[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	United Kingdom Supreme Court
You are here: BAILII >> Databases >> United Kingdom Supreme Court >> Ministry of Defence v AB & Ors [2012] UKSC 9 (14 March 2012) URL: http://www.bailii.org/uk/cases/UKSC/2012/9.html Cite as: [2012] PIQR P13, [2013] 1 AC 78, [2012] UKSC 9, [2012] WLR(D) 79, 125 BMLR 69, (2012) 125 BMLR 69, [2012] Med LR 306, [2012] 2 WLR 643, [2012] 3 All ER 673

[New search] [Printable PDF version] [Buy ICLR report: [2013] 1 AC 78] [View ICLR summary: [2012] WLR(D) 79] [Buy ICLR report: [2012] 2 WLR 643] [Help]

Summary




Hilary Term
[2012] UKSC 9
On appeal from: [2010] EWCA Civ 1317

JUDGMENT

Ministry of Defence (Respondent) v AB and others (Appellants)

before

Lord Phillips, President
Lord Walker
Lady Hale
Lord Brown
Lord Mance
Lord Kerr
Lord Wilson

JUDGMENT GIVEN ON

14 March 2012

Heard on 14-17 November 2011


 


Appellants James Dingemans QC Catherine Foster Nadia Whittaker Mark James (Instructed by Rosenblatt Solicitors)

Respondent Charles Gibson QC Leigh-Ann Mulcahy QC David Evans Adam Heppinstall (Instructed by Treasury Solicitors)

LORD WILSON

1. I consider that each of the nine appeals should be dismissed. In my respectful view the approach of Lord Phillips, Lady Hale and Lord Kerr to the meaning of the word "knowledge" in sections 11(4) and 14(1) of the Limitation Act 1980 ("the Act") is misconceived and would throw the practical application of the subsections into disarray. I also consider that any exercise of the discretion under section 33 so as to permit any of the nine actions to proceed would be aberrant in circumstances in which they have no real prospect of success.
2. What is the nature of the exercise which the court conducts when asked by a defendant to rule that an action in respect of personal injuries is time-barred under section 11 of the Act? Subsection (4) provides that the action shall not be brought after the expiration of three years from

"(a) the date on which the cause of action accrued; or

(b) the date of knowledge (if later) of the person injured."

The subsection refers, at (a), to "the cause of action" notwithstanding that, if the action is to continue, it may well transpire that the claimant has no cause of action. When the subsection turns, at (b), to "the date of knowledge (if later)" and so requires the court to appraise the claimant's knowledge of the four "facts" specified in section 14(1), which relate to, although do not comprise all elements of, his cause of action, the assumption that indeed he has a cause of action remains. That explains why sections 11(4)(b) and 14(1) refer to "knowledge" (which can be only of matters which are true) rather than to "belief" (which can be in matters which are untrue as well as in those which are true). Knowledge of the second of the four facts specified in section 14(1) is "that the injury was attributable in whole or in part to the act or omission which is alleged to constitute negligence, nuisance or breach of duty". This knowledge of attributability (as it is convenient to describe it) is predicated upon the assumption that the claimant has a valid cause of action and thus would be able to establish among other things, even in the teeth of opposition from the defendant, not just attributability (which means only that there is a real possibility that the act or omission caused the injury: Spargo v North Essex District Health Authority [1997] PIQR P235 at P242, Brooke LJ) but, rather, that his act or omission actually caused the injury in the legally requisite sense. In the decision of the Court of Appeal in Halford v Brookes [1991] 1 WLR 428 the trial judge, Schiemann J, is quoted, at p 442H, as having referred to "the bizarre situation when a defendant asserts that the plaintiff had knowledge of a fact which the plaintiff asserts as a fact but which the defendant denies is a fact". The situation may indeed seem bizarre until one remembers that, at the stage of an inquiry under section 11, the exercise requires the existence of the fact to be assumed. Were the action to continue, the defendant might well deny it; but he does not do so at that stage.

3. The assumption that, in an inquiry under sections 11(4) and 14(1) of the Act, the cause of action exists leads me, with inevitable discomfiture, to a profound disagreement with one of the essential foundations for the conclusions of the minority in this court. It was the preferred view of Foskett J, upon which he would have acted had he not felt constrained by authority to act otherwise, that the veterans who issued their claim on 23 December 2004 acquired the requisite knowledge of attributability only on a later date, namely 29 June 2007, when the Rowland report was privately presented to them. It is the conclusion of the minority that

(a) it is indeed possible for a claimant to lack knowledge of attributability at the time when he issues his claim and, if so, time will not have begun to run against him; and

(b) irrespective of whether the later presentation to them of the Rowland report then led them to acquire it, the veterans lacked such knowledge when they issued their claim, with the result that none of them is time-barred.

In my view, however, it is a legal impossibility for a claimant to lack knowledge of attributability for the purpose of section 14(1) at a time after the date of issue of his claim. By that date he must in law have had knowledge of it. Pursuant to CPR 22.1(1)(a) and (4), he must verify his claim form by a statement that he "believes" that the facts stated in it are true. The word in the statement of truth is "believes" rather than "knows" only because – of course – the assumption that the cause of action exists does not apply to the claim form. That it exists is indeed only a claim. Although the statement of truth covers wider ground, it can in my view be regarded as an explicit recognition by the claimant that he then has knowledge of attributability for the purpose of section 14(1).

4. Irrespective, however, of the degree of significance to be attached to the statement of truth, it is clear to me that the inquiry mandated by section 14(1) is retrospective, namely whether the claimant first had knowledge of it (and of the other specified facts) within or outside the period of three years prior to the date of issue. As Lord Mance said of an analogous section of the Act in Haward v Fawcetts [2006] UKHL 9, [2006] 1 WLR 682, at para 106,

"Under section 14A the onus is on a claimant to plead and prove that he first had the knowledge required for bringing his action within a period of three years prior to its bringing."

And, see, to similar effect, the judgment of the Court of Appeal in Nash v Eli Lilly & Co [1993] 1 WLR 782 at p 796H.

5. Lord Phillips cites, at para 143 below, the view of Waite LJ in Whitfield v North Durham Health Authority [1995] 6 Med LR 32 in support of the proposition that lack of knowledge of attributability can survive the issue of a claim. In 1987, thus prior to the claim issued in 1992 with which the Court of Appeal was there concerned, the claimant had issued a claim which had never been served. Waite LJ observed that her issue of the claim in 1987 did not necessarily betoken that she had knowledge under section 14(1). But the observation was an aside in that the court proceeded to find that she had had the requisite knowledge in 1985. In the Eli Lilly case, cited above, by contrast, the Court of Appeal, in a passage at 795H-796A cited with approval by Judge LJ in Sniezek & Bundy (Letchworth) Ltd [2000] PIQR P213 at P228, observed that it had "difficulty in perceiving how in any case where a claimant has sought advice and taken proceedings, it can rightly be held that the claimant had not then had relevant knowledge". It follows that I prefer the latter approach. Yet, frankly, I doubt whether in any of those three cases the Court of Appeal was afforded the leisurely consideration of the nature of "knowledge" for the purpose of sections 11(4) and 14(1) which has been afforded to this court in the present case.
6. The statutes of limitation, which stretch back to 1540, have been in place for two main reasons. One is to protect defendants from being vexed by stale claims. They are Acts of peace: see A'Court v Cross (1825) 3 Bing 329, 332 (Best CJ). The other is to require claims to be put before the court at a time when the evidence necessary for their fair adjudication is likely to remain available, or, in the words of the preamble to the 1540 Act, at a time before it becomes "above the Remembrance of any living Man...to...know the perfect Certainty of such Things". Conventionally, therefore, they have required the assertion, by claim, of a cause of action within a specified period following its accrual. The modification of the conventional requirement now reflected in sections 11 and 14 of the Act was born of the injustice suffered by a claimant who lost his right to claim damages for personal injuries before he knew of its existence: see para 17 of the Report of the Committee on Limitation of Actions in Cases of Personal Injury dated September 1962, Cmnd 1829, chaired by Edmund Davies J. But, in para 30, the committee also expressed its concern not to encourage actions of a speculative character. In the event it set out, at para 34, its conclusion that the conventional requirement should not apply so as to bar a claimant if

"(a) the first occasion on which he discovered, or could reasonably have been expected to discover, the existence of his injury, or the cause to which it was attributable, was such that it was not reasonably practicable for him to start proceedings in time; and

(b) he has in fact started proceedings within a certain period (which we consider should be twelve months) after such occasion."

The committee recommended that, additionally, such a claimant should need the leave of the court. The result was section 1 of the Limitation Act 1963, the terminology of which was to prove troublesome (see Smith v Central Asbestos Co Ltd [1973] AC 518) and thus to lead to the improvements first included in section 1 of the Limitation Act 1975 and soon consolidated in sections 11 and 14 of the Act. For present purposes the only importance of section 1 of the 1963 Act is that, reflective of the recommendations in the Edmund Davies report, it referred to facts being outside the knowledge of the claimant "until a date which.... was a date not earlier than twelve months before the date on which the action was brought" (italics supplied). So, by the latter date, the claimant was taken to have acquired the knowledge; and the only question was whether he had issued his claim within the specified period after having done so. This was in my view an essential boundary of the scheme by which the conventional requirement was relaxed; and I see no reason to doubt that it so remains. It is in my view heretical that a claimant can escape the conventional requirement to assert his cause of action for personal injuries within three years of its accrual by establishing that, even after his claim was brought, he remained in a state of ignorance entirely inconsistent with it. Indeed it is, as Smith LJ observed in the course of argument in the Court of Appeal, "a bit Alice in Wonderland".

