TUI UK Ltd v Griffiths

Mr and Mrs Griffiths and their youngest son went on a package holiday to a resort in Turkey. While staying at a hotel, which offered an inclusive package of meals and facilities, Mr Griffiths suffered a serious stomach upset which has left him with long term problems. He sued the travel company. At trial Mr and Mrs Griffiths gave uncontested evidence as to fact. Mr Griffiths also led the evidence of an expert who opined that, on the balance of probabilities, the food or drink served at the hotel was the cause of Mr Griffiths’ stomach upset. The travel company defendant did not require the expert to attend for cross-examination and did not lead any evidence of its own. In his closing submissions, the travel company’s counsel argued, and persuaded the judge, that deficiencies in the expert’s report meant that the claimant had failed to prove his case on the balance of probabilities.

The appeal raises a question of the fairness of the trial. The question is whether the trial judge was entitled to find that the claimant had not proved his case when the claimant’s expert had given uncontroverted evidence as to the cause of the illness, which was not illogical, incoherent or inconsistent, based on any misunderstanding of the facts, or based on unrealistic assumptions, but was criticised as being incomplete in its explanations and for its failure expressly to discount on the balance of probabilities other possible causes of Mr Griffiths’ illness.

Factual background

Mr Griffiths entered into a package holiday contract with TUI UK Ltd (“TUI”), which is a well-known tour operator, for himself, his wife and their youngest son. The holiday package included return flights from the United Kingdom to Turkey and 15-nights’ all-inclusive accommodation at the Aqua Fantasy Aquapark Hotel in Turkey between 2 and 16 August 2014. Mr Griffiths fell ill on the evening of 4 August 2014 suffering from stomach cramps and diarrhoea. He spent two days in his bedroom before his symptoms began to lessen but they did not settle completely. On 7 August 2014, on the advice of a tour representative, Mr Griffiths, his wife and his son took a hotel shuttle bus to the local town to obtain medication from a pharmacy. While in the town, the Griffiths family went to a local restaurant. Mr Griffiths ordered a meal but could not eat much as he did not have much of an appetite.

After 8 August Mr Griffiths felt that he was beginning to recover. But on 10 August 2014 Mr Griffiths began to feel unwell again. He suffered from diarrhoea and needed to visit the bathroom approximately every hour. He spoke to a doctor, who advised him that he needed hospital treatment. He was admitted to Kusadasi hospital on 13 August where he remained for three days and two nights. He was diagnosed as suffering from acute gastroenteritis and was given intravenous fluids and antibiotics. A stool sample was taken, which on analysis showed multiple pathogens, both parasitic and viral. He continued to feel unwell but was able to travel home with his wife and son on 16 August 2014.

Before Mr Griffiths went on holiday he had eaten food, including a Burger King meal at Birmingham airport. Between 2 and 4 August Mr Griffiths ate only at the hotel. The only food which he ate in Turkey outside the hotel was when he and his family ate at the local restaurant mentioned in paragraph 3 above.

At the time of his trial in June 2019 Mr Griffiths was still suffering from stomach churning and bubbling, cramping pains in his stomach, increased stomach bloating and an increased frequency in bowel movements with urgency and episodes of diarrhoea. Those symptoms are likely to be permanent and affect his ability to undertake social outings. He has concerns about long car journeys. As explained more fully below, the trial judge, HHJ Truman, dismissed his claim. The trial judge expressed the view that the appropriate level of compensation would have been £29,000 for pain, suffering and loss of amenity (including the spoiling of the holiday), plus damages for care and medication costs.

The legal proceedings

Pre-trial

In August 2017 Mr Griffiths commenced his action in the County Court. He pursued his claim on two bases: first, he claimed damages as a consumer against TUI under the Package Travel, Package Holiday and Package Tour Regulations 1992 (SI 1992/3288); secondly, he pursued a claim under sections 4 and 13 of the Supply of Goods and Services Act 1982.

TUI lodged a defence, in which it denied that the illness had been caused by the consumption of food or drink in the hotel and put Mr Griffiths to proof as to the cause of his illness. Thereafter, the claim was allocated to the multitrack under Part 29 of the Civil Procedure Rules on 12 January 2018. Mr Griffiths obtained medical reports from a gastroenterologist, Dr Linzi Thomas, and a microbiologist, Professor Hugh Pennington. Under the case management by the court which that procedure provides, both parties were given permission to rely on expert evidence from a gastroenterologist and a microbiologist.

TUI failed to serve a report from a gastroenterologist within the time specified by the court. It chose not to serve a report by a Dr Gant, a consultant microbiologist, which addressed causation. TUI confirmed that it did not intend to rely on expert evidence from a microbiologist. Mr Griffiths’ lawyers served Professor Pennington’s report. TUI’s lawyers had by then applied for permission to rely on a report by a gastroenterologist and for relief from sanctions. The court refused that application with the result that TUI went to trial without the support of any expert evidence. TUI lodged witness statements by witnesses as to fact who TUI had intended would give evidence by video link but, in the event, they were not called or cross-examined. Their evidence was accordingly discounted. Further, TUI did not seek to have Professor Pennington attend the trial for cross-examination with the result that his evidence was accepted on paper. His expert evidence was therefore uncontroverted in the sense that it was not in conflict with any other evidence led at the trial and was not subjected to challenge by cross-examination.

The trial

HHJ Truman heard the evidence and the submissions of the parties in a one-day trial on 20 June 2019. She accepted the evidence of Mr and Mrs Griffiths, who were cross-examined and whom she described as patently honest and straightforward witnesses. Mr Griffiths in his witness statement criticised the hygiene standards of the hotel, and in particular the buffet restaurants in the open air, and Mrs Griffiths confirmed the contents of his statement. The trial judge made no findings of fact on those matters, but recorded the allegations made in the Statement of Claim. The trial judge recorded the evidence of Dr Thomas including that relating to causation, which drew on the Griffiths’ witness statements about the hygiene standards in the hotel. Dr Thomas was asked questions under CPR Pt 35.6, which she answered, but she did not attend trial . In the event, Mr Griffiths’ counsel relied on Dr Thomas’s report in relation to diagnosis and prognosis but did not rely on her conclusions in relation to causation. As a result, the only expert evidence on causation before the trial judge at the conclusion of the trial was the uncontroverted expert report of Professor Pennington and his answers to questions posed by the defendant’s solicitors under CPR Pt 35.6.

Counsel for TUI made specific criticisms of Professor Pennington’s evidence which were set out in a skeleton argument served on the afternoon before the trial. Those criticisms formed the basis of counsel’s submissions at trial and of the trial judge’s decision. It is therefore necessary to set out in full the substantive parts of Professor Pennington’s report and the CPR Pt 35.6 questions and answers which followed it.

After Professor Pennington recorded his professional qualifications and the materials with which he had been provided for his report, he described his instructions as being to comment on the chronology of events, provide a detailed commentary on the issue of gastric illness and any breaches of health and safety procedures in place at the hotel, and express an opinion as to whether on the balance of probabilities Mr Griffiths’ illnesses were caused by staying at the hotel and a breakdown of health and hygiene practices there. After a brief summary of Mr Griffiths’ symptoms, Professor Pennington recorded the results of the tests on stool samples taken in the Turkish hospital. He stated (para 2):

“According to the discharge report of 16 August 2014 by Dr Yusuf Tuna, Entamoeba histolytica cysts and Giardia intestinalis were said to be seen on microscopy, and rotavirus, adenovirus, E. histolytica and Giardia antugen (sic) tests were positive. However, [according to] the Witness statement of Ibrahim Kocaoglu, the hotel doctor, the stool testsshowed Entamoeba histolytica and Giardia intestinalis cysts, but the Rota, Adeno and Noro virus tests were negative. His statement says that Peter Griffiths was seen on 13 August 2014 with a history of 6 days sickness, abdominal cramps, and diarrhoea, which complaints started after dinner in Kusadasi town center (sic) on 6 August 2014. Self medication partially relieved the symptoms, but diarrhoea started again on 11 August 2014.”

Professor Pennington then briefly stated his opinion as to the cause of Mr Griffiths’ symptoms. He stated:

“3.

