Ipsofactoj.com: International Cases [2002] Part 12 Case 11 [HL]


HOUSE OF LORDS

Coram

Fairchild

- vs -

Glenhaven Funeral

Services Ltd

LORD BINGHAM OF CORNHILL

LORD NICHOLLS OF BIRKENHEAD

LORD HOFFMANN

LORD HUTTON

LORD RODGER OF EARLSFERRY

20 JUNE 2002


Judgment

Lord Bingham of Cornhill

My Lords,

  1. On 16 May 2002 it was announced that these three appeals would be allowed. I now give my reasons for reaching that decision.

  2. The essential question underlying the appeals may be accurately expressed in this way. If

    (1)

    C was employed at different times and for differing periods by both A and B, and

    (2)

    A and B were both subject to a duty to take reasonable care or to take all practicable measures to prevent C inhaling asbestos dust because of the known risk that asbestos dust (if inhaled) might cause a mesothelioma, and

    (3)

    both A and B were in breach of that duty in relation to C during the periods of C's employment by each of them with the result that during both periods C inhaled excessive quantities of asbestos dust, and

    (4)

    C is found to be suffering from a mesothelioma, and

    (5)

    any cause of C's mesothelioma other than the inhalation of asbestos dust at work can be effectively discounted, but

    (6)

    C cannot (because of the current limits of human science) prove, on the balance of probabilities, that his mesothelioma was the result of his inhaling asbestos dust during his employment by A or during his employment by B or during his employment by A and B taken together,

    is C entitled to recover damages against either A or B or against both A and B? To this question (not formulated in these terms) the Court of Appeal (Brooke, Latham and Kay LJJ), in a reserved judgment of the court reported at [2002] 1 WLR 1052, gave a negative answer. It did so because, applying the conventional "but for" test of tortious liability, it could not be held that C had proved against A that his mesothelioma would probably not have occurred but for the breach of duty by A, nor against B that his mesothelioma would probably not have occurred but for the breach of duty by B, nor against A and B that his mesothelioma would probably not have occurred but for the breach of duty by both A and B together. So C failed against both A and B. The crucial issue on appeal is whether, in the special circumstances of such a case, principle, authority or policy requires or justifies a modified approach to proof of causation.

  3. It is common ground that in each of the three cases under appeal conditions numbered (1) to (5) above effectively obtained. During his working life the late Mr. Fairchild worked for an employer (whose successor was wrongly identified as the first-named defendant) who carried out sub-contract work for the Leeds City Council in the early 1960s and may have built packing cases for the transportation of industrial ovens lined with asbestos. He also worked for a builder, in whose employment he cut asbestos sheeting both to repair various roofs and while renovating a factory for Waddingtons plc. In the course of his work Mr. Fairchild inhaled substantial quantities of asbestos dust containing asbestos fibre which caused him to suffer a mesothelioma of the pleura, from which he died on 18 September 1996 at the age of 60. Waddingtons plc accepted at trial that it had exposed Mr. Fairchild to the inhalation of asbestos fibres by a breach of the duty owed to him under section 63 of the Factories Act 1961. (Waddingtons plc was not an employer, but nothing turns on this distinction with the other cases.) It thereby admitted that he had been exposed to a substantial quantity of dust or had been exposed to dust to such an extent as was likely to be injurious to him. After the death of Mr. Fairchild his widow brought this action, originally against three defendants (not including the builder). She discontinued proceedings against the first-named defendant, and on 1 February 2001 Curtis J dismissed her claim against Waddingtons plc and the Leeds City Council. The Court of Appeal dismissed her appeal against that decision in the judgment already referred to, finding it unnecessary (because of its decision on causation) to reach a final decision on all aspects of her common law claim against the Leeds City Council. She challenges that causation decision on appeal to the House.

  4. The late Mr. Fox was employed as a lagger by Spousal (Midlands) Ltd (then known by a different name) for 1½ - 2 years between about 1953 and 1955. In the course of this employment he worked at various different premises. Typical lagging work involved the removal of old lagging, the mixing of lagging paste, the cutting of lagging sections and the sweeping up of dust and debris. Asbestos materials were used on a daily basis. The activities of laggers generated high levels of dust containing asbestos. In these circumstances Mr. Fox was exposed to large amounts of asbestos dust, often for many hours each day. He was described by a witness as being covered in dust from head to foot. No measures were taken to protect him from such exposure. From 1955-1989 he worked as a docker/holdsman in the Liverpool Docks. Until the late 1960s or early 1970s asbestos fibre was imported into Liverpool Docks in sacks. Mr. Fox told his wife that he was regularly involved in moving asbestos cargo and that asbestos was regularly released into his breathing area. The work of handling asbestos cargoes would have exposed Mr. Fox to substantial amounts of dust and it is unlikely that any measures would have been taken to protect him from such exposure. But there is no evidence of when and for how long and how frequently Mr. Fox handled cargoes containing asbestos, nor of what cargoes he handled, nor of the identity of his employers when he was engaged in handling asbestos. Spousal do not dispute that they were in breach of duty in exposing Mr. Fox to substantial amounts of asbestos dust in the course of his employment by them. In 1995 he developed symptoms of mesothelioma and he died on 24 April 1996 at the age of 63. It is accepted that his condition was caused by exposure to asbestos dust. After his death his widow brought these proceedings against Spousal. Her claim was dismissed by Judge Mackay, sitting as a judge of the Queen's Bench Division in Liverpool on 27 March 2001. Her appeal against that decision was dismissed by the Court of Appeal in the judgment already referred to. She challenges that decision on appeal to the House.

  5. Mr. Matthews was employed by Associated Portland Cement Manufacturers (1978) Ltd from 1973-81 at their factory in Strood, Kent. He was exposed to asbestos during the last four years of this employment when working as a boilerman. Each day he spent some time (up to about an hour) in the boilerhouse where the boiler and ancillary pipework were lagged with asbestos material. On a number of occasions (adding up to about 2 days in all) he was in close proximity to men removing lagging from pipes, and such work created large amounts of asbestos dust. On a daily basis he was exposed to dust and debris from the lagging. He walked across pipework disturbing the lagging. He regularly swept the floor in the boilerhouse, stirring up asbestos dust and debris. No effective measures were taken to protect him from exposure to asbestos dust. For 5-6 weeks in January and February 1973 Mr. Matthews was employed by British Uralite plc at their factory in Higham, Kent, where the company manufactured pipes from asbestos material, and Mr. Matthews worked on this process. Large amounts of dust containing asbestos fibres were created by the manufacturing process and such dust permeated the atmosphere of the factory. During each working day Mr. Matthews had prolonged and substantial exposure to asbestos dust. No measures were taken to protect him against such exposure. Between 1965 and 1967 Mr. Matthews was employed by Maidstone Sack and Metal and was again exposed to significant quantities of asbestos dust. For 12 months of this period he operated a scrap metal press and some of the items fed into the press had asbestos linings. For about 2 weeks he worked in a boilerhouse in Chatham Dockyard dismantling a boiler and pipework, during which time he spent a day removing asbestos lagging from the boiler and pipes, which was dusty work. Maidstone Sack and Metal can no longer be sued. Mr. Matthews consulted his doctor complaining of chest pain in March 1999. In February 2000 a diagnosis of mesothelioma was made. His condition has continued to deteriorate, and his life expectancy is now measured in months. Associated Portland Cement and British Uralite admit that Mr. Matthews' mesothelioma was caused by exposure to asbestos dust, and that each of them exposed Mr. Matthews to asbestos dust in breach of duty. Mr. Matthews issued proceedings against both these companies in April 2001. On 11 July 2001 Mitting J gave judgment in his favour against both defendants and awarded damages. The defendants appealed against that decision, and the Court of Appeal allowed their appeal and set aside the award in Mr. Matthews' favour. He has appealed against that decision. It should be recorded that, before the hearing of his appeal in the House, the defendants agreed to pay Mr. Matthews the sum awarded by the judge with interest and costs, without prejudice to the issues in the appeal.

  6. It has been recognised for very many years, at any rate since the "Report on Effects of Asbestos Dust on the Lungs and Dust Suppression in the Asbestos Industry" by Merewether and Price in 1930 and the making of the Asbestos Industry Regulations 1931, that it is injurious to inhale significant quantities of asbestos dust. At first, attention was focused on the risk of contracting asbestosis and other pulmonary diseases. It is a characteristic of asbestosis that the disease, once initiated, will be influenced by the total amount of dust thereafter inhaled. Thus in the case of asbestosis the following situation may arise. C may contract asbestosis as a result of exposure to asbestos dust while employed by A, but without such exposure involving any breach of duty by A. C may then work for B, and again inhale quantities of asbestos dust which will have the effect of aggravating his asbestosis. If this later exposure does involve a breach of duty by B, C will have no claim against A but will have a claim against B. B will not escape liability by contending that his breach of duty is not shown to have had any causative effect.

  7. From about the 1960s, it became widely known that exposure to asbestos dust and fibres could give rise not only to asbestosis and other pulmonary diseases, but also to the risk of developing a mesothelioma. This is a malignant tumour, usually of the pleura, sometimes of the peritoneum. In the absence of occupational exposure to asbestos dust it is a very rare tumour indeed, afflicting no more than about one person in a million per year. But the incidence of the tumour among those occupationally exposed to asbestos dust is about 1,000 times greater than in the general population, and there are some 1,500 cases reported annually. It is a condition which may be latent for many years, usually for 30-40 years or more; development of the condition may take as short a period as 10 years, but it is thought that that is the period which elapses between the mutation of the first cell and the manifestation of symptoms of the condition. It is invariably fatal, and death usually occurs within 1-2 years of the condition being diagnosed. The mechanism by which a normal mesothelial cell is transformed into a mesothelioma cell is not known. It is believed by the best medical opinion to involve a multi-stage process, in which 6 or 7 genetic changes occur in a normal cell to render it malignant. Asbestos acts in at least one of those stages and may (but this is uncertain) act in more than one. It is not known what level of exposure to asbestos dust and fibre can be tolerated without significant risk of developing a mesothelioma, but it is known that those living in urban environments (although without occupational exposure) inhale large numbers of asbestos fibres without developing a mesothelioma. It is accepted that the risk of developing a mesothelioma increases in proportion to the quantity of asbestos dust and fibres inhaled: the greater the quantity of dust and fibre inhaled, the greater the risk. But the condition may be caused by a single fibre, or a few fibres, or many fibres: medical opinion holds none of these possibilities to be more probable than any other, and the condition once caused is not aggravated by further exposure. So if C is employed successively by A and B and is exposed to asbestos dust and fibres during each employment and develops a mesothelioma, the very strong probability is that this will have been caused by inhalation of asbestos dust containing fibres. But C could have inhaled a single fibre giving rise to his condition during employment by A, in which case his exposure by B will have had no effect on his condition; or he could have inhaled a single fibre giving rise to his condition during his employment by B, in which case his exposure by A will have had no effect on his condition; or he could have inhaled fibres during his employment by A and B which together gave rise to his condition; but medical science cannot support the suggestion that any of these possibilities is to be regarded as more probable than any other. There is no way of identifying, even on a balance of probabilities, the source of the fibre or fibres which initiated the genetic process which culminated in the malignant tumour. It is on this rock of uncertainty, reflecting the point to which medical science has so far advanced, that the three claims were rejected by the Court of Appeal and by two of the three trial judges.

    PRINCIPLE

  8. In a personal injury action based on negligence or breach of statutory duty the claimant seeks to establish a breach by the defendant of a duty owed to the claimant, which has caused him damage. For the purposes of analysis, and for the purpose of pleading, proving and resolving the claim, lawyers find it convenient to break the claim into its constituent elements: the duty, the breach, the damage and the causal connection between the breach and the damage. In the generality of personal injury actions, it is of course true that the claimant is required to discharge the burden of showing that the breach of which he complains caused the damage for which he claims and to do so by showing that but for the breach he would not have suffered the damage.

  9. The issue in these appeals does not concern the general validity and applicability of that requirement, which is not in question, but is whether in special circumstances such as those in these cases there should be any variation or relaxation of it. The overall object of tort law is to define cases in which the law may justly hold one party liable to compensate another. Are these such cases? A and B owed C a duty to protect C against a risk of a particular and very serious kind. They failed to perform that duty. As a result the risk eventuated and C suffered the very harm against which it was the duty of A and B to protect him. Had there been only one tortfeasor, C would have been entitled to recover, but because the duty owed to him was broken by two tortfeasors and not only one, he is held to be entitled to recover against neither, because of his inability to prove what is scientifically unprovable. If the mechanical application of generally accepted rules leads to such a result, there must be room to question the appropriateness of such an approach in such a case.

  10. In March v E & MH Stramare Pty Ltd (1991) 171 CLR 506 at 508, Mason CJ, sitting in the High Court of Australia, did not "accept that the 'but for' (causa sine qua non) test ever was or now should become the exclusive test of causation in negligence cases" and (at p 516) he added:

  11. The 'but for' test gives rise to a well-known difficulty in cases where there are two or more acts or events which would each be sufficient to bring about the plaintiff's injury. The application of the test 'gives the result, contrary to common sense, that neither is a cause': Winfield & Jolowicz on Tort, 13th ed (1989), p. 134. In truth, the application of the test proves to be either inadequate or troublesome in various situations in which there are multiple acts or events leading to the plaintiff's injury: see, e.g., Chapman v Hearse, Baker v Willoughby [1970] AC 467; McGhee v National Coal Board; M'Kew (to which I shall shortly refer in some detail). The cases demonstrate the lesson of experience, namely, that the test, applied as an exclusive criterion of causation, yields unacceptable results and that the results which it yields must be tempered by the making of value judgments and the infusion of policy considerations.

  12. In Snell v Farrell [1990] 2 SCR 311 at 320, Sopinka J, delivering the judgment of the Supreme Court of Canada, said:

  13. The traditional approach to causation has come under attack in a number of cases in which there is concern that due to the complexities of proof, the probable victim of tortious conduct will be deprived of relief. This concern is strongest in circumstances in which, on the basis of some percentage of statistical probability, the plaintiff is the likely victim of the combined tortious conduct of a number of defendants, but cannot prove causation against a specific defendant or defendants on the basis of particularized evidence in accordance with traditional principles. The challenge to the traditional approach has manifested itself in cases dealing with non-traumatic injuries such as man-made diseases resulting from the widespread diffusion of chemical products, including product liability cases in which a product which can cause injury is widely manufactured and marketed by a large number of corporations.

  14. McLachlin J, extra-judicially ("Negligence Law - Proving the Connection", in Torts Tomorrow, A Tribute to John Fleming, ed Mullany and Linden, LBC Information Services 1998, at p 16), has voiced a similar concern:

    Tort law is about compensating those who are wrongfully injured. But even more fundamentally, it is about recognising and righting wrongful conduct by one person or a group of persons that harms others. If tort law becomes incapable of recognising important wrongs, and hence incapable of righting them, victims will be left with a sense of grievance and the public will be left with a feeling that justice is not what it should be. Some perceive that this may be occurring due to our rules of causation.

    In recent years, a conflation of factors have caused lawyers, scholars and courts to question anew whether the way tort law has traditionally defined the necessary relationship between tortious acts and injuries is the right way to define it, or at least the only way. This questioning has happened in the United States and in England and has surfaced in Australia. And it is happening in Canada. Why is this happening? Why are courts now asking questions that for decades, indeed centuries, did not pose themselves, or if they did, were of no great urgency? I would suggest that it is because too often the traditional 'but-for', all-or-nothing, test denies recovery where our instinctive sense of justice - of what is the right result for the situation - tells us the victim should obtain some compensation.

  15. My noble and learned friend Lord Hoffmann has, on more than one occasion, discouraged a mechanical approach to the issue of causation. In Environment Agency (formerly National Rivers Authority) v Empress Car Co. (Abertillery) Ltd [1999] 2 AC 22 at 29, he said:

    The first point to emphasise is that common sense answers to questions of causation will differ according to the purpose for which the question is asked. Questions of causation often arise for the purpose of attributing responsibility to someone, for example, so as to blame him for something which has happened or to make him guilty of an offence or liable in damages. In such cases, the answer will depend upon the rule by which responsibility is being attributed.

    More recently, in Kuwait Airways Corporation v Iraqi Airways Co (Nos 4 & 5) [2002] 2 WLR 1353, p 1388, para [128], he said:

    There is therefore no uniform causal requirement for liability in tort. Instead, there are varying causal requirements, depending upon the basis and purpose of liability. One cannot separate questions of liability from questions of causation. They are inextricably connected. One is never simply liable; one is always liable for something and the rules which determine what one is liable for are as much part of the substantive law as the rules which determine which acts give rise to liability.

    Laws LJ was reflecting this approach when he said in Rahman v Arearose Ltd [2001] QB 351 at 367-368:

    So in all these cases the real question is, what is the damage for which the defendant under consideration should be held responsible. The nature of his duty (here, the common law duty of care) is relevant; causation, certainly, will be relevant - but it will fall to be viewed, and in truth can only be understood, in light of the answer to the question: from what kind of harm was it the defendant's duty to guard the claimant? .... Novus actus interveniens, the eggshell skull, and (in the case of multiple torts) the concept of concurrent tortfeasors are all no more and no less than tools or mechanisms which the law has developed to articulate in practice the extent of any liable defendant's responsibility for the loss and damage which the claimant has suffered.

  16. I do not therefore consider that the House is acting contrary to principle in reviewing the applicability of the conventional test of causation to cases such as the present. Indeed, it would seem to me contrary to principle to insist on application of a rule which appeared, if it did, to yield unfair results. And I think it salutary to bear in mind Lord Mansfield's aphorism in Blatch v Archer (1774) 1 Cowp 63 at 65, quoted with approval by the Supreme Court of Canada in Snell v Farrell (above):

    It is certainly a maxim that all evidence is to be weighed according to the proof which it was in the power of one side to have produced, and in the power of the other to have contradicted.

    AUTHORITY

  17. In Bonnington Castings Ltd v Wardlaw [1956] AC 613, the pursuer contracted pneumoconiosis as a result of inhaling silica dust. The dust came from two sources, a pneumatic hammer and swing grinders, both in the dressing shop where he worked. The dust emanating from the pneumatic hammer involved no breach of duty by the employer, but that from the swing grinders did. In a claim against his employer he succeeded before the Lord Ordinary, Lord Wheatley, and by a majority in the First Division of the Court of Session, the Lord President (Lord Clyde) dissenting. The issue on appeal was whether the employer's admitted breach of duty in relation to the swing grinders had caused the pursuer's disease. In his leading opinion, Lord Reid made plain that "the employee must in all cases prove his case by the ordinary standard of proof in civil actions: he must make it appear at least that on a balance of probabilities the breach of duty caused or materially contributed to his injury" (p. 620). He pointed out that pneumoconiosis is caused by a gradual accumulation in the lungs of minute particles of silica inhaled over a period of years (p. 621), and he regarded the real question as "whether the dust from the swing grinders materially contributed to the disease" (p. 621). He considered that any contribution which was not de minimis must be material. The evidence showed that even if more dust came from the pneumatic hammer than from the swing grinders, there was enough dust from the grinders to make a substantial contribution towards the pursuer's disease (p. 622). The pursuer was accordingly entitled to succeed. With these conclusions, Viscount Simonds, Lord Tucker, Lord Keith of Avonholm and Lord Somervell of Harrow agreed, Lord Keith laying stress at p. 626 on the nature of pneumoconiosis as a disease of gradual incidence and on the cumulative effect of inhalation of dust from the grinders over a period, which might be small in proportion but substantial in total quantity. The case differs from the present in two obvious respects.

    • First, the pursuer had only one relevant employer, who was not legally liable for producing some of the dust which the pursuer inhaled but was potentially liable for the balance. 

    • Secondly, pneumoconiosis is, like asbestosis, a condition which is aggravated by the inhalation of increased quantities of dust so that, even if the "innocent" dust had been the first and major cause of the condition, the "guilty" dust, if in significant quantities, could properly be said to have made it worse.

  18. Nicholson v Atlas Steel Foundry & Engineering Co Ltd [1957] 1 WLR 613 was factually a variant of Wardlaw's case. The claim was made by the widow and children of Mr Nicholson, who had worked in the dressing shop of the defenders' steel foundry, had inhaled dust containing minute siliceous particles while doing so, had contracted pneumoconiosis and had died. The complaints made in the action related not to the creation of dust in the dressing shop but to the defenders' failure to provide adequate ventilation to extract the dust. It was common ground that the deceased must inevitably have inhaled a quantity, even a large quantity, of noxious particles about which he could have no cause of complaint, and the only question was whether, in addition to those particles, he was, owing to the fault of the defenders in failing to provide adequate ventilation, bound to have inhaled a number of other particles which made a material contribution to his illness (p 616). The Lord Ordinary found for the family, but his decision was reversed by the First Division. In the House the argument centred on the statutory duty to provide proper ventilation imposed by section 4(1) of the Factories Act 1937, and Viscount Simonds said (at p 618):

    .... if the statute prescribes a proper system of ventilation by the circulation of fresh air so as to render harmless, so far as practicable, all fumes, dust and other impurities that may be injurious to health, generated in the course of work carried on in the factory, and if it is proved that there is no system or only an inadequate system of ventilation, it requires little further to establish a causal link between that default and the illness, due to noxious dust, of a person employed in the shop. Something is required as was held in Wardlaw's case. I was a party to that decision and would not in any way resile from it. But it must not be pressed too far. In the present case there was, in my opinion, ample evidence to support the appellants' case.

  19. Since the family could not complain of the production of dust, and the deceased had been forced to inhale some noxious particles without having any legal complaint, it was doubly incumbent on the employer to safeguard him against any additional risk (p 616). Viscount Simonds' conclusion was clearly expressed (at pp 619-620):

    For it appears to me that [the evidence] clearly established that dust containing dangerous particles of silica was emitted into the air by the operation of pneumatic hammers on the castings, that this dust hung about in concentrated form longer than it would have if there had been better ventilation, and that improved roof ventilators were practicable and would have effectively improved the conditions. It follows that owing to the default of the respondents the deceased was exposed to a greater degree of risk than he should have been, and, though it is impossible, even approximately, to quantify the particles which he must, in any event, have inhaled and those which he inhaled but need not have, I cannot regard the excess as something so negligible that the maxim 'de minimis' is applicable. Accordingly, following the decision in Wardlaw's case, I must hold the respondents liable.

    Lord Oaksey and Lord Morton of Henryton agreed. Lord Cohen agreed and said (at p 622):

    Pneumoconiosis is a progressive disease. The longer a workman is exposed to an intense cloud the graver must be the risk of infection. In the present case it is clearly established by the evidence that at any rate down to 1949 the tool with which the deceased was working on dirty castings created a thick cloud of dust which must have necessarily included siliceous particles to an extent which cannot classed as 'de minimis'. The respondents are admittedly not to blame for the generation of this cloud, but any failure to provide proper ventilation must, I think, lengthen the period during which the cloud remains intense. It seems to me to follow that the respondents' failure to provide adequate ventilation must increase the risk to which the workmen are exposed. Reading the evidence as a whole, I think it establishes that (to use the language of Lord Reid in Wardlaw's case) 'on a balance of probabilities the breach of duty caused or materially contributed to' the injury.

    Lord Keith of Avonholm regarded it as common sense that better ventilation would have appreciably diminished the dust which was in the air for the deceased to inhale (p 627) and accordingly concluded that his death from pneumoconiosis should be ascribed at least partially to the fault of the defenders. Again the case involved a single employer: but the dust, although "innocent" when first produced became, in effect, "guilty" because of the employer's conduct in allowing it to remain in the air for an excessive period. It is noteworthy that two members of the House (Viscount Simonds and Lord Cohen) attached significance to the exposure of the deceased to an increased risk.

  20. Gardiner v Motherwell Machinery & Scrap Co Ltd [1961] 1 WLR 1424, another Scottish case, concerned a pursuer who had worked for the defenders for a period of some three months, demolishing buildings, and had contracted dermatitis. In an action against the defenders he claimed that they should have provided him with washing facilities but had failed to do so and that their failure had caused him to suffer from dermatitis. This contention was upheld by the Lord Ordinary (Lord Kilbrandon) who awarded him damages. The defenders did not on appeal challenge the finding of breach but contended that the pursuer had failed to prove any connection between his disease and the work which he had been doing. The First Division accepted this argument and found for the defenders, a decision against which the pursuer appealed. In his leading opinion in the House, Lord Reid considered at some length the conflict of medical evidence at the trial and its treatment by the First Division, and expressed his conclusion (at p 1429):

    In my opinion, when a man who has not previously suffered from a disease contracts that disease after being subjected to conditions likely to cause it, and when he shows that it starts in a way typical of disease caused by such conditions, he establishes a prima facie presumption that his disease was caused by those conditions. I think that the facts proved in this case do establish such a presumption. That presumption could be displaced in many ways. The respondents sought to show, first, that it is negatived by the subsequent course of the disease and, secondly, by suggesting tinea pedis as an equally probable cause of its origin. I have found the case difficult, but on the evidence as it stands I have come to the opinion that they have failed on both points. If the appellant's disease and consequent loss should be attributed to the work which he was doing in the respondents' service, it was not argued that they are not liable.

    Lord Cohen and Lord Guest agreed, as did Lord Hodson although with some initial hesitation. Lord Guest described the question as a pure question of fact whether on the balance of probabilities the dermatitis had arisen from the pursuer's employment (p 1431). The House would seem to have regarded the pursuer as establishing a prime facie case which the defenders had failed to displace.