7. What, then, is comprised in the knowledge of attributability which section 14 (1) of the Act requires? In articulating his preferred view Foskett J made no bones about it: he considered that one of the constituents of the knowledge should be evidence, specifically that the veteran should appreciate that there was "credible evidence" that, as a result of the tests, he had been exposed to ionising radiation at a level above that to which all human beings are exposed and that his injury was capable of having been caused by his exposure to it. Lord Phillips states, at para 141, that the preferred view of Foskett J was in principle correct. But, no doubt because Mr James Dingemans QC concedes on behalf of the veterans that "evidence" is no part of "knowledge" for the purpose of the subsection, Lord Phillips reformulates the preferred view of Foskett J so as to require that the claimant's belief be "based" (para 137) or "founded" (para 141) on "known fact". For her part, Lady Hale suggests, at paras 168 and 170, that the belief should have a reasonable basis either in evidence or, alternatively, in "objective fact". And, for his part, Lord Kerr, who adheres strictly to the word "knowledge", concludes at para 209 that it exists only when founded on "objectively verifiable facts".
8. In reality, however, all three of these formulations in my view remain requirements that the claimant must, actually or constructively, have evidence before he is to be fixed with the knowledge which will set time running. Indeed, in paras 140 and 142, Lord Phillips suggests that, although the appellants may for long have believed that their injuries were attributable to the exposure, they lacked knowledge of attributability because, at least until presentation to them of the Rowland report, there was "no scientific evidence available that provided significant support to the belief". And, in para 172, electing not there to apply her difficult alternative requirement of a basis in "objective fact" (for which facts are other than objective?), Lady Hale explains her conclusion that the appellants have lacked knowledge because they have lacked "evidence". If, indeed, upon a preliminary issue as to limitation, the court is required to weigh the nature, strength and verifying quality of the evidence as to the attributability of the injuries which became available, actually or constructively, to the claimant, and to identify the time when it did so, the determination of the issue will in my view expand into a preliminary trial entirely contrary to the intention of Parliament as expressed in the subsection.
9. This court should not readily jettison the welter of jurisprudence about the meaning of knowledge in section 14(1) of the Act which has accumulated over more than 20 years. Lord Phillips has helpfully charted it in paras 112 to 121. His analysis is, at para 117, that the Eli Lilly case, cited above, is the first of a series of decisions which equated "knowledge" with "subjective belief" and, at para 141, that the equation was wrong. In fact the phrase "subjective belief" is not to be found in any of the decisions. The concepts of "belief" and indeed of "knowledge" are inherently subjective. Even when under section 14(3) it fixes a claimant with constructive knowledge, the law deems him to have subjective knowledge. So I take Lord Phillips' phrase to be no more than a convenient shorthand for the antithesis of what in his view is connoted by the word knowledge, namely that it must be belief which is founded on fact.
10. In the early case of Davis v Ministry of Defence, 26 July 1985, [1985] CLY 2017, the Court of Appeal took a narrow view of the meaning of knowledge in section 14(1) of the Act. May LJ said that "reasonable belief" was not enough. But in the Halford case, cited above, Lord Donaldson MR said, at p 443F, that "reasonable belief" would normally suffice and that Davis had been an exceptional case. For twenty years Lord Donaldson's approach has prevailed. It was specifically endorsed by Judge LJ in the Sniezek case, cited above, at P228 and ultimately also in the House of Lords, namely in the Haward case, cited above, in the passage in the speech of Lord Nicholls quoted by Lord Phillips in para 121 below.
11. Lord Phillips is therefore correct to point out that when, in the present proceedings, it accepted that the belief had to be held with a degree of confidence but, as an aside, declined to accept that it had to be reasonable, the Court of Appeal was, apparently without so realising, disagreeing with a statement of Lord Nicholls in the House of Lords as well as with that of Lord Donaldson. Had I been offering a view of the meaning of knowledge in section 14(1) in circumstances in which I had been unassisted by authority, I think that I might have ventured the phrase "reasoned belief" rather than "reasonable belief". The word "reasoned" might even better have conveyed the need for the belief not only to be held with a degree of confidence (rather than to be little more than a suspicion) but also to carry a degree of substance (rather than to be the product of caprice). But the distinction between the phrases is a matter of little more than nuance. In the resolution of marginal issues, and even at the level of this court, there is a lot to be said for maintaining consistency in the law. So I consider that this court should reiterate endorsement for Lord Donaldson's proposition that a claimant is likely to have acquired knowledge of the facts specified in section 14 when he first came reasonably to believe them. I certainly accept that the basis of his belief plays a part in the inquiry; and so, to that limited extent, I respectfully agree with para 170 of Lady Hale's judgment. What I do not accept is that he lacks knowledge until he has the evidence with which to substantiate his belief in court. Indeed we should not forget that, if the action is to continue, the court will not be directly interested in evidence about mere attributability; it will require proof of actual causation in the legally requisite sense.
12. What then is the degree of confidence with which a belief should be held, and of the substance which it should carry, before it is to amount to knowledge for the purpose of the subsection? It was, again, Lord Donaldson in the Halford case, cited above, who, in the passage quoted by Lord Phillips in para 115 below, offered guidance in this respect which Lord Nicholls in the Haward case, cited above, was, at para 9, to describe as valuable and upon which, at this level of generality, no judge has in my view yet managed to improve: it is that the belief must be held "with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking legal and other advice and collecting evidence". In Broadley v Guy Clapham & Co [1994] 4 All ER 439 Hoffmann LJ, in the passage quoted by Lord Phillips at para 118 below, paraphrased Lord Donaldson's guidance in terms of a search for the moment at which the claimant knows enough to make it reasonable for him to begin to investigate whether he has a "case" against the defendant. I respectfully agree with the analysis by Lord Phillips of what Hoffmann LJ meant. The investigation upon which the claimant should reasonably embark is into whether in law he has a valid claim (in particular whether the act or omission of the defendant involves negligence or other breach of duty, being a matter of which the claimant is specifically not required to have had knowledge under section 14(1)) and, if so, how that claim can be established in court. So it is an investigation likely to be conducted with the assistance of lawyers; but, in the light of their advice, it may well also embrace a search for evidence, including from experts. The focus is upon the moment when it is reasonable for the claimant to embark on such an investigation. It is possible that a claimant will take legal advice before his belief is held with sufficient confidence and carries sufficient substance to make it reasonable for him to do so. Thus, as Judge LJ pointed out in the Sniezek case, cited above, at P229 and P232, it does not automatically follow that, by the date when he first took legal advice, the claimant will have acquired the requisite knowledge; but such an inference may well be justified.
13. I hasten however to attach an obvious rider. From the fact that a claimant may well need to consult experts after he has acquired the requisite knowledge, it in no way follows that he will have acquired such knowledge by the date when he first consults an expert. Section 14(3) expressly recognises that the facts which he is required to know may be ascertainable by the claimant only with the help of experts and deems him to have acquired such knowledge at the point at which he might, with their help, reasonably have been expected to acquire it. In my view the date upon which the claimant first consulted an expert is not, on its own, likely to assist the court in determining whether by then he had the requisite knowledge. Instead the court will have regard – broadly – to the confidence with which the claimant held the belief, and to the substance which it carried, prior to his consulting the expert (and in particular, no doubt, the reasons which induced the claimant to consult him) and also, if the conclusion is that at that prior stage the claimant lacked belief of the requisite character, the effect upon the claimant's belief of his receipt of the expert's report.
14. In short the assistance given to a claimant by an expert in this respect can be of two kinds. One is assistance in his acquiring "knowledge" of the "facts" required by section 14. He may, for example, advise the claimant that he has a medical condition, of which he was previously unaware, which provides him with a substantive basis for believing that his injury is attributable to an act or omission of the defendant. The other is the provision of evidence which will, in court, help him to substantiate the claim which, in the light (among other things) of his knowledge of the limited matters specified by section 14(1), he proposes to bring.
15. To the above, at its level of generality, I find it impossible to make useful addition. In principle, and subject to the fact that the assault by the appellants in this court upon the reasoning of the Court of Appeal is of a generic character, there is no escape from turning at this stage to attend - at least broadly - to the individual facts of the nine appeals before the court. They are nine out of a large number of claims which have been made subject to a Group Litigation Order because they "give rise to common or related issues of fact or law": see CPR 19.10. Indeed, along with the claim of the late Mr Sinfield continued by his widow, which is not time-barred, they were no doubt chosen because, so it was considered, they had material similarities with many of the other claims in the group and thus their determination would inform that of many of the others. But, with respect to Lord Phillips, I cannot subscribe to his conclusion, at paras 159 and 160 below, that the existence of the other claims in the group should affect determination of the nine appeals. CPR 19.12(1)(a) provides that this court's judgment or order is binding on the parties to all other claims presently within the group "unless the court orders otherwise". If, as to which I have no view, there would be any particular injustice in visiting adverse judgments in the nine appeals upon other, materially similar, claims within the group, the quoted clause would cater for it.

Mr Ayres

16. Mr Ayres' claim was issued on 1 February 2007. He died on 29 November 2010. Under section 11(4) of the Act his claim was barred if he first had the requisite knowledge prior to 1 February 2004. He had served with the RAF as an aircraft fitter on Christmas Island in 1957 when detonations had taken place off Malden Island, and again in 1958, when others had taken place off Christmas Island itself. He was well aware that the detonations caused radiation and that the aeroplanes upon which he worked had had to be decontaminated for that reason. In the late 1990s he developed haematuria (blood in the urine). By then he was aware of the existence of the British Nuclear Tests Veterans Association ("BNTVA"). He knew that it was an action group committed to secure compensation for veterans who suffered injuries believed to have been caused by radiation and that, to that end, it and three of its members had taken proceedings in the European Court of Human Rights ("the ECtHR") in which they had alleged exposure of the servicemen to ionising radiation and consequential illnesses. He kept newspaper articles about the campaign. Mr Ayres said in evidence that, when he developed haematuria, he firmly believed that it was capable of having been caused by his exposure to radiation. But the injury upon which his action is founded is prostate cancer, with which he was diagnosed on 2 December 2003. His evidence, unsurprising in the light of what he already believed, was that, on receiving the diagnosis, he knew that there was a real possibility that the cancer had also been caused by his exposure to radiation.

Mr Brothers

17. Mr Brothers died on 13 June 2000 and his widow's claim under the Fatal Accidents Act 1976 was issued on 23 December 2004. Under section 12(2) of the 1980 Act her claim was barred if she first had the requisite knowledge prior to 23 December 2001. Mr Brothers had served with the RAF as a navigator on "sniffer" planes in 1956 and 1957 which collected radioactive samples from clouds generated by the detonations. In 1997 he was diagnosed as suffering from cancer of the oesophagus. For at least the previous 20 years it had been his practice to smoke 20 cigarettes a day. At the time of the diagnosis, however, Mrs Brothers knew that cancer was capable of being caused by radiation and that Australian veterans had claimed damages for illnesses, including cancer, which they alleged to have been the result of exposure to radiation. Although in evidence she explained that Mr Brothers denied to her that it was possible that his cancer had been the result of his exposure, she added that she believed that it was capable of having caused it. Indeed in letters to two doctors in March 2002 she wrote that she had "always believed" that his cancer had been caused by it.

Mr Clark

18. Mr Clark died on 28 September 1992. His widow's claim under the 1976 Act was issued on 31 March 2008. Under section 12(2) of the 1980 Act her claim was barred if she first had the requisite knowledge prior to 31 March 2005. Mr Clark had done part of his National Service on Christmas Island as a Sapper with the Royal Engineers in 1957 and 1958 when detonations had taken place in the vicinity. In February 1991 he was diagnosed with the lung cancer from which, within two years, he was to die. From his teenage years he had smoked 20 cigarettes a day. Days prior to the diagnosis, however, he had mentioned his service on Christmas Island to the doctors and had told them that he had been unprotected. He got in touch with the BNTVA. In March 1991 he signed a home-made statement, for possible reference in future court proceedings, in which he described his exposure to the tests, his subsequent suffering from various conditions and the recent diagnosis of his terminal cancer. Of those actions on his part Mrs Clark was aware. Shortly after his death she made an unsuccessful application for a war pension on the basis that his cancer had been linked to his service on Christmas Island. In 2002 she consulted solicitors about bringing a claim under the 1976 Act.

Mr Dickson

19. Mr Dickson's claim was issued on 23 December 2004. He died of heart disease in May 2006. Under section 11(4) of the Act his claim was barred if he first had the requisite knowledge prior to 23 December 2001. He had served as a lance corporal in the Royal Engineers on Christmas Island in 1958 when detonations had taken place in the vicinity. Soon afterwards he began to suffer skin disorders and, by 1990, he had begun to suffer a variety of other illnesses, including colitis. In 1986 he became a member of the BNTVA and embarked upon a tireless public campaign on its behalf for veterans to be compensated for injuries alleged to have been sustained by exposure to radiation. He had, so Foskett J found, a genuine belief, which he communicated to his doctor, that his own ill-health had been caused by exposure to it; and he also believed that the respondent's denials about the level of exposure to servicemen had been untrue. In 1989 he applied for a war pension on the ground that his exposure to it in 1958 had damaged his immune system; as in the case of all the other applications for war pensions to which I will refer, the respondent denied that the detonations had caused any significant exposure to radiation and the application was refused. In 1990 Mr Dickson was for some reason expelled from the BNTVA but continued his campaign alone. In 1992 he was quoted in "The People" as saying, on a basis which is unclear, that his army files demonstrated his exposure to a high level of radiation; he added that he wanted action.

Mr Hart

20. Mr Hart's claim was issued on 23 December 2004. Under section 11(4) of the Act his claim was barred if he first had the requisite knowledge prior to 23 December 2001. He had performed his National Service as an engineer mechanic with the Royal Navy and in 1956 had served aboard HMS Diana near the Monte Bello islands when its function had been to monitor the fall-out from two nuclear explosions conducted there. In 1959, following his return to civilian life, he first developed a lipoma (a benign fatty lump on the skin) and, during the next decades, developed numerous further lipomas which spread all over his body. From 1973, if not before, he considered that they might have been caused by his exposure to radiation in 1956. In 1988 he joined the BNTVA and well understood the link which it was asserting between radiation and the injuries of its members. In 1991 he applied for a war pension on the basis that his lipomas had been caused by exposure to radiation. Much later, namely in July 2002, Mr Hart was diagnosed with bowel cancer, whereupon he consulted solicitors. Thereafter his claim was issued reasonably promptly. But he claimed damages for the lipomas as well as for the cancer and it is now common ground that the inquiry is into his knowledge of their attributability.

