Luttrell & Ors, R v

These cases raise points of practical importance in relation to lip-reading evidence. Luttrell, Dawson and Hamberger all appeal, with leave of the single judge, against the ruling of their trial judges that such evidence should be admitted. Questions arise as to how a judge should consider such evidence and as to the appropriate directions to the jury if it is admitted. All members of the constitution have contributed substantially to this judgment.

Luttrell was tried, with a number of other defendants, at Reading Crown Court before His Honour Judge Playford QC. On 10^th September 2002, before the jury was sworn, Dhaliwal and Sahota changed their pleas to guilty to a conspiracy to handle stolen goods on count 3. On 8^th January 2003, following a seventy-day trial, Luttrell, Beagley, Jheeta, Keshwala and Shergil were convicted on the same count. On 27^th February 2003 Luttrell was sentenced to 3½ years imprisonment, Beagley to 7 years, Jheeta to 7 years, Keshwala to 5 years 9 months, Shergil to 5 years 9 months, Dhaliwal to 5 years 3 months and Sahota to 10 years. Subsequently, confiscation orders were made against Sahota in a sum in excess of £700,000, with 5 years consecutively in default of payment, against Beagley in the sum of £27,000 with 15 months consecutive and against Keshwala in a small sum with 7 months consecutive. No orders were made in the confiscation proceedings against Jheeta, Shergil and Dhaliwal and the confiscation hearing against Luttrell was adjourned until after his appeal to this court. Shergil was deprived of his interest in a Toyota motorcar. Co-accused called Sandal and Acar, who pleaded guilty to count 3 before the jury was sworn, were each sentenced to 4 years and 9 months imprisonment and confiscation orders were made against them. Farry who had pleaded guilty in December 2001 was sentenced to 3½ years imprisonment. Five other co-accused were either acquitted by the jury or found not guilty on the judge’s direction. Beagley renews his application for leave to appeal against conviction and sentence, following refusal by the single judge. Sahota, Jheeta, Keshwala, Shergil and Dhaliwal all renew their applications for leave to appeal against sentence following refusal by the single judge, necessary extensions of time having been granted by the Registrar to Sahota and Shergil.

In outline, the circumstances were that, between the beginning of December 2000 and late June 2001, at various places in the United Kingdom, 9 thefts of high value computer and electrical equipment took place by way of armed robbery, burglary or theft from lorries in transit or whilst being stored. The value of the goods stolen was £6-£7 million. The goods were rapidly disposed of and it was said that the conspirators were close to those who committed the primary offences. It was the Crown case that the principal organisers of the disposal were Sahota and Beagley, directing operations behind others in more exposed positions lower down the chain. Sandal, Acar and Farry played major roles acting on the directions of Beagley and/or Sahota. Dhaliwal and Jheeta performed specific tasks, particularly for Sahota whose lieutenants they were. Shergil was the director of a courier company Vanguard Logistics of Slough. Keshwala assisted others higher up: there was no evidence of wealth in his case and he was regarded by several of his co-accused as being “a bit dim”. Luttrell joined in the conspiracy when he was offered goods and sold them on to innocent purchasers. The warehouse used at 919 Yeovil Road Slough, from which Acar’s company Silverplus traded, was central to the conspiracy. There were no company records or books of accounts, bank accounts or VAT returns. A lot of invoices passed to Silverplus from Luttrell’s company DVD. Sahota occasionally, Jheeta and Dhaliwal frequently, visited the unit. Under-cover officers posed as potential buyers. They wore recording devices and there was evidence of meetings and telephone calls between mid-February and early July 2001, surveillance logs being maintained between 24^th April and 5^th July. Between 29^th May and 27^th June audio visual surveillance was carried out. There were photographs, videos and transcripts before the court.

A company called Woodbury Systems Ltd was set up and used to sell the stolen goods. It was registered for VAT purposes on 7^th April 2000 but its office was a boarded-up flat in Tulse Hill. Its invoices named 2 telephone numbers, calls to which were diverted to a mobile telephone. In December 2000 Dhaliwal, using a forged driving licence, set up an account for Woodbury Systems at Barclays in Uxbridge. There was no supporting documentation for the 7 principal deals done by Woodbury. After laptop computers had been stolen in a burglary at Reading on 4^th/ 5^th December 2000, Woodbury invoices dated 5^th to 14^th December showed the sale of 97 laptops to a company in Solihull for over £70,000 but they were not delivered and no money was paid. According to Beagley 41 computers were supplied by Woodbury to Prima Computers and 21 other laptops, which were probably stolen, were sent out by Woodbury.

Following an armed robbery at Sunbury on Thames on 22^nd December 2000, Dhaliwal offered memory chips then stolen to a Mr Patel who paid a total of £40,000 for them. Jheeta offered electrical goods to a Mr Shakil who paid money for goods, which were never received. On 17^th May, Acar or Sandal sold 100 stolen Samsung monitors, which had come from a burglary at a lorry park in Teeside on 24^th April. In June 2001, following a robbery the previous month, £27,000 worth of Sony Discmans were invoiced by Woodbury but never delivered. On 5^th June, at Dhaliwal’s request, Beagley doctored a Woodbury invoice so that it looked like a blank and faxed it to Dhaliwal: he told an accountant that some of the stuff sold was “hot” and, once it was sold, it must never come back. In July 2001, there were recovered from a trade unit in Vauxhall four boxes of memory chips stolen in May, together with Samsung monitors which Acar and Sandal had sent there in June. In mid-July 2000 Jheeta set up a Woodbury bank account at Nat West in Gerrards Cross, claiming to be an importer of computer hardware with a turnover in excess of £1 million. He also had a personal account there and, for both accounts, Jheeta gave Acar’s address.

A telephone schedule before the jury showed many calls between Jheeta and Unit 919 and between Jheeta and Luttrell’s company DVD. Surveillance logs recorded Jheeta’s presence several times at Unit 919, between late April and late May. Sahota, Dhaliwal, Acar, Sandal, Luttrell and Keshwala were also there from time to time. There were telephone calls between Keshwala and Sandal, Keshwala and Sahota and Jheeta, Sandal and Acar. On 15^th June Keshwala, Sahota and Dhaliwal were together in a people carrier. On 23^rd June, shortly after Keshwala had gone to India, between £3 and £4 million pounds worth of Nokia mobile phones were taken in a robbery and delivered to Unit 919 some hours later. On 26^th June, Jheeta was discussing those telephones and, the following day, discussed them and a similar number of legitimate phones in telephone calls with Sandal and Acar. Jheeta said he wanted £195 per phone and, on 28^th June, they were offered to an undercover officer called Tom.

Between early February and the beginning of July, undercover police officers Paul, Tom, Alec and Neal had dealings with a number of the defendants. In February Paul met Farry and Beagley and Farry referred to his associates as having “their fingers in a lot of pies” and to an Indian guy (Sahota) who was “loaded” and very clever: he used courier companies for storage and would hit 3 or 4 lorries, loaded with £2 million worth of goods, at a time. Paul, Tom, and Alec met Farry on 28^th March, the day after £675,000 worth of French laptop computers had been stolen at Membury service station on M4. On 10^th April, Farry provided Paul with a sample of the laptops stolen on the 27^th. On 24^th April, 600 monitors were stolen from Samsung in Teeside in a burglary and, 2 days later, Farry offered these to Tom. On 12^th May £500,000 worth of Toshiba memory upgrade chips were stolen by deception in Colnbrook. Three days later, Farry and Acar discussed with Tom and Alec a new parcel which was to become available. Next day Acar gave Tom a four-page list of available property and, on the same day, officers went to Unit 919 and collected 33 monitors for which they paid £2,500. Also on the same day there was another robbery at Sony, Colnbrook, when £400,000 worth of mini-disc recorders and DVD players were stolen and Tom received a text message from Sandal offering Sony mini-discs at £42 each. On 17^th May Farry discussed with Alec the memory upgrade chips which had been obtained by deception on 12^th May and, on the 18^th, £45,500 was offered by Tom for these. On the 22^nd, Tom, Alec and Paul met Acar and Sandal at Unit 919 and paid £12,500, in an Ellesse bag, for the chips. Sahota, Dhaliwal and others arrived and a few minutes later Sahota was carrying a yellow plastic bag, not the Ellesse bag. A video of the interior of the warehouse, taken on 24^th May, showed boxes of Samsung monitors piled up and, on the same day, Sandal gave Tom a sample of the Toshiba memory chips. On 29^th May, video recording equipment having been set up, a meeting between Alec, Tom, Neal, Sandal and Dhaliwal was recorded. The officers paid the £45,500 agreed for the Toshiba memory chips, obtained on 12^th May. Sandal took £3,500 as his share. Keshwala arrived, driving Sahota’s Porsche. He gave Sandal a list of property for sale and he and Sandal took the balance of £42,000 to Sahota’s house in Gerrards Cross. On 8^th June, Farry said 600 Nokia phones would soon be available. On the night of 8^th/9^th June a trailer with £40,000 worth of household goods was stolen in a burglary at Wollaston and, on the 11^th, Acar offered these items to Alec. On the 28^th Acar offered Tom 5000 phones stolen on the 23^rd and 24^th June. On 3^rd July, Tom and Alec were introduced by Sandal to Shergil, who was described as Sahota’s right hand man. The competence of the robbery team and future deals were discussed. A price was agreed for the stolen phones, subject to Sahota making the final decision. On 5^th July, Paul and Sam met Acar , Sandal, Sahota and Shergil. At Leicester, Tom and Alec met them. Tom handed over £19,500 for the first tranch of phones and Sahota arranged for their delivery. A little later, Dhaliwal arrived and handed over the first batch. Sahota had a rucksack of money which he gave to Sandal. There followed a second similar meeting at which £28,500 was handed over. In the evening, at yet another meeting, a third handover took place for which the officers paid £104,500. At that stage, the police arrested all those involved, apart from Jheeta who was not arrested until October. Most said nothing in interview, though, in a later interview, Jheeta gave a short statement to the police. The defendants’ mobile telephones were seized and their calls analysed. Computers were recovered from their homes and from Unit 919.