I do not think that Peter Griffiths had amoebic dysentery caused by Entamoeba histtolytica (sic). Entamoeba cysts (which were found in his stools) are not diagnostic on their own because they cannot be distinguished routinely from the far commoner cysts of the harmless Entamoeba dispar. The onset of amoebic dysentery is usually gradual or intermittent; acute colitis is uncommon. Vomiting is not a feature and the diarrhoea is almost always bloody. Cases of amoebic dysentery most commonly have an incubation period of 2 to 4 weeks. None of these features lend support to a diagnosis of amoebic dysentery contracted in Turkey in Peter Griffith’s (sic) case. I consider it to be statistically improbable that he had been infected simultaneously with Giardia, adenovirus and rotavirus. I note that a microscopic diagnosis of Giardia is not straightforward. However, it is much more likely as a cause of gastroenteritis in this case than any of the other pathogens.

4.

The possibility cannot be ruled out that Peter Griffiths had two infections, one starting on 4 August, and a second starting on 11 August.

It is not possible to make an accurate aetiological diagnosis in cases of gastroenteritis from symptoms alone. On the balance of probabilities the absence of vomiting as a symptom make (sic) a virus cause much less likely than a bacterial one. The commonest recorded bacterial causes of acute gastroenteritis in places like Turkey are Campylobacter, Shigella and Salmonella. Giardia is considered to be reasonably common. Campylobacter is more commonly recorded in travellers returning to the UK from holidays abroad than Salmonella or Shigella. Enterotoxigenic E.coli (ETEC) and its relatives are considered to be common causes of diarrhoea in countries such as Turkey. For technical reasons they are not routinely tested for in the UK.

The incubation period for Giardia ranges from 1 to 14 days. It averages 7 days. Peter Griffith (sic) had been at the hotel for 2 days before he fell ill, and 9 days before his diarrhoea returned. Campylobacter has an average incubation period of 3 days. For ETEC it ranges from 12 to 72 hours. On the balance of probabilities Peter Griffiths acquired his gastric illnesses following the consumption of contaminated food or fluid from the hotel.”

On receiving Professor Pennington’s report, TUI’s lawyers asked the following CPR Pt 35.6 questions:

“1.

You refer to ‘contemporaneous evidence’ in Paragraph 1 of your report. Please can you set out exactly what ‘contemporaneous evidence’ you relied upon when writing your report?

2.

Please confirm whether you examined the Claimant or interviewed him prior to writing your report.

3.

Do you agree with the proposition that stool samples taken and analysed at the time of an illness complain (sic) are the most reliable form of ascertaining or determining the types of pathogen that may be causing that illness?

4.

You offer opinion that the Claimant suffered gastric illness caused by consumption of contaminated food or fluid from the hotel. In relation to your opinion on causation, to what extent do you consider that there would be:

a.

A range of opinion on causation amongst appropriate experts?

b.

If there is a range, what is it?

c.

What is your position within that range?

d.

What facts and matters have your (sic) relied upon in adopting your position within that range?

5.

To what extent were you able to identify the exact source of contaminated food or fluid that caused the illness? If so, please state what exactly was contaminated and provide supporting evidence of the contamination.

6.

If the Court finds as a fact that the Claimants ate outside of the hotel in the days/weeks leading to inset (sic) of illness, to what extent would that impact on the opinions you express in relation to causation?

7.

If the court finds as fact that others on this holiday who had consumed the food provided by the hotel, were not similarly afflicted, to what extent would that impact on the opinions you express in relation to causation?

8.

If the court finds as fact that the hotel was applying high standards in relation to hygiene and monitoring of food, to what extent would that impact on the opinions you express in relation to causation?

9.

Is Rotavirus a viral infection?”

In the tenth question TUI’s lawyers referred to four official publications in the United Kingdom on Giardia and Rotavirus and asked whether he considered the content of the publications to be reliable sources of information.

Professor Pennington responded as follows:

“1.

Flight and hotel bookings.

2.

I did not interview the claimant.

3.

I agree that stool sample testing done by an accredited laboratory is the most reliable way to ascertain the microbial cause of gastroenteritis.

4.

a-d Regarding causation etc, the appropriate experts would consider the gastroenteritis symptoms, their possible infective cause, the commonness of possible microbial causes in Turkey and their modes of transmission, their incubation periods, and the length of time the claimant had been at the hotel. I did the same. (Emphasis added)

5.

In single cases of infective gastroenteritis it is usually not possible (as in this case) to determine the exact source of contaminated food that led to the infection. To determine the exact source under these circumstances it would be necessary for suspect foods to be tested for the possible pathogens; this is usually impossible because the suspected foods will have been consumed. It is highly unlikely that any will have been retained in a condition suitable for microbiological testing.

6.

If the claimant had eaten outside the hotel the nature of the food and the date(s) of its consumption and the frequency of its consumption would be taken into account in assessing the probability that such food was more or less likely than hotel food to have been the source of the pathogen that caused the gastroenteritis. (Emphasis added)

7.

The great majority of cases of food borne infective gastroenteritis are sporadic and do not occur in outbreaks. So if no other cases similarly affected had been reported, this would not affect my conclusions regarding causation.

8.

I would expect the court to take into account the hotel HACCP plan and its implementation with all its associated documentation in determining its food hygiene standards; if high quality I would take it into account regarding causation. I would put much less weight on food monitoring itself as a food safety measure because of its inherent statistical limitations.

9.

Rotavirus is a virus.

10.

I consider these publications to be reliable sources of information.”

I have emphasised the answers to questions 4 and 6 as they are important to the understanding of Professor Pennington’s reasoning which I discuss in my analysis below.

The trial judge recorded in some detail the challenges which TUI’s counsel made to Professor Pennington’s report. Those criticisms were in summary: (i) there had been a failure to discount the occurrence of two separate infections and the meal in the local town as the cause of a second infection; (ii) Professor Pennington had set out the incubation periods but had given no explanation as to why he concluded that the illness was caused by food or fluid in the hotel; (iii) he had failed to mention the meal in Birmingham airport or the meal in the local town and to exclude them as causes; (iv) he had failed to comment on possible breaches of health and hygiene procedures in the hotel; and (v) he had failed to discount the methods of transmission of the illness which were not related to food which TUI’s counsel had listed.

The trial judge also criticised the report for failing to explain why the adenovirus and rotavirus found in the claimant had no effect or were otherwise discounted. She acknowledged Professor Pennington’s stated opinion that a viral cause was less likely than a bacterial one because of the lack of vomiting, but was unclear how that fitted the facts (i) that only parasites and viruses were isolated in the samples and not bacteria, and (ii) that the pathogens found were known to cause stomach upsets. She also observed that Professor Pennington had not expressly excluded the possible causes, other than food, which were listed in the Statement of Claim, such as the air conditioning system and leakage from a baby’s nappy in the swimming pool.

The trial judge was also critical of Professor Pennington’s responses to the CPR Pt 35.6 questions. In relation to the professor’s answer to question 4, she observed that he had not given any formal range of opinion or stated where within the range his opinion might fall. The judge also recorded Professor Pennington’s answer to questions 6, 7 and 10. The judge referred to and quoted the judgment of this court in Kennedy v Cordia (Services) LLP [2016] UKSC 6; [2016] 1 WLR 597 on the role of the expert witness and the judgment of the Court of Appeal in Wood v TUI Travel plc [2017] EWCA Civ 11; [2018] QB 927, which I discuss below. She criticised the report for not explaining why the pre-flight meal and the meal in the local town had been discounted as causes and why other possible causes and methods of transmission had been considered and excluded.

The trial judge held that Mr Griffiths had not proved his case and dismissed the claim. She summarised her reasons for so doing in para 29 of her judgment:

“Dr Thomas and Professor Pennington are undoubtedly experienced practitioners. They may both well consider, with their years of experience, that the Claimant had infective gastroenteritis caused by eating hotel food, but it seems to me that reports prepared after Wood v TUI need to deal with those matters the Court of Appeal specified. These reports do not do that. In some instances, they do not comply with CPR 35 (the failure to supply a range of opinion). They certainly do not provide me with sufficient information to be able to say that there is a clear train of logic between, for example, the incubation periods and the onset of illness, so that the pre-flight meal can be excluded or that the hotel food is a more likely cause; similarly for the ‘second’ illness – it is not said why it is more likely to be a relapse rather than a second infection, especially where the expert has said that it would be unlikely to have all the identified pathogens from one episode of eating contaminated food. It is thus not clear why the eating out in the local town can be discounted.”