  21. In the course of the present appeals much argument was directed to the decision of the House in McGhee v National Coal Board [1973] 1 WLR 1. The earlier stages of that case are reported at 1973 SC(HL) 37 and are important in understanding what the House decided. Mr McGhee had been employed by the National Coal Board for about 15 years, almost always working in pipe kilns. For some 4½ days he then worked at a brick kiln, giving up because of a dermatitic condition which had by then developed. The work inside the kiln was very hot and very dusty. The heat made men sweat profusely and the operation of the fan caused them to be covered in dust and grit. The pursuer contended that his dermatitis had been caused by his period of working in the brick kiln, short though it had been. The employers contended that his work had not caused the dermatitis and that it was non-occupational in origin. There was at the trial a conflict of medical evidence but the Lord Ordinary (Lord Kissen) held that the pursuer had contracted the dermatitis in the course of his work at the brick kiln and as a result of his exposure to dust and ashes when working there (p 39). Counsel for the pursuer accepted at trial that he could not establish a breach of statutory duty nor a breach of common law duty based on a failure to ventilate, but relied on two alleged breaches by the employers: of a duty to take care that the kiln had cooled sufficiently before men went in to work in it and of a duty to take reasonable care to provide adequate showers to enable men to remove dust from their bodies. The Lord Ordinary rejected the first of these complaints on a number of grounds, including the lack of proof that the breach of duty, even if established, had caused or materially contributed to the dermatitis: it was not enough that a reduction of heat would have lessened the risk (p 41). The Lord Ordinary concluded that the employers were at fault in failing to provide showers (p 42) but found against the pursuer on the basis of evidence given by two expert dermatologists, Dr Hannay and Dr Ferguson, called by the pursuer and the employers respectively. He said (at pp 42-43):

    As I have maintained earlier, the pursuer, in order to succeed, must also establish, on a balance of probabilities, that this fault on the part of the defenders 'caused or materially contributed to his injury', that is to his contracting dermatitis. Dr Hannay's evidence was that he could not say that the provision of showers would probably have prevented the disease. He said that it would have reduced the risk materially but he would not go further than that. Dr Ferguson said that washing reduces the risk. Pursuer's counsel maintained that a material increase in the risk of contracting a disease was the same as a material contribution to contracting the disease and that Dr Hannay established this by his evidence. I think that defenders' counsel was correct when he said that the distinction drawn by Dr Hannay was correct and that an increase in risk did not necessarily mean a material contribution to the contracting of the disease. The two concepts are entirely different. A material increase in risk may refer only to possibilities and may not make a possibility into a probability. It may strengthen the possibility but that cannot mean that in all such cases the possibility has become a probability. What the pursuer has to show is that, as he avers, he would not have contracted the disease but for the defenders' breach of duty. He has to show that this was probable and the degrees of risk have no relevance unless they make the contraction of the disease more probable than not contracting the disease. He cannot succeed if the only inference from the evidence is that lack of shower baths is a possibility as a cause of his having contracted the disease and the provision of shower baths would have increased the possibility but not made it a probability. That is the only inference which I can draw from Dr Hannay's evidence and that was the best evidence for the pursuer. Causal connection between fault and the contraction of the disease has not been established.

  22. The pursuer appealed to the First Division against the dismissal of his claim. The medical evidence given at the trial was reviewed in detail, and in particular an exchange between cross-examining counsel and Dr Hannay ( pp 43-44, 47, 50):

    Q.

    Do I understand you to say you are not in a position to say that the provision of showers would probably have prevented his contracting this skin trouble?

    A.

    No one could say that that would prevent that man developing the condition. It would be likely to reduce the chances.

    In answer to further questions the doctor repeated his opinion that he could only say that the provision of showers would have reduced the chances of the pursuer contracting dermatitis and that that was as far as he was able to go. In the course of his judgment the Lord President, Lord Clyde, considered the pneumoconiosis cases and expressed his conclusion (at p 44):

    But in contrast to the pneumoconiosis cases, the present case is essentially concerned with proof of the causal connection between the fault alleged (i.e. inadequate washing facilities) and the development of dermatitis. Even if the pursuer had established (as he did not) that the absence of washing facilities increased the risk of the pursuer getting dermatitis, that would clearly not prove that the absence of these facilities caused the disease, nor indeed would it go any distance towards proving it. For risk of dermatitis and causation of dermatitis are two quite separate matters.

    Lord Migdale was of the same opinion (at pp 47-48):

    Counsel for the pursuer contended that as it was now accepted that the failure to provide a shower was a breach of the duty which the defenders owed to the pursuer to take reasonable steps for his well-being, the doctors' evidence that it would have materially reduced the risk of dermatitis is enough to link the failure with the injury. Counsel for the defenders, on the other hand, contended that the test of causal connection between the breach and the injury is whether the provision of a shower would, on a balance of probabilities, have prevented the dermatitis. The Lord Ordinary says an increase in risk does not mean a material contribution to the contracting of the disease. A material increase in risk may refer only to possibilities and it does not make a possibility into a probability. 'What the pursuer has to show is that, as he avers, he would not have contracted the disease but for the breach of duty'. He has to show this on a balance of probabilities.

    In my opinion this is correct. Unless the pursuer can point to evidence that shows that a shower would more probably have avoided the disease than not, he cannot succeed and I do not find that evidence in this case.

    Lord Johnston was more hesitant, but in view of the other opinions did not feel inclined to take the view that the evidence was sufficient to allow him to hold that the test of the balance of probability had been satisfied (p 50).

  23. On appeal to the House counsel for the pursuer faced the problem, as he had at trial and in the First Division, that his own evidence precluded a finding that the absence of a shower had probably caused the pursuer's dermatitis. Mr Davidson QC accordingly relied on the evidence that provision of a shower would have materially reduced the risk to contend that he had made out a prima facie case. The contrary argument for the employers was advanced by Mr James Mackay QC, as reported at p 51:

    It was accepted that [the provision of washing facilities] would have been a reasonable precaution, but it did not follow that this would have eliminated the risk. The employee might have developed dermatitis in any event. If the precaution would not have prevented the disease, the appellant was not entitled to damages. In the case of pneumoconiosis the inhalation of dangerous dust inevitably created a basis for the disease by accumulation, whereas in the case of dermatitis a particle of grit would cause the disease only if there were an abrasion which opened up the layer below the horny outer layer of the skin. In the case of pneumoconiosis all the particles could be blamed. It was not so in the case of dermatitis. The mere fact that shower baths would have reduced the chances of the contraction of the disease did not mean that what was probable would thereby have been rendered improbable.

    Thus the issue, as presented to the House, was whether the pursuer could succeed despite his inability to show that he would probably not have suffered dermatitis but for the defenders' failure to provide the showers which they should have provided.

  24. In the House, opinions were given by all five members of the Appellate Committee which heard the appeal and the appeal was allowed: [1973] 1 WLR 1. Lord Reid, giving the first opinion, described the pursuer's complaint based on the failure to provide shower facilities as raising "a difficult question of law" (p 3). He pointed out that the breach of duty in relation to showers was admitted, and it was admitted that the disease was attributable to the work which the pursuer had performed in the brick kiln, but it was contended that the pursuer had not proved that the defenders' failure to carry out the admitted duty had caused the onset of the disease (p 3). Lord Reid's understanding of the evidence, and his view of the proper approach to it, appear from the following passage of his opinion (at pp 4-5):

    In the present case the evidence does not show - perhaps no one knows - just how dermatitis of this type begins. It suggests to me that there are two possible ways. It may be that an accumulation of minor abrasions of the horny layer of the skin is a necessary precondition for the onset of the disease. Or it may be that the disease starts at one particular abrasion and then spreads, so that multiplication of abrasions merely increases the number of places where the disease can start and in that way increases the risk of its occurrence.

    I am inclined to think that the evidence points to the former view. But in a field where so little appears to be known with certainty I could not say that that is proved. If it were, then this case would be indistinguishable from Wardlaw's case. But I think that in cases like this we must take a broader view of causation. The medical evidence is to the effect that the fact that the man had to cycle home caked with grime and sweat added materially to the risk that this disease might develop. It does not and could not explain just why that is so. But experience shows that it is so. Plainly that must be because what happens while the man remains unwashed can have a causative effect, though just how the cause operates is uncertain. I cannot accept the view expressed in the Inner House that once the man left the brick kiln he left behind the causes which made him liable to develop dermatitis. That seems to me quite inconsistent with a proper interpretation of the medical evidence. Nor can I accept the distinction drawn by the Lord Ordinary between materially increasing the risk that the disease will occur and making a material contribution to its occurrence.

    There may be some logical ground for such a distinction where our knowledge of all the material factors is complete. But it has often been said that the legal concept of causation is not based on logic or philosophy. It is based on the practical way in which the ordinary man's mind works in the everyday affairs of life. From a broad and practical viewpoint I can see no substantial difference between saying that what the defender did materially increased the risk of injury to the pursuer and saying that what the defender did made a material contribution to his injury.

  25. Lord Wilberforce acknowledged the need for the pursuer to establish both a breach of duty and a causal connection between the default and the disease complained of (p 5), and also the difficulties of proof which the pursuer faced (pp 5-6):

    [The pursuer's medical expert] could not do more than say that the failure to provide showers materially increased the chance, or risk, that dermatitis might set in.

    Lord Wilberforce accepted that merely to show that a breach of duty led to an increase of risk was not enough to enable a pursuer to succeed, but continued (at page 6):

    But the question remains whether a pursuer must necessarily fail if, after he has shown a breach of duty, involving an increase of risk of disease, he cannot positively prove that this increase of risk caused or materially contributed to the disease while his employers cannot positively prove the contrary. In this intermediate case there is an appearance of logic in the view that the pursuer, on whom the onus lies, should fail - a logic which dictated the judgments below. The question is whether we should be satisfied, in factual situations like the present, with this logical approach. In my opinion, there are further considerations of importance. First, it is a sound principle that where a person has, by breach of a duty of care, created a risk, and injury occurs within the area of that risk, the loss should be borne by him unless he shows that it had some other cause. Secondly, from the evidential point of view, one may ask, why should a man who is able to show that his employer should have taken certain precautions, because without them there is a risk, or an added risk, of injury or disease, and who in fact sustains exactly that injury or disease, have to assume the burden of proving more: namely, that it was the addition to the risk, caused by the breach of duty, which caused or materially contributed to the injury? In many cases, of which the present is typical, this is impossible to prove, just because honest medical opinion cannot segregate the causes of an illness between compound causes. And if one asks which of the parties, the workman or the employers, should suffer from this inherent evidential difficulty, the answer as a matter of policy or justice should be that it is the creator of the risk who, ex hypothesi must be taken to have foreseen the possibility of damage, who should bear its consequences.

    Having referred to Wardlaw's and Nicholson's cases Lord Wilberforce concluded (at page 7):

    The present factual situation has its differences: the default here consisted not in adding a material quantity to the accumulation of injurious particles but by failure to take a step which materially increased the risk that the dust already present would cause injury. And I must say that, at least in the present case, to bridge the evidential gap by inference seems to me something of a fiction, since it was precisely this inference which the medical expert declined to make. But I find in the cases quoted an analogy which suggests the conclusion that, in the absence of proof that the culpable addition had, in the result, no effect, the employers should be liable for an injury, squarely within the risk which they created and that they, not the pursuer, should suffer the consequence of the impossibility, foreseeably inherent in the nature of his injury, of segregating the precise consequence of their default.

  26. Lord Simon of Glaisdale considered that Wardlaw's and Nicholson's cases established a rule (at page 8)

    that where an injury is caused by two (or more) factors operating cumulatively, one (or more) of which factors is a breach of duty and one (or more) is not so, in such a way that it is impossible to ascertain the proportion in which the factors were effective in producing the injury or which factor was decisive, the law does not require a pursuer or plaintiff to prove the impossible, but holds that he is entitled to damages for the injury if he proves on a balance of probabilities that the breach or breaches of duty contributed substantially to causing the injury. If such factors so operate cumulatively, it is, in my judgment, immaterial whether they do so concurrently or successively.

    Lord Simon then continued (page 8):

    The question, then, is whether on the evidence the appellant brought himself within this rule. In my view, the failure to take steps which would bring about a material reduction of the risk involves, in this type of case, a substantial contribution to the injury. In this type of case a stark distinction between breach of duty and causation is unreal. If the provision of shower baths was (as the evidence showed) a precaution which any reasonable employer in the respondents' position would take, it means that such employer should have foreseen that failure to take the precaution would, more probably than not, substantially contribute towards injury: this is sufficient prima facie evidence.

    Lord Simon regarded "material reduction of the risk" and "substantial contribution to the injury" as mirror concepts. Any other conclusion would mean that the defenders were under a legal duty which they could, on the present state of medical knowledge, ignore (page 9).

  27. Lord Kilbrandon appears to have adopted a more orthodox approach to tortious liability. He said (at page 10):

    When you find it proved (a) that the defenders knew that to take the precaution reduces the risk, chance, possibility or probability of the contracting of a disease, (b) that the precaution has not been taken, and (c) that the disease has supervened, it is difficult to see how those defenders can demand more by way of proof of the probability that the failure caused or contributed to the physical breakdown ... In the present case, the pursuer's body was vulnerable, while he was bicycling home, to the dirt which had been deposited on it during his working hours. It would not have been if he had had a shower. If showers had been provided he would have used them. It is admittedly more probable that disease will be contracted if a shower is not taken. In these circumstances I cannot accept the argument that nevertheless it is not more probable than not that, if the duty to provide a shower had not been neglected, he would not have contracted the disease. The pursuer has after all, only to satisfy the court of a probability, not to demonstrate an irrefragable chain of causation, which in a case of dermatitis, in the present state of medical knowledge, he could probably never do.

  28. In Lord Salmon's opinion the question before the House was whether the pursuer's dermatitis was proved to have been caused or materially contributed to by the defenders' negligence (page 11). He rejected the view, expressed by the Lord President (see para [20] above) that to increase the risk of injury was not, in the circumstances of this case, to cause the injury. In such a case he regarded it as unrealistic and contrary to ordinary common sense to hold that the negligence which materially increased the risk of injury did not materially contribute to causing it (pages 11-12). He observed (at page 12):

    I think that the approach by the courts below confuses the balance of probability test with the nature of causation. Moreover, it would mean that in the present state of medical knowledge and in circumstances such as these (which are by no means uncommon) an employer would be permitted by the law to disregard with impunity his duty to take reasonable care for the safety of his employees.

    Lord Salmon's conclusion (pages 12-13) was expressed in these terms:

    In the circumstances of the present case, the possibility of a distinction existing between (a) having materially increased the risk of contracting the disease, and (b) having materially contributed to causing the disease may no doubt be a fruitful source of interesting academic discussions between students of philosophy. Such a distinction is, however, far too unreal to be recognised by the common law.

  29. This detailed review of McGhee permits certain conclusions to be drawn.

    • First, the House was deciding a question of law. Lord Reid expressly said so (page 3). The other opinions, save perhaps that of Lord Kilbrandon, cannot be read as decisions of fact or as orthodox applications of settled law.

    • Secondly, the question of law was whether, on the facts of the case as found, a pursuer who could not show that the defender's breach had probably caused the damage of which he complained could nonetheless succeed.

    • Thirdly, it was not open to the House to draw a factual inference that the breach probably had caused the damage: such an inference was expressly contradicted by the medical experts on both sides; and once that evidence had been given the crux of the argument before the Lord Ordinary and the First Division and the House was whether, since the pursuer could not prove that the breach had probably made a material contribution to his contracting dermatitis, it was enough to show that the breach had increased the risk of his contracting it.

    • Fourthly, it was expressly held by three members of the House (Lord Reid at page 5, Lord Simon at page 8 and Lord Salmon at pages 12-13) that in the circumstances no distinction was to be drawn between making a material contribution to causing the disease and materially increasing the risk of the pursuer contracting it. Thus the proposition expressly rejected by the Lord Ordinary, the Lord President and Lord Migdale was expressly accepted by a majority of the House and must be taken to represent the ratio of the decision, closely tied though it was to the special facts on which it was based.

    • Fifthly, recognising that the pursuer faced an insuperable problem of proof if the orthodox test of causation was applied, but regarding the case as one in which justice demanded a remedy for the pursuer, a majority of the House adapted the orthodox test to meet the particular case. The authority is of obvious importance in the present appeal since the medical evidence left open the possibility, as Lord Reid pointed out at page 4, that the pursuer's dermatitis could have begun with a single abrasion, which might have been caused when he was cycling home, but might equally have been caused when he was working in the brick kiln; in the latter event, the failure to provide showers would have made no difference. In McGhee, however, unlike the present appeals, the case was not complicated by the existence of additional or alternative wrongdoers.

  30. In Wilsher v Essex Area Health Authority a problem of causation arose in a different context. A prematurely-born baby was the subject of certain medical procedures, in the course of which a breach of duty occurred. The baby suffered a condition (abbreviated as RLF) of a kind which that breach of duty could have caused, and the breach of duty increased the risk of his suffering it. But there were a number of other factors which might have caused the injury. In the Court of Appeal ([1987] QB 730 at 771-772) Mustill LJ concluded a detailed review of McGhee by making this statement of principle:

    If it is an established fact that conduct of a particular kind creates a risk that injury will be caused to another or increases an existing risk that injury will ensue; and if the two parties stand in such a relationship that the one party owes a duty not to conduct himself in that way; and if the first party does conduct himself in that way; and if the other party does suffer injury of the kind to which the risk related; then the first party is taken to have caused the injury by his breach of duty, even though the existence and extent of the contribution made by the breach cannot be ascertained.

    Omitted from this statement is any reference to condition (5) in the composite question formulated in para [2] at the outset of this opinion. It was on this omission that Sir Nicolas Browne-Wilkinson V-C founded his dissenting opinion (at page 779):

    To apply the principle in McGhee v National Coal Board [1973] 1 WLR 1 to the present case would constitute an extension of that principle. In the McGhee case there was no doubt that the pursuer's dermatitis was physically caused by brick dust: the only question was whether the continued presence of such brick dust on the pursuer's skin after the time when he should have been provided with a shower caused or materially contributed to the dermatitis which he contracted. There was only one possible agent which could have caused the dermatitis, viz, brick dust, and there was no doubt that the dermatitis from which he suffered was caused by that brick dust.

    In the present case the question is different. There are a number of different agents which could have caused the RLF. Excess oxygen was one of them. The defendants failed to take reasonable precautions to prevent one of the possible causative agents (e.g. excess oxygen) from causing RLF. But no one can tell in this case whether excess oxygen did or did not cause or contribute to the RLF suffered by the plaintiff. The plaintiff's RLF may have been caused by some completely different agent or agents, e.g. hypercarbia, intraventicular haemorrhage apnoea or patent ductus arteriosus. In addition to oxygen, each of those conditions has been implicated as a possible cause of RLF. This baby suffered from each of those conditions at various times in the first two months of his life. There is no satisfactory evidence that excess oxygen is more likely than any of those other four candidates to have caused RLF in this baby. To my mind, the occurrence of RLF following a failure to take a necessary precaution to prevent excess oxygen causing RLF provides no evidence and raises no presumption that it was excess oxygen rather than one or more of the four other possible agents which caused or contributed to RLF in this case.

    The position, to my mind, is wholly different from that in the McGhee case [1973] 1 WLR 1, where there was only one candidate (brick dust) which could have caused the dermatitis, and the failure to take a precaution against brick dust causing dermatitis was followed by dermatitis caused by brick dust. In such a case, I can see the common sense, if not the logic, of holding that, in the absence of any other evidence, the failure to take the precaution caused or contributed to the dermatitis. To the extent that certain members of the House of Lords decided the question on inferences from evidence or presumptions, I do not consider that the present case falls within their reasoning. A failure to take preventative measures against one out of five possible causes is no evidence as to which of those five caused the injury.

  31. On the defendants' appeal to the House, this passage in the Vice-Chancellor's judgment was expressly approved by Lord Bridge of Harwich, who gave the only opinion, with which Lord Fraser of Tullybelton, Lord Lowry, Lord Griffiths and Lord Ackner concurred, and the appeal was allowed: [1988] AC 1074,1090-1092. It is plain, in my respectful opinion, that the House was right to allow the defendants' appeal in Wilsher, for the reasons which the Vice-Chancellor had given and which the House approved. It is one thing to treat an increase of risk as equivalent to the making of a material contribution where a single noxious agent is involved, but quite another where any one of a number of noxious agents may equally probably have caused the damage. The decision of the Court of Appeal did indeed involve an extension of the McGhee principle, as Mustill LJ recognised: [1987] QB 730, 771-772. Lord Bridge was also, as I respectfully think, right to describe the observations of Lord Wilberforce on reversal of the burden of proof (see para [22] above) as expressing a "minority opinion" (p 1087), if Lord Wilberforce was suggesting more than that the proof of an increased risk can found a prima facie case which casts an evidential burden on the defendant. But much difficulty is caused by the following passage in Lord Bridge's opinion in which, having cited the opinions of all members of the House in McGhee, he said (p 1090):

    The conclusion I draw from these passages is that McGhee v National Coal Board [1973] 1 WLR 1 laid down no new principle of law whatever. On the contrary, it affirmed the principle that the onus of proving causation lies on the pursuer or plaintiff. Adopting a robust and pragmatic approach to the undisputed primary facts of the case, the majority concluded that it was a legitimate inference of fact that the defenders' negligence had materially contributed to the pursuer's injury. The decision, in my opinion, is of no greater significance than that and to attempt to extract from it some esoteric principle which in some way modifies, as a matter of law, the nature of the burden of proof of causation which a plaintiff or pursuer must discharge once he has established a relevant breach of duty is a fruitless one.

    This is a passage to which the Court of Appeal very properly gave weight ([2002] 1 WLR 1052, 1080, para [103]), and in argument on these appeals counsel for the respondents strongly relied on it as authority for their major contention that a claimant can only succeed if he proves on the balance of probabilities that the default of the particular defendant had caused the damage of which he complains. As is apparent from the conclusions expressed in para [27] above, I cannot for my part accept this passage in Lord Bridge's opinion as accurately reflecting the effect of what the House, or a majority of the House, decided in McGhee, which remains sound authority. I am bound to conclude that this passage should no longer be treated as authoritative.

    THE WIDER JURISPRUDENCE

  32. The problem of attributing legal responsibility where a victim has suffered a legal wrong but cannot show which of several possible candidates (all in breach of duty) is the culprit who has caused him harm is one that has vexed jurists in many parts of the world for many years. As my noble and learned friend Lord Rodger of Earlsferry shows (see paras [175]-[178] below) it engaged the attention of classical Roman jurists. It is indeed a universal problem calling for some consideration by the House, however superficially, of the response to it in other jurisdictions.

  33. Professor Christian von Bar (The Common European Law of Torts, 2000, vol 2, pp 441-443) has written:

    The phenomenon of double causation is thus an insufficient argument against the but - for test. It is merely peripheral. Of greater importance are the many cases in which, although one cannot speak of a scientifically ascertainable or explicable 'cause' and 'effect', courts have awarded compensation on the basis of fault and probability. German law on medical negligence provides the example of the reduced burden of proof of causation in cases of grave treatment errors. Recent environmental legislation has reacted to the problem of scientifically uncertain causal relationships in a similar manner. The reversal of the burden of proof regarding causation is no more than a reduction of the probability required for attribution. A further development has arisen in the Netherlands regarding liability for medicines in the context of DES liability. The Hoge Raad's solution, on the basis of Art 6.99BW, of holding jointly liable to cancer sufferers the manufacturers of all those carcinogenic medicines available at the time when the victims' mothers had taken those substances is clearly unjustifiable under the conditio sine qua non rule. If the issue is seen to be whether the victim or the person posing the risk is better positioned to bear that risk, the solution is more comprehensible.

  34. In similar vein, Professor Walter van Gerven (van Gerven, Lever and Larouche: Cases, Materials and Text on National, Supranational and International Tort Law, 2000, p 441), surveying the tort law of, in particular, Germany, France and Britain, wrote:

    In many cases, it will be possible for the victim to show that he or she has suffered injury, that it has been caused by someone who must have been at fault, but the author of that fault will not be identifiable. The best that the victim will be able to achieve is to define a class of persons of which the actual tortfeasor must be a member. Strictly speaking, however, the basic conditio sine qua non test will not be met, since it cannot be said of any member of the class that the injury would not have happened 'but for' his or her conduct, given that in fact any other member could have caused the injury. Nonetheless, all the legal systems studied here have acknowledged that it would be patently unfair to deny recovery to the victim for that reason.

    At p 461, after reference to McGhee's case ([1973] 1 WLR 1) and Wilsher's case ([1988] AC 1074) van Gerven added:

    McGhee had put English law on the same path as German law, albeit with a different and arguably stronger rationale (negligent creation of risk instead of impossibility for the plaintiff to prove causation). Furthermore, it must be noted that some French legal writers are advocating that French law moves away from perte d'une chance towards a reversal of the burden of proof on the basis of the negligent creation of risk. It is unfortunate that the House of Lords retreated from McGhee at a time when laws were converging. In the end, the sole relief for the plaintiff under English law is that it suffices for the purposes of causation to show that the conduct of the defendant made a material contribution to the injury, even if it was not its sole cause.

    He concluded (at p 465):

    In certain cases, the plaintiff can show that he or she suffered injury, that it was caused by some person and that the other conditions of liability are otherwise fulfilled but for the fact that the actual tortfeasor cannot be identified among the members of a class of persons. In these cases, the strict application of the 'but for' test would result in the claim of the plaintiff being dismissed, but all systems under study here make an exception to the rules of causation to provide the plaintiff with compensation.

  35. In Germany cases of this kind have been held to be covered by the second sentence (to which emphasis has been added) of BGB §830.1 which provides:

    If several persons have caused damage by an unlawful act committed in common each is responsible for the damage. The same rule applies if it cannot be discovered which of several participants has caused the damage by his act.

    Of this provision Markesinis and Unberath (The German Law of Torts, 4th ed, 2002, p 900) states:

    §830.1, second sentence, applies the same rule to a different situation where several persons participate in a course of conduct which, though not unlawful in itself, is potentially dangerous to others. The difference between this and the previous situation lies in the fact that whereas in the former case of joint tortfeasors the loss is caused by several persons acting in consort, in the latter case only one person has caused the loss but it is difficult if not impossible to say which one has done so. (The classic illustration is that of the huntsmen who discharge their guns simultaneously and the pellets from one unidentifiable gun hit an innocent passer-by.) In this case, as well, §830 BGB adopts the same rule and makes all the participants liable to the victim for the full extent of damage.

    It is evident that this approach has been applied in Germany in a number of different situations: see Palandt, Bürgerliches Gesetzbuch, 61 ed 2002, para [830]. Thus cases decided on this basis have included personal injury caused by several individuals throwing stones in a fight; personal injury of a pedestrian who passes a construction site operated by different companies; personal injury by a New Year's Eve rocket launched by a group of individuals who were all firing rockets; liability of several hunters for the personal injury of a passer-by; liability of a houseowner for the personal injury of a pedestrian who walks on a dangerous path and is hurt on the borderline between the houseowner's land and an unfinished street owned by the local authority; and personal injury of a patient after various interrelated operations conducted by different doctors. The extent to which this approach to causation may be subject to certain restrictive conditions appears to be a matter of some uncertainty; see van Gerven, op. cit, p 446.

  36. The BGB is not alone in expressly recognising the problem of the indeterminate defendant. Article 926 of the Greek Civil Code, entitled "Damage caused by several persons" provides:

    If damage has occurred as a result of the joint action of several persons, or if several persons are concurrently responsible for the same damage, they are all jointly and severally implicated. The same applies if several persons have acted simultaneously or in succession and it is not possible to determine which person's act caused the damage.