The case against Beagley, who, it will be recalled, renews his application for leave to appeal against conviction, was that, trading as Slimline Computers, he brokered deals for Woodbury in relation to stolen goods. There was no proper documentation. Beagley had run a computer sales company with Sahota which went into liquidation in 1997. There were telephone calls between Beagley and Sahota during the surveillance period. One was on 25^th June 2001, the day after the £3 to £4 million worth of 8850 Nokia mobile telephones were stolen in a robbery. The document with 500 8850 on it was found in Beagley’s home. Dhaliwal was Beagley’s main contact at Woodbury. On the night of 4^th/5^th December 2000, it will be recalled that half a million pounds worth of laptops were stolen in the burglary at Reading. On the day before, there was contact between Beagley and Prima Computers and, on the day after, ten of the computers were sold via Woodbury and Slimline to Prima Computers for £10,500. The money went into the Woodbury account which Dhaliwal opened at Barclays in Uxbridge on 11^th December. Beagley sold a further batch of computers to Prima at Christmas some of which, definitely, and the rest, probably, were stolen. Payment of over £10,000 for these also went to the Dhaliwal Woodbury account, from which large sums of cash amounting to almost £90,000 were withdrawn between 20^th December and 20^th March. Beagley, with Farry, met Paul on 8^th February 2001 and sold him 6 computers. Hi-jacks were discussed. On 17^th May, a few days after the theft of the Toshiba memory chips, Beagley, on his computer, sought information about such chips. Two weeks later, he supplied Dhaliwal with the doctored blank invoice. On 25^th June, Sahota called Beagley briefly. On 6^th July, Beagley said, on the advice of his solicitor, that he would not answer questions because there had been inadequate disclosure by the police as to what the offence was about. He did not mention that he had bought computers from Dhaliwal and Woodbury, that he knew Sahota and Farry and that he had worked with Sahota and Dhaliwal. When the police searched Beagley’s home on 6^th July, they found no Woodbury invoices. Beagley produced some at court which the prosecution said were concocted. In evidence, Beagley said he knew Sahota and Dhaliwal and had set up Slimline Computers in September 2000. Dhaliwal from Woodbury supplied him with stock which he knew where to sell. He paid Dhaliwal in cash because he asked for it. No receipts were asked for or given. He never went to any place where Dhaliwal carried on business and he did not know Unit 919 existed. He assumed he was operating from its registered office in Tulse Hill. He agreed there was a substantial amount of business between Woodbury and Slimline though there were only two recorded telephone calls between him and Dhaliwal. On 5^th December he had substantial contact with Farry with whom, in the next two months, there were 137 calls, which stopped after 3^rd February. He then wanted no more to do with Farry. On 5^th December, Dhaliwal had offered him stock which he sold on to Prima. He said he was an innocent go-between with no reason to suspect anything. The purchase ledger account which he produced was not a contemporaneous document. He worked out that he had paid Woodbury almost £790,000, including over £48,000 in December and January. The Woodbury invoices he produced were genuine and had been at his accountants at the time of the police search. Calls on Christmas Day to Sahota, Dhaliwal and Farry were social and unconnected with criminal activity. On 4^th June he had sent the doctored fax to Dhaliwal. He did not recall the short call from Sahota on 25^th June. His explanation of the note about 8850 phones was that he may have been offered them by someone in the industry, but this was not Sahota; or he may have been asked if he could find some.

Beagley’s grounds of appeal relate to the treatment of statements made by Farry, the judge’s refusal to stay the proceedings because of the prosecution’s failure to investigate a man called Nicholas Miller and the adequacy of the summing-up in relation to Farry’s statements and the investigation of Miller.

We turn to the case in relation to Luttrell. Luttrell’s company DVD Technology Ltd was a successful business and the main distributor of CD and DVD products. Luttrell looked after the financial components side and his partner Keith Davies dealt with the technical side. He had known Sandal of Silverplus for several years. He knew of Acar who had worked with Sandal at a company called Conex which went into liquidation in late 1999. In January 2001, DVD made a dealership agreement with Silverplus but, the following month, Luttrell decided not to do any more business with them. On 22^nd January 2001, an armed robbery took place in which 5700 CPUs worth about £700,000 were stolen. The following day, 3 boxes of the stolen chips were sold by Luttrell to Bond House Systems Ltd in Yorkshire, who passed them on to Dimension Cargo and Global Risk Management. Mr Yates of Bond House said their first contact with DVD was in early January when they faxed over their price lists. When they found the chips were stolen, they contacted the police, who saw Luttrell on 26^th January: he made a statement on 29^th January. Luttrell suggested that the police check the CCTV on the building opposite. It was of no use because it had not been working. Dr Cheetham, a director of Bond House, having been told by the police about the robbery and having been given serial numbers of the stolen goods, checked and discovered that 3 boxes from DVD were among the stolen items. He spoke to Luttrell on 29^th January as a result of which Luttrell issued a credit note and took back the 3 boxes. Luttrell told him that he had bought the boxes for cash from 2 or 3 men whom he did not know and who gave no contact details. He had withdrawn the cash from the bank. The men said they would fax an invoice later. He realised he had acted stupidly and it could cost him £130,000. Dr Cheetham said cash deals of this kind were not normal in the industry. However, he continued to do business with Luttrell subsequently. There was evidence of telephone calls between Silverplus and DVD on 18^th, 19^th and 22^nd January including a 4-minute call from Luttrell’s mobile to Sandal’s mobile. On 23^rd January there were 3 calls from Sandal’s home to DVD followed by 8 calls from Luttrell’s mobile to Sandal’s mobile. One of them was made whilst Luttrell was on his way back from the bank with the cash and another at a time which Luttrell said was when the CPUs were being delivered to him. On 24^th January, the day of the sale to Bond House, there were 8 calls from Luttrell’s mobile to Sandal’s mobile. On 27^th January, the day after Luttrell had arranged with the police to make a statement on the 29^th, Luttrell’s mobile rang Sandal’s mobile and the call lasted 2 minutes 44 seconds. There were also calls on 21^st March from Jheeta and Kumar, who was acquitted on count 3. On 23^rd May Luttrell and Davies went to Unit 919 and Luttrell had a 10-minute conversation with Sandal, recorded on CCTV. This gave rise to the lip reading evidence to which we shall later return. On 1^st June there was a call from DVD to Sandal following a message from Sandal which the prosecution said Luttrell had made. Luttrell was arrested on 5^th July and interviewed the following day. On 26^th July he was again interviewed, making no comment on his solicitors advice. On neither occasion did he say, as he did later in evidence, that he went to Unit 919 to offer Sandal a job.

In evidence, Luttrell said that an article in a computer magazine prompted him to decide that DVD could profit by selling in bulk. On 18^th or 19^th January 2001 he received a call from someone called John, whom he assumed was already dealing with a DVD account manager. John said he had 3 boxes of CPUs available and they negotiated over a six-figure price. John came on 22^nd January about lunchtime and briefly met Keith Davies. John said the price had gone up to £125,000. He did not have the CPUs with him. The first two or three deals would have to be in cash. Although surprised, Luttrell said he was not worried. He telephoned the bank and arranged to withdraw £100,000. The balance was to come from a sort of company float. He did not ask for a business card. The following day John arrived with 3 boxes which Luttrell inspected. The labels were fine. The wraps were a bit rough and parts of the tape were broken. He handed over £125,000 in cash. The possibility of other deals was discussed. There was no paper work. John said he would fax the paper work as soon as he got back. He did not do so. Luttrell said he expected to do long-term business with John. He sold the goods to Bond House next day. The prosecution case was that John did not exist and the deal described by Luttrell was incredible. The defence claimed that Luttrell’s suggestion that the police check the CCTV camera showed that he was willing for the incident to be checked. The phone calls with Sandal on 23^rd and 24^th January related to ongoing business between Silverplus and DVD though the mobile calls were personal, involving the chasing of Sandal for payment for an order. Luttrell said he had a piece of paper with every number where he could reach Sandal. The call to Sandal’s mobile at 1.24 am on 27^th January was because Luttrell had a call which woke him up. He pressed his recall button and Sandal’s number came up. He phoned and the number answered but all he could hear was screaming and music. There was no information as to Sandal’s number stored on Luttrell’s phone but he denied that he had deleted it. He denied discussing what he was to tell the police on the Monday. He denied knowledge of Jheeta and Kumar. The only time he visited Unit 919 was on 23^rd May. He went with Davies for a quiet word with Sandal to whom he had previously offered a job because he liked him. When Sandal arrived he spoke outside to him privately about the Silverplus debt of over £3,000 and he offered him a job as agent, whereby Sandal could pay off the debt. Sandal said he had an insurance cheque coming and would pay DVD out of that. Luttrell did not know anyone called Raj and the word Raj was not mentioned. No criminality was discussed. He spoke to Sandal because Davies would not have put it as well as he would. This was the conversation recorded on CCTV cameras. Luttrell denied calling Silverplus on 1^st June although his plane left much later than the call was made.

Keith Davies gave evidence that, on 22^nd January, he had just returned from Taiwan and Luttrell brought someone over whom he introduced as a new customer. He did not know Jheeta or Kumar. In February 2001, Silverplus owed about £3,200. It would be Luttrell who chased Sandal. They both called at Silverplus to collect what money they could as they were en route to Windsor to buy clothes. There was evidence from a Mr Lewis of Luttrell’s upright and trustworthy character.

We come to the lip-reading evidence. On the voire dire, Jessica Rees gave evidence for the prosecution as to what was said at the video meeting between Luttrell and Sandal on 23^rd May. She was 39 and had been completely deaf since the age of four. She passed many O levels and A levels and obtained a degree at Oxford University. She was a skilled lip reader and relied totally on lip reading in her daily life, not using sign language. She had been doing lip-reading semi-professionally for 10 to 12 years and professionally for 5 or 6 years. Her skills had been tested by Professor Summerfield (who gave evidence on the voire dire and before the jury for the defence) in 1983 and 2000, on the latter occasion at the request of the Northumbria police. She was not told what the tape in this case was about when it was given to her. She played it and came up with a few words and phrases and told DC Todd that it was lip- readable. Her transcript described the speakers by initials. She saw the words “Raj”and “computers”. She spent 12½ hours viewing the tape and prepared a report on what was said. If a word was stressed she underlined it. Pauses were indicated by a comma and a series of dots indicated that she could not see the speakers face or was not sufficiently sure of the words. If she was less than 98% sure she would not include a word. If she was only 98% or 99% sure she would put a question mark. She did not agree with Professor Campbell that lip reading based on a silent video was intrinsically unreliable. There was a difference between adult and childhood deaf lip readers. It may be that the adult deafened could not give a reliable account of silent speech but she, having been deafened at an early age, had gained her qualifications through lip reading. In cross-examination she said that, in 1998, she was instructed by the Staffordshire police but did not give evidence. Her lip reading was used for intelligence purposes. When an audiotape became available it showed there were errors in her transcription. She insisted that her transcription had led to the discovery of other evidence. She had given evidence on some 6 occasions and had had her evidence excluded twice to the best of her knowledge, once under PACE because the defence could not get an equivalent reader and in a case called Topkaya where it was excluded for very specific reasons. She was as sure of the word “Raj” as if a hearing person had heard it. She might mis-lip-read something, just as a hearing person might mishear, but she had looked at the tape a number of times and was pretty sure what she had seen. She conceded that, in Professor Summerfield’s test, she got 452 out of 890 words right and introduced 224 unspoken words. But, since that test, new methods have been introduced so such mistakes did not happen. Lip reading for legal purposes was in its infancy and ways to improve it were constantly sought. The criteria for including in the transcripts words of which she was certain were much stricter now. She should be tested again under the new criteria. She was not aware that the police told her the case was about armed robberies or handling. On her initial viewing, she told the policeman the man said “Raj” and “computers” and he was pleased, because she had no prior information. She received 8 or 9 requests for transcriptions each week but remembered this one because the quality of this video was better than Professor Summerfield’s test video: the speakers were unaware that they were being filmed and spoke naturally. There was a clear view of the lips. Normal speech patterns were interrupted in the test video, which made it artificial. Although lip reading evidence is potentially unreliable, if a person is careful and adopts safeguards and it is done properly it can be regarded as reliable. It should not stand alone because a case would not be very strong if there was only one piece of evidence. In re-examination she said she was instructed by prosecution authorities, defence lawyers, in civil cases and by police in Europe, Canada and America.

DC Todd gave evidence of delivering the tape to Miss Rees and giving a brief explanation what the enquiry was about, namely an investigation into lorry hi-jackings. Miss Rees looked at the tape for a couple of minutes and then said half a sentence or something from it and that it was workable. He did not remember her giving a name but she said something useful and he was pleased. He was not aware of a lot of details in the case. He did not give her the name Raj Sahota. There would have been no names on the video.