The appeal to the High Court

Mr Griffiths appealed to the High Court with permission granted by Pepperall J. In a judgment dated 20 August 2020 ([2020] EWHC 2268 (QB)) Martin Spencer J allowed his appeal. Martin Spencer J stated that the appeal raised a fundamental question concerning the proper approach of the court towards uncontroverted expert evidence.

Martin Spencer J noted that HHJ Truman had accepted the evidence of Mr and Mrs Griffiths and that the only expert evidence on causation before the judge was that of Professor Pennington, whose report he described as “minimalist”. However, he distinguished the case from the dicta in Wood v TUI on the basis that Mr Griffiths was not relying on the mere fact of illness to establish his claim but also on the evidence of potential pathogens revealed by the stool samples.

On the judge’s analysis two questions had to be answered. The first was whether a court is obliged to accept an expert’s uncontroverted opinion even if that opinion could properly be characterised as an ipse dixit; and, if not, what were the circumstances in which the court would be justified in rejecting such evidence. His answer was that the court could reject an uncontroverted expert report if it were, literally, a bare ipse dixit, such as a one-sentence report stating the expert’s conclusion. Such a report was difficult to imagine in view of CPR Pt 35 and the well-publicised duties of experts. He continued (para 33):

“However, what the court is not entitled to do, where an expert report is uncontroverted, is subject the report to the same kind of analysis and critique as if it was evaluating a controverted or contested report, where it had to decide the weight of the report in order to decide whether it was to be preferred to other, controverting evidence such as an expert on the other side or competing factual evidence. Once a report is truly uncontroverted, that role of the court falls away. All the court needs to do is decide whether the report fulfils certain minimum standards which any expert report must satisfy if it is to be accepted at all.”

Martin Spencer J found those minimum standards in para 3 of the Practice Direction which accompanies CPR Pt 35.6 (“CPR PD 35”), which, so far as relevant, states:

“Form and Content of an Expert’s Report

…

3.2

An expert’s report must:

(1)

give details of the expert’s qualifications;

(2)

give details of any literature or other material which has been relied on in making the report;

(3)

contain a statement setting out the substance of all facts and instructions which are material to the opinions expressed in the report or upon which those opinions are based;

(4)

make clear which of the facts stated in the report are within the expert’s own knowledge;

(5)

say who carried out any examination, measurement, test or experiment which the expert has used for the report, give the qualifications of that person, and say whether or not the test or experiment has been carried out under the expert’s supervision;

(6)

where there is a range of opinion on the matters dealt with in the report –

(a)

summarise the range of opinions; and

(b)

give reasons for the expert’s own opinion;

(7)

contain a summary of the conclusions reached;

(8)

if the expert is not able to give an opinion without qualification, state the qualification; and

(9)

contain a statement that the expert –

(a)

understands their duty to the court, and has complied with that duty; and

(b)

is aware of the requirements of Part 35, this practice direction and the Guidance for the Instruction of Experts in Civil Claims 2014.”

As can be seen from the text set out above, para 3 of CPR PD 35 addresses the content of an expert’s report but, other than in para 3.2(6)(b), does not expressly require the expert to set out his reasoning for his conclusion. Martin Spencer J stated (para 36) that a failure to set out such reasoning might diminish the weight to be attached to the report but the weight to be attached to the report “is not a consideration: that only arises once the report is controverted.” If there had been controverting expert evidence, Professor Pennington would have expanded upon his reasoning, for example in a meeting of experts and a joint statement. But there was none; nor was the reasoning which lay behind his conclusions challenged by cross-examination. While recognising the trial judge’s criticisms of the report as strong, Martin Spencer J rejected those criticisms as irrelevant as they went to the weight of the report. The court, he said, must assume that there was some reasoning which lay behind Professor Pennington’s conclusion.

The second question was whether Professor Pennington’s report was a bare ipse dixit or otherwise so deficient as to have entitled the court to reject it. He answered this in the negative: despite the “serious deficiencies” identified by the trial judge, Professor Pennington had gone a long way towards substantiating his opinion. The report substantially complied with the Practice Direction and was not a bare ipse dixit. He therefore allowed Mr Griffiths’ appeal.

The appeal to the Court of Appeal

TUI appealed to the Court of Appeal. In a judgment dated 7 October 2021 ([2021] EWCA Civ 1442; [2022] 1 WLR 973) the majority of the Court of Appeal (Asplin LJ and Nugee LJ) allowed the appeal, Bean LJ dissenting. Asplin LJ delivered the leading judgment and Nugee LJ agreed, adding a brief commentary on points of principle.

Asplin LJ addressed the four grounds of appeal which Howard Stevens KC advanced. The first ground was that Martin Spencer J had erred in law in holding that where an expert’s report is uncontroverted (in the sense that there is no factual evidence undermining the factual basis of the report, no conflicting expert evidence and no cross-examination) the court is not entitled to evaluate the report but need only to ask itself whether the report meets the minimum standards prescribed by CPR PD 35. Asplin LJ upheld this ground of appeal. She noted that there were no authorities which supported the bright line rule which the High Court judge had adopted. She stated (para 40) that it all depends upon the circumstances of the case, the nature of the report itself and the purpose for which it was being used in the case. She observed that the judgment of this court in Kennedy v Cordia did not support such a rule and that the judge had been wrong to find an ambiguity in the discussion of expert reports in that judgment. She discussed in some detail the judgment of Clarke LJ in Coopers Payen Ltd v Southampton Container Terminal Ltd [2003] EWCA Civ 1223; [2004] 1 Lloyd’s Rep 331, on which the High Court judge had relied. The case concerned the evidence of a single joint expert which was contradicted by the evidence of an eyewitness. The expert witness had given oral evidence and had been cross-examined. Clarke LJ, who gave the leading judgment, accepted that the court did not have to accept the evidence of the joint expert but could consider it in the light of all the other evidence. Clarke LJ stated (para 42):

“All depends upon the circumstances of the particular case. For example, the joint expert may be the only witness on a particular topic, as for instance where the facts on which he expresses an opinion are agreed. In such circumstances it is difficult to envisage a case in which it would be appropriate to decide this case on the basis that the expert’s opinion was wrong.”

Asplin LJ did not dissent from Clarke LJ’s conclusions but stated that they provided no support for the contention that in all circumstances the court is bound to accept uncontroverted expert evidence which complies with CPR PD 35.

Asplin LJ then discussed several criminal cases to which the Court of Appeal had been referred but did not find them of great assistance. They were concerned with the role of a jury in a criminal trial and were not dealing with expert reports which were deficient in any way. They provided no support for the approach which Martin Spencer J had adopted.

In her discussion of this first ground of appeal Asplin LJ addressed the submission of Robert Weir QC that it was unfair to challenge an expert’s evidence only in closing submissions. She addressed several cases which Mr Weir put forward and which I discuss below: Browne v Dunn (1893) 6 R 67; Markem Corpn v Zipher Ltd [2005] EWCA Civ 267; [2005] RPC 31 (“Markem”); and the recent judgment of the Judicial Committee of the Privy Council in Chen v Ng [2017] UKPC 27. She distinguished the cases on the basis that they had been concerned with circumstances in which there was a challenge to the credibility of a witness in relation to a significant part of the witness’s evidence: fairness required that the witness be given an opportunity to give an explanation when his or her honesty was being impugned. In the present case Professor Pennington’s credibility was not in issue. Asplin LJ stated that there was nothing inherently unfair in challenging expert evidence in closing submissions. HHJ Truman did not hold that Professor Pennington’s report was necessarily wrong; she held that the report was insufficient to enable Mr Griffiths to satisfy the burden of proof as to causation. There was no basis for overturning her evaluative judgment on this matter.

Asplin LJ expressed her conclusion on the first ground of appeal at para 69: there is no strict rule that prevents the court from considering the content of an expert’s report which complies with CPR PD 35 where it has not been challenged by contradictory evidence and where there is no cross-examination.