    The second sentence of this provision would appear to cover the contingency under consideration, although the contrary has been asserted: Unification of Tort Law: Causation, ed J Spier, 2000, at p 77. A somewhat similar provision is to be found in the Austrian Civil Code:

    1302

    In such a case, if the injury is inadvertent, and it is possible to determine the portions thereof, each person is responsible only for the injuries caused by his mistake. If, however, the injury was intentional, or if the portions of the individuals in the injury cannot be determined, all are liable for one and one for all; however, the individual who has paid damages is granted the right to claim reimbursement from the others.

    A similar provision is also found in the Netherlands Civil Code (Article 6.99 BW: see para [39] below).

  37. The problem of attribution has repeatedly arisen in the context of shooting incidents, described by Professor Markesinis as "The classic illustration" in the passage quoted in para [33] above. In Litzinger v Kintzler (Cass. civ. 2e, 5 June 1957, D 1957 Jur. 493), where a group of huntsmen fired a salvo to mark the end of a deer hunt and the plaintiff was shot, the French Cour de Cassation held the whole group liable since "the whole group created a risk through its negligent conduct, and the injury to the plaintiff constituted a realisation of that risk" (see van Gerven, op cit. at pp 442-444). It appears that this reasoning (depending on a broad view of acting in concert) has been reaffirmed in later cases (ibid). In Spain, a provision of the Hunting Act 1970 provides that "In the case of hunting with weapons, if the author of the personal injury is not known, all members of the hunting party shall be jointly and severally liable." This is not a problem which is confined to Europe. In Summers v Tice 199 P 2d 1 (1948) each of two defendants at or about the same time shot at a quail and in doing so fired towards the plaintiff who was struck by shot. The Supreme Court of California referred (at p 3) to earlier authority and said:

    These cases speak of the action of defendants as being in concert as the ground of decision, yet it would seem they are straining that concept and the more reasonable basis appears in Oliver v Miles, supra. There two persons were hunting together. Both shot at some partridges and in so doing shot across the highway injuring plaintiff who was travelling on it. The court stated they were acting in concert and thus both were liable. The court then stated ....

    We think that .... each is liable for the resulting injury to the boy, although no one can say definitely who actually shot him. To hold otherwise would be to exonerate both from liability, although each was negligent, and the injury resulted from such negligence.

    [Emphasis added]

    The conclusion of the court was expressed on p 4, para [5], and p 5, para [7] of the judgment:

    When we consider the relative position of the parties and the results that would flow if plaintiff was required to pin the injury on one of the defendants only, a requirement that the burden of proof on that subject be shifted to defendants becomes manifest. They are both wrongdoers - both negligent toward plaintiff. They brought about a situation where the negligence of one of them injured the plaintiff, hence it should rest with them each to absolve himself if he can .... We have seen that for the reasons of policy discussed herein, the case is based upon the legal proposition that, under the circumstances here presented, each defendant is liable for the whole damage whether they are deemed to be acting in concert or independently.

  38. The Summers alternative liability theory was incorporated in the Restatement Second of Torts, section 433B, subdivision (3) pages 441-447, which provides:

    Where the conduct of two or more actors is tortious, and it is proved that harm has been caused to the plaintiff by only one of them, but there is uncertainty as to which one has caused it, the burden is upon each such actor to prove that he has not caused the harm.

  39. The Supreme Court of Canada confronted this situation in Cook v Lewis [1951] SCR 830, in which Cartwright J, with whom a majority agreed, said at p 842:

    I do not think it necessary to decide whether all that was said in Summers v Tice should be accepted as stating the law of British Columbia, but I am of opinion, for the reasons given in that case, that if under the circumstances of the case at bar the jury, having decided that the plaintiff was shot by either Cook or Akenhead, found themselves unable to decide which of the two shot him because in their opinion both shot negligently in his direction, both defendants should have been found liable.

  40. In a Norwegian case (see Nils Nygaard, Injury/Damage and Responsibility, 2000, pp 342-343) F was a passenger on a motor scooter and was injured in an accident caused either by a cable stretched across the street by a construction company or by the motor scooter falling onto him or as a result of collision with a truck, or by any combination of these factors. In giving judgment (RG 1969 p 285 at 293) the Norwegian court said:

    As stated in the beforementioned conclusions made by experts, they could not conclude whether the situation that resulted in crushed bones in F's left hip region, was a result of falling on the cobble stones in the street or from the truck's front tyre, that ended up on top of F's left hip region. It is possible that the injuries were partially a result of the fall and being hit by the truck. We cannot say anything definite about this. The court finds that it cannot conclude whether it is the fall or being hit by the truck or a combination of both these factors that caused the injury. After a collective evaluation of the whole event the court finds that A (construction company), the scooter and the truck each have a part in F getting injured and each of them must naturally be seen as adequate cause of injury.

    This decisions bears a strong resemblance to that reached by the English Court of Appeal in Fitzgerald v Lane [1987] QB 781, decided after the decision of the Court of Appeal in Wilsher [1987] QB 730 but before the decision of the House of Lords ([1988] AC 1074).

  41. Increasingly in recent years, the problem of attribution has arisen in more complicated factual situations. Sindell v Abbott Laboratories 26 Cal. 3d 588 (1980) was a class action for personal injuries said to have resulted from pre-natal exposure to the anti-miscarriage drug diethylstilbestrol (DES) which had been manufactured by one of a potentially large number of defendants. The plaintiff could not identify which particular defendant had manufactured the drug responsible for her injuries. However, her complaint alleged that the defendants were jointly and individually negligent in that they had manufactured, marketed and promoted DES as a safe drug to prevent miscarriage without adequate testing or warning of its dangerous side effects; that they had collaborated in their marketing methods, promotion and testing of the drug; that they had relied on each others' test results; that they had adhered to an industry-wide safety standard; and that they had produced the drug from a common and mutually agreed generic formula. The court distinguished Summers on the basis that in that case all the parties who were or could have been responsible for the harm to the plaintiff were joined as defendants, whereas in Sindell there were approximately 200 drug companies which had made DES, any of which might have manufactured the injury-producing drug. The court held that it would be unfair, in such circumstances, to require each defendant to exonerate itself. Further, it said that there might be a substantial likelihood that none of the five defendants joined in the action had made the DES which caused the injury, and that the offending producer, not named, would escape liability. The court surmounted this problem by adapting the Summers rule so as to apportion liability on the basis of the defendant's market share. See pp 593-595, 602-603, 604-605 and 612-613. A very similar case concerning the same drug arose in the Netherlands in B v Bayer Nederland BV (Hoge Raad, 9 October 1992, NJ 1994, 535) which turned on Article 6.99 BW. That provision is in these terms:

    Where the damage may have resulted from two or more events for each of which a different person is liable, and where it has been determined that the damage has arisen from at least one of these events, the obligation to repair the damage rests upon each of these persons, unless he proves that the damage is not the result of the event for which he himself is liable.

    In para 3.7.1 of its judgment the Hoge Raad held:

    The facts of the present case fall under that provision, on the assumption that it can be proved that [i] each of the firms which put DES in circulation in the relevant period was at fault in so doing and could thus be held liable, [ii] the total injury of each victim could have been caused by any of these 'events' - i.e. putting DES in circulation - and [iii] the injury occurred because of at least one of these 'events' ....

    [Article 6.99BW] aims to remove the unfairness arising from the fact that the victim must bear his or her own damage because he or she cannot prove whose action caused his or her harm. The victims in the present case are faced with such an evidentiary difficulty ....

    In para 3.7.5 of its judgment the court said:

    It is sufficient for each DES daughter to establish .... in relation to each of the pharmaceutical companies:

    (i)

    that the pharmaceutical company in question put DES in circulation during the relevant period and can therefore be found liable because it committed a fault;

    (ii)

    that another or several other producers - regardless of whether they are parties to the proceedings or not - also put DES in circulation during the relevant period and can therefore also be found liable because it (they) committed a fault; and

    (iii)

    that she suffered injury that resulted from the use of DES, but that it is no longer possible to determine from which producer the DES originated.

    In principle the burden of proof on these issues rests on the DES daughter concerned.

    See van Gerven, op cit, at pages 447-448.

  42. Cases decided in the High Court of Australia do not disclose a clear ratio on which the appellants were able to rely before the House, although they drew attention to dicta which were helpful to them. For example, in Chappel v Hart (1998) 195 CLR 232 at 244, para [27], McHugh J said, in a passage reminiscent of McGhee, although without referring to that case:

    Before the defendant will be held responsible for the plaintiff's injury, the plaintiff must prove that the defendant's conduct materially contributed to the plaintiff suffering that injury. In the absence of a statute or undertaking to the contrary, therefore, it would seem logical to hold a person causally liable for a wrongful act or omission only when it increases the risk of injury to another person. If a wrongful act or omission results in an increased risk of injury to the plaintiff and that risk eventuates, the defendant's conduct has materially contributed to the injury that the plaintiff suffers whether or not other factors also contributed to that injury occurring. If, however, the defendant's conduct does not increase the risk of injury to the plaintiff, the defendant cannot be said to have materially contributed to the injury suffered by the plaintiff. That being so, whether the claim is in contract or tort, the fact that the risk eventuated at a particular time or place by reason of the conduct of the defendant does not itself materially contribute to the plaintiff's injury unless the fact of that particular time or place increased the risk of the injury occurring.

  43. In that case McHugh J dissented on the facts but in Naxakis v Western General Hospital (1999) 197 CLR 269 both Gaudron J (at p 279, para [31]) and Callinan J (at p 312, para [127]) quoted what he had said with approval. In Canada, Sopinka J, speaking for the Supreme Court in Snell v Farrell [1990] 2 SCR 311 said, at pp 326-327:

    I have examined the alternatives arising out of the McGhee case. They were that the plaintiff simply prove that the defendant created a risk that the injury which occurred would occur. Or, what amounts to the same thing, that the defendant has the burden of disproving causation. If I were convinced that defendants who have a substantial connection to the injury were escaping liability because plaintiffs cannot prove causation under currently applied principles, I would not hesitate to adopt one of these alternatives. In my opinion, however, properly applied, the principles relating to causation are adequate to the task. Adoption of either of the proposed alternatives would have the effect of compensating plaintiffs where a substantial connection between the injury and the defendant's conduct is absent. Reversing the burden of proof may be justified where two defendants negligently fire in the direction of the plaintiff and then by their tortious conduct destroy the means of proof at his disposal. In such a case it is clear that the injury was not caused by neutral conduct. It is quite a different matter to compensate a plaintiff by reversing the burden of proof for an injury that may very well be due to factors unconnected to the defendant and not the fault of anyone.

  44. Sopinka J suggested, at p 328, that dissatisfaction with the traditional approach to causation stemmed to a large extent from its too rigid application by the courts in many cases, and that causation need not be determined by scientific precision. Despite this judgment the Manitoba Court of Appeal, in Webster v Chapman (1997) 155 DLR (4th) 82 held, relying on McGhee, that no distinction should be made in that case between materially increasing the risk of damage and materially contributing to the damage.

  45. There is a small but important body of authority on the problem of attribution in mesothelioma cases where the plaintiff has been exposed to asbestos during employment by more than one employer. Bendix Mintex Pty Ltd v Barnes (1997) 42 NSWLR 307 was such a case. A majority of the Court of Appeal of New South Wales held against the plaintiff on the causation issue, relying on Wilsher [1988] AC 1074 among much other authority. Stein JA dissented, citing with approval the following passage from the judgment of King CJ in Birkholtz v R J Gilbertson Pty Ltd (1985) 38 SASR 121 at 130:

    .... the law's view of causation is less concerned with logical and philosophical considerations than with the need to produce a just result to the parties involved. Where a defendant is under a legal duty to take precautions to protect the plaintiff from the risk of contracting disease, and, by omitting those precautions he substantially increases the risk of the plaintiff contracting that disease, the law treats that increase in risk as a sufficient basis, in the absence of evidence showing how the infection occurred, for an inference that the omission of the precautions materially contributed to the contracting of the disease. Justice requires such an approach to the problem of causation and it is the approach which was taken by the House of Lords in McGhee v National Coal Board [1973] 1 WLR 1.

  46. The majority decision in Bendix was followed in Wallaby Grip (BAE) Pty Ltd v Macleay Area Health Service (1998) 17 NSWCCR 355. A different result was reached in E M Baldwin & Son Pty Ltd v Plane (1999) Aust Torts Reports 81-499, but on different medical evidence. A different view of the law was also expressed in Rutherford v Owens-Illinois Inc 67 Cal. Rptr. 2d 16 (1997). In a judgment with which the Chief Justice and all save one member of the Supreme Court of California concurred, Baxter J observed (at p 19):

    Proof of causation in such cases will always present inherent practical difficulties, given the long latency period of asbestos-related disease, and the occupational settings that commonly exposed the worker to multiple forms and brands of asbestos products with varying degrees of toxicity. In general, however, no insuperable barriers prevent an asbestos-related cancer plaintiff from demonstrating that exposure to the defendant's asbestos products was, in reasonable medical probability, a substantial factor in causing or contributing to his risk of developing cancer. We conclude that plaintiffs are required to prove no more than this. In particular, they need not prove with medical exactitude that fibers from a particular defendant's asbestos-containing products were those, or among those, that actually began the cellular process of malignancy.

    Baxter J reviewed earlier cases such as Summers 119 P 2d 1 (1948) and Sindell 26 Cal. 3d 588 (1980) and specifically addressed the factual possibility that a mesothelioma may be caused by inhalation of a single fibre, acknowledging (at pp 30-31) that the single fibre theory raised an apparently unanswerable question: "which particular fibre or fibres actually caused the cancer to begin forming[?]". He observed (at p 31) that plaintiffs could not be expected to prove the scientifically unknown details of carcinogenesis or trace the unknowable path of a given asbestos fibre. For reasons given very clearly but at some length Baxter J rejected a burden-shifting approach to cases of this kind. At p 36 the judgment concluded:

    In the context of a cause of action for asbestos-related latent injuries, the plaintiff must first establish some threshold exposure to the defendant's defective asbestos-containing products, and must further establish in reasonable medical probability that a particular exposure or series of exposures was a 'legal cause' of his injury, ie, a substantial factor in bringing about the injury. In an asbestos-related cancer case, the plaintiff need not prove that fibers from the defendant's product were the ones, or among the ones, that actually began the process of malignant cellular growth. Instead, the plaintiff may meet the burden of proving that exposure to defendant's product was a substantial factor causing the illness by showing that in reasonable medical probability it was a substantial factor contributing to the plaintiff's or decedent's risk of developing cancer.

    The dissent of Mosk J related solely to the court's decision on shifting of the burden: he considered (p 38) that the decision of the majority would deprive numerous plaintiffs suffering from latent diseases caused by exposure to asbestos in the workplace from recovering full compensation.

  47. This survey shows, as would be expected, that though the problem underlying cases such as the present is universal the response to it is not. Hence the plethora of decisions given in different factual contexts. Hence also the intensity of academic discussion, exemplified by the articles of the late Professor Fleming ("Probabilistic Causation in Tort Law" 68 Canadian Bar Review, No 4, December 1989, 661) and Professor Robertson ("The Common Sense of Cause in Fact", 75 Tex L Rev (1996-1997), 1765). In some jurisdictions, it appears, the plaintiff would fail altogether on causation grounds, as the Court of Appeal held that the present appellants did. Italy, South Africa and Switzerland may be examples (see Unification of Tort Law: Causation, ed J Spier, 2000 at pp 90, 102 and 120). But it appears that in most of the jurisdictions considered the problem of attribution would not, on facts such as those of the present cases, be a fatal objection to a plaintiff's claim. Whether by treating an increase in risk as equivalent to a material contribution, or by putting a burden on the defendant, or by enlarging the ordinary approach to acting in concert, or on more general grounds influenced by policy considerations, most jurisdictions would, it seems, afford a remedy to the plaintiff. Development of the law in this country cannot of course depend on a head-count of decisions and codes adopted in other countries around the world, often against a background of different rules and traditions. The law must be developed coherently, in accordance with principle, so as to serve, even-handedly, the ends of justice. If, however, a decision is given in this country which offends one's basic sense of justice, and if consideration of international sources suggests that a different and more acceptable decision would be given in most other jurisdictions, whatever their legal tradition, this must prompt anxious review of the decision in question. In a shrinking world (in which the employees of asbestos companies may work for those companies in any one or more of several countries) there must be some virtue in uniformity of outcome whatever the diversity of approach in reaching that outcome.

    POLICY

  48. The present appeals raise an obvious and inescapable clash of policy considerations. On the one hand are the considerations powerfully put by the Court of Appeal ([2002] 1 WLR 1052 at 1080, para [103]) which considered the claimants' argument to be not only illogical but

    also susceptible of unjust results. It may impose liability for the whole of an insidious disease on an employer with whom the claimant was employed for quite a short time in a long working life, when the claimant is wholly unable to prove on the balance of probabilities that that period of employment had any causative relationship with the inception of the disease. This is far too weighty an edifice to build on the slender foundations of McGhee v National Coal Board [1973] 1WLR 1, and Lord Bridge has told us in Wilsher v Essex Area Health Authority [1988] AC 1074 that McGhee established no new principle of law at all. If we were to accede to the claimants' arguments, we would be distorting the law to accommodate the exigencies of a very hard case. We would be yielding to a contention that all those who have suffered injury after being exposed to a risk of that injury from which someone else should have protected them should be able to recover compensation even when they are quite unable to prove who was the culprit. In a quite different context Lord Steyn has recently said in Frost v Chief Constable of Yorkshire [1999] 2 AC 455, 491 that our tort system sometimes results in imperfect justice, but it is the best the common law can do.

  49. The Court of Appeal had in mind that in each of the cases discussed in paras [14]-[21] above (Wardlaw, Nicholson, Gardiner, McGhee) there was only one employer involved. Thus there was a risk that the defendant might be held liable for acts for which he should not be held legally liable but no risk that he would be held liable for damage which (whether legally liable or not) he had not caused. The crux of cases such as the present, if the appellants' argument is upheld, is that an employer may be held liable for damage he has not caused. The risk is the greater where all the employers potentially liable are not before the court. This is so on the facts of each of the three appeals before the House, and is always likely to be so given the long latency of this condition and the likelihood that some employers potentially liable will have gone out of business or disappeared during that period. It can properly be said to be unjust to impose liability on a party who has not been shown, even on a balance of probabilities, to have caused the damage complained of. On the other hand, there is a strong policy argument in favour of compensating those who have suffered grave harm, at the expense of their employers who owed them a duty to protect them against that very harm and failed to do so, when the harm can only have been caused by breach of that duty and when science does not permit the victim accurately to attribute, as between several employers, the precise responsibility for the harm he has suffered. I am of opinion that such injustice as may be involved in imposing liability on a duty-breaking employer in these circumstances is heavily outweighed by the injustice of denying redress to a victim. Were the law otherwise, an employer exposing his employee to asbestos dust could obtain complete immunity against mesothelioma (but not asbestosis) claims by employing only those who had previously been exposed to excessive quantities of asbestos dust. Such a result would reflect no credit on the law. It seems to me, as it did to Lord Wilberforce in McGhee [1973] 1 WLR 1 at 7, that

    the employers should be liable for an injury, squarely within the risk which they created and that they, not the pursuer, should suffer the consequence of the impossibility, foreseeably inherent in the nature of his injury, of segregating the precise consequence of their default.

    CONCLUSION

  50. To the question posed in para [2] of this opinion I would answer that where conditions (1) - (6) are satisfied C is entitled to recover against both A and B. That conclusion is in my opinion consistent with principle, and also with authority (properly understood). Where those conditions are satisfied, it seems to me just and in accordance with common sense to treat the conduct of A and B in exposing C to a risk to which he should not have been exposed as making a material contribution to the contracting by C of a condition against which it was the duty of A and B to protect him. I consider that this conclusion is fortified by the wider jurisprudence reviewed above. Policy considerations weigh in favour of such a conclusion. It is a conclusion which follows even if either A or B is not before the court. It was not suggested in argument that C's entitlement against either A or B should be for any sum less than the full compensation to which C is entitled, although A and B could of course seek contribution against each other or any other employer liable in respect of the same damage in the ordinary way. No argument on apportionment was addressed to the House. I would in conclusion emphasise that my opinion is directed to cases in which each of the conditions specified in (1) - (6) of para [2] above is satisfied and to no other case. It would be unrealistic to suppose that the principle here affirmed will not over time be the subject of incremental and analogical development. Cases seeking to develop the principle must be decided when and as they arise. For the present, I think it unwise to decide more than is necessary to resolve these three appeals which, for all the foregoing reasons, I concluded should be allowed.

  51. For reasons given above, I cannot accept the view (considered in the opinion of my noble and learned friend Lord Hutton) that the decision in McGhee was based on the drawing of a factual inference. Nor, in my opinion, was the decision based on the drawing of a legal inference. Whether, in certain limited and specific circumstances, a legal inference is drawn or a different legal approach is taken to the proof of causation, may not make very much practical difference. But Lord Wilberforce, in one of the passages of his opinion in McGhee quoted in para [22] above, wisely deprecated resort to fictions and it seems to me preferable, in the interests of transparency, that the courts' response to the special problem presented by cases such as these should be stated explicitly. I prefer to recognise that the ordinary approach to proof of causation is varied than to resort to the drawing of legal inferences inconsistent with the proven facts.

    Lord Nicholls of Birkenhead

    My Lords,

  52. I have no hesitation in agreeing with all your Lordships that these appeals should be allowed. Any other outcome would be deeply offensive to instinctive notions of what justice requires and fairness demands. The real difficulty lies is elucidating in sufficiently specific terms the principle being applied in reaching this conclusion. To be acceptable the law must be coherent. It must be principled. The basis on which one case, or one type of case, is distinguished from another should be transparent and capable of identification. When a decision departs from principles normally applied, the basis for doing so must be rational and justifiable if the decision is to avoid the reproach that hard cases make bad law. I turn therefore to consider the departure from the normal, and the basis of that departure, in the present appeals.

  53. In the normal way, in order to recover damages for negligence, a plaintiff must prove that but for the defendant's wrongful conduct he would not have sustained the harm or loss in question. He must establish at least this degree of causal connection between his damage and the defendant's conduct before the defendant will be held responsible for the damage.

  54. Exceptionally this is not so. In some circumstances a lesser degree of causal connection may suffice. This sometimes occurs where the damage flowed from one or other of two alternative causes. Take the well-known example where two hunters, acting independently of each other, fire their guns carelessly in a wood, and a pellet from one of the guns injures an innocent passer-by. No one knows, and the plaintiff is unable to prove, from which gun the pellet came. Should the law of negligence leave the plaintiff remediless, and allow both hunters to go away scot-free, even though one of them must have fired the injurious pellet?

  55. Not surprisingly, the courts have declined to reach such an unjust decision: see Summers v Tice (1948) 199 P 2d 1, a decision of the Supreme Court of California, and Cook v Lewis [1951] SCR 830, a decision of the Supreme Court of Canada. As between the plaintiff and the two hunters, the evidential difficulty arising from the impossibility of identifying the gun which fired the crucial pellet should redound upon the negligent hunters, not the blameless plaintiff. The unattractive consequence, that one of the hunters will be held liable for an injury he did not in fact inflict, is outweighed by the even less attractive alternative, that the innocent plaintiff should receive no recompense even though one of the negligent hunters injured him. It is this balance ('.. outweighed by ..') which justifies a relaxation in the standard of causation required. Insistence on the normal standard of causation would work an injustice. Hunting in a careless manner and thereby creating a risk of injury to others, followed by injury to another person, is regarded by the law as sufficient causal connection in the circumstances to found responsibility.

  56. This balancing exercise involves a value judgment. This is not at variance with basic principles in this area of the law. The extent to which the law requires a defendant to assume responsibility for loss following upon his wrongful conduct always involves a value judgment. The law habitually limits the extent of the damage for which a defendant is held responsible, even when the damage passes the threshold 'but for' test. The converse is also true. On occasions the threshold 'but for' test of causal connection may be over-exclusionary. Where justice so requires, the threshold itself may be lowered. In this way the scope of a defendant's liability may be extended. The circumstances where this is appropriate will be exceptional, because of the adverse consequences which the lowering of the threshold will have for a defendant. He will be held responsible for a loss the plaintiff might have suffered even if the defendant had not been involved at all. To impose liability on a defendant in such circumstances normally runs counter to ordinary perceptions of responsibility. Normally this is unacceptable. But there are circumstances, of which the two hunters' case is an example, where this unattractiveness is outweighed by leaving the plaintiff without a remedy.

  57. The present appeals are another example of such circumstances, where good policy reasons exist for departing from the usual threshold 'but for' test of causal connection. Inhalation of asbestos dust carries a risk of mesothelioma. That is one of the very risks from which an employer's duty of care is intended to protect employees. Tragically, each claimant acquired this fatal disease from wrongful exposure to asbestos dust in the course of his employment. A former employee's inability to identify which particular period of wrongful exposure brought about the onset of his disease ought not, in all justice, to preclude recovery of compensation.

  58. So long as it was not insignificant, each employer's wrongful exposure of its employee to asbestos dust and, hence, to the risk of contracting mesothelioma, should be regarded by the law as a sufficient degree of causal connection. This is sufficient to justify requiring the employer to assume responsibility for causing or materially contributing to the onset of the mesothelioma when, in the present state of medical knowledge, no more exact causal connection is ever capable of being established. Given the present state of medical science, this outcome may cast responsibility on a defendant whose exposure of a claimant to the risk of contracting the disease had in fact no causative effect. But the unattractiveness of casting the net of responsibility as widely as this is far outweighed by the unattractiveness of the alternative outcome.

  59. I need hardly add that considerable restraint is called for in any relaxation of the threshold 'but for' test of causal connection. The principle applied on these appeals is emphatically not intended to lead to such a relaxation whenever a plaintiff has difficulty, perhaps understandable difficulty, in discharging the burden of proof resting on him. Unless closely confined in its application this principle could become a source of injustice to defendants. There must be good reason for departing from the normal threshold 'but for' test. The reason must be sufficiently weighty to justify depriving the defendant of the protection this test normally and rightly affords him, and it must be plain and obvious that this is so. Policy questions will loom large when a court has to decide whether the difficulties of proof confronting the plaintiff justify taking this exceptional course. It is impossible to be more specific.