Professor Summerfield also gave evidence on the voire dire for the defence. He was Professor at the Medical Research Council’s Institute of Hearing. In 1984 Miss Rees had been a patient there and it was clear that she was an exceptionally good lip reader. He gave her a general reference as to the excellence of her skills for forensic lip reading in the mid 1990s. He compared the accuracy of her transcription with an audio recording at the request of MI5. She produced a disciplined and informative transcription containing the essence of what was said and many actual spoken words. She indicated where she was uncertain but introduced some words that had not been said. In 1999 when he tested her accuracy of transcription at the request of the Northumbria police, he gave her a video recording to transcribe and compared it with an audiotape. In that test she got the gist of what was said and introduced 224 new words out of 820 and therefore, in some respects, she failed completely to convey the meaning. She made three types of errors: relatively minor changes; phrases introduced as a result of having been told the broad context of the conversation; and where an unusual word was used. She would have to have some regard to context for accurate transcription. There was nothing to check her transcription against. If unsure of a word she would leave a blank or put a question mark, so she was certain of the words she had put down. Some sounds such as P, B or M are formed in the mouth and it is impossible for a lip reader to see them. If given the context to the conversation it could be misapplied and a plausible but inaccurate transcription made. He was puzzled by Miss Rees’s criticism of the test tape because it was good and clear speech. He agreed with Professor Campbell that lip reading evidence was unreliable. On occasions there could be more than one plausible interpretation. Accuracy could not be guaranteed. He agreed in cross-examination that his report to Northumbria police described Miss Rees as an exceptionally good lip reader who performed better than 98% of people tested. Transcription by expert lip reading could provide intelligence and corroborative evidence, but, because of her inaccuracies, her evidence was unlikely to be capable of standing alone. There was potential for unreliability in such evidence. In the tests he concluded that she was exceptionally good and outstanding compared to other deaf people, but she was fallible. By repeatedly looking at a test video she increased her accuracy to over 90%. The speech on the test video was no more contrived than speaking in court. He could not comment on the accuracy of her transcription in this case. It was impossible to say if she got the word “Raj” right even if she had not been told the relevance of the name. He would want to see if the word was transcribed several times throughout the tape and was backed by the opinion of more than one lip reader: then it could provide corroboration.

Professor Campbell also gave evidence on the voire dire for the defence. She was a Professor of Communication Disorders in adults and children. She had written two reports on the admissibility of lip reading evidence. The intrinsic unreliability of it was so strong that it should never be used evidentially. What was seen by lip readers was always less than what was said. Even a very good lip reader could not overcome that limitation. There was an art in the ability to lip read but the understanding of speech was a science and the scientific community was well able to assess the validity of claims made about lip reading. There was room for error. The Association of Teachers of Lip Reading for Adults (ATLA) strongly recommend that members should not undertake forensic work because lip reading was unreliable. The eye alone could not give speech understanding as good as that of a hearing person. A person who could not hear could not make use of cultural things relating to hearing. She believed Miss Rees was the only person prepared to serve as an expert lip reader. (It is common ground before us in these appeals that that is not so: Gillian Hadfield had already given evidence at the trial of Dawson and Hamberger and there are others). In cross-examination she agreed that a tape could be slowed down and that Miss Rees had spent 12½ hours looking at 13 minutes of tape which gave plenty of time to look at the mouth. But there was no reliable indicator of the spoken conversation, just one interpretation and it would be safer not to admit it in evidence. Lip reading may be useful as corroborative evidence but should almost never be used in court. It could be useful for general intelligence gathering. Miss Rees’s lip reading, which was very good, may even be right sometimes but it is not possible to test it against anything else. Those who could transcribe, including herself, were unwilling to do so because it was a risky enterprise. In re-examination she said that occasional words could be lip-read but it was not possible to be sure of the word “Raj”. It could be any number of things for example “large”. A good lip reader uses scraps of context, such as the presence of an Asian person. It would be harder to lip-read “Raj” than to guess it.

The judge ruled that Miss Rees’s lip reading evidence was admissible in principle. He reviewed the relevant authorities. He rejected the defence submission that there was no sufficiently well established field of lip reading expertise for it to be accepted as a reliable body of knowledge or experience. There was a sufficiently defined body of quite well established opinion in particular Miss Rees, Professor Summerfield, Professor Campbell and ATLA. In 6 previous cases Miss Rees’s evidence had been ruled admissible in principle and the judge had transcripts of three of those cases including Dawson. The jury had seen the video and it was relevant to know what was being said. The jury would need expert assistance. As to reliability, Miss Rees was very experienced and, although Professor Campbell, supported by the professional body, considered all lip reading intrinsically unreliable and that it should not be used in court, Professor Summerfield found Miss Rees an exceptionally good lip reader in 1984, gave her a reference in the 1990s and reported to the Northumbria police on her in January 2000. The judge rehearsed Professor Summerfield’s evidence. He concluded that Miss Rees’s evidence supported plenty of other evidence implicating Luttrell in the conspiracy. The fact that an expert may be wrong is no reason to deprive the jury of such assistance as may be gleaned from the evidence. The judge did not accept Professor Campbell’s view: she seemed personally involved in trying to make a stand against improper forensic use of lip reading. The transcript indicated that Luttrell discussed a debt: this was accepted by the defence thereby supporting the reliability of the transcription. It was inconceivable that “Raj” could be guessed. There was no basis for excluding the evidence as unreliable.

The judge having admitted the evidence of Miss Rees, she, DC Todd, Professor Summerfield and Professor Campbell gave evidence before the jury in terms similar to their evidence on the voire dire.

We turn to the appeals of Dawson and Hamberger. On 23^rd July 2002 at Lewes Crown Court, following a trial before Miss Recorder Barnes, they were both convicted on count 1 of conspiracy to supply ecstasy and on count 2 of conspiracy to supply amphetamine. On 22^nd August 2002, Dawson was sentenced to 12 years imprisonment on count 1 and 2 years concurrently on count 2 and Hamberger to 14 years on count 1 and 3 years concurrently on count 2. A confiscation order in the sum of £700 was made against Dawson with 28 days imprisonment consecutively in default of payment. They appeal against conviction with the leave of the single judge in relation to the admissibility of lip reading evidence. Hamberger seeks to renew a ground on which the single judge refused leave, as to the terms of the summing-up precluding the jury from speculating about the scale of the conspiracy in relation to other consignments of drugs. Both renew their applications for leave to appeal against sentence following refusal by the single judge.

The primary issue in the appeal is the status of lip-reading evidence. Miss Rees gave evidence for the Crown and Gillian Hadfield for the defence. They differed in their interpretations of what was said.

The circumstances were that, in February or March 2000, Hamberger, a self-employed wheeler-dealer in his mid-fifties, was introduced to Dawson, a man in his mid-twenties who worked as a shipping clerk for a company called KDG which manufactured and exported scientific instruments. Up to the end of April 2000, there were over 100 telephone calls between the two of them and they met several times. Hamberger was also in telephone contact with Godden-Wood who was suspected of being higher in the organisational chain of the conspiracy. On or before 14^th April 2000, Dawson received at work a large box which contained 35,000 ecstasy tablets and 15 kilograms of powder containing 1½ kilograms of amphetamine. He placed it under his desk. On 20^th April he sent it with the accompanying paperwork via DHL to Bestfast Printing in Australia, who were not expecting it. The prosecution case was that Hamberger recruited Dawson to arrange for the box, which Dawson knew contained drugs, to be shipped to Australia. The case depended on observations of meetings between the appellants and of a conversation between them in a car park in Crawley which was the subject of the disputed lip reading evidence. There was also a conversation overheard by police officers on 17^th April at the Little Chef in Fontwell. In addition, there were calls from a telephone kiosk on 27^th April involving Godden-Wood and there were a series of other inter-linked telephone calls from 18^th March onwards. The defence case was that the appellants met innocently, liked each other’s company and thought they might do business together selling second-hand clothing or duty free cigarettes over the internet. Dawson had been used as an innocent dupe. He thought the box contained promotional literature and he used his own name and contact details when despatching it. Hamberger’s case was that the box had nothing to do with him.

There was evidence that, on 19^th March, Dawson met Hamberger and his girlfriend for drinks at the Ship Hotel Chichester and, two days later, Dawson sent Hamberger a fax from a sweet shop, ostensibly to substantiate Hamberger’s status at the hotel as owning a shipping business. In April there were telephone calls between Hamberger and Dawson and Hamberger and Godden-Wood, repeatedly on 9^th and 12^th April. On 12^th April Dawson booked the following day as a day’s leave from work. On 13^th he rang Hamberger, who had gone to Crawley with Godden-Wood. Dawson met Hamberger, but not Godden-Wood, and the two appellants had a conversation in a car park where their movements were recorded on CCTV cameras. Hamberger was, for the most part, facing the camera. Jessica Rees and Gillian Hadfield gave evidence about what was said. On 14^th April the parcel was seen under Dawson’s desk. From 8.10 am that morning there had been several telephone calls between Godden-Wood and Hamberger and between Dawson and Hamberger. Dawson left work about 1 pm. There were no more phone calls between Godden-Wood and Hamberger and, at 2.30 pm, police officers removed the box for testing and then replaced it. The receptionist at KDG said that Dawson had telephoned her to tell her that a package would be arriving: he asked her to sign for it and tell him as soon as it arrived. This was unusual but she did not think it untoward. When it arrived she put it under her desk and called Dawson. He came immediately and took it away. On 15^th April there were 7 calls between Godden-Wood and Hamberger and a call from Dawson to Hamberger which lasted 7½ minutes. On 16^th there were three calls between Godden-Wood and Hamberger. On 17^th Hamberger made a call from a public telephone saying the line was bad. At 6.42 pm Godden-Wood rang Hamberger. By 7.55 pm Dawson had joined Hamberger at the Little Chef at Fontwell. At the next table were DC Macbean and DC Yeats. They overheard “parcel must go by Thursday so that it marries-up with two other parcels”. There was talk of “trying to get rid of 35” and Hamberger said he would take a further 5 if needed if Dawson could only get rid of 30. There was mention of property in Ireland, caravans and meeting in The Bear on Friday and a telephone number was given by Dawson which ended 229. The appellants then left. There was no talk heard about the Internet business or clothing. On 18^th April there were more telephone calls between Hamberger and Dawson and between Hamberger and Godden-Wood. On 19^th April there were three calls from Hamberger to Dawson and one from Hamberger to Godden-Wood. The parcel was despatched in the late afternoon of 20^th April. That morning there were calls between Hamberger and Dawson and between Gooden-Wood and Hamberger. As soon as Dawson left work he rang Hamberger who called him back and they spoke for 8½ minutes. Hamberger immediately called Godden-Wood. There was evidence from KDG employees that Bestfast Printing were not their customers and there was no documentation in relation to the box, although two other entries for carriage by DHL were made that day. There were many further telephone calls between the 21^st and 25^th April between Dawson and Hamberger and Hamberger and Godden-Wood. The Crown case was that this was because delivery of the box was delayed by reason of incorrect paperwork. On 25^th April, by which time the goods should have arrived in Australia, and on 26^th April there were many more similar telephone calls. The box was delivered to Bestfast Printing on 27^th April. It was not expected. The intended recipients were not traced. On the same day there was a call between Dawson in a public telephone box in Barking and Godden-Wood’s mobile. Dawson spoke to Godden-Wood on the telephone. This was the only time they spoke. On 19^th July 2000 the appellants were arrested. Following Dawson’s arrest Hamberger’s mobile called Dawson’s mobile seven times. In interview Dawson, denied sending the fax. to Hamberger on 19^th March. That was a lie, as handwriting evidence was to prove.