In view of that conclusion, Asplin LJ recorded her views on the other three grounds of appeal very briefly. She accepted the challenge that the High Court judge had erred in law in concluding that the report complied with CPR PD 35 because it did not give the range of opinion in response to question 4. In relation to the third ground of appeal she held, citing Kennedy v Cordia and para 62 of the Civil Justice Council’s “Guidance for the Instruction of Experts in Civil Claims” (2014), that a report had to provide some reasoning in support of an expert’s conclusion. Finally, on the fourth ground of appeal, which overlapped with the first ground, she accepted the challenge to the contention that there was a rigid test that a judge must accept uncontroverted expert evidence if it meets the minimum standards established by CPR PD 35.

Nugee LJ agreed with Asplin LJ’s analysis and conclusions and added some comments on points of principle. In summary, he held that trial judges had to evaluate all the evidence, including uncontroverted expert evidence, and decide what weight to give to that evidence in reaching their conclusions on the factual issues. He stated (para 81): “Uncontroverted evidence may be compelling, but it may not be: it may be inherently weak or unhelpful or of little weight for other reasons.”

Bean LJ issued a strongly worded dissent. He described as trite law the statement in Phipson on Evidence, which I quote below, that, in general, a party is required to challenge on cross-examination the evidence of any witness of the opposing party if it wishes to submit to the court that that evidence should not be accepted. While recognising inadequacies in the reasoning in Professor Pennington’s report, he expressed the view that TUI should have challenged his conclusion as to causation by cross-examination. He stated (para 94) that other than in exceptional circumstances, “a judge is generally bound to accept the evidence of an expert if it is not controverted by other expert or factual evidence and the opposing party could have cross-examined the expert on the point but chose for tactical reasons not to do so.” He concluded forcefully (para 99):

“Asplin LJ, with whom Nugee LJ agrees, says at para 65 that ‘as long as the expert’s veracity is not challenged, a party may reserve its criticisms of a report until closing submissions if it chooses to do so’, and that she can see nothing which is inherently unfair in that procedure. With respect, I profoundly disagree. In my view Mr Griffiths did not have a fair trial of his claim. The courts should not allow litigation by ambush.”

The questions raised on this appeal

The principal questions raised on this appeal are: (i) what is the scope of the rule, based on fairness, that a party should challenge by cross-examination evidence that it wishes to impugn in its submissions at the end of the trial? (ii) in particular, does the rule extend to attacks in submissions on the reliability of a witness’s recollection and on the reasoning of an expert witness? and (iii) if the rule does so extend, was there unfairness in the way in which the trial judge conducted the trial in this case?

Mr Weir for Mr Griffiths submits that the rule, which is based on the fairness of the trial, applies generally in civil proceedings in relation to both witnesses of fact and expert witnesses. The court should apply the rule flexibly having regard to the criterion of the fairness of the trial; but it should be applied unless the witness’s evidence is incredible and can be dealt with by an application for reverse summary judgment without the need to cross-examine. Poverty of reasoning does not suffice to exclude the rule. Mr Stevens defends the approach of the majority of the Court of Appeal; it is the task of the court to assess the weight of all evidence led in a trial, including uncontroverted evidence; and the weight which the judge attaches to an expert’s testimony depends upon the strength of the expert’s reasoning. The appellant’s submissions involve an unwarranted extension of the law set out in Browne v Dunn and later cases, which limit the rule to challenges to the honesty of a witness and analogous attacks on the witness’s character, and to when the opposing party is putting forward a positive case which concerns the witness. The appellant’s submission, if accepted, would erode the judicial function. The ultimate question is the overall fairness of the trial.

Analysis: the law

In this judgment I address civil proceedings and leave to one side questions of criminal procedure. It is trite law that as a generality in civil proceedings, the claimant bears the burden of proof in establishing his or her case. It is trite law that the role of an expert is to assist the court in relation to matters of scientific, technical or other specialised knowledge which are outside the judge’s expertise by giving evidence of fact or opinion; but the expert must not usurp the functions of the judge as the ultimate decision-maker on matters that are central to the outcome of the case. Thus, as a general rule, the judge has the task of assessing the evidence of an expert for its adequacy and persuasiveness. But it is trite law that English law operates an adversarial system, and the parties frame the issues for the judge to decide in their pleadings and their conduct in the trial. It is also trite law that, in that context, it is an important part of a judge’s role to make sure that the proceedings are fair. At the heart of this appeal lies the question of the requirements of a fair trial.

Because an expert’s task is to assist the judge in matters outside the judge’s expertise, and it is the judge’s role to decide the case, the quality of an expert’s reasoning is of prime importance. This court gave guidance on the role of the expert in Kennedy v Cordia,in which, in the judgment of Lord Reed and Lord Hodge with whom the other Justices agreed, it was stated:

“48.

An expert must explain the basis of his or her evidence when it is not personal observation or sensation; mere assertion or ‘bare ipse dixit’ carries little weight, as the Lord President (Cooper) famously stated in Davie v Magistrates of Edinburgh 1953 SC 34, 40. If anything, the suggestion that an unsubstantiated ipse dixit carries little weight is understated; in our view such evidence is worthless. Wessels JA stated the matter well in the Supreme Court of South Africa (Appellate Division) in Coopers (South Africa) (Pty) Ltd v Deutsche Gesellschaft für Schädlingsbekämpfung mbH 1976 (3) SA 352, 371:

‘an expert’s opinion represents his reasoned conclusion based on certain facts or data, which are either common cause, or established by his own evidence or that of some other competent witness. Except possibly where it is not controverted, an expert’s bald statement of his opinion is not of any real assistance. Proper evaluation of the opinion can only be undertaken if the process of reasoning which led to the conclusion, including the premises from which the reasoning proceeds, are disclosed by the expert.’

As Lord Prosser pithily stated in Dingley v Chief Constable, Strathclyde Police 1998 SC 548, 604: ‘As with judicial or other opinions, what carries weight is the reasoning, not the conclusion.’”

The courts below observed that Wessels JA left open the possibility that a bald statement of opinion might assist the court if it is uncontroverted. I doubt that it would, unless the opposing party could in the circumstances be taken to have admitted the accuracy of the conclusion, thereby relieving the judge of the task of evaluating the opinion or statement. I agree with the judges of the Court of Appeal in this case that in Kennedy v Cordia, this court was not endorsing Wessels JA’s reference to that possibility. I agree with the statement of Jacob J in Routestone Ltd v Minories Finance Ltd [1997] BCC 180, 188 (cited by Lewison LJ in Kingley Developments Ltd v Brudenell [2016] EWCA Civ 980 (“Kingley Developments”) at para 30 and in this case by Nugee LJ at para 83):

“What really matters in most cases is the reasons given for the opinion. As a practical matter a well-constructed expert’s report containing opinion evidence sets out the opinion and the reasons for it. If the reasons stand up the opinion does, if not, not.”

Martin Spencer J opined that the minimum standards for an expert report were to be found in CPR PD 35. He suggested that that practice direction and the law did not require an expert to set out his or her reasoning. I respectfully disagree. Para 3.2 (9) of CPR PD 35 requires the expert to state his or her awareness of “the requirements of Part 35, this practice direction and the Guidance for the Instruction of Experts in Civil Claims 2014.” That Guidance makes clear that an expert report should set out the expert’s reasoning. Para 62 of the Guidance, which makes mandatory a summary of conclusions, states: “Generally the summary should be at the end of the report after the reasoning.” (Emphasis added) This cannot surprise given the respective roles of the expert and the judge set out in the case law.

Mr Stevens sought to support the approach of the majority of the Court of Appeal in this case by referring to the Scottish case of Davie v Magistrates of Edinburgh 1953 SC 34. That case, like the present case, raised a question of causation, which was whether damage to houses had been caused by blasting operations in the construction of a sewer. The court had to evaluate the expert opinion evidence on explosives led on behalf of the defender, that the operations could not have damaged the pursuer’s property, against the evidence of factual and expert witnesses led on behalf of the pursuer, including an architect, an engineer and a builder, that the pursuer’s property and adjacent properties had been damaged by the vibration caused by the operations. The defender’s experts were cross-examined but the pursuer did not adduce any expert evidence on explosives to contradict them. The Inner House upheld the judgment of the Lord Ordinary (Strachan) which favoured the evidence led on behalf of the pursuer. The Lord President (Cooper)’s reasoning on the separate roles of the expert and the judge is unquestionably correct and contributed to the approach of this court in Kennedy v Cordia, including in the passage which I have quoted above. His critique of the bare ipse dixit or oracular pronouncement of the expert witness was that it could not be tested by cross-examination or independently appraised (p 40). Davie v Magistrates of Edinburgh and Kingley Developments and Armstrong v First York Ltd [2005] EWCA Civ 277; [2005] 1 WLR 2751, which were referred to in the courts below, were cases where the court evaluated an expert’s evidence and preferred contradictory evidence of witnesses who did not have that expertise. The case was not concerned with a circumstance where, as in this case, the expert was neither cross-examined nor contradicted by evidence led by the opposing party. For the correct approach to the scope for judicial evaluation of an expert’s report in such a circumstance one must look elsewhere in the case law and observe how the law has been developed over time. I do so now.

