1 Cr App R 369, 372-374. The fallacy is that which is known as “the prosecutor’s fallacy”, although – as their Lordships have said – it was not a fallacy that was propounded in this case by the prosecutor. It can be explained in this way. Let it be assumed that the evidence about the random occurrence ratio is that one person in 50,000 has a DNA profile which matches that which was obtained from the crime scene. The fact that the defendant has that profile tells us that he is one of perhaps fifty thousand people who share that characteristic. One can then say, having regard to the population of the area, what the statistical probability is that he was the perpetrator. But that is all that can be said about it. The question whether the statistic points to the defendant as the actual perpetrator will depend on what else is known about him. If it is plain from the other evidence that he could not have committed the crime because he was elsewhere at the time, the fact that the defendant’s DNA profile matches that on the sample taken from the crime scene cannot be said to show that he did commit it. That proposition will have been negatived by the other evidence. So the probative effect of the DNA evidence must depend on the question whether there is some other evidence which can demonstrate its significance. And it is for the jury, not the person who gives the DNA evidence, to assess its significance in the light of that other evidence.

It was, of course, open to Dr Cruickshank to identify the random occurrence ratio. It has not been suggested that the tests which she carried out, so far as they went, were defective in any way. The basic facts were there, and they have not been either challenged or criticised It was within the province of her expertise to say what the statistical likelihood was of the same sections or bands of DNA being found in the male fraction of the vaginal swab as was found in the appellant’s blood sample. But it was not for her to express an opinion as to the probability that was his spermatozoa that were found in the deceased’s vagina. Her evidence that she could say that this was so “with a high degree of certainty” went beyond what was permissible. The risk of the jury being misled by this evidence was compounded by her reference in this context to a 99.999 per cent probability. It was made all the greater by the judge’s statement in his summing up that she had come to the conclusion “that the spermatozoa there came from Pringle”.

The problems which were created by Dr Cruickshank’s evidence do not stop there. She was asked questions during her cross-examination both by the judge and by defence counsel about her calculation of the random occurrence ratio. It was quite proper that she should be asked about this matter, as she did not deal with it in her evidence in chief. But if she was not to mislead the jury she had to be accurate. The effect of her evidence was summarised by the prosecutor in his re-examination. It was as follows:

“Q: You said, doctor, the probability in relation to the combined tests of two persons having the same reading as Michael Pringle were, in Jamaica, four in ten thousand?

His Lordship: The combined?

Mr Hibbert: The combined probability.

Witness: We are talking about persons, not male.

Mr Hibbert: Persons.

Witness: Four in ten thousand refer not to male.

Q: But just to persons? So, if you were just using the same ratio, one man to one woman, you would have how many males to this ten thousand?

A: It would be two in ten thousand.”

The conclusion which Dr Cruickshank expressed in this passage was based on the following steps in her evidence. She said first that, in the case of the HLADQa test, eight in 100 Jamaican males shared the same match as in the sample taken from the vaginal swab. For the D1S80 test the figure was six in 1,000 males. No criticism is made of these results. But for each marker she then halved the random occurrence ratio on the assumption that the same match would be spread equally between men and women, with the result that the figures became four in 100 and three in 1,000 males. She then arrived at what she described as the combined probability of both the HLADQa test and the D1S80 test, which she said was a figure of four in 10,000 Jamaican males before making the further calculation in her re-examination which reduced this figure to two in 10,000.

The assumption that the figures could be halved because the match would be spread equally between men and women was incorrect. Their Lordships have been shown a report by Mr M P J Appleby of The Forensic Science Service, Chepstow, which Mr Guthrie does not dispute, in which Mr Appleby states that the DNA markers which were being used in this case are not sex specific. In other words the frequency of occurrence which they will demonstrate of any particular character is the same both for men and women. Mr Appleby also states that, in addition to this error, the combining of figures for two separate DNA systems requires extensive statistical investigation to determine that there is no association between the systems as an association might lead to the figures given after their combination which greatly exaggerated their significance. He says that he is not aware that this type of study has been done. Mr Guthrie said that he did not accept this criticism, but he did accept that Dr Cruickshank had made a mathematical error when she was combining these results. She had already halved the random occurrence ratio for the HLADQa and D1S80 markers to give the figure for the male population in Jamaica. She then halved the combined figure again when she was being re-examined to arrive at her final random occurrence ratio of two in 10,000. The trial judge reminded the jury of this evidence in his summing up. He said that the effect of the two tests, when they were combined was that two men in every 10,000 would have the same characteristics as those found in the accused.