On the voire dire, Jessica Rees gave evidence of her O and A level qualifications and her Oxford degree. She had been deaf from the age of four. She had worked as a journalist but from 1990 her main occupation was studying CCTV videos. In the last two years the method of studying had been tightened and unified. She watched each video a minimum of seven times before starting to translate. She used top of the range equipment which enabled lightening and darkening or enhancing of any aspect of the tape, slowing it down and freezing the frame. Where there were difficulties she would watch a frame or frames up to forty times. She would only put something in her report if she was sure it was right. If she was less than absolutely sure she would put a question mark before the word. She would add alternatives in brackets and put dots if there were gaps. In 2001 she dealt with 200 cases. In May 2002 there were thirty to forty enquiries. She had given evidence in court scores of times. She was an accredited lip-reading expert. The CCTV video in the Crawley car park was of poor quality but not the worst: she would give it four out of ten. Hamberger sometimes turned from the camera and there was a lot of foliage impeding the view of his face from time to time. (The members of this court having seen the video it seems worth commenting that, although there were a few tree branches, there was no apparent foliage impeding the view). She spent eight and a half hours replaying the tape. She was not given any indication of what the men might be speaking about. She was able to interpret homophones (words which sounded the same such as “sun” and “son”) from the context in the same way as a hearing person. Homophenes, (words which look the same on the lips such as “pills” and “mills”) could be confusing but she watched each segment of the video up to forty times to make sure she had the right word. She believed she saw the words “pills”, with a query, “police”, “bringing in the stuff” and “millions”, all of which Miss Hadfield disputed. In cross-examination, she agreed that Hamberger mumbled on occasions but that was not much of a problem. It was possible she may have made mistakes but, in view of her experience, a mistake was highly unlikely.

On the voire dire, Gillian Hadfield gave evidence for the defence. She was 53 years old and had been deaf from the age of 13 and completely deaf from the age of 22. She left school at the age of 16 and worked first in a medical lab and then for the Inland Revenue where she became a tax inspector managing staff of over forty people. She was an accredited lip reading expert and examiner of lip speakers. From time to time she worked for government agencies, solicitors, barristers and TV companies, but only when pressed because of her other commitments and full time job. She had been doing such work for two years. She had fifteen to thirty cases in a year and at the time of the trial she had two or three cases. Reading from the side rather than face on would affect accuracy. Poor lip shapes and unexpressive faces could be difficult to read. Sometimes teeth were a problem. With homophones a certain amount of guess work was involved. With homophenes, the lip reader had to look carefully. Her equipment consisted of a twenty-inch TV screen and a good quality home VCR which she could slow down. She did not think the freeze frame was beneficial, as the free flow of normal speech was lost. This tape was not good quality.

The Recorder said that both were lip reading experts. She applied the Bonython test (to which we shall come) and ruled the evidence admissible. The video tape was played to the jury. Miss Rees and Miss Hadfield gave evidence substantially in accordance with what they had said on the voire dire. There were significant differences between them as to the interpretation of a number of passages in the tape to which the Recorder drew attention in the course of her summing up and a schedule of which, prepared by Miss Hadfield, was before the jury. We shall return later to the terms of the summing up.

The appellant Hamberger gave evidence that he was 56, had never had anything to do with drugs, had been in severe debt and could not read or write. He was a wheeler-dealer. He met Godden-Wood, who had public houses and was a bookkeeper, seven or ten years before, but they lost touch. They resumed contact in early 1999. Godden-Wood was an alcoholic whose house had been repossessed. Hamberger was introduced to Dawson sometime before February 2000. He liked him and met his family. They telephoned each other a lot because they did not live near each other. They decided to go into business together. Dawson was educated and had a lot of contacts in his company to whom he could sell goods. Hamberger produced two documents about selling cigarettes on the Internet but one of them was dated after the conspiracy. He thought the staff at the Ship Hotel were looking down their noses at him, so he got Dawson to fax something making him look like a big shot. On 13^th April Godden-Wood gave him a lift to Crawley as a favour. Dawson stopped for a drink on the way to London. He walked him back to his car, but there was no conversation about drugs in the car park or elsewhere. On 17^th April he had a cup of tea in the Little Chef and a conversation with Dawson which the officers could not have overheard. He could not remember what they talked about, but it was not drugs. He may have mentioned caravans and the Bear but he would not have gone there. On 27^th April he met Dawson at the Royal Oak. He used a landline to call Godden-Wood and let Dawson speak to him. He continued to see Dawson until his arrest and called him on 19^th July with a view to playing golf. He kept ringing after Dawson’s arrest because he didn’t know why he did not answer. He was mildly curious because he thought Dawson was helping the police with their enquiries. In cross-examination, he said he was not concerned that Dawson was with the police on the 19^th July because he had no reason to think that Dawson had done anything wrong. The telephone calls between the two of them were in connection with their business. Although nothing had actually happened in business, other than one fax dated the 14^th June, he did not know what they had talked about in twenty-three calls made over eight days. He had never been to KDG and the parcel was nothing to do with him. He had no idea why Dawson rang him as soon as the parcel left KDG or why he then rang Godden-Wood.

Dawson gave evidence that he had no idea the box contained drugs. He thought it was advertising brochures. He did not know Godden-Wood and had never heard of him. He had talked to Hamberger a lot about business and social matters. He had booked a day’s holiday for the 13^th April because he was going to an interview with a Japanese firm. He had to go past Crawley. He had no idea what the car park meeting was about: it was certainly not about a box of drugs. He had collected the box from reception but he did not expect it and did not remember any conversation warning the receptionist that the box was coming. He just sent it on its way to DHL. On 17^th April it was noisy in the Little Chef and he and Hamberger had moved away from the officers. He did not recall their conversation. “Caravan” was mentioned. There was no reason for the numbers thirty, thirty-five or five to come up. He did not remember the conversation in the phone box on 27^th April. He gave handwriting samples to the police. He had made a mistake when saying he did not send the fax. He remembered nothing about the conversation in the car park. It was just chitchat. He had signed the customs declaration for the box which he sent to Australia. He did not think there was anything suspicious about it. He could have sent the box to dispatch but he had repacked it because it was bulging and he thought it was going to burst open. He was not supposed to repack goods but he was trying to help. The fact that there was no record of the box was just an oversight.

We turn to the grounds of appeal. Mr James Guthrie QC on behalf of Dawson, Mr Peter Griffiths QC on behalf of Hamberger and Mr Forlin on behalf of Luttrell, in mutually supportive and adoptive submissions, submitted, first, that the proper test in relation to the lip reading evidence was and is reliability. As such evidence has never been shown to be reliable it should never be admitted. The submission was later qualified, acknowledging that, although face to face lip-reading might be admissible, lip-reading video footage is novel and unreliable. Alternatively, if such evidence is admissible it must be carefully screened by the judge who, on the voire dire, must direct himself that lip reading generally, even in ideal circumstances by the most skilful, is a skill which can never be entirely accurate or exact, that many sounds forming the whole or part of words uttered in speech do not involve lip movements at all and that many different sounds forming the whole or part of entirely different words involve the same or similar lip movements. Furthermore, where the subject matter of the lip reading evidence is video or film footage, the judge should direct himself and, if the evidence is admitted, in due course direct the jury in summing up, as to potential weaknesses in the evidence arising from the quality of the facial images, any physical features of the speaker rendering reading of his speech more difficult, the completeness or otherwise of the lip reading transaction, whether any information concerning the footage has been provided to the lip reader before or during the transcription process, the extent to which any other lip reader of appropriate skill agrees or disagrees with the transcription relied on by the prosecution and any further specific weaknesses relied upon by the defence. The judge must also exercise his discretion at common law and under section 78 of PACE in deciding whether to admit the evidence.

As to reliability, this court, without objection, heard fresh evidence on behalf of the appellants from Professor Moore and we were referred to a number of English and other authorities.

Professor Moore is professor of auditory perception at Cambridge University and is a distinguished expert in relation, among other things, to the perception of sound, the mechanisms of hearing and the relationship of auditory abilities to speech perception. He produced a report dated 19^th November 2003, the joint author of which was Allen Hirson, a senior lecturer in phonetics and head of the speech acoustics laboratory at City University. He said that, in relation to the accuracy of speech reading, the information provided from visible sources is incomplete. Important sources of information are lacking to even the best speech readers, so speech reading inevitably involves inference from incomplete information. He described the different ways in which specific speech sounds are produced, some of which, caused by the vibration of vocal folds, are not visible. The formant frequencies, which determine the identity of vowel sounds, are related to the position of the tongue, lips and jaw so are, to some extent, visible, but vowels produced by narrowing towards the back of the mouth are harder to speech read than vowels produced by narrowing towards the front of the mouth. Many pairs of sounds will look identical to a lip reader whose skill can assist to a limited extent. Sometimes context helps. In isolation, it is difficult if not impossible to distinguish between, for example, “pa” and “ba”. In normal running speech most of the information is carried by the consonants, very little by the vowels. It is difficult to estimate the overall proportion of words that can be identified by vision alone. Even the best speech readers achieve only up to 80% correctness when the words are presented in clearly spoken sentences. The best that can be achieved from repeated viewing is consistency- perhaps 85% - not certainty. Transcription is susceptible to unconscious bias arising, for example, from knowledge of the source of a tape or from promptings from the movements of another present in the room during transcription. The only safe way to transcribe a video is if no-one else is in the room and the transcriber has been told nothing about the case. The accuracy of speech reading can be decreased by reductions in illumination, viewing from oblique angles and increase in distance between speaker and speech reader. Single syllable words with little context are very difficult to interpret. Because speech readers rely heavily on context, proper nouns, particularly the names of individuals, are unlikely to be speech read accurately. On its own, speech reading evidence cannot provide a firm basis for establishing what a talker has said based on a video recording. It is an advantage in lip reading to have been deaf early in life. Commenting on Professor Summerfield’s test of Miss Rees in 1999, Professor Moore said that he did not doubt that Miss Rees is one of the best lip readers around.

Although we accept Professor Moore’s evidence that the information provided for speech reading from visible sources is incomplete, for the reasons which he explained, and that care must be taken not to prompt a transcriber by prior information, it does not follow that speech reading evidence is so unreliable that it should not be admissible in court. Furthermore, we do not accept that Professor Moore was right to assent to a proposition put to him in cross-examination by Mr Griffiths QC that speech reading can never be “entirely accurate” in the sense that a lip-reader cannot reach an understanding of speech which, even if incomplete, is free from error. There is ample material before us to demonstrate, as is to be expected, that, depending, in particular, on the quality of the video and the visibility of the lip, jaw and other facial movements of the speaker, speech reading may be free from error. Indeed it may be possible to make a complete and accurate transcription. One obvious example is provided by a defendant, of whose videoed speech Miss Rees made a transcript, who was so impressed by its accuracy that he believed, wrongly, that he had been bugged by an audio device. Professor Moore accepted that the best speech readers, of whom Miss Rees is undoubtedly one, can achieve correctness of 80%, or higher on repetition of videoed sequences. Of course, this is significantly short of perfection. It may also be less than Miss Rees claims for herself, in only putting down a word when 100% sure and putting a query when 98 to 99% sure, though caution needs to be exercised when comparing percentages of this kind. Whether speech reading evidence is admissible and to what extent it can be relied on to establish the case against a defendant depends on the application of legal principles to which we now turn.