The starting point is in the adversarial system of litigation and arbitration in English law. In Chilton v Saga Holidays plc [1986] 1 All ER 841 the Court of Appeal ordered a rehearing of a case after a registrar, to whom the case had been referred for arbitration, refused to allow the lawyer of the defendant to cross-examine the claimant who did not have legal representation. Sir John Donaldson MR (p 844) stated:

“… both courts and arbitrators in this country operate on an adversarial system of achieving justice. It is a system which can be modified by rules of court; it is a system which can be modified by contract between the parties; but, in the absence of one or the other, it is basically an adversarial system, and it is fundamental to that that each party shall be entitled to tender their own evidence and that the other party shall be entitled to ask questions designed to probe the accuracy or otherwise, or the completeness or otherwise, of the evidence which has been given.”

In an adversarial system, subject to the constraints of case management, the parties frame the issues which the court is to determine; it is not normally part of the court’s business to investigate admitted facts: Akhtar v Boland [2014] EWCA Civ 872; [2015] 1 All ER 664, para 16 per Sir Stanley Burnton. The trial judge’s role is normally limited to determining the disputed issues which the parties present and to determining those issues based on the evidence which the parties adduce. The trial judge does justice between the parties in so doing: see Air Canada v Secretary of State for Trade [1983] 2 AC 394, 438 per Lord Wilberforce; Al Medenni v Mars UK [2005] EWCA Civ 1041, para 21 per Dyson LJ.

It is the task of a judge in conducting a trial in an adversarial system to make sure that the trial is fair. It is the task of the judiciary in developing the common law, and the makers of the procedural rules, to formulate rules and procedures to that end. One such long-established rule is usefully set out in the current edition of Phipson on Evidence 20^th ed (2022). Bean LJ quoted the previous edition, which was in materially the same terms, at the start of his dissenting judgment. At para 12-12 of the 20^th edition the learned editor states:

“In general a party is required to challenge in cross-examination the evidence of any witness of the opposing party if he wishes to submit to the court that the evidence should not be accepted on that point. The rule applies in civil cases … In general the CPR does not alter that position.

This rule serves the important function of giving the witness the opportunity of explaining any contradiction or alleged problem with his evidence. If a party has decided not to cross-examine on a particular important point, he will be in difficulty in submitting that the evidence should be rejected.”

This statement is supported by case law, some of which I discuss below, and has often been cited with approval by the Court of Appeal. See, for example, recently, In reB (A Child) [2018] EWCA Civ 2127; [2019] 1 FCR 120, para 18 per Peter Jackson LJ; and Edwards Lifesciences LLC v Boston Scientific Scimed Inc. [2018] EWCA Civ 673; [2018] FSR 29 (“Edwards Lifesciences”), para 62 per Floyd LJ. An earlier version of the text from the 12^th edition of Phipson (1976) was cited in Markem, para 59 (p 786) in which the court quoted with approval from the judgment of Hunt J in the Australian case of Allied Pastoral Holdings Pty Ltd v Federal Commissioner of Taxation (1983) 44 ALR 607 (“Allied Pastoral”), in which Phipson was cited.

I am satisfied that the statement in Phipson is correct and, as explained below, it summarises a longstanding rule of general application. It is not simply a matter of extensive legal precedents in the case law. It is a matter of the fairness of the legal proceedings as a whole. While many of the cases may have been concerned with challenges to the honesty of a witness, I see no rational basis for confining the rule to such cases or those analogous categories, such as allegations of bad faith or aspersions against a witness’s character, as Mr Stevens suggests.

Although the rule of professional practice is often referred to as the rule in Browne v Dunn, taking its name from the case in the House of Lords in 1893, the rule is of considerably greater antiquity. The Queen’s Case (1820) 2 Brod & Bing 284; 129 ER 976 is an example of a closely analogous rule. In that case, which concerned the trial in the House of Lords of Queen Caroline for adultery when King George IV sought to annul his marriage to his estranged wife, several legal questions arose which resulted in the House of Lords posing questions to be answered by the Lord Chief Justice and the consulted judges. One of the questions was whether, where a prosecution witness had been examined in chief, and had not been questioned on cross-examination as to allegations that the witness had attempted corruptly to procure others to give evidence for the prosecution, it would be competent for the accused party to lead the evidence of defence witnesses to prove such attempts without first recalling the witness to be examined on those allegations. Abbott CJ (the Lord Chief Justice of the King’s Bench) gave the unanimous opinion of the consulted judges that the proposed evidence could not be adduced without the prior cross-examination of the witness about the matter. He stated (pp 313-314; ER p 988):

“The legitimate object of the proposed proof is to discredit the witness. Now the usual practice of the courts below, and a practice, to which we are not aware of any exception, is this; if it be intended to bring the credit of a witness into question by proof of any thing that he may have said or declared, touching the cause, the witness is first asked, upon cross-examination, whether or no he has said or declared, that which is intended to be proved. …[I]f evidence of this sort could be adduced on the sudden and by surprise, without any previous intimation to the witness or to the party producing him, great injustice might be done; and, in our opinion, not unfrequently would be done both to the witness and to the party; …and one of the great objects of the course of proceeding established in our Courts is the prevention of surprise, as far as practicable, upon any person who may appear therein.” (Emphasis added)

The Lord Chief Justice and the consulted judge gave a similar opinion on the need first to put a point in cross-examination in response to a question whether, when a witness stated that he did not remember the cause of a quarrel, the defendant’s counsel could lead evidence that the witness had on an earlier occasion stated the cause of the quarrel: pp 300-301, ER 982-983.

Moving forward 73 years, Browne v Dunn involved an action for libel against a solicitor and the assertion of legal professional privilege in relation to a document which the solicitor prepared for signature by his proposed clients, containing complaints about the claimant’s behaviour and instructing the solicitor to act for them in relation to those complaints. What is relevant to this appeal is that counsel asked the jury to disbelieve the evidence of the clients that they had instructed the solicitor to act on their behalf against the claimant without having challenged the veracity of that evidence on cross-examination. Lord Herschell LC (at pp 70-71) stated his understanding of the rule:

“I cannot help saying that it seems to me to be absolutely essential to the proper conduct of a cause, where it is intended to suggest that a witness is not speaking the truth on a particular point, to direct his attention to the fact by some questions put in cross-examination showing that that imputation is intended to be made, and not to take his evidence and pass it by as a matter altogether unchallenged, and then, when it is impossible for him to explain, as perhaps he might have been able to do if such questions had been put to him, the circumstances which it is suggested indicate that the story he tells ought not to be believed, to argue that he is a witness unworthy of credit. My Lords, I have always understood that if you intend to impeach a witness you are bound whilst he is in the box, to give him an opportunity of making any explanation which is open to him; and, as it seems to me, that is not only a rule of professional practice in the conduct of a case, but is essential to fair play and fair dealing with witnesses.”

Lord Herschell went on to say that there was no need to waste time by cross-examining a witness where it is perfectly clear that he had prior notice that the opposing party intended to impeach the credibility of the story which he was telling. He continued: “All I am saying is that it will not do to impeach the credibility of a witness upon a matter on which he has not had any opportunity of giving an explanation by reason of there having been no suggestion whatever in the course of the case that his story is not accepted.”