  60. I should comment briefly on the much discussed case of McGhee v National Coal Board [1973] 1 WLR 1. As I understand it, the decision of your Lordships' House is an example of the application of the approach discussed above. In the circumstances of that case the House departed from the usual threshold 'but for' test of causal connection and treated a lesser degree of causal connection as sufficient. The novelty in the decision lay in the adoption of this approach in this country and, further, in the type of claim to which this approach was applied: there, as with the present appeals, the field of industrial disease. Given the medical evidence in McGhee, it was not open to the House, however robustly inclined, to draw an inference that the employer's negligence had in fact caused or materially contributed to the onset of the dermatitis in the sense that, but for that negligence, the dermatitis would not have occurred. Instead, a less stringent causal connection was regarded as sufficient. It was enough that the employer had materially increased the risk of harm to the employee: see Lord Reid, at page 5, Lord Wilberforce, at page 7, Lord Simon of Glaisdale, at page 8 ('failure to take steps which would bring about a material reduction of the risk'), and Lord Salmon, at pages 12-13.

  61. In an area of the law already afflicted with linguistic ambiguity I myself would not describe this process of legal reasoning as a 'legal inference' or an 'inference of causation'. This phraseology tends to obscure the fact that when applying the principle described above the court is not, by a process of inference, concluding that the ordinary 'but for' standard of causation is satisfied. Instead, the court is applying a different and less stringent test. It were best if this were recognised openly.

    Lord Hoffmann

    My Lords,

  62. In one of the three appeals before your Lordships, the appellant Mr Matthews is suffering from a cancer (mesothelioma) caused by exposure to asbestos. Between 1973 and 1981 he was exposed to substantial quantities of asbestos dust in the course of successive employments by the two respondents, Associated Portland Cement and British Uralite. It is accepted that both employers were in breach of duty - the link between asbestos and cancer was well known by the mid 1960s - and that both exposures contributed substantially to the risk that he would contract the disease. But the precise mechanism by which asbestos causes cancer is unknown. It may be caused by the mutation of a single cell caused by a single asbestos fibre. At any rate, it is impossible to say that it was more likely to have been caused by the exposure to asbestos during employment with the one respondent rather than the other. And on that ground the Court of Appeal has held that Mr Matthews' claim must fail. It was said to be a rule of law that in order to succeed against either respondent, he must prove that, but for its breach of duty, he would not have contracted the disease. It applied the same rule to the other two appeals before your Lordships, which raise a similar point.

  63. My Lords, in my opinion the rule applied by the Court of Appeal is not a correct statement of the causal connection between breach of duty and damage which the law requires in a case such as this. I think it is sufficient, both on principle and authority, that the breach of duty contributed substantially to the risk that the claimant would contract the disease.

  64. I shall first consider the question in principle. It is axiomatic that the law will not impose liability to pay compensation for damage unless there is a relevant causal connection between the damage and the defendant's tort, breach of contract or statutory duty. But what amounts to a relevant causal connection?

  65. Everyone agrees that there is no scientific or philosophical touchstone for determining the relevant causal connection in any particular case. The relevance of a causal connection depends upon the purpose of the inquiry. In the present case, the House is required to say what should be the relevant causal connection for breach of a duty to protect an employee against the risk of contracting (among other things) mesothelioma by exposure to asbestos.

  66. It is frequently said that causation is a question of fact or a matter of common sense. Both of these propositions are true but they need to be analysed with some care in order to avoid confusion.

  67. First, in what sense is causation a question of fact? In order to describe something as a question of fact, it is necessary to be able to identify the question. For example, whether someone was negligent or not is a question of fact. What is the question? It is whether he failed to take reasonable care to avoid such damage as a reasonable man would have foreseen might result from his conduct. That question is formulated by the law. It is the law which says that failure to take reasonable care gives rise to liability. And the question is then answered by applying the standard of conduct prescribed by the law to the facts.

  68. The same is true of causation. The question of fact is whether the causal requirements which the law lays down for that particular liability have been satisfied. But those requirements exist by virtue of rules of law. Before one can answer the question of fact, one must first formulate the question. This involves deciding what, in the circumstances of the particular case, the law's requirements are. Unless one pays attention to the need to determine this preliminary question, the proposition that causation is a question of fact may be misleading. It may suggest that one somehow knows instinctively what the question is or that the question is always the same. As we shall see, this is not the case. The causal requirements for liability often vary, sometimes quite subtly, from case to case. And since the causal requirements for liability are always a matter of law, these variations represent legal differences, driven by the recognition that the just solution to different kinds of case may require different causal requirement rules.

  69. Then there is the role of common sense. Of course the causal requirements for liability are normally framed in accordance with common sense. But there is sometimes a tendency to appeal to common sense in order to avoid having to explain one's reasons. It suggests that causal requirements are a matter of incommunicable judicial instinct. I do not think that this is right. It should be possible to give reasons why one form of causal relationship will do in one situation but not in another.

  70. In my opinion, the essential point is that the causal requirements are just as much part of the legal conditions for liability as the rules which prescribe the kind of conduct which attracts liability or the rules which limit the scope of that liability. If I may repeat what I have said on another occasion, one is never simply liable, one is always liable for something - to make compensation for damage, the nature and extent of which is delimited by the law. The rules which delimit what one is liable for may consist of causal requirements or may be rules unrelated to causation, such as the foreseebility requirements in the rule in Hadley v Baxendale (1854) 9 Exch 341. But in either case they are rules of law, part and parcel of the conditions of liability. Once it is appreciated that the rules laying down causal requirements are not autonomous expressions of some form of logic or judicial instinct but creatures of the law, part of the conditions of liability, it is possible to explain their content on the grounds of fairness and justice in exactly the same way as the other conditions of liability.

  71. In the law of negligence, for example, it has long been recognised that the imposition of a duty of care in respect of particular conduct depends upon whether it is just and reasonable to impose it. Over vast areas of conduct one can generalise about the circumstances in which it will be considered just and reasonable to impose a duty of care: that is a consequence of Donaghue v Stevenson [1932] AC 562. But there are still situations in which Lord Atkin's generalisation cannot fairly be applied and in which it is necessary to return to the underlying principle and inquire whether it would be just and reasonable to impose liability and what its nature and extent should be: see Caparo Industries plc v Dickman [1990] 2 AC 605.

  72. The same is true of causation. The concepts of fairness, justice and reason underlie the rules which state the causal requirements of liability for a particular form of conduct (or non-causal limits on that liability) just as much as they underlie the rules which determine that conduct to be tortious. And the two are inextricably linked together: the purpose of the causal requirement rules is to produce a just result by delimiting the scope of liability in a way which relates to the reasons why liability for the conduct in question exists in the first place.

  73. Across most grounds of liability, whether in tort, contract or by statute, it is possible to generalise about causal requirements. These generalisations are explored in detail by Hart and Honoré (Causation in the Law, 2nd ed 1985). They represent what in ordinary life would normally be regarded as the reasonable limits for attributing blame or responsibility for harm: for example, that the defendant's conduct was a necessary condition for the occurrence of the harm (the "but for" test), that it was not caused by the informed and voluntary act of another responsible human being and so on. To that extent, these causal requirements are based upon common sense. But, as Hart and Honoré also point out, there are situations in which these generalisations would fail to give effect to the reasons why it was thought just and reasonable to impose liability. For example, if it is thought just and reasonable to impose a duty to take care to protect someone against harm caused by the informed and voluntary act of another responsible human being, it would be absurd to retain a causal requirement that the harm should not have been so caused. An extreme case of this kind was Reeves v Commissioner of Police of the Metropolis [2000] 1 AC 360, in which the defendant accepted that in the circumstances of the case, he owed a duty to take reasonable care to prevent a responsible human being from causing injury to himself. Your Lordships decided that in those circumstances it would be contradictory to hold that the causal requirements of the tort excluded liability for harm so caused. Thus the causal requirements are always adapted to conform to the grounds upon which liability is imposed. Again, it may be said that this is no more than common sense. But it is capable of rational explanation.

  74. The same link between the grounds of liability and the causal requirements can be seen in cases of statutory liability. Sometimes the causal requirements are expressly stated; if not, the courts will construe the statute as requiring the causal connection which best gives effect to its policy. In Environment Agency v Empress Car Co. (Abertillery) Ltd. [1999] 2 AC 22 this House decided that the causal requirements of a statutory duty not to "cause ... polluting matter … to enter any controlled waters" did not exclude liability in cases in which the immediate cause of the pollution was the deliberate act of a responsible third party. It based this conclusion on the policy of the statute to impose a strict liability for the protection of the environment.

  75. My Lords, even the much derided "last opportunity rule" which was applied in cases of contributory negligence before the Law Reform (Contributory Negligence) Act 1945 was in my opinion a perfectly rational attempt by the courts to fashion the causal requirements of the tort of negligence so as best to achieve a just and fair result in a situation in which the law insisted that any contributory negligence was a complete bar to recovery. To say that it was illogical or inelegant seems to me neither here nor there. If it is recognised that the causal requirements are a matter of law and that the last opportunity rule was a development of the common law to mitigate the rigour of the contributory negligence rule, there was nothing illogical about it. As for inelegance, it was the best which could be done in the circumstances.

  76. The problem in this appeal is to formulate a just and fair rule. Clearly the rule must be based upon principle. However deserving the claimants may be, your Lordships are not exercising a discretion to adapt causal requirements to the individual case. That does not mean, however, that it must be a principle so broad that it takes no account of significant differences which affect whether it is fair and just to impose liability.

  77. What are the significant features of the present case?

    • First, we are dealing with a duty specifically intended to protect employees against being unnecessarily exposed to the risk of (among other things) a particular disease.

    • Secondly, the duty is one intended to create a civil right to compensation for injury relevantly connected with its breach.

    • Thirdly, it is established that the greater the exposure to asbestos, the greater the risk of contracting that disease.

    • Fourthly, except in the case in which there has been only one significant exposure to asbestos, medical science cannot prove whose asbestos is more likely than not to have produced the cell mutation which caused the disease.

    • Fifthly, the employee has contracted the disease against which he should have been protected.

  78. In these circumstances, a rule requiring proof of a link between the defendant's asbestos and the claimant's disease would, with the arbitrary exception of single-employer cases, empty the duty of content. If liability depends upon proof that the conduct of the defendant was a necessary condition of the injury, it cannot effectively exist. It is however open to your Lordships to formulate a different causal requirement in this class of case. The Court of Appeal was in my opinion wrong to say that in the absence of a proven link between the defendant's asbestos and the disease, there was no "causative relationship" whatever between the defendant's conduct and the disease. It depends entirely upon the level at which the causal relationship is described. To say, for example, that the cause of Mr Matthews' cancer was his significant exposure to asbestos during two employments over a period of eight years, without being able to identify the day upon which he inhaled the fatal fibre, is a meaningful causal statement. The medical evidence shows that it is the only kind of causal statement about the disease which, in the present state of knowledge, a scientist would regard as possible. There is no a priori reason, no rule of logic, which prevents the law from treating it as sufficient to satisfy the causal requirements of the law of negligence. The question is whether your Lordships think such a rule would be just and reasonable and whether the class of cases to which it applies can be sufficiently clearly defined.

  79. So the question of principle is this: in cases which exhibit the five features I have mentioned, which rule would be more in accordance with justice and the policy of common law and statute to protect employees against the risk of contracting asbestos-related diseases? One which makes an employer in breach of his duty liable for the claimant's injury because he created a significant risk to his health, despite the fact that the physical cause of the injury may have been created by someone else? Or a rule which means that unless he was subjected to risk by the breach of duty of a single employer, the employee can never have a remedy? My Lords, as between the employer in breach of duty and the employee who has lost his life in consequence of a period of exposure to risk to which that employer has contributed, I think it would be both inconsistent with the policy of the law imposing the duty and morally wrong for your Lordships to impose causal requirements which exclude liability.

  80. My Lords, I turn from principle to authority. The case which most closely resembles the present is McGhee v National Coal Board [1973] 1 WLR 1, which my noble and learned friend Lord Bingham of Cornhill has analysed in some detail.

    • There too, the employer was under a duty (to provide washing facilities) specifically intended to protect employees against being unnecessarily exposed to the risk of (among other things) a particular disease, namely dermatitis. 

    • Secondly, the duty was one intended to create a civil right to compensation for injury relevantly connected with its breach. 

    • Thirdly, it was established that the longer the workman exerted himself while particles of dust adhered to his skin, the greater was the risk of his contracting dermatitis. 

    • Fourthly, the mechanism by which dust caused the disease was unknown, so that medical science was unable to prove whether the particular dust abrasions which caused the dermatitis were more likely than not to have occurred before or after the dust would have been removed if washing facilities had been provided. All that could be said was that the absence of facilities added materially to the risk that he would contract the disease. 

    • Fifthly, the employee contracted the disease against which he should have been protected.

  81. My Lords, in these circumstances, which in my opinion reproduce the essential features of the present case, the House decided that materially increasing the risk that the disease would occur was sufficient to satisfy the causal requirements for liability. It is true that Lord Wilberforce spoke of reversing the burden of proof and imposing liability unless the employer could prove that washing would have made no difference. But I respectfully think that it is artificial to treat the employer as having a burden of proof in a case in which ex hypothesi the state of medical knowledge is such that the burden cannot be discharged. There are also passages in the speeches which suggest that materially increasing the risk of disease is being treated as equivalent to materially contributing to the injury - making the illness worse than it would otherwise have been. But this, as Lord Wilberforce pointed out (at p. 7), is precisely what the doctors did not say. They refused to say that it was more likely than not that the absence of washing facilities had had any effect at all. So when some members of the House said that in the circumstances there was no distinction between materially increasing the risk of disease and materially contributing to the disease, what I think they meant was that, in the particular circumstances, a breach of duty which materially increased the risk should be treated as if it had materially contributed to the disease. I would respecfully prefer not resort to legal fictions and to say that the House treated a material increase in risk as sufficient in the circumstances to satisfy the causal requirements for liability. That this was the effect of the decision seems to me inescapable.

  82. The grounds upon which the House was willing to formulate a special causal requirements rule in McGhee seem to me equally applicable in this case. So Lord Wilberforce said (at p. 6):

    If one asks which of the parties, the workman or the employers, should suffer from this inherent evidential difficulty [i.e. the absence of knowledge of the mechanism by which dust caused the disease], the answer as a matter of policy or justice should be that it is the creator of the risk who, ex hypothesi must be taken to have foreseen the possibility of damage, who should bear its consequences.

    Lord Simon of Glaisdale said (at p. 9):

    To hold otherwise would mean that the respondents were under a legal duty which they could, in the present state of medical knowledge, with impunity ignore.

    Lord Salmon said (at p. 12):

    It would mean that in the present state of medical knowledge and in circumstances such as these (which are by no means uncommon) an employer would be permitted by the law to disregard with impunity his duty to take reasonable care for the safety of his employees.

  83. I therefore regard McGhee as a powerful support for saying that when the five factors I have mentioned are present, the law should treat a material increase in risk as sufficient to satisfy the causal requirements for liability. The only difficulty lies in the way McGhee was explained in Wilsher v Essex Area Health Authority [1988] AC 1074. The latter was not a case in which the five factors were present. It was an action for clinical negligence in which it was alleged that giving a premature baby excessive oxygen had caused retrolental fibroplasia, resulting in blindness. The evidence was that the fibroplasia could have been caused in a number of different ways including excessive oxygen but the judge had made no finding that the oxygen was more likely than not to have been the cause. The Court of Appeal [1987] QB 730 held that the health authority was nevertheless liable because even if the excessive oxygen could not be shown to have caused the injury, it materially increased the risk of the injury happening.

  84. The Court of Appeal reached this conclusion by treating the causal requirement rule applied in McGhee as being of general application. Mustill LJ said, at pp 771-772:

  85. If it is an established fact that conduct of a particular kind creates a risk that injury will be caused to another or increases an existing risk that injury will ensue; and if the two parties stand in such a relationship that the one party owes a duty not to conduct himself in that way; and if the first party does conduct himself in that way; and if the other party does suffer injury of the kind to which the risk related; then the first party is taken to have caused the injury by his breach of duty, even though the existence and extent of the contribution made by the breach cannot be ascertained.

  86. The House of Lords, in a speech by Lord Bridge of Harwich with which all other noble Lords concurred, rejected this broad principle. I would respectfully agree. The principle in McGhee's case is far narrower and I have tried to indicate what its limits are likely to be. It is true that actions for clinical negligence notoriously give rise to difficult questions of causation. But it cannot possibly be said that the duty to take reasonable care in treating patients would be virtually drained of content unless the creation of a material risk of injury were accepted as sufficient to satisfy the causal requirements for liability. And the political and economic arguments involved in the massive increase in the liability of the National Health Service which would have been a consequence of the broad rule favoured by the Court of Appeal in Wilsher's case are far more complicated than the reasons given by Lord Wilberforce for imposing liability upon an employer who has failed to take simple precautions.

  87. I therefore think that Wilsher was correctly decided. The appellants have not made any submission to the contrary. But the grounds upon which McGhee was distinguished are unsatisfactory. Lord Bridge of Harwich said (at p. 1090) that it represented a "robust and pragmatic approach" to the facts which had enabled the House to draw a "legitimate inference of fact" that the "defenders' negligence had materially contributed to the pursuer's injury." My Lords, however robust or pragmatic the tribunal may be, it cannot draw inferences of fact in the teeth of the undisputed medical evidence. My noble and learned friend Lord Bingham of Cornhill has demonstrated that such an analysis of McGhee's case is untenable.

  88. An alternative ground of distinction is to be found in a passage in the dissenting judgment of Sir Nicolas Browne-Wilkinson V-C in the Court of Appeal, which was approved by the House. He said that the difference was that in McGhee's case the agent of injury was the same - brick dust - and the only question was whether it happened before or after it should have been washed off. In Wilsher, the fibroplasia could have been caused by a number of different agencies.

  89. That distinction would leave the present case on the right side of the line because the agent of injury was the same - asbestos dust. But I do not think it is a principled distinction. What if Mr Matthews had been exposed to two different agents - asbestos dust and some other dust - both of which created a material risk of the same cancer and it was equally impossible to say which had caused the fatal cell mutation? I cannot see why this should make a difference.

  90. The question is how narrowly the principle developed in McGhee's case and applied in this case should be confined. In my opinion, caution is advisable. Wilsher's case shows the dangers of over-generalisation. In Rutherford v Owens-Illinois Inc. (1997) 67 Cal Rptr.2d 16 the Supreme Court of California, in a valuable and lucid judgment, said that in cases of asbestos-related cancer, the causal requirements of the tort were satisfied by proving that exposure to a particular product was a substantial factor contributing to the "plaintiff's or decedent's risk of developing cancer": see p.32. That is precisely the rule your Lordships are being invited to apply in this case. The Californian Supreme Court stated the principle specifically in relation to asbestos-related cancer cases. No doubt it could also apply in other cases which were thought to have sufficient common features, but that was left for decision on a case-by-case basis. Likewise I would suggest that the rule now laid down by the House should be limited to cases which have the five features I have described.

  91. That does not mean that the principle is not capable of development and application in new situations. As my noble and learned friend Lord Rodger of Earlsferry has demonstrated, problems of uncertainty as to which of a number of possible agents caused an injury have required special treatment of one kind or another since the time of the Romans. But the problems differ quite widely and the fair and just answer will not always be the same. For example, in the famous case of Sindell v Abbott Laboratories (1980) 607 P.2d 924 the plaintiff had suffered pre-natal injuries from exposure to a drug which had been manufactured by any one of a potentially large number of defendants. The case bears some resemblance to the present but the problem is not the same. For one thing, the existence of the additional manufacturers did not materially increase the risk of injury. The risk from consuming a drug bought in one shop is not increased by the fact that it can also be bought in another shop. So the case would not fall within the McGhee principle. But the Supreme Court of California laid down the imaginative rule that each manufacturer should be liable in proportion to his market share. Cases like this are not before the House and should in my view be left for consideration when they arise. For present purposes, the McGhee principle is sufficient. I would therefore allow the appeals.

    Lord Hutton

    My Lords,

  92. In these three cases the deceased husbands of two of the appellants, Mrs Fox and Mrs Fairchild, contracted the disease of mesothelioma from the conditions in which they worked and the disease caused their deaths. The third appellant, Mr Matthews, also contracted that disease from the conditions in which he worked and he is now very gravely ill.

  93. Mesothelioma is caused by the inhalation of tiny fibres of asbestos contained in asbestos dust. The nature of the disease, the way in which it develops and the extent of medical knowledge in relation to it, and a description of the working conditions which caused the disease in these three cases are set out in the judgment of my noble and learned friend Lord Bingham of Cornhill and I gratefully adopt that account.

  94. In order for a claimant to recover damages against his employer for injury or disease suffered by him he must prove a breach of duty owed to him by his employer and that the breach was a cause of the injury or disease. The defendants have not disputed that they were each in breach of the respective duties which they owed to the husbands of Mrs Fox and Mrs Fairchild and to Mr Matthews. It is the requirement of the law that a claimant must prove causation in order to recover damages which has given rise to the issue in these cases. If Mr Fox, Mr Fairchild and Mr Matthews had each been employed throughout their working lives by a single employer each of them (or his widow) would have been entitled to recover damages against that employer. The dispute in these cases arises because Mr Fox, Mr Fairchild and Mr Matthews were each employed by a number of different employers throughout their working lives and in the present state of medical knowledge and on the basis of the medical evidence given in the High Court it is possible that the disease suffered by each employee could have been caused solely by the breach of duty of one employer. Accordingly each of the defendants has argued that each claimant cannot establish that it was its breach of duty which was a cause of the disease and that it is as probable as not that the disease was caused by the breach or breaches of duty of one or more other employers.

  95. In considering the argument advanced on behalf of the defendants there are two parts of the medical evidence (which did not differ in any material respect between the three cases) which are of particular importance and which are summarised as follows in the agreed statement of facts in the case of Matthews:

    5.

    The risk of mesothelioma increases in proportion to the dose of asbestos received and each successive period of exposure augments the risk that mesothelioma will occur. There is no evidence of a threshold dose below which there is no risk but at low levels of exposure the risk is very small, whereas a lagger with his high level of exposure has a 10% chance of dying from mesothelioma. Mesothelioma usually occurs in persons who have inhaled millions of asbestos fibres in the course of their work.

    6.

    7.

    ....

    The mechanism by which asbestos causes mesothelioma being unknown, the transformation of a normal mesothelial to a cancerous cell could be due to the action of a single fibre, a few fibres or multiple fibres. It is believed that carcinogenesis of mesothelioma is a multi-stage process involving a sequence of 6 or 7 genetic changes which together result in a normal mesothelial cell being transformed into a malignant mesothelioma cell. Asbestos is probably involved at one or more stages of this process.

  96. The claimants rely on the decision of this House in McGhee v National Coal Board [1973] 1 WLR 1; 1973 SC(HL) 37 to counter the arguments of the defendants on the issue of causation and submit that they are entitled to succeed by reason of that decision. In that case a workman contracted dermatitis after working in the hot and dusty brick kilns of his employer. The dermatitis was caused by quantities of abrasive brick dust adhering to his skin softened by sweating, and exertion would cause the dust to injure the horny layer exposing the tender cells below to injury or infection from which dermatitis would develop. Washing was the only practicable way of removing the danger, and if the skin was not thoroughly washed as soon as the workman ceased work the process could continue for some considerable time. The workman, cycling home caked with sweat and dust, was liable to further injury until he could wash himself thoroughly.

  97. The Lord Ordinary found that there was no breach of duty by the employer in requiring the pursuer to work in the hot and dusty brick kilns, but that there was a breach of duty in failing to provide a shower for him to use after his work. On the issue of causation the pursuer's doctor said in evidence that he could not say that the provision of showers would probably have prevented the dermatitis. He said that showers would have reduced the risk materially but he would not go further than that. On this evidence the Lord Ordinary found that the pursuer had failed to prove that the breach of duty had caused the dermatitis and he held that he could not succeed if the only inference from the evidence was that lack of shower baths was a possibility as a cause of his having contracted the disease. His decision was upheld by the First Division, the Lord President stating that risk of dermatitis and causation of dermatitis are two quite separate matters.

  98. On appeal to this House the pursuer relied on the decision of the House in Bonnington Castings Ltd v Wardlaw [1956] AC 613 where it was held that if there are two causes of the disease each materially contributing to it such as dust from two sources, and the defendant company is responsible for only one of them, it is liable notwithstanding that the dust for which it was responsible was not in itself sufficient to cause the disease. Part of the pursuer's argument is reported at 1973 SC(HL) 37, 51:

    Reliance was placed on Wardlaw v Bonnington Castings & Nicholson v Atlas Steel Foundry & Engineering Co. The dicta in these cases were followed in Gardiner v Motherwell Machinery & Scrap Co which applied the principles of the two earlier cases. Where there were a number of contributory factors to an injury, it was enough to bring the case within the dicta if the employers were guilty in relation to one.

    The defender sought to distinguish Wardlaw's case by arguing that in that case every particle of dust inhaled played its part in causing the onset of the disease whereas in the case of McGhee the disease might have been started by a single abrasion.

  99. In considering this point Lord Reid stated in his speech at [1973] 1 WLR 1, p 4F that the evidence did not show how the dermatitis began. It might be a necessary precondition for the onset of the disease that there should be an accumulation of minor abrasions of the skin, or it might be that the disease started at one particular abrasion and then spread, so that multiplication of abrasions merely increased the number of places where the disease could start and in that way increased the risk of its occurrence. Lord Reid stated that he could not hold that it was probable that an accumulation of minor abrasions was a necessary precondition; if this had been proved the case would have been indistinguishable from Wardlaw's case. He then stated, at pp 4-5:

    But I think that in cases like this we must take a broader view of causation. The medical evidence is to the effect that the fact that the man had to cycle home caked with grime and sweat added materially to the risk that this disease might develop. It does not and could not explain just why that is so. But experience shows that it is so. Plainly that must be because what happens while the man remains unwashed can have a causative effect, though just how the cause operates is uncertain. I cannot accept the view expressed in the Inner House that once the man left the brick kiln he left behind the causes which made him liable to develop dermatitis. That seems to me quite inconsistent with a proper interpretation of the medical evidence. Nor can I accept the distinction drawn by the Lord Ordinary between materially increasing the risk that the disease will occur and making a material contribution to its occurrence.

    There may be some logical ground for such a distinction where our knowledge of all the material factors is complete. But it has often been said that the legal concept of causation is not based on logic or philosophy. It is based on the practical way in which the ordinary man's mind works in the everyday affairs of life. From a broad and practical viewpoint I can see no substantial difference between saying that what the defender did materially increased the risk of injury to the pursuer and saying that what the defender did made a material contribution to his injury.