For expert evidence to be admissible, two conditions must be satisfied: first, that study or experience will give a witness’s opinion an authority which the opinion of one not so qualified will lack; and secondly the witness must be so qualified to express the opinion. The first was elucidated in Bonython (1984) 38 SASR 45, where King CJ (at p.46) said that the question “may be divided into two parts: (a) whether the subject matter of the opinion is such that a person without instruction or experience in the area of knowledge or human experience would be able to form a sound judgment on the matter without the assistance of witnesses possessing special knowledge or experience in the area, and (b) whether the subject matter of the opinion forms part of a body of knowledge or experience which is sufficiently organised or recognised to be accepted as a reliable body of knowledge or experience, a special acquaintance with which by the witness would render his opinion of assistance to the court.”

If these two conditions are met the evidence of the witness is admissible, although the weight to be attached to his opinion must of course be assessed by the tribunal of fact: Robb (1991) 93 Cr App R 161, 165; Dallagher [2002] EWCA 1903 [2003] 1 Cr App R 12 para 23. (We use the term “opinion” when referring to the evidence of experts, although, as observed in Phipson on Evidence (15^th Ed. para 37-01 and emphasised by Mr Guthrie, its use is simply one of convenient terminology, and sometimes, as in this case and as is reflected in s.30(4) of the Criminal Justice Act 1986, the expert’s evidence can readily be regarded as evidence of fact.) It might be added that, as with any evidence, expert testimony will not be admitted unless it is relevant in the sense that “it is logically probative or disprobative of some matter that requires proof”: per Lord Simon in Kilbourne, [1973] AC 729, 756D.

As we have indicated, the appellants argued that evidence should not be admitted unless it passes a further test, that the evidence can be seen to be reliable because the methods used are sufficiently explained to be tested in cross-examination and so to be verifiable or falsifiable. Where, as here, the Crown is seeking to adduce the evidence in a criminal trial, this could properly be considered by the court when deciding whether to refuse to allow the evidence, under s.78 of the Police and Criminal Evidence Act 1984 or otherwise, in order to ensure a fair trial. We cannot accept that this is a requirement of admissibility. In established fields of science, the court may take the view that expert evidence would fall beyond the recognised limits of the field or that methods are too unconventional to be regarded as subject to the scientific discipline. But a skill or expertise can be recognised and respected, and thus satisfy the conditions for admissible expert evidence, although the discipline is not susceptible to this sort of scientific discipline. Thus, in In re Pinion decd., [1965] Ch 85 the court was willing, indeed felt obliged, to hear expert evidence on the question whether a collection of paintings and other objects had aesthetic worth so that their display would be of educational value and for the public benefit, notwithstanding, as Harman LJ observed, “de gustibus non est disputandum”.

In some cases, the reliability of the evidence might be relevant to whether the conditions of admissibility are satisfied. Thus in Gilfoyle, [2001] 2 Cr App R 5 at para 25, it was observed that English law will not consider expert evidence properly admissible if it is “based on a developing new brand of science or medicine….until it is accepted by the scientific community as being able to provide accurate and reliable opinion”. In Robb (cit sup), a case concerning the admissibility of evidence of voice identification, which was acknowledged to be “an expert field”, the court had to decide whether the particular witness’s techniques were insufficiently recognised within his profession for him to be properly qualified to give expert evidence. Similarly, evidence might be so lacking in “prima facie reliability” that it has no probative force or its probative force is too slight to influence a decision: Clarke, [1995] 2 Cr App R 425, 432.

However, while reliability of evidence can be relevant to whether the conditions of admissibility are met, in itself reliability goes to its weight. In the Scottish case of Davie v Magistrates of Edinburgh, (1953) SC 34, to which Mr Guthrie referred, Lord Cooper, President, rejecting a submission that the court was bound to accept the evidence of an expert witness in the absence of contrary evidence, put it as follows (at p.40):

“[Expert witnesses’] duty is to furnish the Judge or jury with the necessary scientific criteria for testing the accuracy of their conclusions, so as to enable the Judge or jury to form their own independent judgment by the application of these criteria to the facts proved in evidence. The scientific opinion evidence, if intelligible, convincing and tested, becomes a factor (and often an important factor) for consideration along with the whole other evidence in the case, but the decision is for the Judge or jury. In particular the bare ipse dixit of a scientist, however eminent, upon the issue in controversy, will normally carry little weight, for it cannot be tested by cross-examination nor independently appraised, and the parties have invoked the decision of a judicial tribunal and not an oracular pronouncement by an expert. ”

Lip-reading evidence from a video, like facial mapping is, in our view, a species of real evidence ( see per Steyn LJ in Clarke at 429). Although at one time a more conservative approach had been adopted, the policy of the English courts has been to be flexible in admitting expert evidence and to enjoy “the advantages to be gained from new techniques and new advances in science”: Clarke, at p.430. (It appears that there has been a similar trend elsewhere: see Cross and Tapper on Evidence (9^th Ed) p.523, but cf Ormerod, “Sounding out Expert Voice Identification”, [2002] Crim LR 771 at p.774, about the position in the USA) The preferred view, and in our judgment the proper view, is “that so long as a field is sufficiently well-established to pass the ordinary tests of relevance and reliability, then no enhanced test of admissibility should be applied, but the weight of the evidence should be established by the same adversarial forensic techniques applicable elsewhere”: Cross and Tapper (loc cit).

When these principles are applied in the present cases, we are entirely satisfied that lip-reading evidence as to the contents of a videoed conversation is capable of passing the ordinary tests of relevance and reliability and therefore being potentially admissible in evidence. Lip-reading is a well recognised skill and lip-reading from video footage is no more than an application of that skill. It may increase the difficulty of the task, as may the speaker’s facial features and the angle of the observation, but the nature of the skill remains the same. Furthermore, the judge in Luttrell and others and the Recorder in Dawson and Hamberger were entitled to conclude on the evidence which they had heard on the voir dire that the evidence of Miss Rees, and of Miss Hadfield in Dawson and Hamberger’s case, was admissible. They were also entitled, in the exercise of their discretion at common law and under s78 of PACE, to admit that evidence. It does not of course follow that, in every case where lip-reading evidence is tendered, it will be admissible. The decision in each case is likely to be highly fact sensitive. For example, a video may be of such poor quality or the view of the speaker’s face so poor that no reliable interpretation is possible. There may also be cases where the interpreting witness is not sufficiently skilled. A judge may properly take into account: whether consistency with extrinsic facts confirms or inconsistency casts doubt on the reliability of an interpretation; whether information provided to the lip-reader might have coloured the reading; and whether the probative effect of the evidence depends on the interpretation of a single word or phrase or on the whole thrust of the conversation. In the light of such considerations, (which are not intended to be exhaustive) a judge may well rule on the voire dire that any lip- reading evidence proffered should not be admitted before the jury. As to the skill of such a witness, we have been told that there are presently only four witnesses in this country (including Miss Rees and Miss Hadfield) who undertake this kind of forensic work. As and when new witnesses appear, it will be entirely appropriate, when they first give evidence, for their expertise to be challenged and tested by reference, in appropriate cases, to disclosed material bearing on their skill or lack of it. But, so far as Miss Rees and Miss Hadfield are concerned, if they give evidence in future cases, we would not expect a trial judge to permit extensive trawling through their past “successes” and “failures”. As we have indicated, the material before us establishes conclusively that Miss Rees is one of the very best lip-readers. And, although her background and experience differs, nothing in this case leads us to believe that Miss Hadfield is other than skilled for the purpose of giving lip-reading evidence.

The evidence having been admitted, the next question which arises is whether it was dealt with appropriately in the summing-up in each of these two cases. On behalf of the appellants essentially three issues have been raised:

The appellants submitted, first, that a special warning is always necessary when lip- reading evidence is introduced and, secondly, that the directions in these two cases were inadequate. On behalf of the prosecution, Mr Russell Flint Q.C. submitted that no such direction was called for in either of these cases and that, in any event, the directions to the juries were sufficient.

With certain limited categories of evidence it is necessary for the judge to give the jury a special warning, highlighting both the potential dangers involved in relying on the evidence in question, and the need to exercise particular caution, e.g. eyewitness evidence identifying a defendant: R v Turnbull [1976] 63 Cr App R 132, 1977 1 Q.B. 224; the identification of a defendant by his voice: R v Hersey [1998] Crim LR 281; R v Gummerson and Steadman [1999] Crim LR 680; R v Roberts [2000] Crim LR 183; and alleged “cell confessions” by defendants: R v Pringle UKPC 17 of 2002 unreported, 27 January 2003; Benedetto V The Queen and Labrador v The Queen [2003] 1 WLR 1545; Pollitt v The Queen [1991-1992] 174 C.L.R. 558.

The general principle derived from those and other cases is that a “special warning” is necessary if experience, research or common sense has indicated that there is a difficulty with a certain type of evidence that requires giving the jury a warning of its dangers and the need for caution, tailored to meet the needs of the case. This will often be the case where jurors may be unaware of the difficulty, or may insufficiently understand it.

The strength of the warning and its terms will depend on the nature of the evidence, its reliability or lack of it, and the potential problems it poses. For instance, it has been recognised that identification of a suspect by voice is less reliable than visual identification evidence, and accordingly usually requires a warning that is couched in stronger terms: see Hersey; Gummerson and Steadman and Roberts.

We have no doubt that lip reading evidence requires a warning from the judge as to its limitations and the concomitant risk of error, not least because it will usually be introduced through an expert who may not be completely accurate: as we have indicated above, the material before this court indicates that lip reading evidence will, on occasion, fall significantly short of perfection. That imperfection does not render the material inadmissible, for the reasons we have already explained, but it does necessitate a careful and detailed direction. As with any “special warning”, its precise terms will be fact-dependent, but in most, if not all cases, the judge should spell out to the jury the risk of mistakes as to the words that the lip reader believes were spoken; the reasons why the witness may be mistaken; and the way in which a convincing, authoritative and truthful witness may yet be a mistaken witness. Furthermore, the judge should deal with the particular strengths and weaknesses of the material in the instant case, carefully setting out the evidence, together with the criticisms that can properly be made of it because of other evidence. The jury should be reminded that the quality of the evidence will be affected by such matters as the lighting at the scene, the angle of the view in relation to those speaking, the distances involved, whether anything interfered with the observation, familiarity on the part of the lip-reader with the language spoken, the extent of the use of single syllable words, any awareness on the part of the expert witness of the context of the speech and whether the probative value of the evidence depends on isolated words or phrases or the general impact of long passages of conversation. However, as we have indicated, the precise terms of the direction will depend on the facts of the case, and the instruction to the jury in this, as in many other areas, should never be given mechanistically.

In Luttrell’s case Judge Playford QC rehearsed with impeccable accuracy about which no complaint is or could be made, the evidence given by Miss Rees, Professor Summerfield and Professor Campbell. At volume 3 of the transcript of the summing-up at page 7D said

“I feel that it is best to direct you, even if you accept Miss Rees’ evidence, the safe course for you is to use her evidence as confirmation only in accordance with what Professor Summerfield said seeing the absence of objective testing since 1998. That means that you should not use Miss Rees’ evidence, even if you accept it, as the primary basis for a finding of guilt in the case of Mr Luttrell. That is not in any way, you understand, to cast aspersions on Miss Rees’ skill or accuracy, it is simply that having regard to the absence of up-to-date testing and in the light of Professor Campbell’s views as well as Professor Summerfields it may be unsafe to rely on that sort of evidence for proof on its own of a criminal offence. And as already mentioned, of course, what one has to bear in mind, as I said before, is that forensic lip reading is very much in its infancy so it is better to be safe than sorry”.