The appellant submits that an appropriate adjustment to the random occurrence ratio to correct this error would result in a ratio of two in 1,000 of the population of Jamaica. This calculation, which is not disputed, shows that the effect of Dr Cruickshank’s evidence was greatly to exaggerate the ratio. Other corrections that might have to be made are not agreed. But it is not necessary to determine what the appropriate figure would have been for the purposes of this appeal. It is sufficient to say that there were sufficient errors and inaccuracies in this evidence, and in the judge’s summing up about it, for the jury to have been seriously misled as to the likelihood that it was the appellant who murdered the deceased. Their Lordships have concluded that on this ground alone the appellant’s conviction was unsafe.

The cell confession

The problem as to how to deal with evidence of a cell confession is not new. There has long been an obligation on judges to warn a jury about the special need for caution in cases which are analogous to those of accomplices. These include cases where the witness’s evidence may have been tainted by an improper motive: R v Spenser [1987] AC 128, 134E per Lord Hailsham LC; Archbold 2002, paras 4-404m, 4-404o. It has been held by the Supreme Court of Canada that a warning was necessary in a case where evidence was given by two prison informants who had a strong motivation to lie and who had approached the police when they perceived that some benefit could be exchanged for their testimony: Bevan and Griffith v The Queen (1993) 82 CCC (3d) 310. The High Court of Australia has held that it would only be in exceptional cases that a prison informer would not fall into the category of witnesses about whom a warning should be given by the trial judge of the dangers of convicting on evidence which is potentially unreliable: R v Pollitt (1992) 174 CLR 558. Although Ackner LJ does not appear to have had have prison informers at the forefront of his mind when he delivered his judgment in R v Beck [1982] 1 WLR 461, he recognised the same general rule at p 469A when he said:

“While we in no way wish to detract from the obligation upon a judge to advise a jury to proceed with caution where there is material to suggest that a witness’s evidence may be tainted by an improper motive, and the strength of the evidence must vary according to the facts of the case, we cannot accept that there is any obligation to give the accomplice warning with all that entails, when it is common ground that there is no basis for suggesting that the witness is a participant or in any way involved in the crime the subject matter of the trial.”

The appellant suggested in his written case, against this background, that there were four possible alternative ways the courts should approach the evidence of tainted witnesses. The first was to refuse to admit the evidence. The second was to require it to be corroborated. The third was to give an express warning that the evidence must be treated with caution. The fourth was to draw attention to its potential fallibility and unreliability in the summing up. He did not, of course, suggest that a full accomplice warning was required in this case. Mr Fitzgerald QC did not pursue the argument that the evidence was inadmissible, nor did he press the point that corroboration was required. He directed his argument to the third and fourth alternatives.

The first question is whether there was evidence in this case to suggest that Simmonds’s testimony was of such a character as to require the judge to draw the jury’s attention to the possibility that it was tainted. Mr Fitzgerald drew attention to four features which he said showed that his testimony was of that character. He referred first to the fact that when Simmonds was placed in the same cell as the appellant the appellant was introduced to him by name by a police officer. Secondly, there was his conversation with Mr Bailey in which, after saying that he was bad and wanted to go home, he was asked whether he had heard the appellant saying anything and told him what he had heard and agreed to give a statement about it to Detective Sergeant Coleman. Thirdly, there is the fact that he was given a suspended sentence when he pled guilty to the offence of being in possession of a stolen motor vehicle. Fourthly, the account which Simmonds gave of the confession appeared to contain nothing of significance which was not already known to the police. All these features, said Mr Fitzgerald, showed that there was a risk that Simmonds had an improper motive in giving evidence and the judge should have drawn the jury’s attention to this possibility in his summing up.

The possibility that Simmonds’s evidence was tainted was not fully explored with him in cross-examination. The responsibility for this may well lie with the trial judge. Some preliminary questions were being put to Simmonds by defence counsel about the disposal of the case against him, in reply to which he said that he pleaded guilty and the following week he came out of custody. But at this point the judge intervened:

“His Lordship: Wait a minute. Coming out of custody – He has come out of custody by virtue of a court order, it has nothing to do with the police, so let us get that clear.

Mr Lyn-Cook: I am not saying anything, my Lord. It may be coincidence, sir.

His Lordship: Let us get that clear. He came out of custody because of a court order.”