Lord Halsbury agreed with the Lord Chancellor’s statements as to how a trial should be conducted, and said (pp 76-77):

“To my mind nothing would be more absolutely unjust than not to cross-examine witnesses upon evidence which they have given, so as to give them notice, and to give them an opportunity of explanation, and an opportunity very often to defend their own character, and, not having given them such an opportunity, to ask the jury afterwards to disbelieve what they have said, although not one question has been directed either to their credit or to the accuracy of the facts they have deposed to.” (Emphasis added)

Lord Morris (pp 78-79) concurred with those two speeches but stated that he wished to guard himself against laying down any hard and fast rule as regards cross-examining a witness as a necessary preliminary to impeaching his credit. Lord Bowen agreed in the dismissal of the appeal but made no statement of general principle on the need to cross-examine a witness.

It is clear, as Mr Stevens submits, that the case was concerned with a challenge to the credibility of witnesses. But the passages in the speech of Lord Halsbury, which I have emphasised, envisage a rule of wider scope, giving a witness the opportunity to explain his or her evidence if it is to be impugned on other grounds.

Moving on many years, in Deepak Fertilizers & Petrochemical Ltd v Davy McKee (UK) London Ltd [2002] EWCA Civ 1396, a case concerning a claim for damages for breach of contract following an explosion in a chemical plant in Mumbai, the Court of Appeal, in the judgment of Latham LJ, addressed a question whether the judge had been entitled to reject the evidence of a witness as to the measures which the appellants would have taken if they had been properly advised by the respondents. The witness, Mr Kotwal, was not cross-examined on the matter nor was he asked any questions by the judge concerning the reliability of his evidence. Nonetheless, the judge rejected Mr Kotwal’s evidence on causation. The Court of Appeal held that he was wrong to do so. Latham LJ stated the general rule in these terms (para 49):

“The general rule in adversarial proceedings, as between the parties, is that one party should not be entitled to impugn the evidence of another party’s witness if he has not asked appropriate questions enabling the witness to deal with the criticisms that are being made.”

He cited in support of this proposition a passage from the 15^th edition of Phipson (2000)which was worded differently from the text which I have quoted in para 42 above from the 20^th edition but nonetheless articulated the rule which Latham LJ stated. He continued (para 50):

“So long as a matter remains clearly in issue, it is the judge’s task to determine the facts on which the issue is to be decided. However it seems to me that where, as in the present case, an issue has been identified, but then counsel asks no questions, the judge should be slow to conclude that it remains an issue which has to be determined on the basis of an assessment of reliability or credibility without enquiry of the parties as to their position. The judge should be particularly cautious of doing so if he or she has not given any indication of concern about the evidence so as to alert the witness or counsel acting on the side calling the witness, to the fact that it may be that further explanation should be given in relation to the issue in question.”

Latham LJ recognised that each case depends upon the way in which the issue arose and was dealt with in the evidence. He concluded that unfairness had arisen in that case. Hart J and Brooke LJ agreed with his judgment, subject to qualifications which are not relevant to the issue of unfairness.

In Markem, which was decided in 2005, Jacob LJ, with whom Kennedy and Mummery LJJ agreed, addressed several appeals concerning entitlement to patents in a dispute between a company and its former employees. One of the issues which arose was whether the trial judge had been entitled to disbelieve the evidence of a witness, Mr Buckby, concerning his lack of knowledge about a memorandum when there had been no suggestion on cross-examination that his evidence in chief was false. The judge’s adverse findings about the evidence of other witnesses called by the defendant were challenged on the same basis. In para 56 Jacob LJ described the challenge, which was upheld, in these terms:

“procedural fairness not only to the parties but to the witnesses requires that if their evidence were to be disbelieved they must be given a fair opportunity to deal with the allegation.”

Before the appellate hearing, the Court of Appeal had alerted the parties to the decision in Browne v Dunn and the Australian case of Allied Pastoral. Jacob LJ observed that Browne v Dunn, which had been reported only in “a very obscure set of reports”, was not well known to practitioners in the United Kingdom although practitioners in Australia and Canada were very alive to the rule. Jacob LJ quoted and applied the conclusion of Hunt J in Allied Pastoral who, having cited Browne and Dunn and Phipson about the correct procedure where counsel proposes to impeach a witness’s credit, stated (p 634):

“I remain of the opinion that, unless notice has already clearly been given of the cross-examiner’s intention to rely upon such matters, it is necessary to put to an opponent’s witness in cross-examination the nature of the case upon which it is proposed to rely in contradiction of his evidence, particularly where that case relies upon inferences to be drawn from other evidence in the proceedings.”

While Markem was concerned with an attack on the credibility of witnesses, the passage in Allied Pastoral which the Court of Appeal approved was, like the speech of Lord Halsbury in Browne v Dunn, framed in broad terms and not confined to attacks on a witness’s credibility.

I also observe that any lack of awareness of the case of Browne v Dunn in the United Kingdom would have been balanced by an awareness of the rule from the leading textbooks, including Phipson and Crosson Evidence. Each of the 11^th to 14^th editions of Phipson in 1970, 1976, 1982 and 1990 referred to the requirement on cross-examiners to put their own case so far as it concerned the particular witness, and to put any suggestion that a witness was not speaking the truth so as to give an opportunity for explanation, and cited Browne v Dunn as authority for those propositions. Similarly Cross on Evidence (after the 7^th edition, Cross & Tapper) in its 6 editions between 1970 and 1999 (the 4^th to 9^th editions) consistently stated that any matter on which it was proposed to contradict a witness must normally be put to that witness so that the witness may have an opportunity of explaining the contradiction and that a failure to do so may be held to be an implied acceptance of the evidence. Browne v Dunn was cited as authority. It may be that the general rule was enforced with greater rigour in Australia, but the rule itself would, I suggest, have been well known in England when Markem was decided.

In Tullow Uganda Ltd v Heritage Oil and Gas Ltd [2013] EWHC 1656 (Comm); [2014] 1 All ER (Comm) 22, Burton J addressed a circumstance where defence counsel had not challenged the veracity of witnesses, Mr Inch and Mr Martin, in cross-examination lasting almost three days and three and a half days respectively, or in relation to Mr Martin in written closing submissions, but had mounted such an attack in his oral closing submissions. In further submissions the court was referred to Browne v Dunn and Markem, and Burton J stated (para 62):

“[I]f there is to be such an onslaught on the honesty and credibility of these two professional witnesses as has been carried out in the closing submissions, challenge to the accuracy of their evidence is plainly insufficient, and it must be necessary and in any event sensible and fair to put to a witness that in certain (in this case apparently numerous) respects he has been dishonest and is not telling the truth. This is not simply out of fairness to the witness, but it is also necessary for the judge, because if I am to conclude that an otherwise apparently honest and respected professional has been deliberately false and misleading, I must have the opportunity to see how the witnesses respond to each such suggestion and see whether I am persuaded by their answer (if any).” (Emphasis added)

The focus of this case was again on the credibility of the witnesses, but Burton J usefully pointed out that what was at stake was not just fairness to the witness (and he might have added fairness to the party who had called the witness) but the integrity of the court process itself in enabling the judge to reach a sound conclusion.

In Chen v Ng, which involved a dispute about the ownership of shares, the Judicial Committee of the Privy Council (“the Board”) considered how to apply the general rule in Browne v Dunn and Markem in a context in which a witness, Mr Ng, had been challenged in cross-examination that he was not telling the truth about the basis on which shares had been transferred but not on the two grounds on which the judge ultimately disbelieved his evidence. The Board, in a judgment delivered by Lord Neuberger and Lord Mance, upheld the decision of the Court of Appeal of the Eastern Caribbean Supreme Court that the judge had acted unfairly in relying on those grounds. The Board reached that view because the ultimate factual dispute was the basis upon which, and the circumstances in which, the transfer of the shares had taken place, and the issue on which Mr Ng had been disbelieved was central to the proceedings. Both grounds could reasonably be expected to have been put in cross-examination and it was possible that Mr Ng would have given believable evidence which weakened or undermined those grounds.

The Board expressed the general rule in these terms (para 53):

“In other words, where it is not made clear during (or before) a trial that the evidence, or a significant aspect of the evidence, of a witness (especially if he is a party in the proceedings) is challenged as inaccurate, it is not appropriate, at least in the absence of further relevant facts, for the evidence then to be challenged in closing speeches or in the subsequent judgment.”