That was a cautious and wise direction in this case. But we comment, in passing, that such a direction will not in our judgment be necessary in by any means all cases where the prosecution rely on lip reading evidence. It is highly unlikely that lip reading evidence will ever stand alone because, unless there is other evidence, it is unlikely that a video recording will ever have been made. But there is in our judgment no reason in principle why lip-reading evidence should not establish a prima facie case. The usual principles will apply in this regard, and the judge will need to consider the strengths and weaknesses of the evidence, particularly in the light of the matters that we have set out above. As we have already indicated, the directions will be fact dependent..

At page 16 of volume 3 of the transcript, having rehearsed Miss Rees’ evidence in relation to identifying the words “Raj” and “computers” and Professor Campbell’s evidence that she could not identify those words, the judge said this

“If you are sure that Miss Rees did spontaneously come up with the word “Raj” and that she is right, then that clearly contradicts Mr Luttrell’s sworn assertion that he knew no Raj, had never come across a person called Raj and did not use that word in conversation then or at all. It would demonstrate, if that be right, a knowledge of Raj Sohata that Mr Luttrell has been at pains in court to deny, and so long as you are satisfied that there is no innocent explanation for what in those circumstances must be a lie, then that is capable of being evidence going to confirm guilt. If, on the other hand, you are not sure that the word Raj was used, whether it be by a mistake or in some way unwittingly or otherwise being suggested to Miss Rees, being put into her mind, if you think that is or may be the case, then that would presuppose serious unreliability in Miss Rees’s transcript. If that is your conclusion, you should ignore the lip reading evidence altogether, since, frankly, there would be no way you could be sure of any other part of that transcript. So that may be an approach which appeals to you, or it may not, members of the jury. It is merely my suggestion and I make absolutely clear that the directions I gave you at the outset in regard to this expert evidence holds good and it is a matter entirely for you what you make of it at the end of the day”.

The direction which the judge had earlier given appears at page 3C of the same volume:

“Expert evidence is permitted in a criminal trial to provide you with information and opinion which is within the witness’s expertise but which is likely to be outside your experience and knowledge. It is by no means unusual for expert evidence to be called, although it is fair to say, as you will have gathered from the evidence in this case, lip reading evidence is, at least as far as the court is concerned, a newish area of expertise and is fairly unusual. It is important that you should see this expert evidence in its proper prospective, which is that it is before you as part of the evidence as a whole to assist you in regard to one particular aspect, namely what was said at the meeting on 23^rd May. A witness called as an expert is entitled to express an opinion and you are entitled, and would no doubt wish, to have regard to that evidence and to the opinions expressed by the experts when coming to your own conclusions. You should bear in mind that if, having given the matter careful consideration, you do not accept the evidence of the experts or any of them you do not have to act upon it. It is for you to decide whose evidence and whose opinions you accept, if any. You should remember that this evidence relates only to part of the case and whilst it may be of assistance to you in reaching a verdict you must reach your verdict having considered all the evidence”.

In the light of those directions, subject to the matter of disclosure to which we will come later, it is impossible to accept Mr Forlin’s submission that the admission of Miss Rees’s evidence rendered Luttrell’s conviction unsafe. Although the judge did not set out a number of the matters which we have indicated should usually form part of the directions on this subject, his instruction that the jury should only treat this evidence as supportive of other evidence establishing guilt, together with his full rehearsal of what had been said by the relevant witnesses, has led us to the firm conclusion that his directions overall were fair and sufficient. Quite apart from the lip-reading evidence, there was abundant evidence of Luttrell’s guilt which we have earlier set out, in paragraphs 10 and 11 in particular, as to the circumstances in which he bought this equipment for a huge cash sum from a man of whose name and provenance he was ignorant and in the absence of any documentation.

We turn to the Recorder’s summing-up in the case of Dawson and Hamberger. It contained a direction in relation to expert evidence in accordance with the JSB recommendation but there was no “special warning” of the kind that we have indicated should usually be given. Moreover, as is apparent from what occurred when leading counsel raised the matter with her towards the end of her summing-up (transcript T7 page 44), she deliberately decided not to remind the jury of the evidence in relation to Professor Summerfield’s testing of Miss Rees in 1999 and the cross-examination in relation to that. Miss Rees had refused to attempt a transcription of one of Professor Summerfield’s videos because of its poor quality and there was evidence that the quality of Hamberger’s video was even worse. Bearing in mind that forensic lip reading was in its infancy we think that this was a mistake by the Recorder. Furthermore, although the Recorder correctly directed the jury at page 10C of volume T7, that they must not try and discern the lip movements in the video, she also directed them in terms which may have been understood to preclude them from looking at the video at all. This was a misdirection, because the quality of the video, in terms of the extent to which Hamberger’s face and lips were visible, was the subject of a dispute which the jury could not properly resolve unless they looked at the video. Accordingly, not without a degree of hesitation, (bearing in mind, first, that Miss Rees’s evidence about “pills” was not certain and her reading of “bringing things in” was at variance with the prosecution case of exporting drugs and, secondly, the large amount of other evidence in relation to both the appellants), we have concluded that it is possible that, had these non-directions and the misdirection not occurred, the jury’s verdicts might not have been the same.

We turn to the question of disclosure. In both cases the appellants complain about inadequate disclosure made by the Crown in respect of material that, they say, casts doubt upon the reliability of Miss Rees’ lip reading skills and reliability. The Crown have disclosed material for the purposes of these appeals some of which (for example, many of the internal CPS memoranda) would not have been disclosable at trial, but much of which the appellants say should have been disclosed and which, they submit, could have been deployed both on the voire dire and in cross-examination of Miss Rees before the juries. It is said that this in turn would, in the Dawson and Hamberger case, have “radically affected the tone and content of the summing up”.

The focus of the complaint by Dawson and Hamberger is material relating to six cases with which Miss Rees had been previously involved. The first is Kenyon and Bowen tried in the Stafford Crown Court in 1999. Two defendants were charged with the murder of the husband of one of them. Before she was charged, the widow of the deceased visited her co-defendant in prison, and video recordings were made of conversations that she and other visitors had with him. These were interpreted by Miss Rees, and her transcripts were instrumental in the deceased’s wife being charged. The police were able to carry out a test of Miss Rees’ reliability in interpreting the conversations because there was an audio recording of one of the prison conversations. A detective constable made a transcript of it and compared it with the transcript made by Miss Rees. He raised concerns about the results saying that Miss Rees’ transcript contained approximately 2,100 words spoken by the accused Kenyon, and continuing: “I examined the transcripts made by Miss Rees and myself. I found a vast difference in the two transcripts. I recorded that she only accurately recorded approximately 234 words spoken by Kenyon. Therefore approximately 1,800 were added.” In ensuing correspondence, Miss Rees pointed out that some of the criticisms made of her work reflected errors on the part of the police in interpreting the audio-tape or mistakes in typing up her notes. She also observed that her task was the more difficult because the subject matter of the conversation kept changing and what she was lip reading lacked context. However, she did acknowledge shortcomings: “... there is one specific area which I have identified …where I have completely failed to grasp the gist of the conversation and this is of extreme concern. There are also other single errors.” In a later letter, she attributed the quality of her work to trying to lip read for too long at a stretch, and, apologising for her work, she stressed that “valuable lessons have been learned”. In the event, Miss Rees was not called to give evidence and her work was not used at the trial. This decision is explained in an advice written by the Crown’s junior counsel. Commenting upon the police’s attempt to use the audio-tape to measure Miss Rees’ reliability, he wrote: “In theory, this was laudable. In practice, I take the view that as a “test” it was unfair and unreliable”. He also made it clear that the decision not to call Miss Rees to give her lip reading evidence was not taken because of the test, but because the Crown considered that they had very strong evidence without it and were concerned that the defendants might exploit challenges to Miss Rees’ evidence to “sidetrack the jury”.

Secondly, Gray, was tried in the Liverpool Crown Court in May 2000: Miss Rees gave evidence for the prosecution of what was said in a videoed conversation in which the defendant took part, a challenge to the admissibility of her evidence having been rejected. In the event, however, the Crown invited the Jury not to place any reliance upon Miss Rees’ evidence. It is apparent from a note that the prosecution, we are told, placed before the judge, that this decision was taken in light of the prosecution learning of the “test” of Miss Rees’ transcription in Kenyon and Bowen and also of the results of another test of Miss Rees undertaken by the Institute of Hearing Research, probably referring to Professor Summerfield’s test. However, the evidence from the prosecutor provided to this court makes it clear that the view was also taken by the prosecution that they had sufficient evidence and did not want to open up unnecessary areas of dispute and possible grounds of appeal.

Next, Mara in the Nottingham Crown Court in 2001, in which defendants were alleged to have disposed of a corpse following a murder: there were video recordings of conversations between the defendants in Lincoln Prison in which they were said to have been discussing dismembering and burying the corpse. Miss Rees provided lip-read interpretation of what was said. The prosecution did not rely upon her evidence and, we understand, offered no evidence against the defendants. A note disclosed for the purposes of this appeal explains the decision: when concerns were first raised about her reliability in light of Kenyon and Bowen, Miss Rees explained that she had changed her working practices since 1999. However, the note continues: “The pros[ecution] were re-assured but sought to prove her reliability by getting a report from Quentin Summerfield who gave a positive evaluation. Then cracks began to appear as to her past reliability none of which appear in her CV. We got the video transcript re Stafford and the correspondence with her, we were concerned. There were serious doubts as to her reliability. It was the Stafford case that has influenced us most significantly. In a case at Liverpool she was to give evidence, doubts arose and she was not used.”

We need refer only briefly to the next two other cases mentioned by Mr Guthrie. In Edgar and Edgar, at Woolwich Crown Court in about June 2000, it appears likely from the material before us that the decision not to proceed against the defendants on the more serious charges was again taken because of concerns engendered by Kenyon and Bowen. Ray and Cornish in the Newcastle Crown Court in about March 2000 was the case which led to Professor Summerfield conducting the tests to which we have referred. Miss Rees’ evidence was not used by the Crown. According to her, the view was taken that the evidence against the defendants was strong enough without having to rely upon her potentially controversial and distracting transcripts.

Finally, Musgrove and Hibbert, involving drugs offences, was heard in the Nottingham Crown Court in 2000: Miss Rees provided a report dated 18 April 2000, together with a transcript of what she claimed she could lip-read from a video. She observed that the videotape was “not of particularly good quality”, although it was by no means the worst that she had viewed. She also said that the conversation was “very incomplete”. Miss Rees gave evidence of her attempts to interpret the video recording. It then came to light that part of the conversation had also been audio-recorded by the video recorder. This appears from a judgment on 18 October 2002 in this court upon an application by Musgrove to appeal against his conviction. In the course of the judgment on that application, the voices are said to have been recorded “in a less than adequate way” and that there had been picked up “something by way of sound, although it is not suggested that anyone could detect exactly what was said”. Nevertheless, a Detective Constable made a transcript of it. The complaint is that, had this material been disclosed, the two transcripts could have been compared and deployed as evidence of her reliability. It is the appellants’ submission that there was “almost no correspondence” between them. It seems likely that the first trial was abandoned and there was a re-trial.