Mr McGinley

21. Mr McGinley's claim was issued on 23 December 2004. Under section 11(4) of the Act his claim was barred if he first had the requisite knowledge prior to 23 December 2001. He had served as a plant operator in the Royal Engineers on Christmas Island in 1958 when detonations had taken place in the vicinity. Soon afterwards he began to suffer from bouts of vomiting and diarrhoea and from blisters on the skin. In 1976 he was diagnosed as infertile. He was a founder member of the BNTVA. He was its Chair from its inception in 1983 until 2000; and he was perhaps its most vociferous spokesman. In 1984 he applied for a war pension on the ground that exposure to radiation in 1958 had caused his infertility. In 1991, with the assistance of a journalist, he wrote a book, entitled "No Risk Involved", in which he set out his experiences on Christmas Island and his subsequent ill-health. In 1991 he launched one of the applications to the ECtHR to which I have referred in para 16. In a document in support of it, signed by him in 1993, he referred to "the realisation that his prolonged and continuing debilitating illnesses and infertility were caused by his deliberate and unprotected exposure by the UK in 1958" to the detonations. In his present claim, however, Mr McGinley does not repeat the allegation that the exposure was deliberate, known as the "guinea-pig" allegation.

Mr Noone

22. Mr Noone's claim was issued on 23 December 2004. Under section 11(4) of the Act his claim was barred if he first had the requisite knowledge prior to 23 December 2001. He had served in the RAF as an air frame mechanic on Christmas Island in 1957 when detonations had taken place in the vicinity. From then onwards he suffered from severe and persistent acne. He soon came to suspect that it had been caused by exposure to radiation. By 1983, notwithstanding contrary advice from a consultant dermatologist, he had come to believe that it had been so caused. On 3 June 1983 he was reported in The Guardian as having stated that the exposure had caused his condition. In the same year he joined the BNTVA and made a similar statement in an application for a war pension. In 1986 and 1989 he suggested likewise to different GPs.

Mr Ogden

23. Mr Ogden died of cancer on 5 August 2004 and his step-daughter's claim for the benefit of his estate was issued on 23 December 2004. Under section 11(4) and (5) of the 1980 Act her claim was barred if Mr Ogden first had the requisite knowledge prior to 5 August 2001. He had served with the RAF as an air wireless fitter on Christmas Island in 1958 when detonations had taken place in the vicinity. In 1986 he suffered a brain tumour, became a member of the BNTVA and promptly applied for a war pension on the basis that the tumour had been caused by his exposure to radiation in 1958. In 1994 he was diagnosed with cancer of the colon and on 12 April 2001, in making a second application for a pension so as to encompass the cancer as well as the tumour, he wrote that they had been caused by his proximity to the detonations.

Mr Rokoratu

24. The Claim of Mr Rokoratu (who has sadly died days prior to the delivery of these judgments) was issued on 23 December 2004. Under section 11(4) of the Act his claim was barred if he first had the requisite knowledge prior to 23 December 2001. He is a citizen of Fiji and had served as a stevedore with the Fijian Royal Naval Volunteer Reserves on Christmas Island in 1958 when detonations had taken place in the vicinity. From 1961 onwards he suffered a variety of illnesses, in particular lipomas. In October 1998 he applied to the ECtHR on the basis that he had suffered injuries as a result of exposure to radiation on Christmas Island. In a report dated 9 November 1998 in support of the application a consultant physician in Fiji wrote that his lipomas were likely to be linked to his exposure to the detonations in 1958. Mr. Rokoratu told Foskett J that the report had confirmed his belief in the link.
25. In my view the Court of Appeal was correct to conclude that all nine of the appellants had the requisite knowledge prior to the period of three years relevant to them. For the facts of each case which I have distilled in the above paragraphs drive a conclusion that, prior to the relevant period, each reasonably believed that the injury was able to be attributed to the nuclear tests conducted by the respondent between 1956 and 1958. Their many private and public statements down the years about the cause of their conditions; the nation-wide campaign for compensation pursued for so long and with such vigour through the BNTVA; the applications for war pensions; and the applications to the ECtHR: all these were the product of reasonable beliefs. The appellants held them with sufficient confidence to have made it reasonable for them to begin to investigate whether they had valid claims against the respondent. In asking the court to allow them further time in which to obtain it, Mr. Dingemans concedes that even today the appellants lack evidence with which to establish a credible case that the injuries were caused by the tests; and so it follows that, irrespective of when they began to investigate whether they had valid claims against the respondent, they would probably have learned that, as remains the position today, their claims had no reasonable prospect of success. But that is entirely irrelevant to an inquiry under sections 11(4) and 14(1): once the requisite knowledge has arisen, the difficulty of actually establishing the claim confers no right thereunder to a further, open-ended, extension of the time within which the action must be brought. In so saying I have returned to the irrelevance of evidence to an inquiry under the subsections.
26. If their actions were to proceed, the nine appellants therefore needed to persuade the court to exercise its discretionary power under section 33 of the 1980 Act to disapply section 11(1). Section 33 requires the court first, by subsection (1), to conduct an inquiry into the degree of prejudice likely to be suffered by the defendant in the event of exercise of the power and by the claimant in the event of a refusal to do so; and second, by subsection (3), to have regard to all the circumstances of the case and in particular to six specified matters. But the appellants' grounds of appeal require this court to consider only one, generic, feature of the reasoning which led the Court of Appeal to decline to exercise the power under section 33 in any of the nine actions; and, at the end of his oral presentation of the appeals which stretched in effect over almost two days, Mr. Dingemans wisely devoted only the final five minutes to the ground referable to section 33. It is that the Court of Appeal wrongly elevated the issue of causation to be the determining factor under the section. It is indeed a fair reading of that court's full judgment upon the issue that it regarded the difficulties which confront the appellants in establishing that their injuries were caused by the tests as determinative against exercise of the power under section 33. I stress, however, that it carefully weighed all the other relevant factors, for example that, as will have been noted, the claim of Mr Ayres was out of time by less than two months whereas the claims of the other eight appellants were out of time by between three years (in the case of Mr Rokoratu) and 18 years (in the case of Mr Noone).
27. It is undesirable that a court which conducts an inquiry into whether a claim is time-barred should, even at the stage when it considers its power under section 33, have detailed regard to the evidence with which the claimant aspires to prove his case at trial. But the ten claims placed before Foskett J were of particular complexity; and the nature of the submissions made to him on behalf of the appellants about the meaning of knowledge for the purpose of section 14(1) of the 1980 Act led him to undertake, over ten days of hearing and expressed in 885 paragraphs of judgment, a microscopic survey of the written evidence available to the parties, in particular to the appellants, in relation to causation. At all events the result was to yield to the Court of Appeal an unusual advantage, namely a mass of material which enabled it with rare confidence to assess the appellants' prospects of establishing causation. It expressed its conclusion in terms of the "very great difficulties" which confronted the appellants in that regard. But, in line with the realistic concession made by Mr Dingemans in this court, the fact is that, for the reasons set out by Lord Phillips in paras 156 to 158 below, their claims have no real prospect of success. In my view it would have been absurd for the Court of Appeal to have exercised the discretion to disapply section 11 so as to allow the appellants to proceed in circumstances in which the next stage of the litigation would be likely to have been their failure to resist entry against them of summary judgment pursuant to CPR 24.2(a)(i). In this regard I do not share the view of Lord Phillips, at para 160 below, about the relevance of the fact that, at least until that next stage, the action brought by the late Mr Sinfield, together no doubt with other actions in the group which do not fall foul of section 11, are to proceed.

LORD WALKER

28. The decision of the House of Lords in Cartledge v E Jopling & Sons Ltd [1963] AC 758 revealed a serious injustice in the law relating to limitation of actions. Workmen suffering from an insidious industrial disease, pneumoconiosis, might find that their rights of action against their employers were statute-barred before they even knew that they were suffering from the disease. To remedy that injustice Parliament enacted the Limitation Act 1963.
29. That Act was severely criticised by the House of Lords in Smith v Central Asbestos Co Ltd [1973] AC 518 and it was repealed and replaced by the Limitation Act 1975, now consolidated as part of the Limitation Act 1980 ("the 1980 Act"). The need for the claimant to know the legal significance of the proposed defendant's acts or omissions (one of the main points of criticism) was removed. But two important general features were reproduced (though in a different form) in the new legislation. One was that the commencement of the limitation period was to be triggered by the claimant's actual or constructive knowledge of certain facts. The other was that these included the fact that the claimant's personal injuries were "attributable" to conduct of the proposed defendant (which was described in the original statute in terms of negligence, nuisance or breach of duty, but in the new statute as "the act or omission which is alleged to constitute negligence, nuisance or breach of duty").
30. The new legislation also has produced difficult problems for the courts. They can be roughly grouped under two general heads. First, what is it that the claimant has to know at the date of knowledge ("the what? question"). Secondly, how must the claimant know what he has to know – that is, what state of mind, assessed subjectively or objectively or by a mixture of the two, amounts to knowledge for this purpose ("the how? question"). The what? question depends on the interpretation and application of section 14(1) of the 1980 Act, and in particular (since it gives rise to most of the problems) section 14(1)(b), which relates to the fact "that the injury was attributable in whole or in part to the act or omission which is alleged to constitute negligence, nuisance or breach of duty". The how? question depends partly on the interpretation and application of section 14(3) of the 1980 Act:

"For the purposes of this section a person's knowledge includes knowledge which he might reasonably have been expected to acquire –

(a) from facts observable or ascertainable by him; or

(b) from facts ascertainable by him with the help of medical or other appropriate expert advice which it is reasonable for him to seek;

but a person shall not be fixed under this subsection with knowledge of a fact ascertainable only with the help of expert advice so long as he has taken all reasonable steps to obtain (and, where appropriate, to act on) that advice."

It also depends on giving a fair and workable meaning to the provisions as a whole.

31. Almost all of the many authorities cited to the Court in this appeal are concerned with one or both of these questions. My perception is that the case law has made more progress in clarification of the what? question than of the how? question. That may be because in some of the leading cases the House of Lords or the Court of Appeal has been able to reach a conclusion on actual knowledge and has not found it necessary to consider constructive knowledge. For instance in Haward v Fawcetts [2006] 1 WLR 682 the defendants (a firm of accountants) relied only on the actual knowledge of the claimant, and the House of Lords found that his actual knowledge of the financial state of the business in which he had invested was sufficient to make it reasonable for him to consider whether his accountants' advice had been flawed. The case was concerned with section 14A of the 1980 Act, added by the Latent Damage Act 1986, but the same principles apply. So the difficulties of constructive knowledge do not feature in Lord Nicholls' admirably brief statement of the relevant principles at paras 7 to 15.
32. In Spargo v North Essex District Health Authority [1997] PIQR P235, P242 Brooke LJ referred to this branch of the law being already "grossly over-loaded with reported cases". That was fifteen years ago, and the overload has increased. But this appeal requires the Court, in the context of heavy group litigation, to grapple with some unresolved difficulties. In view of the differences of opinion in the Court I wish, while conscious of adding to the overload, to set out my reasons in my own words. I start with some observations on the what? question and then address the how? question.

The what? question

33. The case law on the concept of "attributable" has developed in a coherent way. It is not without its difficulties, especially in cases involving specialised and technical areas of expertise (discussed by Lord Mance in Haward v Fawcetts [2006] 1 WLR 682 at paras 114 to 121). But on the whole the case law is consistent and provides a workable test.
34. In Smith v Central Asbestos Co Ltd [1973] AC 518, 543 Lord Pearson quoted the Oxford English Dictionary definition of "attributable" ("capable of being attributed or ascribed, especially as owing to, produced by") and stated that "attributable" refers to causation. This view has been consistently followed in later authorities on the legislation in its present form. In Haward v Fawcetts [2006] 1 WLR 682, para 45, Lord Scott quoted a passage from the judgment of Hoffmann LJ in Hallam-Eames v Merrett Syndicates Ltd [2001] Lloyd's Rep PN 178, 181, which conveniently sets out some of the most important cases:

"In other words, the act or omission of which the plaintiff must have knowledge must be that which is causally relevant for the purposes of an allegation of negligence . . . It is this idea of causal relevance which various judges of this court have tried to express by saying the plaintiff must know the 'essence of the act or omission to which the injury is attributable' (Purchas LJ in Nash v Eli Lilly & Co [1993] 1 WLR 782, 799) or ' the essential thrust of the case' (Sir Thomas Bingham MR in Dobbie v Medway Health Authority [1994] 1 WLR 1234, 1238) or that ' one should look at the way the plaintiff puts his case, distil what he is complaining about and ask whether he had in broad terms knowledge of the facts on which that complaint is based' (Hoffmann LJ in Broadley v Guy Clapham & Co [1993] 4 Med LR 328, 332)."