The effect of this intervention, which their Lordships regard as unfortunate, appears to have been to stop any further questioning about the circumstances which led to Simmonds receiving a suspended sentence and being released from custody. Whether this created any unfairness because it stopped defence counsel from pursuing a material line of inquiry is not something about which their Lordships would wish to speculate. They do not need to do so, as there are other indications that the jury ought to have been told by the judge that they should take special care when they were considering Simmonds’s evidence.

The most important of these is his account of his conversation with Mr Bailey when he was taken back to St Ann’s Bay Police station. He said that Mr Bailey asked him how he was doing, in reply to which he said that he was bad. Mr Bailey then asked him if he had heard the appellant say anything. This might suggest that he was being encouraged to act as an informant. Then there are the contents of the account which the appellant is said to have given to Montgomery. He said that he put on his “Kalcanine” shoes. There is perhaps an indication here the Simmonds had been fed with information by the police. They had found a pair of Karl Kani boots in the appellant’s house was being searched. No mention was made by the appellant that he had had sexual intercourse with the deceased before killing her. But it appears that Dr Cruickshank’s results were not yet available to the police, and that they were unaware that spermatozoa were found to have been present in her vagina. Then there is the fact that Simmonds was an untried prisoner. A person in that position might well have a motive for seeking to ingratiate himself with the authorities.

The next question is what the judge should have said about this in his summing up. It is not possible to lay down any fixed rules about the directions which the judge should give to a jury about the evidence which one prisoner gives against another prisoner about things done or said while they are both together in custody. There may be cases where the correct approach will be to treat the prisoner simply as an ordinary witness about whose evidence nothing out of the usual need be said. Examples of that situation are where the prisoner is a witness to an assault on another prisoner or a prison officer or is a witness to a drugs transaction which has taken place in the place where he is being held.

But a judge must always be alert to the possibility that the evidence by one prisoner against another is tainted by an improper motive. The possibility that this may be so has to be regarded with particular care where, as in this case, a prisoner who has yet to face trial gives evidence that the other prisoner has confessed to the very crime for which he is being held in custody. It is common knowledge that, for various reasons, a prisoner may wish to ingratiate himself with the authorities in the hope that he will receive favourable treatment from them. Of course, as Ackner LJ indicated in R v Beck at p 469A, there must be some basis for taking this view. The indications that the evidence may be tainted by an improper motive must be found in the evidence. But this is not an exacting test, and the surrounding circumstances may provide all that is needed to justify the inference that he may have been serving his own interest in giving that evidence. Where such indications are present, the judge should draw the jury’s attention to these indications and their possible significance. He should then advise them to be cautious before accepting the prisoner’s evidence.

The judge dealt with the matter in this way in his summing up. He reminded the jury that the defence was that Simmonds made up the story and that it had been suggested that he was involved in some sort of conspiracy to ensnare the appellant. He then drew their attention to various points in the evidence. He reminded them that the Crown were asking them to say that there were elements in the story that Simmonds could not have known about, such as the Karl Kani boots and the gully, as he was not from the area. He then said that he had been struck by one of the reasons which Simmonds gave for saying nothing about the confession to anybody for over two weeks. This was that Mr Bailey had sent for him and asked him how he was getting along, in reply to which he said that he wanted to get out of jail. The judge pointed out that Mr Bailey had not been called as a witness and that no reason had been given as to why Mr Bailey should send for Simmonds. He said that this was a factor that the jury might wish to consider. He also said that the person who took the statement from him, D S Coleman, said that Simmonds never complained to him and that he did not express any hope that if he gave his statement he would get any benefits from it.

It is true that the judge drew the jury’s attention to some matters that they might like to consider when they were assessing the veracity of Simmonds’s evidence. But their Lordships consider that there was a significant omission from this part of the judge’s summing up. He ought to have drawn attention also to the factors which might indicate that the witness had an improper motive which tainted his evidence. These were that he was an untried prisoner, that it is not unknown for persons in his position to wish to ingratiate themselves with the police and that to give them information that the appellant had confessed to the crime for which he was being held by them in custody was a convenient and obvious way of doing so. He ought then to have given an express direction to the jury that they should be cautious before they accepted this witness’s evidence.

The Crown’s case against the appellant was based mainly on Simmonds’s evidence about the appellant’s conversation with Montgomery when they were all together as prisoners in the police cell. It was crucial, if the appellant was to receive a fair trial, that the jury should be told that they should be cautious before they accepted this evidence in view of the possibility that it was tainted. As this was not done, their Lordships have concluded that on this ground also the appellant’s conviction must be held to be unsafe.