Miss Rees was also involved in Neil, which was listed in the Southwark Crown Court in June 2003 and in which the Crown offered no evidence. According to a note which we understand to have been written by the counsel representing the defendant in that case, the lip-reading evidence of Miss Rees about a conversation recorded on CCTV footage was “radically different” in its conclusions from the evidence of experts instructed for the defendant, including Miss Hadfield; and the Crown informed the Court that there was “no proper basis to secure a conviction”. That case does not, of course, give rise to any question of disclosure, because it was subsequent to the trials giving use to the present appeals, but it is put forward on the basis that it is another example of the Crown deciding not to rely upon Miss Rees’ evidence.

The appellant Luttrell complains about the failure to disclose not only material relating to these cases but also tests relating to Miss Rees’ lip reading skill that were carried out on two other occasions. One was the test conducted by Professor Summerfield at the request of the Northumbria police, about which he gave evidence during the trial of Luttrell. The second was arranged when, in March 1999, it was thought that Miss Rees might give lip-reading evidence at the Old Bailey and the prosecution sought to test her reliability. It involved her transcribing a conversation between two actors recorded on a video. In light of this, the prosecution decided not to call her to give evidence. Her transcription has been described as “not very accurate” in one note disclosed by the prosecution in relation to this appeal.

The representatives of Dawson and Hamberger had sought disclosure about Miss Rees’ expertise and reliability before the trial started. In his Defence Case Statement dated 21 May 2002, Dawson sought “Details of any case where the Crown have previously instructed Miss Rees, had reports prepared but then withdrawn the evidence and the reasons why” and “Details of any findings of complaints against Miss Rees”. Further, in a letter dated 10 June 2002 the Crown were asked for information about criminal cases in which Miss Rees had provided evidence or in which it had been decided not to use her evidence after she had provided a statement: the express concern was whether her evidence had been doubted or considered unreliable by the Crown. When the case was listed for mention on 10 June 2002, following representations from the defence, the Crown were directed to obtain and disclose material relating to the tests carried out at the request of the Northumbria police, a reference, of course, to Professor Summerfield’s tests.

In Dawson and Hamberger the Crown disclosed the following in relation to Miss Rees: certificates of her proficiency; details of the test that she had taken in March 1999; details of the test conducted by Professor Summerfield; an undated letter from Miss Rees setting out her past experience and commenting upon the tests that she had taken; and a transcript of a ruling in the Croydon Crown Court in a case called Steele, in which the judge rejected an application under section 78 of the Police and Criminal Evidence Act 1984 to disallow Miss Rees’ evidence . The letter referred, perhaps somewhat obscurely, to four of the six cases: Kenyon and Bowen, Gray, Ray and Cornish, and Musgrove and Hibbert. (There was some uncertainty about whether there was disclosed any version of a further document in which Miss Rees wrote a “synopsis” of the cases in which she was involved. Mr Guthrie and Mr Griffiths do not believe that it was, and we proceed on that basis. Certainly no correspondence has been produced that indicates its disclosure.)

Luttrell’s representatives requested by letter dated 7 January 2002 “details of all the cases Ms Rees has been involved in as an expert witness within the last 5 years”. Before the trial there was an application made to have the charge against Luttrell dismissed, and in this context his representatives again made it clear that they wished to have disclosed all other cases where Miss Rees’ evidence had been discredited. In response to these requests, the Crown served a document entitled “Synopsis of Cases to date”. Miss Rees had prepared it herself, and it was a record that she apparently brought up to date from time to time. The version that was served was said to be “last updated 02.05.02”. It contained Miss Rees’ description of the cases in which she was involved, including Kenyon and Bowen, Gray, Edgar and Edgar, and Musgrove and Hibbert. The Crown also served copies of two Crown Court rulings about whether Miss Rees’ evidence should be admitted. One was Liburd in the Southwark Crown Court, in which her evidence was ruled admissible. The second was Topkaya in the Harrow Crown Court, the unusual feature of which was that Miss Rees lip-read from a video a conversation in Turkish, a language with which, we understand, she is unfamiliar. The court disallowed her evidence under s.78 of the Police and Criminal Evidence Act 1984. We are told by Mr Forlin that those representing Luttrell never accepted that this disclosure was adequate, and continued during the trial to press for more. He reinforces his criticism of the Crown by submitting that the prosecutor should have appreciated that there was more material to be disclosed and referring, for example, to an e-mail dated 30 April 2002 sent by the Criminal Justice Policy Division of the CPS in response to the prosecuting solicitor’s enquiries about disclosure in respect of Miss Rees. The writer referred to a “package of papers relating to cases”, and also said, “In most cases in which Miss Rees was instructed by the prosecution, she had been involved in “accuracy tests”.

The appellants in both cases submit that the disclosure was both incomplete and misleading in that the Crown failed to disclose material that should have been disclosed and that what was disclosed gave a distorted picture of how reliable the Crown had found Miss Rees’ lip-reading in the past. It is said to be incomplete because there should, at the least, have been further disclosure (i) of material relating to the six cases to which we have referred; and (ii) of material relating to evaluations of Miss Rees or her transcripts: in the Luttrell case there was a failure to disclose material relating to the tests carried out in March 1999 and those arranged by Professor Summerfield, and in both cases there was a failure in relation to the comparisons made in Kenyon and Bowen and Musgrove and Hibbert. Mr Forlin added that there should also have been disclosure in the Luttrell case of the “package” referred to in the e-mail of 30 April 2002. In Dawson and Hamberger the basis of the complaint that disclosure was misleading is Miss Rees’ undated letter. It is not necessary to set out all the criticisms of the letter made by Mr Guthrie, and the following is sufficient to give their flavour. In the letter, Miss Rees explained what had happened in Kenyon and Bowen by emphasising the difficulty of the task that she faced (for example, because of the quality of the video, and the facts that the speaker’s face was often obscured and he had a beard) and, while acknowledging errors, did not reveal the criticisms of her that were made. She referred to Gray and the guilty verdict, without revealing that her evidence was withdrawn from the Jury, or why the prosecution concluded that they could not rely upon it. She simply mentioned that the Judge had told the Jury of the results of the test in Kenyon and Bowen and that her evidence “could not be taken as gospel”. She referred to Mara, but explained it on the basis that the audiotape gave “a clearer and fuller picture” than what she could read from the videotape and did not refer to the doubt cast on her reliability. She did not mention other cases in which her evidence was not used. She referred to the tests in March 1999 and January 2000, but it is said that she was “explaining away” her shortcomings. In Luttrell, similar complaints are made of the synopsis. Again, it is not necessary to go into great detail. The synopsis does not, perhaps unsurprisingly, record or reflect the questions that had been raised about Miss Rees’ reliability. Thus, for example, of Kenyon and Bowen it simply states that the transcripts that she prepared “corroborated police suspicions and led to other sources of evidence”. Of Gray, the synopsis states that Miss Rees was cross-examined in court and that the defendant was found guilty, but not that the prosecution invited the jury not to rely upon her evidence.

Mr Russell Flint QC argued that the Crown’s disclosure was not defective. He did not submit that the further material about which the appellants complain was not “prosecution material” within the meaning of s.3(2) of the Criminal Procedures and Investigations Act 1996: as he put it, “the Crown accept for the purpose of these appeals that the Crown Prosecution Service is a single entity”. He submitted, however, that it is not material which might undermine the case for the prosecution. We are unable to accept that submission. Miss Rees’ experience and reliability were in issue in both trials. In Dawson and Hamberger, for example, her evidence when she was cross-examined before the jury was that she had experience of “actually giving evidence in front of judge and jury, I would say between twenty and thrity times”, and that she had “done other videos of which a soundtrack has subsequently turned up, and at times I did not know that a sound track was available, and the accuracy rate has been 98/99%”. In Luttrell she told the Jury that she would like to think that her synopsis “is fairly clearly set out and states exactly what happened in each case and what the outcome was and whether [her] evidence was used or not”.

In these circumstances, it does not seem to us that in either of these cases the prosecutors could properly have considered that there was no realistic possibility of doubt being cast upon Miss Rees’ credibility, expertise and reliability by reference to documents recording occasions when other prosecutors had criticised or questioned her reliability or by reference to occasions when her reliability had been tested by or at the request of prosecutors. This material should have been available for the defendants so that they could consider whether they wanted to deploy it during the trial, whether during the voir dire or in cross-examination of Miss Rees before the jury or both. This conclusion is reinforced by the complaints that such disclosure as was made was misleading. The Code of Practice issued under Part II of the Criminal Proceedings and Investigations Act 1996 and the Attorney General’s Guidelines on Disclosure recognise that prosecutors’ responsibilities depend upon what they can reasonably be expected to do in all the circumstances. But we cannot accept that the material about which the appellants complain should not have been available for disclosure, for example through careful enquiries of Miss Rees herself.

However, Mr Russell Flint had a further argument. He submitted that any failure to make disclosure did not materially affect the conduct of either trial and does not jeopardise the safety of these convictions. We agree with this submission in both cases. The material relating to the six cases undoubtedly shows concern among the prosecuting authorities about the reliability of Miss Rees’ lip-reading and concern about using her evidence at trial. Undoubtedly, material of this kind, and in particular, the letters of explanation and apology that Miss Rees wrote in response to the criticisms in Kenyon and Bowen, might have been presented with a sort of forensic flourish in the course of cross-examination. However, Mr Russell Flint rightly points out that, in relation to four of the six cases, namely Gray, Mara, Edgar and Edgar and Ray and Cornish, the material evinces no more than a view formed on the basis of an understanding, and, it seems likely, an imperfect understanding, of the comparison of the transcripts conducted by the police in the Kenyon and Bowen investigation and the implications thereof. As for the other two investigations, Kenyon and Bower itself and Musgrove and Hibbert, in which comparisons of transcripts were carried out to test Miss Rees’ interpretation, the views formed by police officers and others about Miss Rees’ expertise and abilities were no more valuable than the evidence of the tests or comparisons themselves. The argument that there was significant non-disclosure therefore depends upon the material about the tests and comparisons. Here it is important that the defence in both cases had available the results of the tests conducted by Professor Summerfield. In both cases Miss Rees was cross-examined about them and her explanation for the results. In Luttrell, Professor Summerfield gave evidence. These tests, unlike others, were conducted by a well-qualified examiner of high standing. They were not open to uncertainties introduced by inexpert comparison with an audiotape of questionable quality. This being so, we do not accept that it would have affected the outcome of the voire dire hearings if reference had been made to other tests of which the appellants’ representatives were unaware. As for the deployment of material about other tests or comparisons before the jury, each of these three appellants has complained that the time spent on lip-reading issues “hi-jacked” the hearing of their case to their disadvantage. We are unable to accept that it would have been productive for any of them to have aggravated this by cross-examination on the basis of tests inherently less reliable than those conducted by Professor Summerfield. There is no reason to suppose that disclosure of the package about which Mr Forlin complains would have been any more useful to Luttrell than the material to which we have already referred.

Accordingly, in Luttrell’s case, for the reasons given in paragraphs 47 and 63, there is no reason to regard his conviction as unsafe. His appeal is dismissed.

In relation to Dawson and Hamberger there is, for the reasons given in paragraph 63, no substance in their complaint about disclosure. But we cannot be sure, for the reasons given in paragraph 48, that their convictions are safe. Accordingly, their appeals are allowed and their convictions quashed. It is not suggested that a re-trial would be inappropriate. Unless we hear submissions to the contrary, we shall direct a re-trial at Lewes Crown Court, that a new indictment be preferred within 28 days and that, within 28 days thereafter, they be re-arraigned.