  100. Lord Wilberforce stated, at p 6B:

    My Lords, I agree with the judge below to the extent that merely to show that a breach of duty increases the risk of harm is not, in abstracto, enough to enable the pursuer to succeed. He might, on this basis, still be met by successful defences. Thus, it was open to the respondents, while admitting, or being unable to contest, that their failure had increased the risk, to prove, if they could, as they tried to do, that the appellant's dermatitis was 'non-occupational'.

    But the question remains whether a pursuer must necessarily fail if, after he has shown a breach of duty, involving an increase of risk of disease, he cannot positively prove that this increase of risk caused or materially contributed to the disease while his employers cannot positively prove the contrary. In this intermediate case there is an appearance of logic in the view that the pursuer, on whom the onus lies, should fail — a logic which dictated the judgments below. The question is whether we should be satisfied, in factual situations like the present, with this logical approach. In my opinion, there are further considerations of importance. First, it is a sound principle that where a person has, by breach of a duty of care, created a risk, and injury occurs within the area of that risk, the loss should be borne by him unless he shows that it had some other cause. Secondly, from the evidential point of view, one may ask, why should a man who is able to show that his employer should have taken certain precautions, because without them there is a risk, or an added risk, of injury or disease, and who in fact sustains exactly that injury or disease, have to assume the burden of proving more: namely, that it was the addition to the risk, caused by the breach of duty, which caused or materially contributed to the injury? In many cases, of which the present is typical, this is impossible to prove, just because honest medical opinion cannot segregate the causes of an illness between compound causes. And if one asks which of the parties, the workman or the employers, should suffer from this inherent evidential difficulty, the answer as a matter of policy or justice should be that it is the creator of the risk who, ex hypothesi must be taken to have foreseen the possibility of damage, who should bear its consequences.

  101. Lord Simon of Glaisdale stated, at p 8B:

    But Bonnington Castings Ltd v Wardlaw [1956] AC 613 and Nicholson v Atlas Steel Foundry & Engineering Co Ltd [1957] 1 WLR 613 establish, in my view, that where an injury is caused by two (or more) factors operating cumulatively, one (or more) of which factors is a breach of duty and one (or more) is not so, in such a way that it is impossible to ascertain the proportion in which the factors were effective in producing the injury or which factor was decisive, the law does not require a pursuer or plaintiff to prove the impossible, but holds that he is entitled to damages for the injury if he proves on a balance of probabilities that the breach or breaches of duty contributed substantially to causing the injury. If such factors so operate cumulatively, it is, in my judgment, immaterial whether they do so concurrently or successively.

    The question, then, is whether on the evidence the appellant brought himself within this rule. In my view, a failure to take steps which would bring about a material reduction of the risk involves, in this type of case, a substantial contribution to the injury. In this type of case a stark distinction between breach of duty and causation is unreal. If the provision of shower baths was (as the evidence showed) a precaution which any reasonable employer in the respondents' position would take, it means that such employer should have foreseen that failure to take the precaution would, more probably than not, substantially contribute towards injury: this is sufficient prima facie evidence.

  102. Lord Kilbrandon stated, at p 10F:

    In the present case, the pursuer's body was vulnerable, while he was bicycling home, to the dirt which had been deposited on it during his working hours. It would not have been if he had had a shower. If showers had been provided he would have used them. It is admittedly more probable that disease will be contracted if a shower is not taken. In these circumstances I cannot accept the argument that nevertheless it is not more probable than not that, if the duty to provide a shower had not been neglected, he would not have contracted the disease. The pursuer has after all, only to satisfy the court of a probability, not to demonstrate an irrefragable chain of causation, which in a case of dermatitis, in the present state of medical knowledge, he could probably never do.

  103. At pp 11H-12A, Lord Salmon stated:

    In the circumstances of the present case it seems to me unrealistic and contrary to ordinary common sense to hold that the negligence which materially increased the risk of injury did not materially contribute to causing the injury.

    And at pp 12H-13A:

    In the circumstances of the present case, the possibility of a distinction existing between (a) having materially increased the risk of contracting the disease, and (b) having materially contributed to causing the disease may no doubt be a fruitful source of interesting academic discussions between students of philosophy. Such a distinction is, however, far too unreal to be recognised by the common law.

  104. In Wilsher v Essex Area Health Authority [1987] QB 730 the plaintiff was born prematurely and due to the negligence of a doctor excessive amounts of oxygen were administered to him. The plaintiff developed retrolental fibroplasia, a condition of the eyes, which resulted in blindness. The administration of the excessive amounts of oxygen was a possible cause of the retrolental fibroplasia but there were a number of other possible causes of the condition which were in no way related to the negligence of the doctor. At first instance the judge gave judgment for the plaintiff and awarded damages. The defendant appealed and the appeal was dismissed by the Court of Appeal (Mustill and Glidewell LJJ, Sir Nicolas Browne-Wilkinson V-C dissenting). Mustill LJ recognised that the case differed from McGhee in that in the latter case there was only one risk operating, which was that the contact of brick dust with a sweaty skin would lead to dermatitis, whereas in the Wilsher case there was a list of risk factors, any one of which might have caused the injury. However he came to the conclusion that the approach taken in McGhee could be adopted in the case before him and stated, at pp 771H-772C:

    Reading all the speeches together, the principle applied by the House seems to me to amount to this. If it is an established fact that conduct of a particular kind creates a risk that injury will be caused to another or increases an existing risk that injury will ensue; and if the two parties stand in such a relationship that the one party owes a duty not to conduct himself in that way; and if the first party does conduct himself in that way; and if the other party does suffer injury of the kind to which the risk related; then the first party is taken to have caused the injury by his breach of duty, even though the existence and extent of the contribution made by the breach cannot be ascertained. If this is the right analysis, it seems to me that the shape taken by the enhancement of the risk ought not to be of crucial significance. In the McGhee case [1973] 1 WLR 1, the conduct of the employers made it more likely that the pursuer would contract dermatitis, and he did contract dermatitis. Here, the conduct of those for whom the defendants are liable made it more likely that Martin would contract RLF, and he did contract RLF. If considerations of justice demanded that the pursuer succeed in the one case, I can see no reason why the plaintiff should not succeed in the other.

  105. In his dissenting judgment the Vice-Chancellor stated, at p 779F:

    The position, to my mind, is wholly different from that in the McGhee case [1973] 1 WLR 1, where there was only one candidate (brick dust) which could have caused the dermatitis, and the failure to take a precaution against brick dust causing dermatitis was followed by dermatitis caused by brick dust. In such a case, I can see the common sense, if not the logic, of holding that, in the absence of any other evidence, the failure to take the precaution caused or contributed to the dermatitis. To the extent that certain members of the House of Lords decided the question on inferences from evidence or presumptions, I do not consider that the present case falls within their reasoning. A failure to take preventative measures against one out of five possible causes is no evidence as to which of those five caused the injury.

  106. The defendant appealed to this House and the appeal was allowed [1988] AC 1074. The ground on which the appeal was allowed was the ground stated by the Vice-Chancellor in his dissenting judgment and in his speech at pp 1090 and 1091, with which the other members of the House concurred, Lord Bridge of Harwich set out the relevant passage in the judgment of the Vice-Chancellor and stated that he was quite unable to find any fault with it.

  107. The House also considered the passages in the speech of Lord Wilberforce in McGhee at pp 6 and 7 which the House considered constituted a reversal of the burden of proof. Lord Bridge stated, at p 1087F:

    My Lords, it seems to me that both these paragraphs, particularly in the words I have emphasised, amount to saying that, in the circumstances, the burden of proof of causation is reversed and thereby to run counter to the unanimous and emphatic opinions expressed in Bonnington Castings Ltd v Wardlaw [1956] AC 613 to the contrary effect. I find no support in any of the other speeches for the view that the burden of proof is reversed and, in this respect, I think Lord Wilberforce's reasoning must be regarded as expressing a minority opinion.

  108. In his speech Lord Bridge also stated that in his opinion the decision of Lord Reid, Lord Simon, Lord Kilbrandon and Lord Salmon was based on the drawing of an inference. He said, at pp 1087H and 1088A-B:

    A distinction is, of course, apparent between the facts of Bonnington Castings Ltd v Wardlaw, where the "innocent" and "guilty" silica dust particles which together caused the pursuer's lung disease were inhaled concurrently and the facts of McGhee v National Coal Board [1973] 1 WLR 1 where the "innocent" and "guilty" brick dust was present on the pursuer's body for consecutive periods. In the one case the concurrent inhalation of "innocent" and "guilty" dust must both have contributed to the cause of the disease. In the other case the consecutive periods when "innocent" and "guilty" brick dust was present on the pursuer's body may both have contributed to the cause of the disease or, theoretically at least, one or other may have been the sole cause. But where the layman is told by the doctors that the longer the brick dust remains on the body, the greater the risk of dermatitis, although the doctors cannot identify the process of causation scientifically, there seems to be nothing irrational in drawing the inference, as a matter of common sense, that the consecutive periods when brick dust remained on the body probably contributed cumulatively to the causation of the dermatitis. I believe that a process of inferential reasoning on these general lines underlies the decision of the majority in McGhee's case.

    After citing passages from the four speeches he said, at p 1090C:

    The conclusion I draw from these is that McGhee v National Coal Board [1973] 1 WLR 1 laid down no new principle of law whatever. On the contrary, it affirmed the principle that the onus of proving causation lies on the pursuer or plaintiff. Adopting a robust and pragmatic approach to the undisputed primary facts of the case, the majority concluded that it was a legitimate inference of fact that the defenders' negligence had materially contributed to the pursuer's injury. The decision, in my opinion, is of no greater significance than that and to attempt to extract from it some esoteric principle which in some way modifies, as a matter of law, the nature of the burden of proof of causation which a plaintiff or pursuer must discharge once he has established a relevant breach of duty is a fruitless one.

  109. Mr Stewart QC, in an impressive argument on behalf of two of the defendants, advanced submissions which I summarise as follows:

    (1)

    The fundamental principle of English law in relation to industrial injury cases which has been constantly followed and applied by this House is that stated by Lord Reid in Bonnington Castings v Wardlaw at p 620:

    "It would seem obvious in principle that a pursuer or plaintiff must prove not only negligence or breach of duty but also that such fault caused or materially contributed to his injury, and there is ample authority for that proposition both in Scotland and in England. I can find neither reason nor authority for the rule being different where there is breach of a statutory duty. The fact that Parliament imposes a duty for the protection of employees has been held to entitle an employee to sue if he is injured as a result of a breach of that duty, but it would be going a great deal farther to hold that it can be inferred from the enactment of a duty that Parliament intended that any employee suffering injury can sue his employer merely because there was a breach of duty and it is shown to be possible that his injury may have been caused by it. In my judgment, the employee must in all cases prove his case by the ordinary standard of proof in civil actions: he must make it appear at least that on a balance of probabilities the breach of duty caused or materially contributed to his injury.

    (2)

    Wardlaw was a case where the inhalation of "innocent" and "guilty" dust contributed cumulatively over a period of time to the pursuer's lung disease, and therefore the inhalation of the "guilty" dust was properly regarded as making a material contribution to the disease. Mr Stewart submitted that McGhee's case was also a case where it was established that there had been a cumulative causative process because the "innocent" and "guilty" brick dust had been present on the pursuer's body for consecutive periods. In support of this submission he relied on the passage in Lord Bridge's speech in Wilsher, at 1087H to 1088B, where he referred to the cumulative contribution to the causation of the dermatitis. He further submitted that the present case was not governed by the decision in McGhee because in this case the medical evidence was that it was as probable as not that the onset of the disease was not due to a cumulative process occurring during the entire period when an employee was employed by different employers but was due to the inhalation of a single fibre or a number of fibres during the course of employment with only one employer.

    (3)

    This House held in Wilsher that the decision in McGhee had not laid down a new principle of law that proof of a material increase in risk constitutes proof of causation: rather the decision was based on the robust drawing of an inference. But in the present case it was not open to draw such an inference because the medical evidence established that it was as likely as not that the cause of the onset of the disease was the inhalation of a single fibre or a small number of fibres.

    (4)

    If this House were to hold that the decision in McGhee laid down a principle of law that proof of a material increase of risk constitutes proof of causation, such a decision would reverse the burden of proof contrary to the decision in Wardlaw's case where the House overruled the decision of the Court of Appeal in Vyner v Waldenberg Brothers Ltd [1946] KB 50 that the onus is on an employer to show that the breach of a safety regulation was not the cause of an accident. Moreover if the House were to hold that there was such a principle of law, it would not be possible to confine the principle to cases such as the present ones where there is only one possible cause, and the principle would be extended to cases of medical negligence (as was done by the Court of Appeal in the Wilsher case) and to other cases.

  110. I do not accept Mr Stewart's submission that McGhee was a case where there was a cumulative process analogous to that in Wardlaw, and that the present case is not governed by McGhee and is distinguishable from it because in this case it was as probable as not that the disease developed from the inhalation of one fibre or a small number of fibres. Lord Reid stated in his speech in McGhee at p 4F-H that whilst it was possible that the pursuer's disease started with an accumulation of minor abrasions this was not proved and that it was possible that the disease started at one particular abrasion. Therefore it is clear that Lord Reid held that the defender was liable even if the onset of the disease had not been due to an accumulation of abrasions but had started at one particular abrasion.

  111. My Lords, I consider that the question whether the decision in McGhee was based on an inference raises an issue of some difficulty. There does not appear to be a clear statement in the speeches that the House was formulating a new principle of law rather than basing its decision on an inference from the evidence. The case was one in which there was no real conflict between the medical witnesses and therefore the House was in as good a position to evaluate the evidence and to come to a factual conclusion on causation as were the trial judge and the Inner House: see per Lord Reid in Benmax v Austin Motor Co Ltd [1955] AC 370, 376. Moreover the issue of causation is one to be decided ultimately by the tribunal of fact and not by the medical witnesses (see Sentilles v Inter-Caribbean Shipping Corporation 361 US 107 (1959) cited later in para [114] of this opinion), although the tribunal of fact must give proper weight to their opinions.

  112. Although at the commencement of his speech Lord Reid said at p 3D that the allegation that the defender was at fault in not providing adequate washing facilities "raises a difficult question of law", there are other passages in the speeches which I think can be read as pointing to the decision being based on the drawing of an inference.

  113. Lord Simon, referring to Wardlaw & Nicholson v Atlas Steel Foundry & Engineering Co Ltd, said, at p 8C:

    the law does not require a pursuer or plaintiff to prove the impossible, but holds that he is entitled to damages for the injury if he proves on a balance of probabilities that the breach or breaches of duty contributed substantially to causing the injury ....

    The question, then, is whether on the evidence the appellant brought himself within this rule. In my view, a failure to take steps which would bring about a material reduction of the risk involves, in this type of case, a substantial contribution to the injury. In this type of case a stark distinction between breach of duty and causation is unreal. If the provision of shower baths was (as the evidence showed) a precaution which any reasonable employer in the respondents' position would take, it means that such employer should have foreseen that failure to take the precaution would, more probably than not, substantially contribute towards injury: this is sufficient prima facie evidence.

    [emphasis added]

    It is also relevant to note that at the conclusion of his speech, at p 9A, Lord Simon cited with approval the decision in Gardiner v Motherwell Machinery & Scrap Co Ltd [1961] 1 WLR 1424 where Lord Reid stated, at p 1429:

    In my opinion, when a man who has not previously suffered from a disease contracts that disease after being subjected to conditions likely to cause it, and when he shows that it starts in a way typical of disease caused by such conditions, he establishes a prima facie presumption that his disease was caused by those conditions. I think that the facts proved in this case do establish such a presumption. That presumption could be displaced in many ways.

    In my opinion in the context of Gardiner and McGhee there is little, if any, difference between a prima facie presumption and an inference.

  114. Lord Kilbrandon said, at p 10G:

    The pursuer has after all, only to satisfy the court of a probability, not to demonstrate an irrefragable chain of causation, which in a case of dermatitis, in the present state of medical knowledge, he could probably never do.

    And Lord Salmon said, at pp 11H-12A:

    In the circumstances of the present case it seems to me unrealistic and contrary to ordinary common sense to hold that the negligence which materially increased the risk of injury did not materially contribute to causing the injury.

    I incline to the view that these statements by Lord Simon, Lord Kilbrandon and Lord Salmon point more to the drawing of an inference in a common sense way from the evidence than to the formulation of a principle of law.

  115. In Hotson v East Berkshire Area Health Authority [1987] AC 750, Lord Mackay of Clashfern (who was counsel for the defender in McGhee) referring to that case, said, at p 786D:

    the decision of this House may be taken as holding that in the circumstances of that case it was reasonable to infer that there was a relationship between contraction of dermatitis in these conditions and the absence of washing facilities and therefore it was reasonable to hold that absence of washing facilities was likely to have made a material contribution to the causation of the dermatitis.

  116. It is clear that this House in Wilsher considered that a new principle of law was not laid down in McGhee and that the decision of the majority was based on the drawing of an inference from the evidence in a robust and commonsense way. And in the Court of Appeal [1987] QB 730, at p 779 the Vice-Chancellor said:

    To the extent that certain members of the House of Lords decided the question on inferences from evidence or presumptions, I do not consider that the present case falls within their reasoning.

  117. In describing in Wilsher the manner in which, in his opinion, the House drew an inference in McGhee, Lord Bridge said, [1988] AC 1074, at p 1088A-B:

    But where the layman is told by the doctors that the longer the brick dust remains on the body, the greater the risk of dermatitis, although the doctors cannot identify the process of causation scientifically, there seems to be nothing irrational in drawing the inference, as a matter of common sense, that the consecutive periods when brick dust remained on the body probably contributed cumulatively to the causation of the dermatitis.

    I consider that this approach, whereby the layman applying broad common sense draws an inference which the doctors as scientific witnesses are not prepared to draw, is one which is permissible. In the United States Supreme Court in Sentilles v Inter-Caribbean Shipping Corporation 361 US 107 (1959) Brennan J stated, at p 109:

    The jury's power to draw the inference that the aggravation of petitioner's tubercular condition, evident so shortly after the accident, was in fact caused by that accident, was not impaired by the failure of any medical witness to testify that it was in fact the cause. Neither can it be impaired by the lack of medical unanimity as to the respective likelihood of the potential causes of the aggravation, or by the fact that other potential causes of the aggravation existed and were not conclusively negated by the proofs. The matter does not turn on the use of a particular form of words by the physicians in giving their testimony. The members of the jury, not the medical witnesses, were sworn to make a legal determination of the question of causation.

  118. It is apparent from his judgment in the action brought by Mr Matthews in which he found that the breaches of duty by the defendants caused his mesothelioma that Mitting J based his decision on an inference from the evidence. He referred to the judgment of Lord Bridge in Wilsher and stated, at p 15 of his judgment:

    Adopting that robust approach, my firm conclusion in this case is that the claimant's exposure to asbestos fibres in the employment of the two defendants, did materially contribute to, and so cause his mesothelioma. I reach that conclusion by a simple, direct process of reasoning, readily understandable to a layman. The claimant was exposed by each defendant and by both defendants, to asbestos fibres, in quantities sufficient greatly to increase his risk of contracting mesothelioma, on the evidence of Dr. Rudd, by up to a thousand times the general population risk. They owed him statutory and common law duties to take steps to minimise that risk. They failed to do so. He contracted the very disease against which it was their duty to take those steps.

    In those circumstances, like Lord Reid in McGhee, I can see no substantial difference between saying that what the defendant did materially increased the risk of injury to the claimant and saying that what the defendants did made a material contribution to his injury. It seems to me wholly artificial to require a claimant to prove which fibre, or fibres, inhaled in whose employment in precisely what circumstances, caused or set off or contributed to the process by which one or more mesothelial cells became malignant. In principle, it is just as artificial as requiring proof in a case in which a pool of liquid, collected from separate sources, has caused injury, of precisely which molecule was the mechanical cause of injury. Even though the precise mechanism by which the claimant's mesothelioma was caused or set off cannot be established, those conclusions point, inexorably, to the finding that the defendants' breaches of duty materially contributed to his contracting the disease. In other words, the evidence which I have heard proves, on the balance of probabilities, that their breach of duty caused his injury.

  119. The courts in Canada have taken the view that the majority decision in McGhee was based on an inference. In Snell v Farrell (1990) 72 DLR (4th) 289 Sopinka J, in delivering the judgment of the Supreme Court, stated with reference to McGhee, at p 296g:

    Two theories of causation emerge from an analysis of the speeches of the Lords in this case. The first, firmly espoused by Lord Wilberforce, is that the plaintiff need only prove that the defendant created a risk of harm and that the injury occurred within the area of the risk. The second is that in these circumstances, an inference of causation was warranted in that there is no practical difference between materially contributing to the risk of harm and materially contributing to the harm itself.

  120. The judgment of Lambert JA in the British Columbia Court of Appeal in Haag v Marshall (1989) 61 DLR (4th) 371 contains, in my respectful opinion, an illuminating discussion of the decision in McGhee. He stated, at pp 378 and 379:

    McGhee v National Coal Board was considered by the House of Lords in Kay v Ayrshire & Arran Health Board [1987] 2 All ER 417; Hotson v East Berkshire Area Health Authority [1987] 2 All ER 909, and Wilsher v Essex Area Health Authority [1988] 1 All ER 871. It is clear from those decisions that McGhee is not now, and never was, authority for the legally adventurous proposition that if a breach of duty is shown, and damage is proven within the area of risk that brought about the duty, and if the breach of duty materially increases the risk of damage of that type, then the onus of proof shifts from the plaintiff to the defendant to disprove the causal connection. That proposition could be derived only from the speech of Lord Wilberforce and it is now clear that it was never a binding principle emerging from the McGhee case.

    But McGhee remains a worthwhile study. And there is a somewhat more cautious principle underlying the decision in that case. However, it is not an "onus" principle but an "inference" principle ....

    The "inference" principle derived from McGhee, and from the Canadian cases to which I have referred, is this: Where a breach of duty has occurred, and damage is shown to have arisen within the area of risk which brought the duty into being, and where the breach of duty materially increased the risk that damage of that type would occur, and where it is impossible, in a practical sense, for either party to lead evidence which would establish either that the breach of duty caused the loss or that it did not, then it is permissible to infer, as a matter of legal, though not necessarily logical, inference, that the material increase in risk arising from the breach of duty constituted a material contributing cause of the loss and as such a foundation for a finding of liability ....

    Whether the inference of causation should in fact be made in any particular case depends on whether it is in accordance with common sense and justice in that case to say that the breach of duty which materially increased the risk ought reasonably to be considered as having materially contributed to the loss.

  121. Courts in Australia have also taken the view that the decision in McGhee was arrived at on the basis of an inference. In the Supreme Court of South Australia in Birkholz v RJ Gilbertson Pty Ltd (1985) 38 SASR 121, 130 King CJ stated:

    Has the failure to take those precautions been shown to have caused or materially contributed to the contracting of the disease by the appellant? It might be argued as a matter of strict logic, that the fact that given precautions would substantially diminish the risk, does not prove that failure to take those precautions materially contributed to the appellant's infection unless it can be established how that infection occurred. But the law's view of causation is less concerned with logical and philosophical considerations than with the need to produce a just result to the parties involved. Where a defendant is under a legal duty to take precautions to protect the plaintiff from the risk of contracting disease, and, by omitting those precautions, he substantially increases the risk of the plaintiff contracting that disease, the law treats that increase in risk as a sufficient basis, in the absence of evidence showing how the infection occurred, for an inference that the omission of the precautions materially contributed to the contracting of the disease. Justice requires such an approach to the problem of causation and it is the approach which was taken by the House of Lords in McGhee v National Coal Board.

  122. Another illuminating discussion of the proper approach to the issue of causation where a claimant seeks to recover damages for an asbestos-related cancer is contained in the judgment of Baxter J in the Supreme Court of California (with which the majority of the court agreed) in Rutherford v Owens-Illinois Inc (1997) 16 Cal. 4th 953 where he stated at p 977:

    In refining the concept of legal cause we must also ensure that the triers of fact in asbestos-related cancer cases know the precise contours of the plaintiff's burden. The generally applicable standard instructions on causation are insufficient for this purpose. Those instructions tell the jury that every "substantial factor in bringing about an injury" is a legal cause (BAJI No. 3.76), even when more than one such factor "contributes concurrently as a cause of the injury" (BAJI No. 3.77). They say nothing, however, to inform the jury that, in asbestos-related cancer cases, a particular asbestos-containing product is deemed to be a substantial factor in bringing about the injury if its contribution to the plaintiff or decedent's risk or probability of developing cancer was substantial.

    Without such guidance, a juror might well conclude that the plaintiff needed to prove that fibers from the defendant's product were a substantial factor actually contributing to the development of the plaintiff's or decedent's cancer. In many cases, such a burden will be medically impossible to sustain, even with the greatest possible effort by the plaintiff, because of irreducible uncertainty regarding the cellular formation of an asbestos-related cancer. We therefore hold that, in the trial of an asbestos-related cancer case, although no instruction "shifting the burden of proof as to causation" to defendant is warranted, the jury should be told that the plaintiff's or decedent's exposure to a particular product was a substantial factor in causing or bringing about the disease if in reasonable medical probability it was a substantial factor contributing to plaintiff's or decedent's risk of developing cancer.

  123. As I have stated, I think that there is no clear statement in the speeches in McGhee as to the underlying basis of the decision. The preponderance of subsequent judicial opinion has been that the decision was based on an inference from the facts. On the evidence before it in McGhee I think that the House was entitled to draw an inference in the way described by Lord Bridge in Wilsher. But some authorities suggest that in cases where the claimant can prove that a breach of duty materially increased the risk of the contraction of a particular disease and the disease occurred, the law should treat this as giving rise to the inference that the breach of duty was a cause of the disease rather than that the judge as the tribunal of fact should draw a factual inference. I think this was the view of Lambert JA in Haag v Marshall when he said at p 379,

    it is permissible to infer, as a matter of legal, though not necessarily logical, inference, that the material increase in risk arising from the breach of duty constituted a material contributing cause of the loss.

  124. I think that King CJ was expressing the same view when he said in Birkholz v RJ Gilbertson Pty Ltd at p 130 that

    the law treats that increase in risk as a sufficient basis, in the absence of evidence showing how the infection occurred, for an inference that the omission of the precautions materially contributed to the contracting of the disease.

  125. A rule that such a legal inference should or can arise is, in effect, a principle of law. I consider that such a rule is virtually indistinguishable from the approach of the Supreme Court of California in Rutherford where Baxter J stated that a jury should be directed that

    a particular asbestos-containing product is deemed to be a substantial factor in bringing about the injury if its contribution to the plaintiff or decedent's risk or probability of developing cancer was substantial .... the jury should be told that the plaintiff's or decedent's exposure to a particular product was a substantial factor in causing or bringing about the disease if in reasonable medical probability it was a substantial factor contributing to plaintiff's or decedent's risk of developing cancer.