35. In this context, therefore, "attributable" has been interpreted by the courts as directed to a real possibility of a causal link: Lord Nicholls in Haward v Fawcetts at para 11, citing Nash v Eli Lilly & Co at pp 797-798. In that case Purchas LJ (who gave the judgment of the Court) quoted with approval some observations of Hidden J in his second judgment on the preliminary issue:

"The stark strength of the word 'knowledge' does not stand alone. It is knowledge that attribution is merely possible, a real possibility and not a fanciful one, a possible cause as opposed to a probable cause of the injury."

36. At this point the what? question and the how? question come into close proximity, since confident knowledge that there may be some causal link between two events is not dissimilar from a less confident belief that there is indeed a causal link between them. So the way in which "attributable" has been interpreted in the case law eases the Court's task in deciding whether "knowledge" includes more or less firmly held belief. But it does not remove all the difficulties, as this appeal shows.
37. Broadley v Guy Clapham & Co [1994] 4 All ER 439 was an unusual case because it involved a double limitation point. Mrs Broadley had a complaint against a surgeon who had operated on her in August 1980, but she did not consult a solicitor (the defendant) until June 1983. The solicitor arranged for her to see a specialist in July 1983, who gave a favourable oral opinion. But for some unexplained reason nothing was done to pursue the claim and in August 1990, having taken other legal advice, Mrs Broadley sued the solicitor whom she had consulted in 1983. He pleaded that the claim against him was statute barred, because (as he contended) her claim against the surgeon became statute barred in August 1983, and so that was when any cause of action against him arose. So there was an issue as to whether the standard three-year period applied to her original claim against the surgeon, or was to be treated as extended under sections 11 and 14 of the 1980 Act.
38. In his judgment Hoffmann LJ used a colloquial expression, "barking up the wrong tree", which has been repeated in some later cases. He said ([1994] 4 All ER 439, 449):

"Ordinarily it will suffice that he knows that the injury was caused by an act or omission of the defendant. But there may be cases in which his knowledge of what the defendant did or did not do is so vague and general that he cannot fairly be expected to know what he should investigate. He will also not have reached the starting point if, in an unusual case like Driscoll-Varley v Parkside Health Authority, he thinks he knows the acts and omissions he should investigate but in fact he is barking up the wrong tree."

Driscoll-Varley v Parkside Health Authority [1991] 2 Med LR 346 is mentioned a little earlier in the judgment. It was a case in which the plaintiff thought that an injury to her leg had been caused by a surgeon's negligence, but later discovered that the real cause was not the operation but the removal of the leg from traction during subsequent treatment. It seems a rather marginal example of barking up the wrong tree, since the plaintiff's misapprehension was in relation to the causative event in a single course of treatment, although the real complaint was about the after-care rather than the operation itself.

39. The point is relevant in this appeal because Mr Dingemans QC put in the forefront of his case the submission that those of his clients who thought they had been exposed to ionising radiation were barking up the wrong tree, because they were focusing on "prompt" (gamma ray) radiation. Foskett J was inclined to accept that submission (para 515, in the course of the discussion of his "preferred view"). The Court of Appeal (para 86) rejected this, having observed in the previous paragraph that the claimants' contention on this point demonstrated a fundamental misunderstanding of the concept of knowledge for limitation purposes.

The how? question

40. That leads on to the how? question. Many of the authorities which discuss this question are concerned with a range of significant injuries (such as dermatitis, hearing loss or pneumoconiosis) caused by an employer's failure to provide a proper working environment and a safe system of work. One employee may be unaware of even the possibility that his injury is caused by his working environment. Another may be in a state of suspicion, which he would wish to have confirmed by a medical expert. Yet another may be totally convinced, on not wholly rational grounds, that the working environment is the cause of his trouble. One may wait an unreasonable length of time before taking medical advice; another may consult his general practitioner, but get no further; yet another may be referred to a specialist consultant. And where the potential claimant does seek medical advice, whether from a general practitioner or from a specialist, it may on occasion turn out to be wrong. So the courts have had to interpret and apply the provisions of section 14 to a wide variety of factual situations. I shall consider some of them, in chronological order.
41. Like Lord Phillips, I start with Davis v Ministry of Defence 26 July 1985, CA transcript 413 of 1985. The plaintiff worked for the defendant as a welder from 1955 until 1971. In 1973 he started an action for damages for dermatitis which he and his general practitioner thought to have been caused by dust in his working environment. For reasons that are not clear, the first action lapsed but in 1982 Mr Davis started a fresh action. The Court of Appeal allowed his appeal against an order striking out the new action. Lord Phillips sets out two passages from the judgment of May LJ including his much-quoted observation:

"'Knowledge' is an ordinary English word with a clear meaning to which one must give full effect: 'reasonable belief' or ' suspicion' is not enough."

42. I have to say that I find the judgment of May LJ quite puzzling. Early in the judgment he directed himself, correctly, that "attributable" meant "capable of being attributed to". He recorded that at the time when the first action was commenced, Mr Davis firmly believed that his trouble was caused by his work, that his doctor shared that view, and that a doctor who examined Mr Davis on behalf of the Ministry considered his condition to be "not unconnected with the work which he had been doing". On the other hand Mr Davis stated in an affidavit that his specialist medical opinions "were in some respects conflicting and confused" and that having considered counsel's final opinion he was forced to conclude that his dermatitis might have been caused by his own predisposition. That seems to leave open at least the possibility that it had been caused by dust in the workplace. But May LJ referred to "the combined state of mind of the appellant himself, as a layman, and that of his doctors and legal advisers" as not amounting to knowledge in the relevant sense. I am left wondering whether, although asking himself whether they knew that the dermatitis was capable of being attributed to the working environment, May LJ was setting too high a threshold in his interpretation of "capable of being attributed", as compared with the passages referred to in para 35 above.
43. In Halford v Brookes [1991] 1 WLR 428 Lord Donaldson of Lymington MR described the facts of Davis v Ministry of Defence as highly unusual. He summarised the advice given to Mr Davis as more conclusively unfavourable to him than appears from my reading of the transcript. But on any view Halford v Brookes was itself a much more unusual case, in that it was concerned (under section 14(1) (c)) with the identity of the proposed defendant or defendants in a claim arising out of a lethal attack on a teenage girl. That was the context in which Lord Donaldson made his much-quoted observation that "reasonable belief will normally suffice". But in fact he concluded that the plaintiff (the dead girl's mother) "knew (with sufficient confidence to justify embarking on the preliminaries to the issue of a writ against both defendants)" all the facts listed in section 14(1), including that the acts of violence against her daughter were done by one or other, or both, of the proposed defendants. The Court of Appeal held that actual knowledge was established, and that it was not a case in which constructive knowledge had any part to play. The Court of Appeal exercised discretion under section 33 of the 1980 Act to allow the claim to proceed.
44. Neither of those cases can be said to have settled the law, but Halford v Brookes has had much more influence on its development. Nash v Eli Lilly & Co [1993] 1 WLR 782 is the first case bearing any resemblance to the present appeal. The limitation issues arose in class actions alleging injuries caused by a pharmaceutical product for relief of arthritic pain, marketed in the United Kingdom as Opren between October 1980 and August 1982, when it was withdrawn because it was producing unacceptable side-effects including photosensitivity and onycholysis, and sometimes fatal liver and kidney failure. On the trial of preliminary issues Hidden J held that the claims of almost all of the lead plaintiffs were statute-barred, and declined to exercise discretion under section 33 in their favour. The Court of Appeal allowed three of the eighteen appeals, two on the grounds that the claims were not statute-barred, and one by exercising discretion under section 33 (the judgment on the individual appeals is not reported).
45. The judgment of the Court of Appeal had to deal with three aspects of section 14(1): significant injury under para (a), attributability under (b), and (because different companies in the pharmaceutical group were sued) identification of defendants under para (c). It also had to consider the how? question, including constructive knowledge under section 14(3).
46. In a section of the judgment headed "Knowledge" the Court of Appeal discussed Davis v Ministry of Defence and Halford v Brookes and tended to prefer the approach in the latter case (p 792):

"We do not, of course, intend to lay down a definition of the word 'knowledge' for the purposes of a statute in which Parliament left the word to speak for itself. In applying the section to the facts of these cases, we shall proceed on the basis that knowledge is a condition of mind which imports a degree of certainty and that the degree of certainty which is appropriate for this purpose is that which, for the particular plaintiff, may reasonably be regarded as sufficient to justify embarking upon the preliminaries to the making of a claim for compensation such as the taking of legal or other advice.

Whether or not a state of mind for this purpose is properly to be treated by the court as knowledge seems to us to depend, in the first place, upon the nature of the information which the plaintiff has received, the extent to which he pays attention to the information as affecting him, and his capacity to understand it. There is a second stage at which the information, when received and understood, is evaluated. It may be rejected as unbelievable. It may be regarded as unreliable or uncertain."

This was essentially a subjective approach. In relation to the issue of "significant" injury (which was an important issue in that case) a subjective element may appear to be mandated by section 14(2), but the House of Lords has recently shown that approach to be mistaken: see Lord Hoffmann in A v Hoare [2008] AC 844 at paras 33 to 35; compare Lady Hale at paras 56 to 61.

47. Taken together, the unanimous decisions of House of Lords in A v Hoare (on section 14(2)) and Adams v Bracknell Forest Borough Council [2005] 1 AC 76 (on section 14(3); paras 42 to 51 are particularly in point) appear to me to mark a decisive shift away from a subjective approach on these issues. What was within a claimant's actual knowledge is undoubtedly a subjective question. But the notion that "whether a claimant has knowledge depends both upon the information he has received and upon what he makes of it" (Nash v Eli Lilly & Co [1993] 1 WLR 782, 795) can no longer be accepted, at any rate without a lot of qualification. The recent authorities recognise that the policy of the law is for the date of knowledge to be ascertained in the same way for all claimants, without regard to their personal characteristics, which can be taken into account at the later stage of exercising discretion under section 33 of the 1980 Act. As Lord Hoffmann put it in Adams v Bracknell Forest Borough Council [2005] 1 AC 76, para 45:

"The Court of Appeal in Forbes [v Wandsworth Health Authority [1997] QB 402] was right in saying that the introduction of the discretion under section 33 had altered the balance. As I said earlier, the assumptions which one makes about the hypothetical person to whom a standard of reasonableness is applied will be very much affected by the policy of the law in applying such a standard. Since the 1975 Act, the postponement of the commencement of the limitation period by reference to the date of knowledge is no longer the sole mechanism for avoiding injustice to a plaintiff who could not reasonably be expected to have known that he had a cause of action. It is therefore possible to interpret section 14(3) with a greater regard to the potential injustice to defendants if the limitation period should be indefinitely extended."