We turn to Beagley’s renewed application in relation to conviction. Mr Turton submitted first that the judge ought to have excluded evidence of statements made to undercover officers by Farry (who pleaded guilty and did not give evidence) which tended to implicate Beagley in the conspiracy. The judge agreed to excise those parts of Farry’s statements which were not corroborated, but this did not avail the defence who had to introduce them in order to demonstrate Farry’s unreliability and mendacity. In our judgment the judge was entitled to rule that the evidence was admissible in relation to the conspiracy and he correctly directed the jury (at pages 28 and 29 of the transcript volume 2) to exercise the greatest care in relation to Farry’s statements. The weight to be attached to those statements was a matter for the jury. Mr Turton secondly submitted that the judge ought to have stayed the proceedings against Beagley in the light of the prosecution’s failure to investigate the role played by a man called Nicholas Miller, lest there be confusion between his activity and that of Beagley whose first name was also Nicholas. In our judgment the judge was entitled to refuse a stay. The extent of Miller’s involvement, great or small, did not bear on the involvement of Beagley. The jury were properly directed in relation to the evidence against Beagley which, as is evident from paragraph 8 above, was considerable, quite apart from anything said by Farry. There is no reason to regard Beagley’s conviction as being, even arguably, unsafe. Accordingly, his renewed application in relation to conviction is refused.

Finally, we turn to the applications before us in relation to sentence.

In passing sentence, Judge Playford Q.C., who had presided over a three month trial, rightly stated that all the defendants before him stood convicted of a grave crime, not least because of their proximity to those who committed the primary offences. He was satisfied, as he was entitled to be, that there was close liaison with major criminals both before and after they committed their crimes. He was particularly impressed by what he described as the serious and audacious nature of some of the primary offences, which led to the loss of goods worth between £6 and £7 million, although the value of the property handled was of the order of £2 million. The overall sophistication of the operation was described by the judge as follows:

“As the evidence showed, you were tough negotiators, and not much was in general knocked off the true market value. There was, of course, the VAT element on top. Such was the high price that you obtained that innocent and respectable businessmen were able to purchase in good faith and then found themselves in possession of, or having sold on stolen goods.

In effect, there was a fully organised, sophisticated and professional business, run on commercial lines with a sham company, fraudulent invoices, a warehouse and place of business, and even an accountant. Dishonest activities were muddled in with honest dealings and the operation might never have been uncovered had it not been for resourceful and skilled surveillance by the police.”

On behalf of Sahota it was submitted by Mr Katz Q.C. that the sentence of 10 years was manifestly excessive. On the basis that the judge gave him full credit for his plea, it was argued that when his case is compared with R v Webbe [2002] 1 Cr App R (S) 22 and R v Reader [1988] 10 Cr App R (S) 210 the judge’s starting point must have been too high. Alternatively, since the judge did not refer to the issue of credit following Sahota’s plea, insufficient or no credit may have been accorded. In particular we were reminded that the applicant had a proper basis for delaying entering a plea, not least because of the outstanding public interest immunity hearing.

The judge described Sahota as “a shrewd, sophisticated and wary criminal, in a league of your own when compared with your co-conspirators.” He was dealt with as the prime mover and on the basis that the conspiracy was his brainchild. He did not have a substantial criminal record, his previous convictions including only some relatively minor offences of dishonesty. We note that Sahota is doing well in prison, and is attending various courses.

In refusing leave the single judge observed as follows:

“I have considered the papers in your case and your grounds of appeal and have decided that your application must be refused. I accept that on a plea of guilty 10 years was a severe sentence. But the judge (who had a detailed knowledge of the case) took the view that you were the prime mover in a conspiracy involving the handling of goods worth £7 million. He was satisfied that you or someone acting for you was the liaison with the primary criminals who carried out the armed and other robberies. As your counsel accepts, most of the aggravating factors mentioned in Webbe were present in your case. The judge was in my view justified in saying that you merited a sentence near the maximum. You were entitled to credit for your plea of guilty. But you instructed your legal advisors to take PII and other points including a Newton hearing whether “offence 9” was a genuine robbery (as it was found to be). In the circumstances I do not think you were entitled to the full discount. All in all I am not persuaded that your sentence was manifestly excessive.”

On behalf of Beagley, Mr Turton submitted that 7 years imprisonment was both manifestly excessive and wrong in principle. It was argued that the judge wrongly treated this applicant as Sahota’s lieutenant when the evidence suggested that he was only a “minor and occasional player”. In detailed submissions, advanced both orally and in writing, we have been addressed as to the precise nature of the evidence against Beagely, and in particular the three main transactions: handling a total of 27 laptop computers stolen from North America in December 2000; involvement in the Kingston Technology handling in May 2001; and the offer made to him of 500 stolen telephones in June 2001. Additionally, Mr Turton set out the favourable position on the evidence that was secured as regards the man “Nick”. However, the judge, who had presided over this long trial, expressly stated that, whilst he did not consider that Beagley was a controller or principal in the same league as Sahota, he rejected the suggestion that he was only a minor or occasional player. It is to be observed that property passed through Beagley’s hands in December 2000 and January and February 2001, and he was, as indicated, involved with the Kingston Technology memory chips and with the Nokia telephones in May and June. The judge concluded that Beagley was an intelligent and able young man with experience of crime on account of his poor antecedent history, who fell to be treated as Sahota’s lieutenant. The judge expressly gave credit for his medical condition and the position of Beagley’s children.

A similar submission was advanced by Mr Cockings on behalf of Jheeta, by comparing what is said to have been his role with that of Beagley. Mr Cockings submitted that Jheeta was “further down the chain of command”, playing a lesser role than Acar and Sandal, who, together with Dhaliwal, ridiculed him. His slight role was particularly demonstrated, by his less than constant presence at unit 919 and his lack of involvement with any undercover police officers. There was no enrichment on his part nor illegitimate funds passing through his account, and no stolen goods were found at his house.

The judge expressly rejected the suggestion that Jheeta was only a “small-time player”. As he observed in passing sentence, Jheeta was “pervasively present on the phone and at the warehouse in Yeovil Road”. He further highlighted the fact that he dealt with the mobile telephones that were the proceeds of the exceptionally serious armed robbery on 24 June 2000. He concluded that, on the evidence, Jheeta was a lieutenant who had been involved throughout and whose main mitigation was his good character.

On behalf of Keshwala, the principal argument relied on by Mr Davies was that his sentence was too long, particularly given his role. Mr Davies compared the sentences passed on Beagley, Acar and Sandal, who were said to be at the heart of the conspiracy, with that passed on this applicant who was of lesser importance, being only actively involved for some 5 weeks of a conspiracy that lasted 7 months. We were urged to pay particular regard to the fact that the evidence against this applicant was limited to a few particular events: 29 May 2001 (the collection of £42,000 for delivery of stolen computer chips, together with supplying a list of stolen property that was for sale); 31 May 2001 (a visit to the unit); 15 June 2001 (loading a pallet that was connected to a vehicle in which a large number of stolen mobile telephones were discovered); and 5 July 2001 (the sale of stolen telephones to undercover officers during as a carefully co-ordinated operation). Finally, it was argued that his age (25), good character and recent marriage, together with his mother’s dependence on him – she has a disability- meant that the sentence should have been shorter.

This applicant was to be regarded as one of “the family”, and although the judge accepted that his duties were of a menial nature, they were nonetheless essential. The judge considered that his role was not a minor one and even though Dhaliwal appeared to be in a position of authority over him, Keshwala’s role was still important. As the judge observed “your conversation with the undercover police on 5 July confirms that you knew perfectly what you were doing, and driving Sahota’s Porsche to 919 on 29 May, and there collecting £45,000 in cash, which the undercover police had just paid, and taking it to Sahota’s home, shows that you were a trusted and valued member of the gang.” Although he did not have the advantage of a guilty plea, he was a young man of good character and on that basis, together with the somewhat menial role and the effect of this case on his family, the judge reduced the sentence he otherwise would have passed.

Shergill advanced, in written grounds, a disparity argument: it was submitted that his sentence is too long in the light of the sentences passed on Luttrell and Keshwala. it should have been no longer than that passed on Luttrell and shorter than that passed on Keshwala. Shergill was no more than a trainee lieutenant whose involvement was limited to acting on behalf of Sahota at a meeting with undercover officers on 3 July 2001 and being present at another meeting with Sahota, other defendants and police officers two days later. Moreover, it was submitted, his personal mitigation, particularly his relatively good character, his efforts to retain his marriage and the effect of his sentence on his family, should have resulted in a shorter sentence.

The judge’s conclusion as regards this applicant was that he was one of the family, and as such was trusted to undertake delicate negotiations, and to be present on 5 July. The judge concluded that Sahota would not have permitted Shergill to undertaken negotiations in relation to the mobile telephones “that were still scorching hot after the robbery nine days earlier” if he had been anything other than an “entirely reliable member of (Sahota’s) gang”. However the judge accepted that, apart from the events in July in the latter part of the conspiracy, he featured on the evidence hardly at all. Accordingly he was sentenced as something below a “fully fledged lieutenant”.

Similarly, in written grounds, Dhaliwal urged disparity and that the judge was over influenced by his previous convictions, including offences of dishonesty and violence. It was argued that he had a lesser role than some of his co accused, such as Beagley, Acar and Sandal who received lesser sentences. However, the judge concluded that Dhaliwal was deeply involved in the conspiracy and, as such, a major player. Although some of the evidence, in the judge’s estimation, clearly placed Dhaliwal in the position of being a “key-player”, the judge considered that the fairest approach was to treat him as being on the same level as Acar and Sandal, but with an adjustment to reflect his previous record.

We have looked with care at the respective sentences passed by Judge Playford Q.C. Although, for our part, we accept that there can be valid reasons for delaying entering a guilty plea e.g. that there are unresolved legal issues that properly fall for determination by the trial judge, we entirely agree with the analysis of the single judge as to the sentence imposed on Sahota. It was certainly a long sentence, near the top of the range available in these circumstances, but, given the gravity of the offences and weight of the evidence against him, we are unable to describe it as either manifestly excessive or wrong in principle. This was very serious criminality that merited condign punishment, irrespective of his relatively minor previous convictions and his powerful mitigation.

As to the other applicants, we consider this case exemplifies the advantage that a judge has when passing sentence at the end of a long trial. He is in a unique position, having heard the evidence and seen those involved at close quarters over a long period, to determine where in the hierarchy any particular defendant is. Sentencing is an art and, especially when reviewing a sentencing exercise of this kind, this court will only interfere if the judge has clearly fallen into error. This was a complex case, with defendants involved in different ways and for varying lengths of time. The judge was faced with personal mitigation that varied from defendant to defendant, particularly as to pleas and antecedents. His sentencing remarks were careful and precise; he acted, as he was entitled to, on his own conclusions as to the evidence and its significance; and he clearly had well in mind the mitigation advanced on behalf of each applicant. In the event, we have been unable to detect any merit either in the disparity arguments, or in the submissions that the sentences were excessive bearing in mind individual roles and mitigation. We consider that the judge correctly identified the position in the hierarchy and culpability of each accused; he gave proper credit for the various guilty pleas; and he weighed appropriately the personal mitigation.

Accordingly, notwithstanding the persuasive and concise submissions advanced by counsel for each of these applicants, we refuse these renewed applications for leave to appeal the sentences imposed.