  126. Therefore, whilst the decision taken by the House in McGhee may have been based on an inference of fact, I consider that it is in the interests of justice that it should now be held as a matter of law that the approach taken by the House in McGhee is one which should be followed by trial judges in cases such as the present ones where the claimant can prove that the employer's breach of duty materially increased the risk of him contracting a particular disease and the disease occurred, but where in the state of existing medical knowledge he is unable to prove by medical evidence that the breach was a cause of the disease.

  127. Whilst there is very little practical difference between the two views I prefer, with respect, to take the view that the McGhee approach is based on the drawing of a factual or legal inference leading to the conclusion that the breach of duty was a cause of the disease rather than that the decision in McGhee laid down a new principle that, in cases where medical evidence as to the precise nature of the causation cannot be adduced, the material increase in the risk is taken in law to be a cause of the disease without reliance on a factual or legal inference. As well as the passages in the speeches of Lord Simon of Glaisdale, Lord Kilbrandon and Lord Salmon to which I have referred in para [110] and [111] of this opinion I incline to the view that Lord Reid's references to the legal concept of causation being based "on the practical way in which the ordinary man's mind works in the everyday affairs of life" and to a "broad and practical viewpoint" (see p 5 B) point more to the drawing of an inference than to the laying down of a new principle of law. Moreover I think that in his speech in McGhee, at p 7, Lord Wilberforce recognised that there were two ways in which injustice to the pursuer could be avoided: one way was by drawing an inference (which he rejected in that case as a fiction), and the other way was, in effect, by reversing the burden of proof on causation. But as Lord Bridge observed in Wilsher, at p 1087 G, there was no support for the reversal of the burden of proof in the other speeches.

  128. In some cases in other jurisdictions (as in the hunting cases where two hunters fire negligently at the same time and the claimant, struck by one shot, cannot prove which shot struck him) the courts, in order to avoid injustice, have expressly held that the burden of proof on causation should be reversed. In Summers v Tice 199 2d 1 (1948) Carter J delivering the judgment of the Supreme Court of California stated, at page 4 "a requirement that the burden of proof on that subject be shifted to defendants becomes manifest"; see also the majority judgment of the Supreme Court of Canada in Cook v Lewis [1952] 1 DLR 1, 18. In the hunting cases and in cases analogous to them justice may well require that the burden of proof be reversed, but in relation to cases of industrial disease there appears to be no support in the English authorities, other than in Lord Wilberforce's speech in McGhee, for the reversal of the burden of proof as a matter of law.

  129. Where the claimant proves that the breach of duty materially increased the risk of the onset of the disease from which he suffers, it will be open to the defendant to adduce evidence and to argue that even though the breach of duty materially increased the risk, nevertheless the evidence adduced on its behalf displaces the inference of causation. Therefore I turn to consider whether the medical evidence relied on by the defendants displaces the inference which should otherwise be drawn that the prolonged exposure to asbestos dust which increased the risk of contracting mesothelioma materially contributed to the onset of the disease in the three employees. The evidence upon which the defendants rely is that it is as probable as not that the transformation of a normal mesothelial cell to a cancerous cell could be due to the action of a single asbestos fibre or of a few asbestos fibres. In the light of this evidence the defendants submit that it was as probable as not that each employee contracted the disease during the course of his employment with only one employer who cannot be identified. Therefore the evidence that each successive period of exposure augments the risk that the disease will occur should not give rise to the inference that the exposure to asbestos dust created by the default of each employer was the cause of the disease.

  130. I am unable to accept that submission. The other part of the evidence as to the mechanism by which asbestos causes mesothelioma is that it is as probable as not that the onset of the disease could be due to the action of multiple fibres. When this is added to the evidence that each successive period of exposure augments the risk that the disease will occur, and that mesothelioma usually occurs in persons who have inhaled millions of asbestos fibres in the course of their work, I think that the inference of causation is not displaced by the evidence upon which the defendants rely.

  131. In the course of his evidence in the trial of the action brought by Mrs Fairchild her medical witness, Dr Rudd, referred to the common sense conclusion on the question whether Mr Fairchild's mesothelioma was caused by one fibre or by prolonged exposure to many fibres. I set out a portion of the transcript so that his observation on common sense can be seen in its context. I add that I do not think that his evidence departs from the agreed statement of facts in any of the three cases:

    Re-examined by MR HOGARTH

    MR HOGARTH: (To the witness)

    Q.

    Dr Rudd, you are (inaudible) deal with the situation (?) put to you by Mr Stewart, where he suggested that if you assume that there are two equal exposers it is impossible to say which one produced or exposed to the fatal fibre. Can you say that one or other, or both?

    A.

    It's not possible to say it was one or another, or both. It could have been one. It could have been the other. It could have been both if there was more than one fibre involved.

    Q.

    It has to be one, or the other, or both?

    A.

    What is known for certain is that together they were responsible for the malignant transformation. The situation is analogous to causation of lung cancer by cigarette smoking. The cigarette smoke causes a [series] of mutations which, together, give rise to the lung cancer. One could say, theoretically, that last mutation was caused by a molecule of tar from one cigarette, but it would not be in accordance with common sense to suggest that the man who had smoked for thirty years and developed lung cancer got it as a result of smoking one cigarette rather than another. The cancer is the end result of the sum total of smoking. Mesothelioma is the end result of the sum total of asbestos exposure.

    Cross-examined by MR STEWART

    MR STEWART: My Lord, I am a little concerned your Lordship may possibly be misled by that last answer. I wonder if I may clarify it? (To the witness):

    Q.

    Dr Rudd, that last statement, that it is the sum total of the exposure to asbestos, is strictly on the hypothesis that more (inaudible) which was not proven that as asbestos fibres, as opposed to one asbestos fibre, is more likely to have caused it?

    A.

    It's a statement based upon the facts which can be known with certainty. The facts which can be known with certainty are that asbestos causes mesothelioma. The risk that mesothelioma will occur increases with the quantity of asbestos inhaled. Mr Fairchild was exposed to a quantity of asbestos from various sources. As a result, he developed mesothelioma. The rest is speculation.

    Q.

    You are not, I assume, [resiling from] the fact that you agreed that it is possible that the risk lies with exposure simply because each fibre has a chance in initiating the fatal (?) mutation?

    A.

    That's possible.

    Q.

    That is equally likely, as you said before, that it is the one fibre out of all those millions that it is equally likely to be that, as more than one fibre?

    A.

    Yes, we're moving on to mechanisms, now, as opposed to observable facts.

    I consider that the common sense view expressed by Dr Rudd is the one which a court should take and is in conformity with the opinion of Lord Reid in McGhee, at p 5B, that the issue of causation in cases such as these should be assessed from a broad and practical viewpoint.

  132. The application of the McGhee principle in these cases and in similar cases, where successive employers of an employee have been in breach of duty to take steps to guard against an industrial disease which the employee contracts, may mean that an employer may be held liable when in reality, if medical science were able to be certain as to how the employee's disease started, it was not fibres inhaled during the employment with it which caused the onset of the disease. But if the McGhee principle is not applied the consequence will be that an employee who undoubtedly sustained the disease because one or some or all of his employers were in breach of duty to take steps to guard against the onset of the disease, will recover no damages because in the present state of medical knowledge the doctors are unable to say which breach of duty by an employer caused the disease. In these circumstances I have no doubt that justice is better served by requiring an employer, who has been in breach of duty and who has materially increased the risk of its innocent employee incurring the disease, to pay damages than by ruling that the employee who has sustained a grievous disease can recover nothing. Therefore I am in respectful agreement with the opinion of Lord Wilberforce in McGhee, at p 6F:

    And if one asks which of the parties, the workman or the employers, should suffer from this inherent evidential difficulty, the answer as a matter of policy or justice should be that it is the creator of the risk who, ex hypothesi must be taken to have foreseen the possibility of damage, who should bear its consequences.

  133. This view is also well put by Lambert JA in Haag v Marshall, at p 379:

    as between an innocent plaintiff and a defendant who has committed a breach of duty to the plaintiff and by so doing materially increased the risk of loss to the plaintiff, in a situation where it is impossible, as a practical matter, to prove whether the breach of duty caused the loss, it is more in keeping with a common sense approach to causation as a tool of justice, to let the liability fall on the defendant.

  134. In its judgment the Court of Appeal stated, at [2002] 1 WLR 1052, 1080, in para [103]:

    the claimant is wholly unable to prove on the balance of probabilities that that period of employment had any causative relationship with the inception of the disease. This is far too weighty an edifice to build on the slender foundations of McGhee v National Coal Board [1973] 1 WLR 1, and Lord Bridge has told us in Wilsher v Essex Area Health Authority [1988] AC 1074 that McGhee established no new principle of law at all. If we were to accede to the claimants' arguments, we would be distorting the law to accommodate the exigencies of a very hard case. We would be yielding to a contention that all those who have suffered injury after being exposed to a risk of that injury from which someone else should have protected them should be able to recover compensation even when they are quite unable to prove who was the culprit.

    For the reasons which I have given I am unable, with respect, to agree with the opinion of the Court of Appeal that the claimants were unable to prove on the balance of probabilities that there was a causative relationship between a period of employment during which there was exposure to asbestos dust at the onset of the disease. The Court of Appeal also stated, at pp 1080-1081, in para [104]:

    In our judgment, what Lord Reid said in McGhee [1973] 1 WLR 1, at pp 4G to 5B, in the passage of his speech which forms the basis of Mr Langstaff's argument, must be read in the context of that case. There was only one causative agent, brick dust, and only one possible tortfeasor, Mr McGhee's employer. In that situation Lord Reid was prepared to find causation established in the absence of scientific proof. Lord Bridge in Wilsher has made it clear that the same technique cannot be used where there is more than one causative agent. It seems to us that for the same reasons the same technique cannot be used where there is more than one tortfeasor, unless of course the evidence establishes, on the balance of probabilities, that the disease was caused by cumulative exposure.

    I am unable to agree with this view because I consider that where there is only one causative agent (in this case asbestos dust) the McGhee principle can apply notwithstanding that there are a number of tortfeasors.

  135. Therefore applying the McGhee principle I concluded that the breach of duty by each defendant materially increasing the risk of the onset of mesothelioma in Mr Fox, Mr Fairchild and Mr Matthews involved a substantial contribution to the disease suffered by them and it was for this reason that I allowed the appeals.

  136. I observe that no argument was addressed to the House that in the event of the claimants succeeding there should be an apportionment of damages because the breaches of duty of a number of employers had contributed to cause the disease and therefore the damages awarded against a defendant should be a proportion of the full sum of damages which the claimant would have recovered if he (or the claimant's husband) had been employed by only one employer for the whole of his working life. Therefore each defendant is liable in full for a claimant's damages, although a defendant can seek contribution against another employer liable for causing the disease.

  137. In my respectful opinion in Wilsher the House was right to hold that the majority of the Court of Appeal should not have extended the McGhee principle to apply where there were five possible candidates which could have caused the plaintiff's blindness. I consider that, as the Vice-Chancellor observed at [1987] 730, 780E-F, the justification for holding a defendant liable is that the defendant created a risk and that the injury suffered by the plaintiff fell squarely within that risk. Subject to this observation on the decision in Wilsher, I wish to confine my opinion to the circumstances of these cases. It may be necessary in the future to consider whether the McGhee principle should be applied to other cases, but such decisions will have to be taken when such cases arise.

    Lord Rodger of Earlsferry

    My Lords,

  138. These appeals concern the liability of employers for mesothelioma which employees developed as a result of being exposed to asbestos dust in the course of their employment with more than one employer. The Court of Appeal held ([2002] 1 WLR 1052) that the claimants had not proved that the defendants whom they had sued had caused or materially contributed to the mesothelioma afflicting the men in respect of whom the actions had been raised. Their claims accordingly failed, as all similar claims in courts throughout the United Kingdom would have failed if the Court of Appeal's approach had prevailed. On 16 May 2002, however, your Lordships reported to the House that the appeals should be allowed. I agree with the reasons given in the speech of my noble and learned friend, Lord Bingham of Cornhill, for allowing the appeals, but I have thought it right to explain in my own words why I differed from the unanimous judgment of the Court of Appeal on this important matter.

  139. I gratefully adopt the account of the facts and issues given by Lord Bingham of Cornhill. As he has explained, the Court of Appeal reached their conclusion on the basis of the expert medical evidence, which was essentially the same in each of the cases. Mesothelioma arises when one of the mesothelial cells in the pleura is damaged and undergoes a malignant transformation. While it is certain that asbestos fibres play a role in triggering this transformation, it is not known whether the triggering mechanism involves a single asbestos fibre or multiple asbestos fibres. In the state of medical knowledge today, the one is as likely as the other. Once the malignant transformation has been triggered, the malignancy develops in a process that may involve some six or seven genetic changes. Asbestos may have a role to play at one or more of these stages, but again there is no evidence that the action of asbestos is more or less likely at one stage than at another. In summary ([2002] 1 WLR 1052, 1064, para [26]) therefore,

    it could not be said whether a single fibre of asbestos was more or less likely to have caused the disease, alternatively whether more than one fibre was more or less likely to have caused the disease. In the latter event, it could not be shown that it was more likely than not that those fibres came from more than one source. In other words, none of these scenarios could be proved on the balance of probabilities. Similarly, it could not be proved on the balance of probabilities that any one man's mesothelioma was caused cumulatively by exposure to asbestos dust in more than one employment.

  140. Since it is possible that mesothelioma is triggered by a single fibre of asbestos and since the disease, once contracted, is not made worse by the extent of any subsequent exposure to asbestos dust, the disease differs, for example, from pneumoconiosis or asbestosis where the severity of the condition will depend, to some extent at least, on the amount of dust to which the victim has been exposed. In pneumoconiosis and asbestosis, therefore, by contrast with mesothelioma, it can be said in an appropriate case that the severity of the particular victim's condition will have been caused cumulatively by exposure to dust in more than one employment.

  141. Equally importantly, however, "the risk that mesothelioma will occur increases in relation to the total dose of asbestos received" ([2002] 1 WLR 1052, 1064A, para [25]). Why this should be so is not known. It may be because the build-up of asbestos progressively weakens and eventually overwhelms the body's defences, with the result that a mesothelioma then develops. But this is now thought to be unlikely. The alternative explanation would seem to be that the greater the number of asbestos fibres taken into the body, the greater are the chances that one of them will trigger a malignant transformation. Whatever may be the reason for the increase in risk, since all the defendants exposed the men concerned to an atmosphere containing a substantial amount of asbestos dust, they all increased the risk that the men would develop mesothelioma.

  142. All of the men were employed by more than one employer. This is, of course, typical of the employment patterns of men who have been exposed to asbestos, whether in fitting out or stripping buildings and ships or in casual dock labour, handling cargoes of asbestos. Mesothelioma takes at least ten, and up to forty, years to manifest itself. When the disease does eventually show itself, therefore, the men inevitably find it difficult to identify more than a few of the employers for whom they worked; after so many years employment records tend to be relatively few and far between. The claimants in these cases accordingly sued only those employers whom they could identify and who - or whose successors - were still in existence. They could do no more.

  143. Because of the current state of medical knowledge about the aetiology of mesothelioma, it was impossible for the claimants to prove on the balance of probabilities that the men's illness had been triggered by a fibre or fibres inhaled while working with any particular employer and, more especially, while working with the particular defendants whom they had sued. For that reason the Court of Appeal rejected their claims. The claimants thus failed because of the particular stage which medical science has reached. Research has gone far enough for scientists to be able to see, at a microscopic level, what the possible mechanisms may be, but not far enough for them to determine which is the one that actually operates. In future more may be known. As Mr Stewart rightly observed, in the course of submissions that were both helpful and sensitive, this may change the way in which the law treats such cases. But the House must deal with these appeals on the basis of the evidence as to medical knowledge today and leave the problems of the future to be resolved in the future.

  144. At the hearing counsel for the appellants simply contended that the appeals should be allowed and that the respondents should be found liable in damages. The contentions for the respondents were to the opposite effect. No argument was advanced to the effect that, if they were held liable, the defendants' liability should be reduced or apportioned in some way to reflect an assessment of the chances that the particular defendants, rather than other employers, were actually responsible for the exposure that led to the development of the claimants' mesothelioma. Indeed it was said that no such assessment, even on a rough basis, was possible. Counsel accepted accordingly that, if liable at all, the defendants would be jointly and severally liable for the whole of the damage done to the claimants.

  145. Cases on the liability of employers to their employees who develop industrial diseases have been coming before the courts for many years. At the hearing counsel cited a number of well-known decisions of your Lordships' House in this sphere. Although the battles in those cases were hard fought, both for the employees and for the employers, for the most part the descendants of the protagonists now acknowledge in tranquillity what was won and what was lost in the struggle: the resulting decisions have long been accepted as shaping the law that governs these matters. McGhee v National Coal Board [1973] 1 WLR 1; 1973 SC(HL) 37 is an exception. It dealt with the issue of causation in a case where, as here, the medical experts could not fully explain how the pursuer's condition had started. Controversy persists as to why the pursuer, having lost his claim for damages for dermatitis in the Court of Session, won it in this House. In Wilsher v Essex Area Health Authority [1988] AC 1074 the House sought, with only limited success, to put an end to that controversy. Giving the unanimous views of the Appellate Committee, Lord Bridge indicated that in McGhee the pursuer had won his case simply because the House had been prepared to draw from the primary facts an inference that the defenders' wrongful failure to provide showers had materially contributed to the onset of his dermatitis. On that basis McGhee laid down no new legal principle and was nothing more than a decision on a matter of fact, of no relevance for the present case. On behalf of the appellants Sir Sydney Kentridge and Mr Langstaff questioned that explanation of McGhee, however. They submitted that the majority of their Lordships had proceeded on the principle that, in the circumstances of that case, proof that the defenders' wrongdoing had increased the risk of the pursuer developing dermatitis was to be equated with proof that the defenders had materially contributed to his dermatitis. If applied to the facts in the present cases, that principle would determine the appeals in favour of the appellants. For their part, counsel for the respondents frankly, and in my view correctly, acknowledged that if the House applied that approach in these cases, then the appeals must indeed succeed. Counsel contended, however, that the House should follow the interpretation of McGhee adopted in Wilsher. On that basis the appeals must fail since, on the evidence, the claimants had not proved that the defendants had caused or materially contributed to their mesothelioma. The proper interpretation of McGhee is therefore critical to the disposal of these appeals.

  146. To appreciate the novelty of the issue confronting the House in McGhee and the possible significance of that decision for the present proceedings, it is necessary to go back to Bonnington Castings Ltd v Wardlaw [1956] AC 613, the starting-point of much of the law in this field. The pursuer sued his employers for damages for contracting pneumoconiosis as a result of inhaling air containing minute particles of silica. The evidence showed that the dust in the workshop was caused by the operation of two types of equipment, the pneumatic hammer at which the pursuer worked and certain swing grinders. Given the technology available at the time, the defenders were not at fault in the operation of the pneumatic hammer nor in failing to extract the dust which it threw off. They were, however, at fault inasmuch as the apparatus for intercepting the dust from the swing grinders was choked. The First Division of the Court of Session (Lord President Clyde, Lord Carmont and Lord Russell) held that the pursuer could succeed only if he could show that the sole source of his disease was the dust from the swing grinders. The majority (the Lord President dissenting) applied a reverse onus of proof and so held that the pursuer had indeed proved this (1955 SC 320). The defenders appealed and the House held that the majority had been wrong to apply the reverse onus of proof. Lord Reid said ([1956] AC 613, 620):

    In my judgment, the employee must in all cases prove his case by the ordinary standard of proof in civil actions: he must make it appear at least that on a balance of probabilities the breach of duty caused or materially contributed to his injury.

    He went on, however, to hold that the pursuer did not need to show that the dust from the swing grinders had been the sole source of his illness. It appeared to Lord Reid ([1956] AC 613, 621) that:

    the source of his disease was the dust from both sources, and the real question is whether the dust from the swing grinders materially contributed to the disease. What is a material contribution must be a question of degree. A contribution which comes within the exception de minimis non curat lex is not material, but I think that any contribution which does not fall within that exception must be material. I do not see how there can be something too large to come within the de minimis principle but yet too small to be material.

    Lord Keith of Avonholm observed ([1956] AC 613, 626):

    It was the atmosphere inhaled by the pursuer that caused his illness and it is impossible, in my opinion, to resolve the components of that atmosphere into particles caused by the fault of the defenders and particles not caused by the fault of the defenders, as if they were separate and independent factors in his illness, as if they were separate and independent factors in his illness. Prima facie the particles are acting cumulatively, and I think the natural inference is that had it not been for the cumulative effect the pursuer would not have developed pneumoconiosis when he did and might not have developed it at all.

    The House held that it had been established that the swing grinders had indeed contributed a quota of dust that was not negligible and that they had therefore helped to produce his disease. On that footing the House dismissed the appeal.

  147. As Lord Reid noted ([1956] AC 613, 620), there was nothing new in the idea that the pursuer required to prove no more than that the defenders' wrongful act had materially contributed to his injury. Although he did not refer to it, there was ample authority for the proposition in both English and Scots law. For instance, in Senior v Ward (1859) 1 El & El 385, 392, where the owner of a coal mine who superintended its operation had failed to test the cable of a cage as required by statutorily approved rules, Lord Campbell CJ, sitting in the Queen's Bench, would have held him liable, even though the banksman had also been at fault, because the mine owner's own negligence had "materially contributed to the death of the deceased." In fact, however, the court held that the plaintiff's action failed because the deceased's own negligence had "materially contributed to his death". In Wakelin v London & South Western Railway Co (1886) 12 App Cas 41, 47 Lord Watson stated that the liability of a defendant must rest in the first place on there being

    some negligent act or omission on the part of the company or their servants which materially contributed to the injury or death complained of…. Mere allegation or proof that the company were guilty of negligence is altogether irrelevant; they might be guilty of many negligent acts or omissions, which might possibly have occasioned injury to somebody, but had no connection whatever with the injury for which redress is sought, and therefore the plaintiff must allege and prove, not merely that they were negligent, but that their negligence caused or materially contributed to the injury.

  148. Similarly, in Craig v Glasgow Corporation 1919 SC(HL) 1, 6 where the pursuer sought damages for personal injuries caused by a tramcar driven by one of the defenders' employees, Lord Buckmaster said:

    It is a common and familiar principle that in an action seeking such relief it is incumbent upon the pursuer to prove both that the defenders were guilty of negligence, and that such negligence caused, or materially contributed to, the injuries received.

  149. The idea of liability based on wrongful conduct that had materially contributed to an injury was therefore well established long before Wardlaw. But Wardlaw became a convenient point of reference, especially in cases of industrial disease. In such cases this basis of liability is of considerable importance. Since it is enough that the defendant's wrongful act materially contributed to the claimant's injury, the law is not applying the causa sine qua non or "but for" test of causation. In Wardlaw, for instance, the pursuer did not need to prove that, but for the dust from the swing hammers, he would not have developed pneumoconiosis All he needed to prove was that the dust from the swing hammers contributed materially to the dusty atmosphere which he breathed and which caused his illness. As will be seen below, in the Court of Session in McGhee the judges lost sight of this important point.

  150. The House quickly applied this approach in two Scottish appeals that were heard one after the other: Quinn v Cameron & Roberton Ltd [1958] AC 9, 23 per Viscount Simonds, 34 per Lord Morton of Henryton, and Nicholson v Atlas Steel Foundry & Engineering Co Ltd [1957] 1 WLR 613. The second is of particular importance for present purposes. The action was at the instance of the relatives of a workman who had died from pneumoconiosis. The man had worked in the defenders' dressing shop where there was a pneumatic hammer the operation of which, lawfully at the time, gave rise to siliceous dust in the atmosphere. There were also two swing grinders, but it was unclear on the evidence whether they had given off dust. The House therefore proceeded on the basis that all the dust in the atmosphere had been caused by the lawful operation of the pneumatic hammer. It was proved, however, that the ventilation in the dressing shop was inadequate in terms of the relevant statutory provisions. As a result the dust remained in the atmosphere for longer than would have been the case if the ventilation had met the statutory requirements. The House followed the decision in Wardlaw and held the defenders liable on the ground that the inadequate ventilation had materially contributed to the man's pneumoconiosis.

  151. Viscount Simonds, with whom Lord Oaksey and Lord Morton of Henryton concurred, set out his approach in these terms ([1957] 1 WLR 613, 618):

    My Lords, I do not think that it can be wrong to approach this question from the angle that, if the statute prescribes a proper system of ventilation by the circulation of fresh air so as to render harmless, so far as practicable, all fumes, dust and other impurities that may be injurious to health, generated in the course of work carried on in the factory, and if it is proved that there is no system or only an inadequate system of ventilation, it requires little further to establish a causal link between that default and the illness, due to noxious dust, of a person employed in the shop. Something is required as was held in Wardlaw's case. I was a party to that decision and would not in any way resile from it. But it must not be pressed too far. In the present case there was, in my opinion, ample evidence to support the appellants' case. The appellants rightly or wrongly by their pleadings assumed the burden not only of proving that there was a causal link between default and disease but also of proving that it was practicable for the respondents to have reduced the risk by the provision of additional roof ventilators and openings in the walls for the admission of fresh air.

  152. Viscount Simonds thus firmly adhered to the decision in Wardlaw, that there could be no question of reversing the burden of proof which lay on a pursuer to establish the necessary causal link between the employer's default and his illness. Nevertheless, he considered that, where a statutory provision prescribed a system of ventilation to remove elements in the atmosphere that might be injurious to health and the employer had no ventilation or only an inadequate system, these factors were relevant in considering how much more the pursuer required to prove in order to establish the causal link between the employer's default and the pursuer's disease due to noxious dust. In other words, these factors were relevant to what or how much the pursuer had to prove in order to establish the causal link, rather than to the attitude that the court should take when assessing the evidence adduced for that purpose. Viscount Simonds went on to consider whether it was practicable for the defenders to have provided a better system of ventilation. He concluded that it was. Therefore the dust containing dangerous particles had hung about in the air in a concentrated form for longer than it would have done if there had been that better system of ventilation. He continued ([1957] 1 WLR 613, 620):

    It follows that owing to the default of the respondents the deceased was exposed to a greater degree of risk than he should have been, and, though it is impossible, even approximately, to quantify the particles of dust which he must, in any event, have inhaled and those which he inhaled but need not have, I cannot regard the excess as something so negligible that the maxim 'de minimis' is applicable. Accordingly, following the decision in Wardlaw's case, I must hold the respondents liable.