Actual knowledge and constructive knowledge

48. Adams shows that in Nash v Eli Lilly & Co the Court of Appeal was wrong, in para 4 of the summary of its conclusions (at p796) to state that "the temporal and circumstantial span of reasonable inquiry [under section 14(3)] will depend on the factual context of the case and the subjective characteristics of the individual plaintiff involved." But that is not the only point on section 14(3) that calls for examination. As already mentioned, many of the reported cases were decided simply on actual knowledge. It may be that both litigants and judges tend to regard that as a more satisfactory approach, with a focus on the claimant's oral evidence given at the hearing of a preliminary issue, or at trial. The issue of constructive knowledge generally calls for more elaborate pleadings and for expert evidence. Although the general burden of proving that he is entitled to the benefit of a deferred date of knowledge is on the claimant, in practice it is for the defendants to raise issues under section 14(3), as Haward v Fawcetts illustrates (the issue of burden of proof in these cases was fully, and in my view correctly, examined by Mance J in Crocker v British Coal Corporation (1995) 29 BMLR 159, 169-173). So in practice the parties tend to join issue on actual knowledge, and judges to reach a conclusion on that issue, with constructive knowledge being held in reserve, as it were. As Lord Phillips points out in para 119 of his judgment, the well-known survey of the relevant principles made by Brooke LJ in Spargo v North Essex District Health Authority [1997] PIQR P235, P242 does not deal with constructive knowledge at all.
49. But understandable though it is that courts may tend to look first at actual knowledge, that approach does not give full effect to Parliament's purpose in enacting section 14(3). What the statute requires is a single inquiry as to the claimant's knowledge, which under section 14(3) is extended, not only to facts which he could have learned with the help of "medical or other appropriate expert advice", but also more generally to "facts observable or ascertainable by him". There is little authority as to these wide general words, but it was suggested in Nash v Eli Lilly & Co [1993] 1 WLR 782, 800, that they would include any relevant information that had been given wide publicity in the press or on television, for instance as to a drug's unacceptable side-effects, or its withdrawal from the market. In this appeal the Ministry of Defence has pleaded a large number of matters of that sort, starting in 1945 and going down to 1999, in sub-paragraphs (a) to (j) of para 31 of its points of defence on the limitation issue.
50. Adams marks a very important shift towards a more objective approach to the claimant's state of knowledge. This goes a long way to blunt or blur the clear distinction, in ordinary discourse, between knowledge and belief. As Simon Brown LJ said in O'Driscoll v Dudley Health Authority [1998] Lloyd's Rep Med 210, 221, "knowledge and belief inevitably shade into one another." Lord Donaldson's well-known statement that "reasonable belief will normally suffice" is reinforced, but weight must be given to the belief being "reasonable" – or, as Lord Wilson suggests, "reasoned".

The significance of legal advice

51. There is one further problem about the how? question that I must address, before trying to draw some conclusions. It is the significance of the claimant seeking legal advice by consulting a solicitor. This is a topic that crops up repeatedly in the authorities, and judicial opinions have varied a good deal.
52. In Halford v Brookes [1991] 1 WLR 428, 434, Russell LJ rejected the suggestion that "other appropriate expert advice" included legal advice. One of the most important changes, when the Limitation Act 1963 was replaced by the Limitation Act 1975, was to get away from the claimant needing to know about the technicalities of different causes of action. In general, legal advice is not a prerequisite to knowledge within the meaning of the 1980 Act (though this must be qualified in some cases within section 14A concerned with questionable advice on technical matters such as financial services and pensions: Haward v Fawcetts [2006] 1 WLR 682, paras 59 to 62 and 113 to 117).
53. In line with that, in Halford v Brookes (p 443) Lord Donaldson put forward the test of knowledge as "know with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking legal and other advice and collecting evidence." This formulation has been taken up in later cases, notably Nash v Eli Lilly & Co [1993] 1 WLR 784, 796 (point 3), Spargo v North Essex District Health Authority [1997] PIQR P235, P242 (point 3) and Haward v Fawcetts [2006] 1 WLR 682, para 9 (Lord Nicholls).
54. That is a formidable line of authority. But still there is no clear consensus. Most strikingly, in Sniezek v Bundy (Letchworth) Ltd [2000] PIQR P213, Judge LJ (at P229) rejected the notion that time automatically starts to run against a client who has taken legal advice, whereas Simon Brown LJ (at P 234) found it "difficult indeed to imagine a case where, having consulted a solicitor with a view to making a claim for compensation, a claimant could still then be held lacking in the requisite knowledge."
55. I respectfully but unhesitatingly prefer the view of Judge LJ. The typical scenario for a claim for personal injury sustained from a bad working environment (exemplified by Ali v Courtaulds Textiles Ltd (2000) 52 BMLR 129) is for the potential claimant to go for medical advice to his general practitioner. The overworked GP is naturally more interested in diagnosis and treatment than in aetiology, unless his patient presses him. It is often a trade union representative (or in Mr Ali's case a community worker) who at some later date advises the claimant to take legal advice, which at that stage can be no more than preliminary; it generally results in a referral to a medical specialist who is asked to advise on the likely cause of the trouble, as well as on the seriousness of the injury and its prognosis. The facts of Sniezek, as recounted in detail by Bell J at P216 to P217, show how protracted and uncertain that process can be. Mr Sniezek first consulted his union solicitors in 1990; it was 1994 before he obtained favourable medical advice linking the hyposensitivity of his aerodigestive tract with polymer exposure; and further investigations postponed the issue of the writ until 1998 (the reference to 1988 on P217 of the report is one of several obvious errors in editing).
56. So in practice a claimant's first visit to a solicitor may do no more than initiate the process of obtaining expert medical advice. That process may take years, with the solicitor's function limited to the collation of medical and other technical evidence (such as the nature of the polymer in Ali, or the nature of the pesticide in Griffin v Clwyd Health Authority [2001] EWCA Civ 818, [2001] PIQR P31). In the present appeal several different branches of science and medicine are relevant to the what? question under section 14(1)(b), as appears from the different specialisms of the expert witnesses on both sides.
57. To return to the original formulation in Halford v Brookes, it is clear that Lord Donaldson envisaged that the collection of evidence to support the claimant's claim was something which would normally come after the date of knowledge, when the claimant first knows that he has a possible claim. That is how it was understood by Hoffmann LJ in Broadley v Guy Clapham & Co [1994] 4 All ER 439, 449. In a passage just before his reference to "barking up the wrong tree" Hoffmann LJ observed:

"How does one determine the 'essence' of the act or omission? The purpose of section 14(1), as Lord Donaldson MR pointed out in Halford v Brookes [1991] 1 WLR 428 at 443, is to determine the moment at which the plaintiff knows enough to make it reasonable for him to begin to investigate whether or not he has a case against the defendant. He then has three years in which to conduct his inquiries and, if advised that he has a cause of action, prepare and issue his writ."

58. Sniezek shows that in practice three years may not be enough where the claim for personal injury raises difficult issues of causation, and in the present appeal the causation issues are very complex indeed. Nevertheless there is a distinction in principle between a claimant's knowledge (actual or constructive) that he has a real possibility of a claim (Brooke LJ's second point in Spargo), and the assembly by the claimant and his legal team, with the help of experts, of material justifying the commencement of proceedings with a reasonable prospect of success. Of all the difficulties in this anxious appeal, the biggest difficulty of all, to my mind, is in the practical application of this abstract distinction between knowledge of the "essence" of a claim and the evidence necessary to prove it to the requisite legal standard.

The judgments below

59. There has been a good deal of discussion of the judge's "preferred view" referred to in paras 514 to 521 of his judgment. It is apparent from para 521 that it amounts to setting a relatively high "threshold [to] the level of appreciation of the material matters." But it is not entirely clear whether this relates to the degree of specificity of the section 14(1)(b) facts (which in this appeal is much the most important element in the what? question) or to the clarity or confidence of the lead claimants' state of mind (the how? question). Paras 514 to 517 are concerned with the specificity of the facts, but then the judge seems to move on to the claimants' state of mind.
60. The Court of Appeal had no doubt that the preferred view set the threshold too high. It stated (para 85):

"It is clear from the principles set out in Spargo that it is the knowledge of possibilities that matters; a claimant needs only enough knowledge for it to be reasonable to expect him to set about investigation. He can have knowledge even though there is no helpful evidence yet available to him. The claimants' contention that they did not have knowledge of possible attributability until they received the results of the Rowland study demonstrates a fundamental misunderstanding of the concept of knowledge for limitation purposes."

This is to be contrasted with the judge's preferred view (in para 514 of his judgment) that a claimant would not have knowledge unless he appreciated, not only that his injury was capable of being caused by abnormal radiation, but also that "there is some credible evidence that he was exposed to ionising radiation" at an abnormal level during, or shortly after, and in consequence of, the nuclear tests.

61. The Court of Appeal described the judge's preferred view (and the critical importance which it places on the Rowland study) as demonstrating a fundamental misunderstanding. In my respectful view this criticism is too strongly expressed. The judge had, during his ten days of the hearing of the preliminary issue, and the further period when he was writing his very clear and comprehensive judgment, taken on board an enormous mass of complex evidence and some intricate legal submissions. There was no legal test by which he could, as with an alchemist's touchstone, distinguish "essence" from evidential support. It was a matter of considering the voluminous material before him, stepping back, and making an evaluative judgment. It is an exercise on which an appellate court will be slow to differ from the trial judge who has seen and heard several of the lead claimants (or their widows) giving evidence.
62. I respectfully doubt whether the Court of Appeal was right to differ from the judge in his conclusion that the belief of many of the claimants that they had been exposed to prompt radiation was a significant misconception which (had it stood alone) would have amounted to "barking up the wrong tree". But as the Court of Appeal pointed out, it did not stand alone. The facts as to fallout exposure to alpha and beta radiation were readily available and widely known, and exposure to fallout was pleaded as part of the lead claimants' case.
63. More crucially, however, I respectfully consider that the judge was wrong, not only in his preferred view, but also in his evaluation of the state of knowledge at the lower level of appreciation which he instructed himself to apply. Even under the more demanding test adopted on the preferred view, it was common knowledge from the 1980s (indeed, from soon after the bombs dropped on Hiroshima and Nagasaki in 1945) that exposure to fallout radiation could cause leukaemia, many other forms of cancer, infertility and other serious injuries. It was also well known that many of the 22,000 service personnel who took part in the nuclear tests had been exposed to fallout radiation which, while relatively low, was above the normal background radiation to which all living creatures are exposed.
64. The real difficulty for the claimants was to produce cogent evidence, either from their individual medical histories or from epidemiological material, that the dose of radiation was sufficiently high for a causative link with their injuries to be established on the balance of probabilities. The Ministry of Defence adamantly maintained throughout that their exposure was for practical purposes negligible, and this seems to have been confirmed by successive NRPB epidemiological reports in 1988, 1993 and 1999 (apart from a small additional risk in respect of most forms of leukaemia and multiple myeloma) and by the Phelps-Brown study (of cataracts) in 1996-1997. The Rowland study (the results of which were made available to the claimants in 2007, before its full publication in 2008) was seen by the claimants and their advisers as a long-awaited breakthrough in the evidence of causation (the Ministry of Defence are very critical of this report, but that issue lies in the future). All this is carefully recorded, in very much greater detail, in the judge's judgment. But the judge did in my view err in treating the Rowland report as essential rather than evidential. Putting it in the simplest terms (and I am very conscious of the danger of over-simplification in this appeal), I think that the judge erred on the what? question rather than on the how? question.
65. My final position is therefore close to that set out in Lord Wilson's judgment, and I gratefully adopt his summary (at paras 16 to 24) of the particular circumstances of the individual appellants. I agree with Lord Wilson that it was appropriate for the Court of Appeal to make a fresh exercise of discretion under section 33 of the 1980 Act. I also agree that because of the unusual course which the preliminary issue has taken, and the mass of evidence touching on the causation issue, the Court of Appeal was in an unusually good position to exercise that discretion, and this Court should not interfere with its decision not to let any of the appellants' claims proceed.
66. I do however have reservations about Lord Wilson's proposition (concurred in by Lord Mance) that the effect of the statutory provisions is that the claimant is assumed to have a cause of action. No doubt this is correct in the general sense that every claimant who issues a claim form commencing contentious proceedings is assumed to have a cause of action unless and until his particulars of claim are struck out, or the action is discontinued or dismissed. But I do not see that this general assumption helps, and it may actually be a distraction, in understanding the way that sections 11, 14 and 14A of the 1980 Act operate. The putative character of the section 14(1)(b) and section 14A(8)(a) facts depends not on any implicit assumption but on the long-standing and consistent meaning which the courts have given to "attributable." So I am inclined to think that this is a novel and unnecessary refinement.
67. Like Lord Wilson and Lord Mance, I most respectfully disagree with much of Lord Phillips's reasoning. I do not see how a claimant who has issued a claim form claiming damages for personal injury can be heard to suggest that he did not, when it was issued, have the requisite knowledge for the purposes of the 1980 Act. More generally, I consider that the practical result of Lord Phillips's analysis would be a situation that Parliament cannot have intended when it enacted these provisions. It would mean that persons (and sometimes, as in this case, large groups of persons) with a belief that they had suffered personal injuries through the fault of a government department or local authority, or any other public-sector or private-sector body, but with no real prospect of proving legal liability on the balance of probability, would be able to keep their claims on ice, as it were, for an indefinite period, in the hope that one day the right evidence might turn up.
68. Our judgments on this appeal will not, I fear, be an ideal source of guidance to lower courts which regularly have to deal with these difficult problems. There are two reasons for that: the extreme complexity of this group litigation, and the division of opinion in the Court. For my part I would suggest that short summaries like that of Brooke LJ in Spargo (which Lord Phillips rightly describes as a "valiant attempt") may be unhelpful if treated as if they were statutory texts. The words of the 1980 Act themselves must be the starting-point, illuminated where necessary by judicial exposition, of which the opinion of Lord Nicholls in Haward v Fawcetts [2006] 1 WLR 682, paras 8 to 15, is the most authoritative. To that guidance I would tentatively add two points. In a complex case section 14(3) is an essential part of the statutory scheme, not an occasional add-on. And the date of a claimant's first visit to a solicitor is (without more) of very little significance in most cases.