It advised that it was appropriate to take a nuanced approach to the general rule. The Board stated (para 52):

“In a perfect world, any ground for doubting the evidence of a witness ought to be put to him, and a judge should only rely on a ground for disbelieving a witness which that witness has had an opportunity of explaining. However, the world is not perfect, and while both points remain ideals which should always be in the minds of cross-examiners and trial judges, they cannot be absolute requirements in every case. Even in a very full trial, it may often be disproportionate and unrealistic to expect a cross-examiner to put every possible reason for disbelieving a witness to that witness, especially in a complex case, and it may be particularly difficult to do so in a case such as this, where the Judge sensibly rationed the time for cross-examination and the witness concerned needed an interpreter. Once it is accepted that not every point may be put, it is inevitable that there will be cases where a point which strikes the judge as a significant reason for disbelieving some evidence when he comes to give judgment, has not been put to the witness who gave it.”

The Board concluded that the question for an appellate court was the overall fairness of the trial; the Board stated (para 54):

“Ultimately, it must turn on the question whether the trial, viewed overall, was fair bearing in mind that the relevant issue was decided on the basis that a witness was disbelieved on grounds which were not put to him.”

In Edwards Lifesciences the Court of Appeal (Kitchin, McCombe and Floyd LJJ) addressed the application of the rule in Browne v Dunn to the unchallenged evidence of an expert witness in a dispute about the validity and infringement of patents. The defendant, Boston, argued that the court was bound to accept the evidence of Professor Lutter in relation to certain matters as a consequence of the failure by counsel for Edwards to cross-examine him on those matters, or at least that the appellate court should look at the matter again and if persuaded that cross-examination could have made a difference to the outcome, set aside the judge’s conclusion. Edwards’ response was that the rules about what must be put on cross-examination should not be rigidly applied in relation to expert evidence. Edwards submitted that the judge was able to evaluate the reasons set out in the expert’s report. The points had been addressed by another of Boston’s expert witnesses on cross-examination and in a rejoinder report.

In his discussion of the point Floyd LJ quoted the obligation to cross-examine set out in the 19^th edition of Phipson (2018)and referred to Browne v Dunn and Markem. Floyd LJ recognised that the rule is an important one, but, like the Board in Chen v Ng (which appears not to have been cited to the Court of Appeal), he did not consider it to be an inflexible one. In his discussion in paras 63-69 he made six points. First, where, to save time, it is proposed not to cross-examine two witnesses on the same or similar subject matter it was good practice to raise the matter with the judge and obtain his or her directions to ensure fairness. (That suggestion is not relevant to this appeal). Secondly, the purpose of the rule is not only for the benefit of the witness but is to ensure the overall fairness of the proceedings for the parties. Thirdly, the rule applies with particular force where a witness gives evidence of fact of which the witness has knowledge, and it is proposed to invite the court to disbelieve that evidence. Fairness to the witness and to the parties demands that the witness be given the opportunity to respond to the challenge. Fourthly, it was not appropriate to apply the rule rigidly in every situation. Where, as in the case in question, there had been an opportunity to respond to the other side’s case through several rounds of expert evidence which made the position taken by each side’s experts clear, the potential for unfairness to the witness was much reduced. Fifthly, not every part of the evidence of a witness to fact needs to be challenged head-on that it is untrue or simply misguided; the test was fairness; see Various Claimants v Giambrone & Law [2015] EWHC 1946, para 21 per Foskett J. Sixthly, the question for the appellate court is “whether the decision not to cross-examine has led to unfairness to the extent that the judge’s decision on the relevant issue is thereby undermined” (para 69). In that case, there had been no unfairness to the expert witness or the party adducing his evidence as the witness had had the opportunity to respond to the case made against his position.

From this review of the case law it is clear that there is a long-established rule as stated in Phipson at para 12.12 with which practising barristers would be familiar, as Bean LJ suggested in para 87 of his judgment. There are also circumstances in which the rule may not apply. Several come to mind. First, the matter to which the challenge is directed is collateral or insignificant and fairness to the witness does not require there to be an opportunity to answer or explain. A challenge to a collateral issue will not result in unfairness to a party or interfere with the judge’s role in the just resolution of a case; and a witness in such a circumstance needs no opportunity to respond if the challenge is not an attack on the witness’s character or competence.

Secondly, the evidence of fact may be manifestly incredible, and an opportunity to explain on cross-examination would make no difference. For example, there may be no need for a trial and cross-examination of a witness in a bankruptcy application where the contemporaneous documents properly understood render the evidence asserted in the affidavits simply incredible: Long v Farrer & Co [2004] EWHC 1774 (Ch); [2004] BPIR 1218, para 60, in which Rimer J quotes from the judgment of Chadwick J in In re Company (No 006685 of 1996) [1997] 1 BCLC 639, 648.

Thirdly, there may be a bold assertion of opinion in an expert’s report without any reasoning to support it, what the Lord President (Cooper) in Davie v Magistrates of Edinburgh described as a bare ipse dixit. But reasoning which appears inadequate and is open to criticism for that reason is not the same as a bare ipse dixit.

Fourthly, there may be an obvious mistake on the face of an expert report. Bean LJ referred to this possibility in para 94 of his judgment and cited Woolley v Essex County Council [2006] EWCA Civ 753 as a useful example. In Hull v Thompson [2001] NSWCA 359, (“Hull v Thompson”) Rolfe AJA at para 21 expressed the view that such a circumstance would be where the report was ex facie illogical or inherently inconsistent. See also A/S Tallinna Laevauhisus v Estonian State Steamship Line (1946) 80 Ll L Rep 99, 108 (“Tallinna”) where Scott LJ spoke of the court rejecting an expert’s evidence if “he says something patently absurd, or something inconsistent with the rest of his evidence.”

I would add that what is said about the evaluation of expert evidence of foreign law in Tallinna and the other cases cited by the parties in argument in this appeal may now need to be read in the light of the recent guidance of this court in Brownlie v FS Cairo (Nile Plaza) LLC [2021] UKSC 45, [2022] AC 995 and of the Board in Perry v Lopag Trust Reg [2023] UKPC 16; [2023] 1 WLR 3494.

Fifthly, the witnesses’ evidence of the facts may be contrary to the basis on which the expert expressed his or her view in the expert report. Rolfe AJA in Hull v Thompson, para 21, spoke of the report being “based on an incorrect or incomplete history, or where the assumptions on which it is founded are not established.”

Sixthly, as occurred in Edwards Lifesciences, an expert has been given a sufficient opportunity to respond to criticism of, or otherwise clarify his or her report. For example, if an expert faces focused questions in the written CPR Pt 35.6 questions of the opposing party and fails to answer them satisfactorily, a court may conclude that the expert has been given a sufficient opportunity to explain the report which negates the need for further challenge on cross-examination.

Seventhly, a failure to comply with the requirements of CPR PD 35 may be a further exception, but a party seeking to rely on such a failure would be wise to seek the directions of the trial judge before doing so, as much will depend upon the seriousness of the failure.

Because the rule is a flexible one, there will also be circumstances where in the course of a cross-examination counsel omits to put a relevant matter to a witness and that does not prevent him or her from leading evidence on that matter from a witness thereafter. In some cases, the only fair response by the court faced with such a circumstance would be to allow the recall of the witness to address the matter. In other cases, it may be sufficient for the judge when considering what weight to attach to the evidence of the latter witness to bear in mind that the former witness had not been given the opportunity to comment on that evidence. The failure to cross-examine on a matter in such circumstances does not put the trial judge “into a straitjacket, dictating what evidence must be accepted and what must be rejected”: MBR Acres Ltd v McGivern [2022] EWHC 2072 (QB), para 90 per Nicklin J. This is not because the rule does not apply to a trial judge when making findings of fact, but because, as a rule of fairness, it is not an inflexible one and a more nuanced judgment is called for. In any event, those circumstances, involving the substantive cross-examination of the witness, are far removed from the circumstances of a case such as this in which the opposing party did not require the witness to attend for cross-examination.