  153. Similarly Lord Cohen said ([1957] 1 WLR 613, 622):

    The respondents are admittedly not to blame for the generation of this cloud, but any failure to provide proper ventilation must, I think, lengthen the period during which the cloud remains intense. It seems to me to follow that the respondents' failure to provide adequate ventilation must increase the risk to which the workmen are exposed. Reading the evidence as a whole, I think it establishes that (to use the language of Lord Reid in Wardlaw's case) 'on a balance of probabilities the breach of duty caused or materially contributed to' the injury.

  154. Questions of the risk or increased risk of causing harm are more frequently considered in relation to issues of foreseeability and fault. But these passages are important precisely because they show that the House was proceeding on the basis that, in considering whether the pursuers in Nicholson had proved that the defenders had materially contributed to the deceased's condition, it was relevant to consider whether the defenders' wrongful act had exposed him to an increased risk of disease. In other words, in that case proof of increased risk of harm to the deceased was relevant to proof of causation of that harm.

  155. In McGhee Lord Simon of Glaisdale suggested ([1973] 1 WLR 1, 8H) that the same had been said in the opinion of the Lord Ordinary (Kilbrandon) in another dermatitis case, Gardiner v Motherwell Machinery & Scrap Co Ltd 1961 SC(HL) 1, 3. Lord Kilbrandon recorded the argument on behalf of the pursuer in these terms:

    that the washing facilities which were provided were inadequate and primitive, and that, if they had been up to standard, the risk of dermatitis would have been very much reduced.

    As Lord Simon noted, Lord Kilbrandon accepted that argument. He did so towards the end of his opinion (1961 SC(HL) 1, 4):

    In my opinion, the defenders neglected to supply reasonable washing facilities in the circumstances, and thereby exposed the pursuer to an enhanced risk of contracting industrial dermatitis. They are accordingly liable to him in damages.

    It appears, however, that, both when recording the argument and when giving his decision on it, Lord Kilbrandon was concerned with fault rather than with causation, since he had already held (1961 SC(HL) 1, 3), on the basis of the dermatologists' evidence, that the pursuer had established the necessary causal link between his work with the defenders and the onset of his dermatitis. In these later passages Lord Kilbrandon seems to have been making the point that the defenders' failure to provide adequate washing facilities amounted to legally relevant fault where it was foreseeable that this would lead to an enhanced risk of dermatitis.

  156. Be that as it may, the statements in the speeches of Viscount Simonds and Lord Cohen in Nicholson had stood unchallenged for some 15 years when the House came to consider McGhee v National Coal Board, again an appeal from the Court of Session. The case involved a claim for dermatitis which the pursuer had developed after working in a hot and dusty atmosphere, pulling bricks out of the defenders' kiln. At the relevant time there were no showers at the works and so, when his working day was over, the pursuer had to cycle home without washing the sweat, dust and grit from his body. As indeed often happens, the point upon which the decision was ultimately to turn emerged only after the Lord Ordinary had rejected much of the rest of the pursuer's case. The pursuer began the proof before answer with both a common law and a statutory case to the effect that the defenders should have provided a system of ventilating the kiln. He disclaimed the statutory case at the hearing on evidence. The Lord Ordinary (Kissen) rejected the common law case. He also rejected the pursuer's case that the defenders should have taken reasonable care to see that the kiln had cooled sufficiently before the pursuer and his colleagues were sent to work in it.

  157. The pursuer had a further common law case which he set out at the end of article 3 of condescendence in these terms:

    It was their duty to take reasonable care to provide adequate washing facilities including showers, soap and towels to enable men to remove dust from their bodies. In each and all of said duties the defenders failed and so caused said disease. Had the defenders fulfilled said duties incumbent on them the pursuer would not have contracted said disease.

    The final averments in this passage relate, of course, to the alleged failure by the defenders to perform all the common law duties averred by the pursuer, including those that the Lord Ordinary had rejected. The Lord Ordinary held that the pursuer had indeed contracted his dermatitis in the course of his work in the brick kiln and as a result of his exposure to dust and ashes there (1973 SC(HL) 37, 39). He also held that the defenders were at fault in not providing showers but that the pursuer had failed to prove on the balance of probabilities that this fault on the part of the defenders had caused or materially contributed to his injury (1973 SC(HL) 37, 42 - 43). Lord Kissen put the point in this way:

    Dr Hannay's evidence was that he could not say that the provision of showers would probably have prevented the disease. He said that it would have reduced the risk materially but he would not go further than that. Dr Ferguson said that washing reduces the risk. Pursuer's counsel maintained that a material increase in the risk of contracting a disease was the same as a material contribution to contracting the disease and that Dr Hannay established this by his evidence. I think that defenders' counsel was correct when he said that the distinction drawn by Dr Hannay was correct and that an increase in risk did not necessarily mean a material contribution to the contracting of the disease. The two concepts are entirely different. A material increase in risk may refer only to possibilities and may not make a possibility into a probability. It may strengthen the possibility but that cannot mean that in all such cases the possibility has become a probability. What the pursuer has to show is that, as he avers, he would not have contracted the disease but for the defenders' breach of duty. He has to show that this was probable and the degrees of risk have no relevance unless they make the contraction of the disease more probable than not contracting the disease. He cannot succeed if the only inference from the evidence is that lack of shower baths is a possibility as a cause of his having contracted the disease and that the provision of shower baths would have increased the possibility but not made it a probability. That is the only inference which I can draw from Dr Hannay's evidence and that was the best evidence for the pursuer. Causal connection between fault and the contraction of the disease has not been established.

    [emphasis added]

  158. The pursuer reclaimed but the First Division (Lord President Clyde, Lord Migdale and Lord Johnston) refused the reclaiming motion. The Division proceeded on the basis that, in accordance with his averments in article 3 of condescendence, the pursuer had to prove that the provision of showers would have prevented his dermatitis (1973 SC(HL) 37, 43 per the Lord President, 46 - 47 per Lord Migdale). They also took the view that the effects of the heat and dust had come to an end when the pursuer left the kiln so that the failure to provide washing facilities could have played no part in the development of the pursuer's condition (1973 SC(HL) 37, 44 per Lord President Clyde, 47 per Lord Migdale, with whom Lord Johnston somewhat hesitantly agreed). Since all the damage had been done in the kiln, a shower could not have obviated or lessened that damage (1973 SC(HL) 37, 48 per Lord Migdale). In addition, however, in a passage (1973 SC(HL) 37, 45) that Lord Salmon was specifically to disapprove ([1973] 1 WLR 1, 11E - F) the Lord President held that

    Even if the pursuer had established (as he did not) that the absence of washing facilities increased the risk of the pursuer getting dermatitis, that would clearly not prove that the absence of these facilities caused the disease, nor indeed would it go any distance towards proving it. For risk of dermatitis and causation of dermatitis are two quite separate matters.

  159. This House analysed both the evidence and the pursuer's case rather differently from the Court of Session. The medical witnesses were agreed that the defenders' failure to provide washing facilities increased the risk of the pursuer developing dermatitis. In this sense, and in this sense only, the abrasion to the pursuer's skin was cumulative: "the longer a subject is exposed to injury the greater the chance of his developing dermatitis" ([1973] 1 WLR 1, 3H - 4A per Lord Reid). The fact that the risk did increase in this way if no washing facilities were provided was important because it showed that, contrary to the view taken by the First Division, the pursuer had not left behind the causes which made him liable to develop dermatitis when he left the brick kiln ([1973] 1 WLR 1, 4H - 5A per Lord Reid).

  160. On the other hand, the evidence did not show just how the type of dermatitis affecting the pursuer began. Lord Reid summarised the effect of the evidence in this way ([1973] 1 WLR 1, 4F - G):

    It suggests to me that there are two possible ways. It may be that an accumulation of minor abrasions of the horny layer of the skin is a necessary precondition for the onset of the disease. Or it may be that the disease starts at one particular abrasion and then spreads, so that multiplication of abrasions merely increases the number of places where the disease can start and in that way increases the risk of its occurrence.

    I am inclined to think that the evidence points to the former view. But in a field where so little appears to be known with certainty I could not say that it is proved. If it were, then this case would be indistinguishable from Wardlaw's case.

    In this passage Lord Reid goes into more of the detail of the medical evidence than either the Lord Ordinary or the First Division. From what he says here and elsewhere ([1973] 1 WLR 1, 3G) it is clear, of course, that the pursuer's condition would not have developed but for his work in the kiln. It was there that he was exposed to the heat that made him sweat profusely and so softened his skin and made it easily injured. It was there also that he was exposed to the cloud of brick dust, with the result that the particles adhered to his skin. Exertion then caused the dust to injure the horny layer of his skin and so exposed the tender cells below to injury or infection. The most significant exertion would, presumably, have occurred while the pursuer was working in the kiln, pulling out the bricks. But the pursuer would also have exerted himself while cycling home and, perhaps, when he got home. So any abrasions on his softened skin could have been caused either by the exertion in the kiln or else by the exertion during his ride home or by a combination of the two. For the consequences of the exertion in the kiln alone the defenders would not have been liable since they were not at fault in that respect. For the consequences of the exertion during his cycle ride, the defenders would, on the other hand, have been liable since the pursuer would not have been open to injury due to dust on his skin at that stage if the defenders had fulfilled their duty to provide showers.

  161. On this analysis the pursuer could succeed in his claim only if he could connect his dermatitis with his ride home while unwashed, rather than simply with his work in the kiln. Lord Reid pointed out, however, that the pursuer's pleadings in respect of the duty to provide showers were potentially misleading ([1973] 1 WLR 1, 4C). He did not elaborate the point, but he clearly had in mind the standard-form averment at the end of article 3 of condescendence in which the pursuer offered to prove that, if the defenders had fulfilled their duty to provide showers, he would not have contracted dermatitis. Even though both the Lord Ordinary and the First Division had judged his case on that basis, the pursuer's self-imposed goal was too tough. He did not need to go that far. It would be sufficient and relevant in law if the pursuer showed that the defenders' failure had materially contributed to the development of his dermatitis, whether or not the dermatitis would have developed but for that failure. So, despite the terms of the pursuer's averment, Lord Reid considered his case by simply asking whether the defenders' failure to provide showers had materially contributed to the development of the pursuer's dermatitis. It was at this point that the difficulty in the case came into focus.

  162. According to the medical evidence, the pursuer's dermatitis might have occurred either as a result of an accumulation of abrasions or as a result of a single abrasion. If the former were the case, then one could say that the accumulation of abrasions had resulted from the effects of the pursuer's exertions as a whole. Just as it had been right in Wardlaw to consider the effects of the cloud of dust as a whole and not to break it down into "innocent" and "guilty" particles, so in McGhee it would have been right to consider the effects of the exertion as a whole acting on the dust particles on his skin and not to break the exertion down into "innocent" exertion in the kiln and "guilty" exertion on the way home. Provided that the exertion on the way home was material, it would be proper to hold that the effects of that exertion on the dust still on his unwashed skin materially contributed to cause the pursuer's illness. But the pursuer had only proved that the accumulation of abrasions was one possible mechanism for triggering his dermatitis. The evidence showed that another possibility was that the disease had started from a single abrasion. If that were the case, liability could not be based on the cumulative effect of the pursuer's exertions in producing an accumulation of abrasions. So the analogy of Wardlaw would not apply. But, furthermore, the pursuer could not prove that this hypothetical single abrasion had been caused by his exertion when riding home with the dust adhering to his skin, rather than by his exertions in the kiln. The defenders argued that the pursuer's case must therefore fail because he had not proved, and could not prove, that the defenders' failure to provide showers had caused or materially contributed to the onset of his illness.

  163. Lord Reid dealt with that argument in this way ([1973] 1 WLR 1, 4G - 5B):

    But I think that in cases like this we must take a broader view of causation. The medical evidence is to the effect that the fact that the man had to cycle home caked with grime and sweat added materially to the risk that this disease might develop. It does not and could not explain just why that is so. But experience shows that it is so. Plainly that must be because what happens while the man remains unwashed can have a causative effect, though just how the cause operates is uncertain. I cannot accept the view expressed in the Inner House that once the man left the brick kiln he left behind the causes which made him liable to develop dermatitis. That seems to me quite inconsistent with a proper interpretation of the medical evidence. Nor can I accept the distinction drawn by the Lord Ordinary between materially increasing the risk that the disease will occur and making a material contribution to its occurrence.

    There may be some logical ground for such a distinction where our knowledge of all the material factors is complete. But it has often been said that the legal concept of causation is not based on logic or philosophy. It is based on the practical way in which the ordinary man's mind works in the everyday affairs of life. From a broad and practical viewpoint I can see no substantial difference between saying that what the defender did materially increased the risk of injury to the pursuer and saying that what the defender did made a material contribution to his injury.

    Lord Reid holds that, even though the evidence leaves open the two possible causes of his illness, the pursuer succeeds because the court takes a broader view of causation. There is nothing whatever to suggest that, in reaching this conclusion, Lord Reid was simply adopting a robust and pragmatic approach to the primary facts and holding that it could somehow be inferred - in the face of the expert evidence which he had just narrated and which gave rise to the very problem that had to be resolved - that, if the pursuer's dermatitis was caused by a single abrasion, that particular abrasion was caused by his exertion while riding home. What Lord Reid does, rather, is to accept that the pursuer must prove that the defender's conduct materially contributed to the onset of his illness but also, like Viscount Simonds and Lord Cohen in Nicholson, he considers what it is that the pursuer must prove in order to establish that material contribution. Taking the "broader view of causation", he holds that, in these particular circumstances, there is no substantial difference between saying that what the defenders did materially increased the risk of injury to the pursuer and saying that it made a material contribution to his injury. This is his decision on the point of law posed in the case. Proof that the defenders' failure to provide showers had materially increased the risk that the pursuer would develop dermatitis would therefore be, as a matter of law, sufficient to prove that the defenders had materially contributed to the onset of his condition. And, of course, the pursuer had actually proved that their failure had increased that risk. Lord Reid accordingly allowed the appeal.

  164. My Lords, I have analysed the leading speech of Lord Reid at considerable length. Happily, it is unnecessary to go into the other speeches in the same detail, since it is in my view clear that Lord Wilberforce, Lord Simon of Glaisdale and Lord Salmon all took essentially the same approach as Lord Reid. They were prepared to hold, under reference to the speeches of Viscount Simonds and Lord Cohen in Nicholson, that in the particular circumstances, by proving that the defenders' negligent failure to provide showers had increased the risk that he would develop dermatitis, the pursuer had proved that the defenders' failure had materially contributed to his dermatitis. I do not take up space by repeating the relevant passages, all of which have been set out by Lord Bingham of Cornhill. I refer in particular to the long passage towards the end of the speech of Lord Wilberforce ([1973] 1 WLR 1, 6G - 7F), which is by no means predicated on the reference to a possible reversal of the onus of proof that the House detected and disapproved in Wilsher [1988] AC 1074, 1087G. In McGhee indeed any reference to the possibility of proof by the defenders was necessarily obiter since, ex hypothesi, there was no relevant medical evidence which they could have led. Despite his reference to "sufficient prima facie evidence", Lord Simon's approach is similar ([1973] 1 WLR 1, 8D - 9B). The same goes for two passages in Lord Salmon's speech where he specifically rejects the distinction between materially increasing the risk of contracting industrial disease and materially contributing to causing it ([1973] 1 WLR 1, 11H - 12A, 12G - 13A). Although Lord Kilbrandon reached the same conclusion as the rest of the House, his exact approach is harder to determine. Like the judges in the Court of Session, however, he appears to have taken the view that the pursuer needed to prove that, but for the defenders' failure to provide showers, he would not have contracted dermatitis ([1973] 1 WLR 1, 10G). For present purposes it is unnecessary to investigate his minority view in more detail.

  165. While Lord Wilberforce's observations on the burden of proof have been disapproved, the decision itself in McGhee has not been overruled. Nor indeed did counsel for the respondents suggest that it should now be overruled. It may well be that in McGhee the members of the House could have proclaimed more clearly and more openly that they were stating a new principle. If so, the possible virtues of a certain reticence in such matters were famously recognised by Lord Devlin (The Judge (1979), p 12). But the risk was, of course, that opinions couched in this way might be misconstrued. As, in my view, did indeed happen. In Wilsher v Essex Area Health Authority [1988] AC 1074 the House examined McGhee and came to the conclusion that the decision rested not on any legal principle but on nothing more than "a robust and pragmatic approach" to the facts of the case. In my respectful opinion, that conclusion was inconsistent with the terms of the speeches in McGhee.

  166. Lord Bridge, with whom the other members of the House agreed, said ([1988] AC 1074, 1088A - B) that in McGhee

    the consecutive periods when 'innocent' and 'guilty" brick dust was present on the pursuer's body may both have contributed to the cause of the disease or, theoretically at least, one or other may have been the sole cause. But where the layman is told by the doctors that the longer the brick dust remains on the body, the greater the risk of dermatitis, although the doctors cannot identify the process of causation scientifically, there seems to me to be nothing irrational in drawing the inference, as a matter of common sense, that the consecutive periods when brick dust remained on the body probably contributed cumulatively to the causation of the dermatitis. I believe that a process of inferential reasoning on these general lines underlies the decision of the majority in McGhee's case.

    Lord Bridge then quoted the relevant passages from the speeches of their Lordships, apart from Lord Wilberforce, in McGhee and went on ([1988] AC 1074, 1090):

    The conclusion I draw from these passages is that McGhee v National Coal Board [1973] 1 WLR 1 laid down no new principle of law whatever. On the contrary, it affirmed the principle that the onus of proving causation lies on the pursuer or plaintiff. Adopting a robust and pragmatic approach to the undisputed primary facts of the case, the majority concluded that it was a legitimate inference of fact that the defenders' negligence had materially contributed to the pursuer's injury. The decision, in my opinion, is of no greater significance than that and to attempt to extract from it some esoteric principle which in some way modifies, as a matter of law, the nature of the burden of proof of causation which a plaintiff or pursuer must discharge once he has established a relevant breach of duty is a fruitless one.

    In the present case the Court of Appeal naturally had regard to what Lord Bridge said in this passage: [2002] 1 WLR 1052, 1080F - G, para [103].

  167. Although Lord Bridge took the opportunity of the appeal in Wilsher to comment on McGhee, the two cases were very different. In Wilsher the plaintiff had been born prematurely and placed in a special unit in a hospital managed by the defendants. It was necessary for him to be given extra oxygen but, unfortunately, due to mistakes on their part the staff administered too much oxygen. The plaintiff developed retrolental fibroplasia, which resulted in blindness. He claimed damages from the defendants on the basis that his condition had been caused by the unduly high level of oxygen. The evidence in the case showed that exposure to such a high level of oxygen increased the risk that the plaintiff would suffer retrolental fibroplasia. The evidence also showed, however, that the condition could occur in premature babies without any artificial administration of oxygen. More particularly, there was evidence to indicate a correlation between the occurrence of retrolental fibroplasia and four other conditions from which the plaintiff, like many other premature babies, suffered. Medical science had not, however, positively identified any causal mechanism linking those conditions and the development of retrolental fibroplasia.

  168. In the Court of Appeal Mustill LJ, with whom Glidewell LJ agreed, found in favour of the plaintiff. He acknowledged that the facts of Wilsher differed from those of McGhee in an important respect ([1987] QB 730, 771D - 772C):

    In the McGhee case there was only one risk operating, namely that contact of a sweaty skin with brick dust would lead to dermatitis. The fact that such contact did cause the injury was not in dispute. Just as in Bonnington Castings Ltd v Wardlaw [1956] AC 613 the defenders' fault lay in not taking proper steps to reduce that single risk. The uncertainty was whether the fault had tipped the scale. In the present case there is a greater uncertainty. Instead of a single risk factor known to have caused the injury there is a list of factors, which cannot be fully enumerated in the current state of medical science, any one of which might have caused the injury. What the defendants did was not to enhance the risk that the known factor would lead to injury, but to add to the list of factors which might do so. I acknowledge that this is much further from the facts of Bonnington Castings Ltd v Wardlaw, which was the springboard for the McGhee case than were the facts of the McGhee case itself.

    The question is whether this makes a crucial difference. The root of the problem lies in the fact that, for reasons of policy, their Lordships'House mitigated the rigour of the rule that the plaintiff must prove that the breach caused the loss, in the interests of achieving a result which was considered to be just. Given that this was a decision based on policy, rather than a chain of direct reasoning, the difficulty is to know whether a similar approach can properly be adopted in the different circumstances of the present case. After much hesitation I have come to the conclusion that it can. Reading all the speeches together, the principle applied by the House seems to me to amount to this. If it is an established fact that conduct of a particular kind creates a risk that injury will be caused to another or increases an existing risk that injury will ensue; and if the two parties stand in such a relationship that the one party owes a duty not to conduct himself in that way; and if the first party does conduct himself in that way, and if the other party does suffer injury of the kind to which the risk related; then the first party is taken to have caused the injury by his breach of duty, even though the existence and extent of the contribution made by the breach cannot be ascertained. If this is the right analysis, it seems to me that the shape taken by the enhancement of the risk ought not to be of crucial significance. In the McGhee case [1973] 1 WLR 1, the conduct of the employers made it more likely that the pursuer would contract dermatitis, and he did contract dermatitis. Here, the conduct of those for whom the defendants are liable made it more likely that Martin would contract RLF, and he did contract RLF. If considerations of justice demanded that the pursuer succeed in the one case, I can see no reason why the plaintiff should not succeed in the other.

  169. Sir Nicolas Browne-Wilkinson V-C dissented. In a passage with which Lord Bridge was quite unable to find any fault ([1988] AC 1074, 1090E) he pointed out that to apply the "principle" in McGhee to the facts of Wilsher would constitute an extension of that principle. He then summarised the facts of McGhee in this way ([1987] QB 730, 779B - C):

    In the McGhee case there was no doubt that the pursuer's dermatitis was physically caused by brick dust: the only question was whether the continued presence of such brick dust on the pursuer's skin after the time when he should have been provided with a shower caused or materially contributed to the dermatitis which he contracted. There was only one possible agent which could have caused the dermatitis, viz, brick dust, and there was no doubt that the dermatitis from which he suffered was caused by that brick dust.

    Having stated the issues in Wilsher, the Vice-Chancellor continued:

    The position, to my mind, is wholly different from that in the McGhee case [1973] 1 WLR 1, where there was only one candidate (brick dust) which could have caused the dermatitis, and the failure to take a precaution against brick dust causing dermatitis was followed by dermatitis caused by brick dust. In such a case, I can see the common sense, if not the logic, of holding that, in the absence of any other evidence, the failure to take the precaution caused or contributed to the dermatitis. To the extent that certain members of the House of Lords decided the question on inferences from evidence or presumptions, I do not consider that the present case falls within their reasoning. A failure to take preventative measures against one out of five possible causes is no evidence as to which of those five caused the injury.

  170. Adopting the reasoning of the Vice-Chancellor, the House reversed the decision of the Court of Appeal - and rightly so. Mustill LJ's extension of the approach in McGhee to a situation where there were all kinds of other possible causes of the plaintiff's condition, resulted in obvious injustice to the defendants. In particular, there was nothing to show that the risk which the defendants' staff had created - that the plaintiff would develop retrolental fibroplasia because of an unduly highly level of oxygen - had eventuated. That being so, there was no proper basis for applying the principle in McGhee. As Sir Nicolas Browne-Wilkinson decisively observed, a failure to take preventive measures against one of five possible causes was no evidence as to which of those five had caused the injury. The reasoning of the Vice-Chancellor, which the House adopted, provided a sound and satisfactory basis for distinguishing McGhee and for allowing the appeal. The year before indeed, the House had distinguished McGhee on the basis that a pursuer had not proved that the junior house officer's mistake had materially increased the risk of the particular kind of neurological damage suffered by his son: Kay's Tutor v Ayrshire & Arran Health Board 1987 SC(HL) 145; [1987] 2 All ER 417.

  171. Instead of distinguishing the decision in McGhee, however, Lord Bridge chose to gloss it in a way that does not do justice to the reasoning in the speeches, with the possible exception of Lord Kilbrandon's. Contrary to Lord Bridge's view that nothing more was involved than an issue of fact, Lord Reid had expressly stated that the pursuer's case raised "a difficult question of law" ([1973] 1 WLR 1, 3D). Even with the benefit of the further analysis of my noble and learned friend, Lord Hutton, I am satisfied that it was this question of law that Lord Reid and the other judges proceeded to elucidate and to determine, having regard to the various policy issues that arose. Taking account of such matters in formulating a legal test is usual and legitimate. By contrast, using them as a basis for modifying the way in which a court assesses evidence and finds facts, as Lord Bridge appears to suggest, is at best questionable. A judge applying "a robust and pragmatic approach to the undisputed primary facts" may all too readily stray beyond the realm of inference into the wilderness of "mere speculation or conjecture": Caswell v Powell Duffryn Associated Collieries Ltd [1940] AC 152, 169 - 170 per Lord Wright. In particular, even though it is always for the judge rather than for the expert witness to determine matters of fact, the judge must do so on the basis of the evidence, including the expert evidence. The mere application of "common sense" cannot conjure up a proper basis for inferring that an injury must have been caused in one way rather than another when the only relevant evidence is undisputed scientific evidence which says that either way is equally possible. In my respectful opinion, therefore, despite the criticism of Sopinka J in Snell v Farrell [1990] 2 SCR 311, 331f - g, Lord Wilberforce was right to say that using inference to bridge the evidential gap in McGhee would have been something of a fiction since it was precisely that inference which the medical expert had declined to make ([1973] 1 WLR 1, 7E). In any event, the gloss added by the House in Wilsher, highly influential though it has subsequently proved to be, cannot in itself supplant the reasoning of the judges in McGhee. In the circumstances I too take the view that these observations in Wilsher should no longer be regarded as authoritative.

  172. Although counsel for the respondents accepted that, if reinstated, the principle in McGhee would govern the present appeals and would mean that the appeals would have to be allowed, it is worth noticing why that should be so.