LORD BROWN

69. I too would dismiss these appeals for the reasons given by Lord Walker, Lord Mance and Lord Wilson. I do not believe that there are any significant differences between their three judgments but, if there are, and if something approaching a canonical text is required, I would align myself principally with Lord Wilson's reasoning.
70. Perhaps the most critical proposition to which each of the above three judgments commits is (in Lord Wilson's words at para 3): "It is a legal impossibility for a claimant to lack knowledge of attributability for the purpose of section 14(1) at a time after the date of issue of his claim". As Lord Walker puts it at para 67: "I do not see how a claimant who has issued a claim form claiming damages for personal injury can be heard to suggest that he did not, when it was issued, have the requisite knowledge for the purposes of the 1980 Act."
71. Although Lord Walker in the previous paragraph expressed "reservations about Lord Wilson's proposition (concurred in by Lord Mance) that the effect of the statutory provisions is that the claimant is assumed to have a cause of action", I do not myself understand these reservations to amount to any ultimate difference in approach. Rather it seems to me that the only point Lord Wilson (and Lord Mance) are making when they say that, in deciding whether a given claim is statute-barred, the court has to assume that the claimant has knowledge of the facts necessary to support his pleaded cause of action, is that the claimant cannot at that stage be heard to suggest otherwise – ie just what Lord Walker then says in the above-quoted para 67. In short, once a claimant issues his claim, it is no longer open to him to say that he still lacks the knowledge necessary (by reference to sections 11 and 14) to set time running.
72. On Lord Phillips' approach, the more hopeless the claim, the likelier it is that the claimant will be in a position to defeat the Limitation Act defence, and this, indeed, no matter how long ago (some half a century in the present cases) the alleged cause of action arose. With the best will in the world, this simply cannot have been Parliament's intention.
73. I share to the hilt Lord Phillips' view (expressed at para 158) that these claims have no reasonable prospect of success. But I profoundly disagree with his conclusion that on this account, because "there were no known facts capable of supporting a belief that the veterans' injuries were attributable to exposure to ionising radiation" (para 139), even the Rowland report "fall[ing] well short of establishing causation according to established principles of English law" (para 157), time has still to this day not begun to run.
74. Nor do I find any more persuasive Lord Kerr's view that "time [began] to run from the date that [the veterans] became aware of or ought to have been aware of the contents of the Rowland report" (para 211) so as to delay the claimants' date of knowledge until after their claims were issued – presumably until they saw the Rowland report in June 2007.
75. The plain fact is that, despite decades spent urgently trying to assemble a viable case, on the evidence as it presently stands these claims (in which huge costs have already been expended) are doomed to fail. As the claimants' then leading counsel readily accepted in argument for the Court of Appeal, "We haven't got material which gets you near a balance of probabilities" so that "a further policy exception" (to the Fairchild exception) would be needed to allow for a claim based merely on a material increase in risk – a development of which, in the light of this court's judgments in Sienkiewicz v Greif [2011] 2 AC 229, Lord Phillips at para 157 rightly recognised there to be "no foreseeable possibility".
76. Even had I been persuaded that time had not run in any of these cases I would, like Lord Mance, nevertheless have been disposed to dismiss them. As he says at para 88: "If proceedings have no proper basis in fact, they should not be allowed to persist."
77. In short, although the veterans can hardly be expected to recognise this, these appeals now provide the court with the opportunity, rather than yet again to extend, instead once and for all to end, the false hopes on which these claims have for so long rested.

LORD MANCE

78. Lord Phillips and Lord Wilson have expressed radically different views about the concept of knowledge in the Limitation Act 1980. The present appeals concern personal injuries claims, and their disagreement relates to the "knowledge" referred to in sections 11(4)(b), 12(2)(b) and 14. But parallel disagreement must necessarily exist between them with regard to sections 11A(4)(b) and (5)(b) (actions in respect of defective products) and 14A (special time limit for negligence actions, other than for personal injuries, where facts relevant to cause of action are not known at date of accrual). The correct resolution of this disagreement is of general importance.
79. In my opinion, Lord Wilson's analysis is consistent with and correct in the light of prior authority, and is the analysis which makes sense of the statute and its purpose. I agree with his reasoning and conclusions. I shall not repeat his examination of authority, but content myself with a few points. First, the statute assumes that a cause of action has accrued (section 11(4)(a)) and that "facts" exist of which "knowledge" may exist (section 14(1)). Such facts include an injury which is significant (section 14(1)(a)), and which is attributable to an act or omission now alleged to constitute negligence, nuisance or breach of duty (section 14(l)(b)). "Attributable" here means "capable of being attributed" as a "possible cause of the damage, as opposed to a probable one": see Spargo v North Essex District Health Authority [1997] PIQR P235 and Haward v Fawcetts [2006] UKHL 9, [2006] 1 WLR 682, paras 10-11, per Lord Nicholls). The facts further include the identity of the defendant (section 14(1)(c)) and, if the relevant act or omission is of some other person, the identity of that person and the additional facts supporting the bringing of an action against the defendant (section 14(1)(d)).
80. The assumption that a cause of action and relevant facts exist favours the claimant. They are taken as given. There is no investigation at this stage as to whether they can be made good. The facts to which paragraphs (a) to (d) of section 14(1) refer must be ascertained from the way in which the claimant puts his or her case in the proceedings which are being pursued. Hoffmann LJ encapsulated this in a much-quoted sentence in Broadley v Guy Clapham & Co [1994] 4 All ER 439, 448h-j:

"Section 14(1)(b) requires that one should look at the way in which the plaintiff puts his case, distil what he is complaining about and ask whether he had, in broad terms, knowledge of the facts on which that complaint is based".

81. This passage was repeated by Hoffmann LJ, giving the judgment of the court, in Hallam-Eames v Merrett Syndicates Ltd [2001] Lloyd's Rep PN 178, 181. Significantly, it received full approval by the House of Lords in Haward v Fawcetts [2006] 1 WLR 682, a decision under section 14A(5). Lords Nicholls, Lord Walker and I all quoted the passage with approval (paras 10, 62 and 120), with Lord Walker adding: "The court is concerned with the identification of the facts which are the 'essence' or 'essential thrust of the case' or which 'distil what [the claimant] is complaining about' (para 66). Lord Scott accepted and applied "the opinions expressed in Nash v Eli Lilly & Co [1993] 1 WLR 782, Dobbie v Medway Health Authority [1994] 1 WLR 1234 and Hallam-Eames … that the requisite knowledge is knowledge of the facts constituting the essence of the complaint of negligence" (para 49). Lord Brown said that "What the claimant must know to set time running is the essence of the act or omission to which his damage is attributable, the substance of what ultimately comes to be pleaded as his case in negligence" (para 90).
82. The speeches in the House of Lords endorsed guidance regarding the concept of knowledge given in a series of Court of Appeal decisions, going back to Halford v Brookes [1991] 1 WLR 428. Lord Nicholls said (para 9):

"Thus, as to the degree of certainty required, Lord Donaldson of Lymington MR gave valuable guidance in Halford v Brookes [1991] 1 WLR 428, 443. He noted that knowledge does not mean knowing for certain and beyond possibility of contradiction. It means knowing with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking advice, and collecting evidence: 'suspicion, particularly if it is vague and unsupported, will indeed not be enough, but reasonable belief will normally suffice'. In other words, the claimant must know enough for it to be reasonable to begin to investigate further."

Lord Walker noted at para 57 the numerous cases showing that "the starting point may occur at a time when a claimant's knowledge about his complaint is far from complete", that a claimant "may have the requisite knowledge …. 'even though he may not yet have the knowledge sufficient to enable him or his legal advisers to draft a fully and comprehensively particularised statement of claim'", but that "by the time, often years later, that the limitation issue comes to be decided, whether as a preliminary issue or at trial, the claimant's case will have been pleaded, and the defendant's 'act or omission which is alleged to constitute negligence' will (or at any rate should) have been clearly identified". I referred to the same passage as Lord Nicholls (paras 112 and 126).

83. These passages indicate that courts, by using the words "reasonable belief" as part of the description of the requisite knowledge, are focusing not so much on whether or how far the belief is evidence-based, but more on whether it is held with a sufficient degree of confidence to justify embarking on the preliminaries to making a claim including collecting evidence. There is a degree of circularity about such a definition, but this is probably inherent in the concept of knowledge in any context (cf Insurance Corporation of the Channel Islands v Royal Insurance (UK) Ltd (unreported) (Comm Ct, 30 July 1997, BAILII: [1997] EWHC 373 (Comm)), where, in the different context of affirmation, I described it as a "jury question"). If a claimant is pursuing proceedings which he has issued for personal injuries and his state of mind when he issued them was in substance no different from his state of mind for more than the three prior years, then, in agreement with Lord Wilson's para 5 and the passages he there cites, I find it difficult to see how he can claim in those proceedings that he lacked sufficient knowledge of the facts asserted for the purposes of the Limitation Act 1980.
84. It is of course for a claimant to put his case as he thinks fit. No-one is bound to commence proceedings, and the position may be different – it is unnecessary to decide - if the claimant has issued proceedings which he is no longer pursuing and in relation to which no limitation issue can therefore arise (as was the case in Whitfield v North Durham Health Authority [1995] 6 Med LR 32). But, if a claimant elects to issue and is pursuing proceedings, he must identify the case made and stand by it. Among the allegations which must, either explicitly or implicitly, be made, is that the case is not time-barred. Once an issue of knowledge is identified as arising under sections 11(4)(b) and 14(1), the onus lies upon the claimant to make good his case on knowledge, as I noted in Haward v Fawcetts, para 106. A claimant bringing proceedings necessarily asserts that he or she has a properly arguable claim. In the present cases, the claims were expressly to the effect that the claimants had suffered personal injuries by reason of the negligence of the defendant in exposing them to radiation, radioactivity or contaminated material in one way or another. In modern procedure, such an assertion is attested by a statement of belief, as Lord Wilson notes in para 3, and so it was here. Once proceedings are begun, it is by reference to the facts asserted as giving rise to the claim that the question of knowledge must be tested. The claimant cannot avoid this. Indeed, it is difficult in normal circumstances to think of a claimant trying to do so.
85. Nor did the claimants originally try to do so in the present case. They pleaded a case of conventional causation. However, shortly before and at trial, the case run acknowledged in effect that causation could not be established as a matter of probability. The argument then was that a material increase in risk was sufficient. The hope was to invoke the principle or an extension of the principles in Bonnington Castings Ltd v Wardlaw [1956] AC 613 and/or Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22, [2003] 1 AC 32. That was and is, however, a hope without prospect of success. During the trial of the present issue, the emphasis shifted to an attempt to show that, by the time of any trial on the merits, the claimants could hope to have acquired evidence to show causation by reference to a balance of probability or a doubling of risk or a synergical effect.
86. As matters stand, the claimants clearly have no case on causation. But that is no answer in my opinion to their limitation problems. They have chosen to bring proceedings on the basis of certain facts. Whether the facts by reference to which their case falls to be assessed for limitation purposes are those pleaded (a straightforward allegation of causation) or those later asserted (an increase in the risk of injury being caused or, now, an admission that the claimants cannot presently establish causation, coupled with a submission that the proceedings should continue in the hope that causation will in future prove possible to establish), the limitation question is not whether those facts give rise to a good claim in law. It is when the claimants first had knowledge of those facts, in accordance with the test indicated in Halford v Brookes. This they did, in each of the nine cases before the Court, more than three years prior to the issue of the proceedings (or, in the case of Mr Ogden, more than three years prior to his death).
87. The opposite view of the law taken by Lord Phillips leads him to the conclusion that the claimants can overcome or avoid any limitation problems, because they have never had and still do not have the knowledge of any facts which could lead to success. The defendant's appropriate response to this situation would, in his view, have been to apply to strike out the claims or for summary judgment. If Lord Phillips is right about this bifurcation of remedies, then, contrary to Lord Phillips' reference to the present case involving "an unusual feature" (para 93) or "unusual facts" (para 147), I think that it could often be relevant. The present case illustrates that the weakness of a claimant's factual case may become apparent in relation to issues arising in conjunction with a limitation defence, such as an application for an extension of time under section 33. Defendants often deny that any factual basis exists for a claim, and, particularly though not exclusively where the claim is said to have been fabricated, it would follow that, on their case, the claimant could never have had knowledge, in the evidence-based sense in which Lord Phillips uses the word, of essential facts on which the claim was based. As a matter of caution, a defendant contemplating the possibility that the claim might be time-barred would be bound to consider the possibility that the court might conclude that the claimant had probably not had evidence-based knowledge of the facts alleged, and so that the claim could not properly be struck out. To cover this possibility the defendant would have to adopt a double-limbed approach. One limb would be based on limitation; in relation to that the onus would be on the claimant. The other limb would involve an application to strike out or for reverse summary judgment, grounded on the absence of any factual basis for the claimant's case; in relation to that the onus would be on the defendant.
88. In the present case, Lord Phillips concludes that the claimants can overcome the limitation problem, because even now they have no evidence for the facts that they need to show in order to succeed. But Lord Phillips refuses to strike out or dismiss the claims, because the defendant has not pursued any formal application to that effect. That is a result which would I think be viewed with some surprise by an observer of the English legal system. It is not one with which I could concur, even if I were otherwise of Lord Phillips' view. If proceedings have no proper basis in fact, they should not be allowed to persist. I agree with Lord Wilson's remarks in this connection, particularly with his indication that the fact that this is a group action should not be allowed to prejudice any other claimant who may show that in his or her particular circumstances there is a viable claim which is not time-barred.
89. On the question whether there should be an extension of time under section 33, the Court of Appeal was in my view right in concluding that the judge erred in the exercise of his discretion and that it was incumbent on it to re-exercise the discretion, as it did on a generic basis. To the reasons it gave, particularly in paras 103 to 111, one might add the judge's under-estimate of the difficulties on causation (evident for example in paras 187 and 230 of his judgment) when linked with references to the claimants genuinely believing "on apparently reasonable grounds" that they have a case and to a "credible fall-out case" (paras 618 and 625). These passages also suggest that he must have approached the issue of discretion on a wrong basis.