In conclusion, the status and application of the rule in Browne v Dunn and the other cases which I have discussed can be summarised in the following propositions:

Analysis: application of the law to the facts

In assessing the fairness of the trial in this case it is important to have regard to the approach which TUI’s legal team adopted in response to the claim. TUI in its defence put Mr Griffiths to proof of his claim. TUI chose not to lodge the report of an expert microbiologist, which it obtained. That report might have put forward a case on causation which differed from that of Professor Pennington. TUI failed to lodge the report of their expert gastroenterologist in a timely manner and called no witnesses as to fact. The CPR Pt 35.6 questions, which I have set out in para 14 above, were not clearly focused on the matters which were the objects of criticism in counsel’s submissions and did not put Professor Pennington on notice of those criticisms. TUI chose not to request that Professor Pennington be made available for cross-examination. TUI’s challenge to his evidence was not intimated to Mr Griffiths’ legal team until the submission of its skeleton arguments on the eve of the trial, by which time it would have been too late for them to seek to have him attend to give evidence.

It is also necessary to consider the factual evidence which was available to Professor Pennington and was before the trial judge. HHJ Truman accepted in full the evidence of Mr and Mrs Griffiths. They were on an all-inclusive package at the hotel and ate there almost exclusively. They gave some evidence of poor hygiene standards at the hotel, which was not contradicted. They spoke of eating out in the local town on one occasion on the evening of 7 August 2014; Mr Griffiths did not eat much. In contrast with the evidence relating to the hotel, there was no evidence as to the hygiene standards at Burger King in Birmingham airport or at the restaurant in the local town.

Professor Pennington’s report related to causation, which was the central issue on which Mr Griffiths had been put to his proof. It was terse and could and should have included more expansive reasoning. It left many relevant questions unanswered. But it was far from a bare ipse dixit. In support of his conclusion that on the balance of probabilities Mr Griffiths acquired his gastric illnesses following the consumption of contaminated food or fluid from the hotel, his reasoning, as gleaned from what he expressly said, appears to be as follows: (i) the stool tests identified giardia as one of the pathogens, (ii) giardia is common in Turkey, (iii) the occurrence of Mr Griffiths’ illness was within the range of the incubation period of giardia so that the onset of the illness was consistent with eating contaminated food in the hotel, (iv) amoebic dysentery and viral infection (including adenovirus and rotavirus) were unlikely to be the causes of the illness for the reasons he gave.

As I have said, the report left many questions unanswered. But in the context of a claim of relatively low value, Professor Pennington may have thought that his full reasoning was implicit. Importantly, he explained an important part of his reasoning in his answers to the CPR Pt 35.6 questions which I have set out in para 15 above. Consistently with the publications which he accepted as reliable (answer 10), he associated giardia with poor hygiene standards and contaminated food or fluid (para 4 of his report and answers 5 and 8). I have highlighted answers 4 and 6 because they point to his having made a simple assessment of the likely cause of the illness. Professor Pennington explained that he had regard to the length of time spent in the hotel, the nature of the food consumed, and the frequency of consumption of food in the hotel as relevant considerations in attributing the cause of the illness to the ingestion of food or fluid in the hotel. In my view, what he was saying was that he was relying in making his assessment of likely causation on the frequency and circumstances of eating in the hotel when set against the single meal at Birmingham airport and the meal in the local town on 7 August 2014 and other possible sources of infection. This assessment of the balance of probabilities is at a high level of generality but it is not irrational and may have been proportionate in the circumstances of the claim. Further, there is no basis for concluding that Professor Pennington would not have explained his reasoning more clearly if challenged on cross-examination.

None of the exceptions identified in paras 61-68 above applied to Professor Pennington’s evidence. In the absence of a proper challenge on cross-examination it was not fair for TUI to advance the detailed criticisms of Professor Pennington’s report in its submissions or for the trial judge to accept those submissions.

Both the trial judge and the majority of the Court of Appeal erred in law in a significant way. The trial judge did not consider the effect on the fairness of the trial of TUI’s failure to cross-examine Professor Pennington. The majority of the Court of Appeal did, but they erred in limiting the scope of the rule to challenges to the honesty of a witness. As a result, neither properly addressed the application of rule to the facts of this case. In my view, in agreement with Bean LJ’s powerful dissent in para 99 of his judgment, Mr Griffiths did not have a fair trial.

I also respectfully disagree with Asplin LJ’s acceptance of Mr Stevens’ attempted distinction between holding that Professor Pennington’s report was wrong and holding that it did not establish Mr Griffiths’ case on causation. This argument is, as Bean LJ stated, hair-splitting. On any view, the trial judge rejected Professor Pennington’s conclusion that on the balance of probabilities the cause of Mr Griffiths’ illness was food or fluid ingested in the hotel.

In view of those errors of law, it falls to this court to make its own assessment of the evidence. TUI failed to challenge Professor Pennington’s report on cross-examination, which was therefore uncontroverted. I have regard to the factual findings of the trial judge summarised in para 72 above (findings of poor hygiene standards in the hotel at which Mr and Mrs Griffiths had almost all their meals during their stay and the absence of evidence of poor standards at the other establishments). I also have regard to Professor Pennington’s report and CPR Pt 35.6 answers summarised in paras 73 and 74 above (the identification of giardia in the stool tests, its commonness in Turkey, its incubation period, his explanation as to why he excluded amoebic dysentery and viral infection as unlikely, and his explanation of the straightforward basis of his assessment in answers to questions 4 and 6). I conclude that, on that evidence, Mr Griffiths has established his case on the balance of probabilities.

Other matters

I should mention briefly the case of Wood v TUI to which HHJ Truman and Martin Spencer J referred. The obiter dicta of Burnett LJ and Sir Brian Leveson P in paras 29 and 34 of that case respectively were made in the context of TUI’s expressed concern that it should not be liable for every upset stomach occurring on a package holiday which it provides. Burnett LJ was unquestionably correct that the burden lies on the claimant to prove that food or drink provided by the hotel included in the package holiday caused his or her illness. Both he and Sir Brian Leveson suggested that it might be very difficult to do so in the absence of evidence that others who consumed the food had been similarly affected. That suggestion may be questionable in the light of Professor Pennington’s evidence in answer 7 to the CPR Pt 35.6 questions that most cases of infective gastroenteritis caused by eating food are sporadic. But his opinion was not explored, as it should have been, on cross-examination in this case, and it is not appropriate to say more. In para 34 Sir Brian Leveson suggested that alternative explanations “would have to be excluded”. If, by the use of the term “excluded”, he meant that the alternative explanations were to be discounted as less likely causes of the illness than the impugned food and drink, I would agree.

Notwithstanding the concerns in Wood v TUI, in the present case, there was no question of inferring from the mere fact of illness that the illness was caused by contaminated food or drink provided by the hotel. Professor Pennington had the advantage of the stool samples to identify the likely pathogen and, as discussed above, a basis in the factual evidence for inferring on the balance of probabilities that the likely cause of the illness was the ingestion of food or fluid in the hotel.

Finally, TUI expressed concern about the adverse consequences to the cost-effective resolution of civil litigation, including low-value holiday sickness claims, if the appeal were to be upheld. The conclusion I have reached does not mean that in most cases of modest value when a claimant presents an inadequately reasoned expert report, a defendant will inevitably have to obtain a detailed expert report and require a claimant’s expert to attend for cross-examination. A defendant may be able to adopt more economic ways of testing the expert’s evidence. It is important and consistent with the ethos of the CPR that there be a proportionate use of resources in the pursuit and defence of such claims. A defendant can ask focused CPR Pt 35.6 questions which articulate clearly the challenge or challenges on which the defendant wishes to make and give the expert the opportunity to explain his or her evidence in response to those challenges, thereby obviating the need to seek the expert’s attendance for cross-examination. In this case TUI’s questions did not give adequate notice of the challenges it ultimately made. Where the defendant has expert advice, a meeting of experts to discuss their positions can lead to a joint report restricting the issues in dispute. In any event, a focused cross-examination making the challenge and giving the expert the opportunity to explain his or her report and CPR Pt 35.6 answers need not be long.

Further, as Dr Julian Fulbrook observed in his insightful case note on the Court of Appeal’s judgment (Journal of Personal Injury Law (2022), C55-C60), if the court were to sanction the detailed critique and demolition of an uncontroverted expert report in closing submissions, that would undermine the CPR’s arrangement for agreeing expert reports in advance of trial and narrowing down the areas of dispute. It might also encourage experts defensively to produce prolix reports and add to the cost of the legal proceedings.

Conclusion

I would allow the appeal. I would invite the parties to make written submissions on the appropriate form of order within 14 days of the date of this judgment.