  173. The parallels between the cases are striking. In McGhee the defenders had negligently failed to provide showers to remove sweat and dust; in these cases the defendants had negligently failed to provide protection from asbestos dust. In McGhee, by removing the sweat and dust, the showers would have been intended to guard the workmen against suffering skin diseases such as dermatitis; here the protection against inhaling asbestos dust would have been intended to guard the workmen against suffering asbestos-related illnesses such as mesothelioma. In McGhee the failure to provide the showers materially increased the risk of the pursuer developing dermatitis; in these cases the failure to protect against inhaling asbestos dust materially increased the risk of the claimants developing mesothelioma. In McGhee the pursuer developed dermatitis due to the presence of dust and sweat on his skin, while in these cases the claimants developed mesothelioma due to inhaling asbestos dust. In McGhee it was not possible in the state of medical knowledge for the pursuer to prove in the usual way whether the dermatitis started because of a single abrasion or because of multiple abrasions. Here in the state of medical knowledge it is not possible for the claimants to prove whether the mesothelioma started from the effect of a single fibre or from the effect of multiple fibres. In McGhee it was simply not possible for the pursuer to prove that his dermatitis was caused by an accumulation of abrasions; similarly, here it is simply not possible for the claimants to prove that their mesothelioma was caused by an accumulation of asbestos fibres. In McGhee it was not possible for the pursuer to prove that the hypothetical single abrasion had been caused at a time after he should have had a shower and was cycling home. Here it is not possible for the claimants to prove that the hypothetical single fibre had been inhaled while they were working with any particular employer and especially while they were working with any of the defendants.

  174. In one respect, of course, the cases diverge. In McGhee the only possible source of the dust and sweat was the National Coal Board's kiln and the only possible wrongdoers were the Coal Board. Here, by contrast, the defendants are simply some among a number of employers who negligently exposed the claimants to asbestos dust. The Court of Appeal attached some importance to this distinction: [2002] 1 WLR 1052, 1080H - 1081A. On closer inspection, however, the distinction does not appear to be material for present purposes. The important point is that in both cases the state of scientific knowledge makes it impossible for the victim to prove on the balance of probabilities that his injury was caused by the defenders' or defendants' wrongdoing rather than by events of a similar nature which would not constitute wrongdoing on their part. Therefore, if the principle applies to permit the pursuer to recover in McGhee, it should similarly apply to allow the claimants to recover in these cases. Indeed, on one view the principle is easier to apply in the present cases than in McGhee since it is not disputed that the men developed mesothelioma as a result of a tort by one of their employers. The claimants thus have all the necessary elements for a successful claim except, it is argued, proof of causation. In McGhee, on the other hand, it was possible that the pursuer's dermatitis had been prompted purely by his exposure to dust in the kiln and by his exertions there, for which the employers would not have been liable in delict. So, application of the principle was crucial to connect the pursuer's illness not just with the defenders' legal wrong but with any legal wrong at all. In that sense these cases are a fortiori McGhee.

  175. The decision in McGhee undoubtedly involved a development of the law relating to causation. The context within which that development falls to be evaluated is described in the observations on principle in the speeches of my noble and learned friends, Lord Nicholls of Birkenhead and Lord Hoffmann, to which I respectfully refer.

  176. As counsel for the defendants submitted, the principle in McGhee involves an element of rough justice, since it is possible that a defendant may be found liable when, if science permitted the matter to be clarified completely, it would turn out that the defendant's wrongdoing did not in fact lead to the men's illness. That consideration weighed with the Court of Appeal ([2002] 1 WLR 1052, 1080E - H, para [103]). It must be faced squarely. The opposing potential injustice to claimants should also be addressed squarely. If defendants are not held liable in such circumstances, then claimants have no claim, even though, similarly, if the matter could be clarified completely, it might turn out that the defendants were indeed the authors of the men's illness. Other considerations colour the picture. The men did nothing wrong, whereas all the defendants wrongly exposed them to the risk of developing a fatal cancer, a risk that has eventuated in these cases. At best, it was only good luck if any particular defendant's negligence did not trigger the victim's mesothelioma. The defendants, in effect, say that it is because they are all wrongdoers that the claimants have no case. In other words: the greater the risk that the men have run at the hands of successive negligent employers, the smaller the claimants' chances of obtaining damages. In these circumstances, one might think, in dubio the law should favour the claimants. Moreover, in McGhee the House did nothing more than set the requirement of proof at the highest that the pursuer could possibly attain - hardly a relaxation in any real sense. He had proved all that he could and had established that the defenders' wrongdoing had put him at risk of the very kind of injury which befell him. To require more would have been to say that he could never recover for his injury - unless he achieved the impossible. Finally, as was recognised in McGhee ([1973] 1 WLR 1, 9B - C per Lord Simon, 12G per Lord Salmon), if the law did indeed impose a standard of proof that no pursuer could ever satisfy, then, so far as the civil law is concerned, employers could with impunity negligently expose their workmen to the risk of dermatitis - or, far worse, of mesothelioma. The substantive duty of care would be emptied of all practical content so far as victims are concerned. In my view considerations of these kinds justified the House in developing the approach of Viscount Simonds and Lord Cohen in Nicholson to fashion and apply the principle in McGhee. A fortiori they justify the application of that principle in the present case where the risk to the men was so much worse.

  177. I derive support for that conclusion from what has been done in other legal systems. In the course of the hearing counsel for both sides referred to authorities from a number of different jurisdictions. It would be impossible to do justice to all of them in this opinion. Broadly speaking, they appear to me to demonstrate two things:

    • first, that other systems have identified the need to adopt special rules or principles to cope with situations where the claimant cannot establish which of a number of wrongdoers actually caused his injury;

    • secondly, that there are considerable divergences of view and indeed uncertainty as to the proper area within which any such special rules or principles should apply.

    I have simply selected a few among the many authorities cited by counsel.

  178. Perhaps the most telling illustration of the universality of the problem with which the House is faced comes from two passages in the Digest cited by Sir Sydney. The issues in the texts have been helpfully analysed in a recent study by Mr Jeroen Kortmann, "Ab alio ictu(s): Misconceptions about Julian's View of Causation" (1999) 20 Journal of Legal History 95. The texts show that, in a certain form, problems with unidentifiable wrongdoers had begun to exercise the minds of Roman jurists not later than the first century BC.

  179. D 9 2 51 Julian 86 digesta contains a substantial extract from one of the most important works on Roman law, written in the second century AD, the high classical period of Roman law. In the principium Julian is discussing chapter 1 of the Lex Aquilia, which gives the owner of a slave the right to claim damages if someone wrongfully "kills" the slave. Julian considers whether someone "kills" a slave for these purposes if he mortally wounds him and later someone else attacks the slave who dies more quickly as a result. Julian takes the view, which was probably not shared by all the jurists, that both persons who attacked the slave should be liable for "killing" him. In support of that view, he says in D 9 2 51 1 that it follows from the authoritative rulings of the old, Republican, jurists who held that, where a slave was wounded by a number of people in such a way that it was impossible to say whose blow had caused his death, then all of them were liable under the lex Aquilia:

    idque est consequens auctoritati veterum qui, cum a pluribus idem servus ita vulneratus esset ut non appareret cuius ictu perisset, omnes lege Aquilia teneri iudicaverunt.

    The old translation in C H Monro, The Digest of Justinian vol 2 (1909), pp 140-141 is reliable, with the addition of the words in square brackets:

    This is in keeping with the view handed down from the old lawyers, who, where the same slave was wounded by several persons under such circumstances that it did not appear by whose hand it was that he died, came to the conclusion that they were all liable [under the lex Aquilia].

    The text does not state expressly whether the case involved a concerted attack or simply a series of blows inflicted by different people, perhaps at different times. But the context of this particular passage, at least as preserved by Justinian's compilers, would suggest that, even if the case involves a concerted attack, Julian is thinking in terms of the liability of the participants as individuals, striking their own separate blows. Otherwise, Julian would not quote it as support for his view on a case which plainly involved blows delivered by different people in the course of quite separate attacks. In any event, the underlying doctrine must be that, in the kind of attack in question, in principle only the actual person who kills the slave is liable for killing. Were that not so, the old jurists would not have needed to make this special ruling that all the participants are liable under the lex Aquilia for killing the slave when it is unclear whose blow actually killed him. The reason for the ruling was, obviously, to make the statutory remedy of damages under chapter 1 available to the owner of the slave in these circumstances.

  180. This passage in Julian's digesta is referred to by the later writer Ulpian in D 9 2 11 2 Ulpian 18 ad edictum:

    sed si plures servum percusserint, utrum omnes quasi occiderint teneantur, videamus. et si quidem apparet, cuius ictu perierit, ille quasi occiderit tenetur: quod si non apparet, omnes quasi occiderint teneri Iulianus ait, et si cum uno agatur, ceteri non liberantur: nam ex lege Aquilia quod alius praestitit, alium non relevat, cum sit poena.

    Again the translation by Monro, The Digest of Justinian vol 2, pp 121 - 122 is satisfactory for present purposes, with the omission of the word in square brackets:

    But if a number of persons struck him, it is a point to consider whether all are not liable, as having [all] killed him. As to this, if it is known who struck the fatal blow, that one is liable, as having killed him; but, if it is not known, then, according to Julianus, all are liable, as having slain, and if an action is brought against one, this does not release the others, because, under the lex Aquilia, payment by one man is no discharge to another, as it is a case of penal damages.

    Ulpian considers whether, if several people strike a slave, all of them are liable for killing him. He says that, if it is clear who struck the blow from which the slave died, that person is liable for killing him. But he reports Julian's view that, if this is not clear, then all of them are liable for killing him. Again the precise factual situation is not spelled out, but it looks as if Ulpian is considering the case of an attack on the slave by several people at once. Since only the actual person whose blow killed the slave is liable if his identity is known, Ulpian must, however, be thinking primarily in terms of the individual liability of the person who does the killing: it is only if you cannot tell whose blow proved fatal that Julian holds that all are liable for killing the slave. A separate rule is adopted for that situation.

  181. I would take from these passages the clear implication that classical Roman jurists of the greatest distinction saw the need for the law to deal specially with the situation where it was impossible to ascertain the identity of the actual killer among a number of wrongdoers. If strict proof of causation were required, the plaintiff would be deprived of his remedy in damages for the death of his slave. In that situation, some jurists at least were prepared, exceptionally, to hold all of the wrongdoers liable and so afford a remedy to the owner whose slave had been killed. The exact scope of these decisions can, of course, no longer be ascertained and it is likely that different jurists held differing views: the sixth-century compilers of the Digest may well have altered the texts to some extent, if only by abbreviation, cutting out the cut and thrust of debate. Nor could the decisions, as recorded, furnish any guidance on the formulation of any equivalent rule today. The point remains, however, that all these centuries ago considerations of policy plainly led to a departure from what the law would usually require by way of proof of causation.

  182. Among slightly more recent authorities the decision of the Supreme Court of California in Rutherford v Owens-Illinois Inc 67 Cal Rptr 2d 16 (1997) stands out both as being directly in point and as providing powerful support for applying the principle in McGhee to the circumstances of the present cases. Mr Rutherford had worked with asbestos products on board ships at Mare Island Naval Shipyard. He developed, and died from, lung cancer caused by inhaling asbestos fibres. Proceedings were raised against some 19 manufacturers and distributors of asbestos products, including Owens-Illinois. Before the issues of liability and damages fell to be determined, all of the defendants except Owens-Illinois settled. In the trial the judge gave the jury an instruction that the burden was on Owens-Illinois to prove that their products were not a legal cause of the deceased's injuries. The Supreme Court held that this direction should not have been given, but further held that the plaintiffs need do no more than prove that exposure to the company's asbestos products was, in reasonable medical probability, a substantial factor in causing or contributing to his risk of developing cancer (67 Cal Rptr 2d 16, 19).

  183. The Supreme Court were satisfied that the plaintiffs bore the burden of proving which exposures to asbestos were substantial factors increasing the risk (67 Cal Rptr 2d 16, 32 - 33). For their part Owens-Illinois conceded that plaintiffs in asbestos cases were not required to identify the manufacturer of specific fibres that caused the cancer. Giving the judgment of the Supreme Court, Baxter J agreed (67 Cal Rptr 2d 16, 31 - 32):

    Plaintiffs cannot be expected to prove the scientifically unknown details of carcinogenesis, or trace the unknowable path of a given asbestos fiber. But the impossibility of such proof does not dictate use of a burden shift. Instead, we can bridge this gap in the humanly knowable by holding that plaintiffs may prove causation in asbestos-related cancer cases by demonstrating that the plaintiff's exposure to defendant's asbestos-containing product in reasonable medical probability was a substantial factor in contributing to the aggregate dose of asbestos the plaintiff or decedent inhaled or ingested, and hence to the risk of developing asbestos-related cancer, without the need to demonstrate that fibers from the defendant's particular product were the ones, or among the ones, that actually produced the malignant growth.

    The judge considered the standard jury instructions but commented (67 Cal Rptr 2d 16, 32):

    They say nothing, however, to inform the jury that, in asbestos-related cancer cases, a particular asbestos-containing product is deemed to be a substantial factor in bringing about the injury if its contribution to the plaintiff or decedent's risk or probability of developing cancer was substantial.

    Without such guidance, a juror might well conclude that the plaintiff needed to prove that fibers from the defendant's product were a substantial factor actually contributing to the development of the plaintiff's or decedent's cancer. In many cases, such a burden will be medically impossible to sustain, even with the greatest possible effort by the plaintiff, because of irreducible uncertainty regarding the cellular formation of an asbestos-related cancer. We therefore hold that, in the trial of an asbestos-related cancer case, although no instruction 'shifting the burden of proof as to causation' to defendant is warranted, the jury should be told that the plaintiff's or decedent's exposure to a particular product was a substantial factor in causing or bringing about the disease if in reasonable medical probability it was a substantial factor contributing to plaintiff's or decedent's risk of developing cancer.

  184. The court held that proof, that exposure to a defendant's product was a substantial factor contributing to the risk of developing cancer, was to be equated with proof that the exposure was a substantial factor in causing or bringing about the disease. Mr Stewart accepted that analysis of the decision. Allowing for slight differences in terminology, the doctrine espoused by the Californian court is similar to the principle to be derived from McGhee which the appellants contend should be applied in these cases. There is nothing in the report of the judgment to show that the Supreme Court of California were referred to McGhee. The fact that, apparently independently, they came to an essentially similar conclusion is powerful corroboration of the approach in McGhee and of its applicability in the present cases. Mr Stewart criticised Rutherford on the ground that the law should not develop principles to be applied only in the case of asbestos-related illness. While the court's opinion is indeed couched solely in relation to cases of that kind, the reasoning itself develops from the impossibility of proof inherent in those cases. There is nothing in that reasoning to suggest that the Supreme Court would refuse to apply a similar approach in other cases where similar irremediable problems of proof of causation arose. I would therefore reject that criticism, which in any event does not strike at the application of the reasoning in McGhee to cases of mesothelioma.

  185. Counsel referred to many cases decided by the courts of Australia and Canada and Lord Bingham of Cornhill has analysed the most important of them. Again, they illustrate the very real problem which confronts the law in cases of this kind. The decisions vary but, overall, they tend to reflect the same rival lines of thinking as are to be found in McGhee and Wilsher. I merely mention some that are favourable to plaintiffs. In the classic case of Cook v Lewis [1951] SCR 830 the Supreme Court of Canada were faced with a situation where the jury in a civil action had been unable to determine which of two huntsmen had fired the shot which injured the plaintiff. For the majority, Cartwright J held ([1951] SCR 830, 842) that, if the shot had been fired negligently, both should have been held liable. The fact that, otherwise, the victim would have been remediless appears to have been a factor in the court's decision. In some more recent cases examples are to be found of courts and judges applying the kind of approach adopted in McGhee: Birkholtz v R J Gilbertson Pty Ltd (1985) 38 SASR 121, 130 per King CJ; Bendix Mintex Pty Ltd v Barnes (1997) 42 NSWLR 307 per Stein JA dissenting; Chappel v Hart (1998) 195 CLR 232, 244, para [27], per McHugh J dissenting. I draw support for my view on the applicability of the McGhee principle from these judgments.

  186. The Commonwealth cases were supplemented, at your Lordships' suggestion, by a certain amount of material describing the position in European legal systems. Again I do not repeat Lord Bingham of Cornhill's survey. The material provides a check, from outside the common law world, that the problem identified in these appeals is genuine and is one that requires to be remedied.

  187. So, for instance, the French courts have been particularly exercised by cases, similar to Cook v Lewis, arising out of hunting accidents where the victim is unable to pinpoint which of a group of huntsmen caused his injury. The law has permitted the victim to recover damages, partly at least by finding that the substantial cause of the victim's injury was some negligence or recklessness of the group as a whole. The decision of the Second Chamber of the Cour de Cassation in Litzinger c Kintzler D 1957 493 note Savatier is an early example. I refer to the discussion in W van Gerven, J Lever, P Larouche, Tort Law (2000) pp 442 - 444.

  188. The German position is even more instructive. Since its inception, Article 830(1) second sentence of the Bürgerliches Gesetzbuch has contained a special provision which in effect provides that, where it cannot be ascertained which of a number of individual wrongdoers caused the victim's injury, each of them is to be held liable:

    Haben mehrere durch eine gemeinschaftlich begangene unerlaubte Handlung einen Schaden verursacht, so ist jeder für den Schaden verantwortlich. Das Gleiche gilt, wenn sich nicht ermitteln lässt, wer von mehreren Beteiligten den Schaden durch seine Handlung verursacht hat.

    The translation is:

    If several persons have caused damage by an unlawful act committed in common, each is responsible for the damage. The same rule applies if it cannot be discovered which of several participants has caused the damage by his act.

    The mere fact that the draftsmen of the Code identified the need for a general provision of this type shows that the problem of injury caused by unidentified wrongdoers was thought to be of some significance. It is, moreover, a problem that is not necessarily to be solved simply by leaving the loss to lie on the victim. On the other hand, it is equally plain that in Germany the policy is to resolve such difficulties in favour of the claimant on a wide front and, therefore, to do so in many cases where the law in this country would simply hold that he had failed to prove his case. See, for instance, Motive zu dem Entwurfe eines Bürgerlichen Gesetzbuches für das Deutsche Reich (1888) vol 2, p 738. That reflects the particular policy choice of German law, which the courts articulate: BGHZ 25, 271, 274; OLG München MDR 1967, 671, 672. A good example of this more extensive approach is the case of the lady who fell on an uneven surface and could not establish whether the precise place where she fell was on an unfinished street belonging to the local authority or on land belonging to an individual. The Bundesgerichtshof held that the provision in Article 830(1) applied and allowed her to recover damages against the individual: BGHZ 25, 271; van Gerven, Tort Law, pp 444 - 445.

  189. At the very least, the cross-check with these systems suggests that it is not necessarily the hallmark of a civilised and sophisticated legal system that it treats cases where strict proof of causation is impossible in exactly the same way as cases where such proof is possible. As I have tried to show, there are obvious policy reasons why, in certain cases at least, a different approach is preferable in English law too. The present are among such cases. Following the approach in McGhee I accordingly hold that, by proving that the defendants individually materially increased the risk that the men would develop mesothelioma due to inhaling asbestos fibres, the claimants are taken in law to have proved that the defendants materially contributed to their illness.

  190. While that is sufficient for the decision of the appeals, Mr Stewart urged that, if minded to apply some version of the principle in McGhee, the House should define its scope. He pointed out that the speeches in McGhee had left doubt as to the scope of the principle that the House had been applying and the decision of the Court of Appeal in Wilsher had shown only too clearly that it could be extended too far. It is indeed plain that, as Lord Nicholls of Birkenhead has observed, considerable restraint is called for in using the principle. Identifying, at an abstract level, the defining characteristics of the cases where it is, none the less, proper to apply the principle is far from easy. The common law naturally and traditionally shies away from such generalisations especially in a developing area of the law. But, having regard to the cases cited by counsel and also, in particular, to the cases and textbooks on the German law referred to in van Gerven's Tort Law, pp 444 - 447 and 459 - 461, I would tentatively suggest that certain conditions are necessary, but may not always be sufficient, for applying the principle. All the criteria are satisfied in the present cases.

    • First, the principle is designed to resolve the difficulty that arises where it is inherently impossible for the claimant to prove exactly how his injury was caused. It applies, therefore, where the claimant has proved all that he possibly can, but the causal link could only ever be established by scientific investigation and the current state of the relevant science leaves it uncertain exactly how the injury was caused and, so, who caused it. McGhee and the present cases are examples.

    • Secondly, part of the underlying rationale of the principle is that the defendant's wrongdoing has materially increased the risk that the claimant will suffer injury. It is therefore essential not just that the defendant's conduct created a material risk of injury to a class of persons but that it actually created a material risk of injury to the claimant himself.

    • Thirdly, it follows that the defendant's conduct must have been capable of causing the claimant's injury.

    • Fourthly, the claimant must prove that his injury was caused by the eventuation of the kind of risk created by the defendant's wrongdoing. In McGhee, for instance, the risk created by the defenders' failure was that the pursuer would develop dermatitis due to brick dust on his skin and he proved that he had developed dermatitis due to brick dust on his skin. By contrast, the principle does not apply where the claimant has merely proved that his injury could have been caused by a number of different events, only one of which is the eventuation of the risk created by the defendant's wrongful act or omission. Wilsher is an example.

    • Fifthly, this will usually mean that the claimant must prove that his injury was caused, if not by exactly the same agency as was involved in the defendant's wrongdoing, at least by an agency that operated in substantially the same way. A possible example would be where a workman suffered injury from exposure to dusts coming from two sources, the dusts being particles of different substances each of which, however, could have caused his injury in the same way. Without having heard detailed argument on the point, I incline to the view that the principle was properly applied by the Court of Appeal in Fitzgerald v Lane [1987] 1 QB 781.

    • Sixthly, the principle applies where the other possible source of the claimant's injury is a similar wrongful act or omission of another person, but it can also apply where, as in McGhee, the other possible source of the injury is a similar, but lawful, act or omission of the same defendant. I reserve my opinion as to whether the principle applies where the other possible source of injury is a similar but lawful act or omission of someone else or a natural occurrence.

  191. For these reasons I was in favour of allowing the appeals and of making the appropriate orders in each of the cases


Cases

March v E & MH Stramare Pty Ltd (1991) 171 CLR 506; Snell v Farrell [1990] 2 SCR 311; Environment Agency (formerly National Rivers Authority) v Empress Car Co. (Abertillery) Ltd [1999] 2 AC 22; Kuwait Airways Corporation v Iraqi Airways Co (Nos 4 & 5) [2002] 2 WLR 1353; Rahman v Arearose Ltd [2001] QB 351; Blatch v Archer (1774) 1 Cowp 63; Bonnington Castings Ltd v Wardlaw [1956] AC 613; Nicholson v Atlas Steel Foundry & Engineering Co Ltd [1957] 1 WLR 613; Gardiner v Motherwell Machinery & Scrap Co Ltd [1961] 1 WLR 1424; McGhee v National Coal Board [1973] 1 WLR 1; Wilsher v Essex Area Health Authority [1987] QB 730; Litzinger v Kintzler (Cass. civ. 2e, 5 June 1957, D 1957 Jur. 493); Summers v Tice 199 P 2d 1 (1948); Cook v Lewis [1951] SCR 830; Fitzgerald v Lane [1987] QB 781; Sindell v Abbott Laboratories 26 Cal. 3d 588 (1980); B v Bayer Nederland BV (Hoge Raad, 9 October 1992, NJ 1994, 535); Chappel v Hart (1998) 195 CLR 232; Naxakis v Western General Hospital (1999) 197 CLR 269; Webster v Chapman (1997) 155 DLR (4th) 82; Bendix Mintex Pty Ltd v Barnes (1997) 42 NSWLR 307; Birkholtz v R J Gilbertson Pty Ltd (1985) 38 SASR 121; Wallaby Grip (BAE) Pty Ltd v Macleay Area Health Service (1998) 17 NSWCCR 355; E M Baldwin & Son Pty Ltd v Plane (1999) Aust Torts Reports 81-499; Rutherford v Owens-Illinois Inc 67 Cal. Rptr. 2d 16 (1997); Hadley v Baxendale (1854) 9 Exch 341; Donaghue v Stevenson [1932] AC 562; Caparo Industries plc v Dickman [1990] 2 AC 605; Reeves v Commissioner of Police of the Metropolis [2000] 1 AC 360; Rutherford v Owens-Illinois Inc. (1997) 67 Cal Rptr.2d 16; Sindell v Abbott Laboratories (1980) 607 P.2d 924; Benmax v Austin Motor Co Ltd [1955] AC 370; Sentilles v Inter-Caribbean Shipping Corporation 361 US 107 (1959); Gardiner v Motherwell Machinery & Scrap Co Ltd [1961] 1 WLR 1424; Hotson v East Berkshire Area Health Authority [1987] AC 750; Haag v Marshall (1989) 61 DLR (4th) 371; Birkholz v RJ Gilbertson Pty Ltd (1985) 38 SASR 121; Senior v Ward (1859) 1 El & El 385; Wakelin v London & South Western Railway Co (1886) 12 App Cas 41; Craig v Glasgow Corporation 1919 SC(HL) 1; Quinn v Cameron & Roberton Ltd [1958] AC 9; Kay's Tutor v Ayrshire & Arran Health Board 1987 SC(HL) 145; [1987] 2 All ER 417; Caswell v Powell Duffryn Associated Collieries Ltd [1940] AC 152

Legislations

Factories Act 1937: s.4(1)

Greek Civil Code: Art.926

Austrian Civil Code: Art.1302

Netherlands Civil Code: Art.6.99 BW

Restatement Second of Torts: s.433B, subdivision (3) 

Authors and other references

Merewether & Price: "Report on Effects of Asbestos Dust on the Lungs and Dust Suppression in the Asbestos Industry" (1930)

McLachlin J: "Negligence Law - Proving the Connection", Torts Tomorrow, A Tribute to John Fleming, ed Mullany & Linden, LBC Information Services (1998)

Prof. Christian von Bar: The Common European Law of Torts (2000) vol 2

Prof. van Gerven, Lever & Larouche: Cases, Materials and Text on National, Supranational and International Tort Law (2000)

Markesinis & Unberath: The German Law of Torts, 4th ed (2002)

Palandt, Bürgerliches Gesetzbuch, 61 ed (2002)

Unification of Tort Law: Causation, ed J Spier (2000)

Nils Nygaard, Injury/Damage and Responsibility (2000)

Prof. Fleming: "Probabilistic Causation in Tort Law" 68 Canadian Bar Review, No 4, (Dec 1989)

Prof. Robertson: "The Common Sense of Cause in Fact", 75 Tex L Rev (1996-1997)

Hart & Honoré: Causation in the Law, 2nd ed (1985)

Jeroen Kortmann: "Ab alio ictu(s): Misconceptions about Julian's View of Causation" (1999) 20 Journal of Legal History 95

D 9 2 51 Julian 86 digesta 

C H Monro: The Digest of Justinian vol 2 (1909)

Ulpian in D 9 2 11 2 Ulpian 18 ad edictum

Motive zu dem Entwurfe eines Bürgerlichen Gesetzbuches für das Deutsche Reich (1888) vol 2


all rights reserved