DISSENTING JUDGMENTS

LORD PHILLIPS

Introduction

90. Between October 1952 and September 1958 the respondent ("MoD") carried out experimental explosions in the atmosphere of a total of 21 thermonuclear devices. This was a mammoth operation. It took place in Australia and the South Pacific and involved approximately 22,000 soldiers, sailors and airmen, many of whom were performing National Service. From these servicemen are drawn the majority of the 1011 claimants, most of whom commenced a group action on 23 December 2004 but a minority of whom have joined the action by claim forms issued on various dates between 16 November 2007 and 29 September 2008. They have become known as "atomic veterans" and I shall call them "the veterans". Some of the claims are brought by the personal representatives of veterans who have died. Each claim alleges breach of duty on the part of the MoD in exposing the veteran to radiation that has caused illness, disability or death. I shall refer to these alleged consequences, which in most cases involve some form of cancer, simply as "injuries".
91. There is an issue in many of the individual cases as to whether the claim is time-barred under the provisions of the Limitation Act 1980. On 5 July 2007 the Senior Master ordered, inter alia, that this question be tried as a preliminary issue. Further to that order the group and the MoD each selected five lead cases for the trial of the issue of limitation. The object of the Senior Master's order was to obtain rulings on issues that are generic to all the cases. It has been common ground that the question of whether their claims are time-barred has to be decided case by case on consideration of the particular facts of each case, but there are issues of law and of the application of the law in a case such as this that are generic. The application of the 1980 Act to claimants involved in group litigation raises particular difficulties that will have to be explored.

The issues

92. Three generic issues arise. Sections 11 and 14 of the 1980 Act ("section 11" and "section 14") together provide that the limitation period within which a claimant must bring a claim in respect of personal injuries that he has suffered is three years from the date when the cause of action accrued or, if later, the date when he acquired "knowledge" that he had sustained an injury that was "attributable" to the "act or omission" which he alleges constituted breach of duty on the part of the defendant. I shall refer to this as "knowledge of attributability" by way of shorthand. A similar provision in relation to knowledge applies in the case of a claim brought by a personal representative or dependant of someone who has died. For the sake of simplicity I shall throughout this judgment treat the veterans as being the claimants. The first generic issue relates to the extent to which knowledge can be equated with belief. So far as concerns the existence of facts, "knowledge" and "belief" are words that can, in some circumstances at least, be used to describe the same state of mind. My knowledge of my birthday is the same as my belief as to the day on which I was born. There is an issue as to whether, in all circumstances, knowledge can be equated with subjective belief for the purposes of sections 11 and 14.
93. The second generic issue arises out of an unusual feature of this case. It is the MoD's case that there are no known facts that support the allegations of breach of duty and causation pleaded by the claimants. It is the veterans' own primary case that they only acquired knowledge of attributability after they had commenced their proceedings. This raises the question of the effect of the 1980 Act and the proper approach of the court if proceedings are commenced before the litigant has acquired the knowledge that would normally cause time to begin to run. That question has to be considered in the context of a group action.
94. The third generic issue relates to section 33 of the 1980 Act ("section 33"). This gives the court power to allow an action to proceed notwithstanding that it has not been commenced within the limitation period. The Court of Appeal declined to exercise this power in relation to any of the veterans. A common reason for the decision in the case of each veteran was that the claim had no realistic prospect of success. There is an appeal against that decision in each case. The question arises of the relevance of individual prospects of success where group litigation is being pursued.

The uncertainties

95. The problems to which this appeal gives rise are due, in large measure, to the absence of evidence of fact that supports the claim that the veterans' injuries are attributable to exposure to ionising radiation. Exposure to radiation can damage your health in one of two ways. If you are close to the explosion you can be exposed to what is called "prompt" radiation from gamma rays. This radiation, while powerful, is short lived. Alternatively you may be exposed to fall-out of alpha and beta particles. These can be carried quite some distance from the seat of the explosion. If they are ingested by breathing or swallowing they can remain within the body for a significant period during which they will continue to radiate, producing a cumulative effect.
96. The master particulars of claim were served on 29 December 2006. The veterans' solicitor has endorsed them with the requisite statement on behalf of the veterans that they believe the facts stated in the particulars of claim to be true. Those facts include:

i) an allegation in para 2 that each veteran was exposed to "radiation, radiation contamination, radioactivity, radioactive fallout and/or biological residue during the conduct of the tests" and their aftermath;

ii) allegations in para 13 of both "external" and "internal" exposure and failure to protect against exposure to ionising radiation;

iii) allegations in para 13 of failure to prevent servicemen from contamination with radioactive fallout as a result of swimming and consuming seafood.

Thus both prompt and fallout exposure is alleged.

The uncertainties

97. The areas of uncertainty were and are twofold, albeit that the two are interlinked. The first is whether the veterans were exposed to radiation as alleged. The second is whether their injuries have been caused by exposure to radiation. Each of these matters is alleged by the veterans and denied by the MoD. Mr Dingemans QC submits that this uncertainty has recently significantly diminished. Although the claimants believed that the veterans had been exposed to ionising radiation there was no objective ground for this belief until the preparation of a report ("the Rowland report") in 2007. The Rowland report gives the results of tests on blood samples taken from 50 New Zealand veterans who had served on ships that were no closer to the site of some of the tests than had been most, if not all, of the claimants. Many, though not all, of the samples showed an abnormal incidence of changes to chromosomes that was indicative of exposure to low dose radioactive fallout. Mr Dingemans submits that these tests provided, for the first time, objective grounds for concluding that the veterans were subjected to similar exposure. I shall deal with the significance of this submission in due course.
98. At the start of the hearing of the limitation issue before Foskett J the veterans abandoned that part of their claim that alleged exposure to prompt radiation. As allegations of exposure to prompt radiation had been at the forefront of their claims this was a dramatic change of stance. The claims are now solely based on alleged exposure to radioactive fallout. If that alleged exposure can be proved, it does not follow that the veterans have viable claims. It will still be necessary to prove that the injuries in respect of which the claims are made were caused by the exposure. The veterans are not currently in a position to prove this. There is scientific evidence that demonstrates that ionising radiation is capable of causing some, at least, of the injuries in respect of which individual claims are brought. There are, however, other potential causes of such injuries. They are experienced by many of the same age as the veterans for a variety of reasons. The most that the evidence currently available can establish is that such low dose exposure as may be proved will have increased the risk to the particular veteran of sustaining the injury in respect of which the claim is made. There is no known basis for concluding that the exposure will have gone so far as to double that risk. On the law as it stands, merely proving an increase in risk will not establish a good cause of action. To succeed a veteran must show that, on balance of probability, the injury would not have been sustained had it not been for the exposure. In the course of argument Mr Dingemans accepted that none of the 9 lead claimants currently has the evidence needed to establish a credible case of causation.

The 1980 Act

99. The following are the material provisions of the 1980 Act.

"11 Special time limit for actions in respect of personal injuries.

(1) This section applies to any action for damages for negligence, nuisance or breach of duty (whether the duty exists by virtue of a contract or of provision made by or under a statue or independently of any contract or any such provision) where the damages claimed by the plaintiff for the negligence, nuisance or breach of duty consist of or include damages in respect of personal injuries to the plaintiff or any other person.

(3) An action to which this section applies shall not be brought after the expiration of the period applicable in accordance with subsection (4) or (5) below.

(4) Except where subsection (5) below applies, the period acceptable is three years from –

(a) the date on which the cause of action accrued: or

(b) the date of knowledge (if later) of the person injured.

(5) If the person injured dies before the expiration of the period mentioned in subsection (4) above, the period applicable as respects the cause of action surviving for the benefit of his estate by virtue of section 1 of the Law Reform (Miscellaneous Provisions) Act 1934 shall be three years from –

(a) the date of death: or

(b) the date of the personal representative's knowledge; whichever is the later.

"14 Definition of date of knowledge for purposes of sections 11 and 12

(1) In sections 11 and 12 of this Act references to a person's date of knowledge are references to the date on which he first had knowledge of the following facts –

(a)that the injury in question was significant; and

(b) that the injury was attributable in whole or in part to the act or omission which is alleged to constitute negligence, nuisance or breach of duty; and

(c) the identity of the defendant; and

(d) if it is alleged that the act or omission was that of a person other than the defendant, the identity of that person and the additional facts supporting the bringing of an action against the defendant;

and knowledge that any acts or omissions did or did not, as a matter of law, involve negligence, nuisance or breach of duty is irrelevant.

…

(2) For the purposes of this section an injury is significant if the person whose date of knowledge is in question would reasonably have considered it sufficiently serious to justify his instituting proceedings for damages against a defendant who did not dispute liability and was able to satisfy a judgment.

(3) For the purposes of this section a person's knowledge includes knowledge which he might reasonably have been expected to acquire –

(a) from facts observable or ascertainable by him; or

(b) from facts ascertainable by him with the help of medical or other appropriate expert advice which it is reasonable for him to seek; but a person shall not be fixed under this subsection with knowledge of a fact ascertainable only with the help of expert advice so long as he has taken all reasonable steps to obtain (and, where appropriate, act on) that advice.

33 Discretionary exclusion of time limit for actions in respect of personal injuries or death.

(1) If it appears to the court that it would be equitable to allow an action to proceed having regard to the degree to which -

(a) the provisions of section 11 … or 12 of this Act prejudice the plaintiff or any person whom he represents; and

(b) any decision of the court under this subsection would prejudice the defendant or any person whom he represents;

the court may direct that those provisions shall not apply to the action, or shall not apply to any specified cause of action to which the action relates."

100. The draftsman of the Act seems to have proceeded on the basis that, by the time that the action was commenced, there would be no doubt that the act or omission alleged had caused the claimant's injury. That impression is further supported by section 33(3) (e). This includes in the matters relevant to the exercise of the discretion that the court enjoys under section 33:

"the extent to which the plaintiff acted promptly and reasonably once he knew whether or not the act or omission of the defendant, to which the injury was attributable, might be capable at that time of giving rise to an action for damages."

Thus the Act does not address what constitutes "knowledge" that an injury is attributable to an alleged act or omission where there is an issue between the parties as to whether the alleged act or omission occurred at all and, if it did, as to whether it caused the claimant's injury.

The approach of the courts below

101. One of the ten lead cases related to a veteran called Sinfield. Foskett J held that he was first diagnosed as having a significant injury less than two years before he commenced proceedings. That finding has not been challenged by the MoD, which now accepts that the claim in relation to Mr Sinfield is not time-barred. It is the other 9 cases that raise the question of the meaning of "knowledge".
102. Foskett J and the Court of Appeal held that the test of knowledge had been laid down by binding authority. The relevant case law demonstrated that the "knowledge" referred to in sections 11 and 14 could be equated with subjective belief. Each veteran had pleaded exposure to atomic radiation causing injury. Each veteran acquired the relevant knowledge at that moment in time when he formed the subjective belief that his injury was "attributable to" exposure to radiation. "Attributable to" did not mean "caused by" but "capable of having been caused by".
103. Thus the courts below held that each case turned on its own facts. The evidence bearing on each veteran's state of mind had to be examined in order to identify when he first had the belief that started time running.
104. The hearing before Foskett J lasted 10 days. The evidence called included expert evidence in relation to the development of scientific knowledge of the effects of ionising radiation. Individual veterans gave evidence of their knowledge and belief in relation to the injuries sustained and their cause. Having analysed the evidence Foskett J delivered a judgment that was 885 paragraphs in length. He held that 5 of the 10 claims had been commenced more than 3 years after the date when the relevant knowledge was acquired. In relation to those claims he exercised his discretion under section 33 in favour of the veteran, so that the claims were permitted to proceed. He held that the other 5 claims had been started within three years of acquiring the relevant knowledge, so that they were in time.
105. The hearing before the Court of Appeal spanned a week. The judgment of the Court, delivered by Smith LJ, was 305 paragraphs in length. The Court of Appeal did not differ in principle from the approach of Foskett J. The test to be applied was one of subjective belief. When looking at the individual cases, however, the Court repeatedly held that Foskett J had applied too high a threshold of knowledge or belief. The Court held that in the case of each of the 9 claimants knowledge had been acquired more than three years before proceedings were commenced. The Court of Appeal held that Foskett J had erred in principle in the exercise of his discretion under section 33. None of the 9 claims should be permitted to proceed. There was one common objection to permitting the claims to proceed. This was that none of them had a realistic prospect of success.
106. We were told that the veterans have been represented in the limitation proceedings under a conditional fee agreement ("CFA") that is restricted to those proceedings and that if they were successful they would seek to recover from the MoD costs in the sum of £17.5m, inclusive of success fee and ATE premium.

The first generic issue: the meaning of knowledge

107. Foskett J had toyed with an alternative test of "knowledge", which he had described as his "preferred view". This introduced into the test of knowledge an objective element. No veteran could acquire the knowledge that started limitation running until there was accessible to him scientific evidence that demonstrated the possibility that his injury was caused by exposure to ionising radiation. That evidence was provided for the first time by the Rowland report. All claims were brought within three years of the publication of that report, indeed most of them, including the claims of all of the lead cases, were brought before it was published. It followed that no claim was out of time. Foskett J concluded, however, that he was precluded by authority from applying his preferred view. The Court of Appeal held that he was right to reach that conclusion.
108. Mr Dingemans has put the preferred view at the forefront of his case before us. He has urged this Court to hold that no veteran acquired knowledge until the Rowland findings were published. By way of alternative submission he has sought to restore the findings of Foskett J, urging that they were correct and not findings with which the Court of Appeal should properly have interfered.
109. Mr Gibson QC for the MoD observed that, if the preferred view is correct, none of the nine claimants had the relevant knowledge when they commenced proceedings. He submitted that the Court of Appeal was correct both in its approach and in its conclusions. Subjective belief in attributability amounts to knowledge of attributability.

What is the test of knowledge?

110. I turn to consider the authorities that led Foskett J, reluctantly, and the Court of Appeal to conclude that knowledge could be equated with subjective belief.
111. Sections 11 and 14 are concerned with knowledge of what section 14(1) describes as "facts". The significant facts are (i) the injury sustained by the claimant (ii) the act or omission of the defendant alleged to constitute a breach of duty and (iii) the fact that the injury is "attributable" to that act or omission. In many claims for personal injury all three will be matters of which a claimant can sensibly say he has "knowledge". The cause of the injury will be known to the claimant from his own observation. There will be some cases, however, where cause and effect are not clear. Primary facts may be in issue. Or the causal nexus between those facts may only be capable of ascertainment by the application of specialist knowledge or expertise that the claimant does not enjoy. Even then, they may only be capable of evaluation on the basis of degree of probability. This is such a case. I turn to consider the cases that address the problem of "knowledge" where the material facts are not clear.
112. My starting point is the unreported case of Davis v Ministry of Defence (July 26 1985, CA; Transcript No 413 of 1985). The plaintiff contracted dermatitis when working for the defendant. He believed that it was caused by his conditions of work. His general practitioner was of the same view. Accordingly he started an action against the defendant in 1973, but he did not pursue it because he received advice, including expert medical advice, that the action had no reasonable prospect of success. He continued to believe, however, that the defendant was responsible. In 1982 after another attack of dermatitis he received fresh medical advice that, contrary to the previous advice, his condition was likely to be caused by his conditions of work. He began a fresh action. An application to strike this out on the ground that his claim was unarguably out of time succeeded at first instance, but was reversed by the Court of Appeal. May LJ said this at pp 7 and 9 of the transcript in relation to knowledge under section 14:

"'Knowledge' is an ordinary English word with a clear meaning to which one must give full effect: 'reasonable belief' or 'suspicion' is not enough. The relevant question merits repetition – 'When did the appellant first know that his dermatitis was capable of being attributed to his conditions at work?'

With all respect to the learned judge, I think that he wrongly assimilated what the appellant firmly believed throughout to what he knew. I have no doubt, as I have said, that the appellant has always believed that his dermatitis was due to his employers' fault and that he had a good claim against them. However, it is clear that he was advised that he did not and the combined state of mind of the appellant himself, as a layman and that of his doctors and legal advisers, which must be attributable to him by section 14(3) of the 1980 Act, cannot, in my opinion so surely be said to have been such that they knew, prior to 10 November 1978 that the dermatitis was capable of being attributed to the appellant's working conditions."

113. This is the first case where the plaintiff knew of both the facts that he alleged had caused his illness and of the illness itself, but where the uncertainty related to causation. It is of particular interest that May LJ applied section 14(3) so as to give the plaintiff constructive knowledge that his illness was not attributable to his conditions of work. Also significant is his interpretation of "attributable" as meaning "capable of being attributed". Where the uncertainty is as to causation of an illness or disease there may be a number of possible causes. In this situation the meaning of "knowledge" raises particular problems. The claimant is unlikely to be in a position to form a considered view of the cause of his illness from his own knowledge. He will need advice on this. The relevant knowledge is thus likely to be constructive, under section 14(3)(b). Applying the approach of May LJ, the claimant will not have knowledge that his illness is attributable to a particular cause unless there is a body of respectable medical opinion that recognises that this is possible.
114. In Halford v Brookes [1991] 1 WLR 428 the plaintiff had issued a writ as administratrix making a civil law claim for damages against two defendants for murdering her daughter nine years before. One defendant had been prosecuted for murder and acquitted, essentially because, although it was obvious that one or other of the defendants had murdered the young girl, it was not clear which had done so. There were strong grounds for suspecting that both defendants had been complicit in the murder. The plaintiff delayed commencing proceedings because she was unaware that it was open to her to make a civil claim. The defendants' defence denied the allegation of murder but, at the same time, contended that the claim was time barred because the plaintiff had had knowledge that the murder was "attributable" to them for well over three years. At first instance Schiemann J remarked upon the paradox of this stance, but held that the claim was time barred.
115. In the Court of Appeal Lord Donaldson of Lymington MR held that May LJ's definition of knowledge in Davis could only be applied to the special facts of that case. He held at p 443:

"'Knowledge' clearly does not mean 'know for certain and beyond possibility of contradiction.' It does, however, mean 'know with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking legal and other advice and collecting evidence.' Suspicion, particularly if it is vague and unsupported, will indeed not be enough, but reasonable belief will normally suffice."

Thus the Master of the Rolls applied an objective test to the quality of the "belief". It had to be sufficiently firm to justify taking the preliminary steps towards the issue of proceedings. Furthermore, although Lord Donaldson referred to "belief", the belief in question was based upon knowledge of facts that gave rise to the inference that the two defendants had been guilty of murder. This was not a case, such as Davis, where the uncertainty was whether medical evidence supported a link between working conditions and the illness contracted. It was uncertainty as to some of the primary facts. Where some of the primary facts are in the exclusive knowledge of the defendant, reasonable belief in the existence of those facts will necessarily be founded on other secondary facts. Thus, on analysis, the test applied in Halford was whether the facts known to the plaintiff should have led a reasonable person with knowledge of the law to take steps to commence legal proceedings.

116. Davis received detailed consideration by the Court of Appeal in Nash v Eli Lilly & Co [1993] 1 WLR 782. This case is of particular importance because, like the present case, it involved a group action in which lead claimants had been selected. The group action was for personal injuries, in the form of unpleasant side effects ("injuries"), in particular photosensitivity, alleged to have been caused by taking the drug known as Opren. This drug was withdrawn from the market in 1982 and by the time of the limitation proceedings it was common ground that Opren was capable of causing the injuries in respect of which the claims were brought. The actions were commenced in 1987 and 1988. Just as in the present case limitation was tried as a preliminary issue. One issue related to the date at which each plaintiff acquired knowledge that he had sustained a "significant injury". The more pertinent issue was the date on which he acquired knowledge that this injury was attributable to Opren. The trial judge held that if a plaintiff's medical practitioner would have advised him that his symptoms could be attributable to Opren, he had constructive knowledge of that fact. The position of each of the lead plaintiffs was explored and it was held that most of the claims were time barred.
117. The hearing in the Court of Appeal lasted 12 days and the judgment of the Court was delivered by Purchas LJ. A particular issue arose in relation to the position of at least one of the claimants who had formed the firm belief that his injuries were attributable to Opren but who did not bring proceedings because he was advised by a specialist that this was not so. The Court gave detailed consideration to Davis in a lengthy portion of its judgment dealing with "the distinction between belief and knowledge". This included the following critical passage:

"It is to be noted that a firm belief held by the plaintiff that his injury was attributable to the act or omission of the defendant, but in respect of which he thought it necessary to obtain reassurance or confirmation from experts, medical or legal, or others, would not be regarded as knowledge until the result of his inquiries was known to him or, if he delayed in obtaining that confirmation, until the time at which it was reasonable for him to have got it. If negative expert advice is obtained, that fact must be considered in combination with all other relevant facts in deciding when, if ever, the plaintiff had knowledge. If no inquiries were made, then, if it were reasonable for such inquiries to have been made, and if the failure to make them is not explained, constructive knowledge within the terms of section 14(3) must be considered. If the plaintiff held a firm belief which was of sufficient certainty to justify the taking of the preliminary steps for proceedings by obtaining advice about making a claim for compensation, then such belief is knowledge and the limitation period would begin to run. "

The last sentence suggests that a firm belief that an injury is attributable to an alleged act or omission can start the limitation period running however unreasoned or ill-informed that belief may be. This was the start of a series of decisions that equated knowledge with subjective belief.

118. Broadley v Guy Clapham & Co [1994] 4 All ER 439 involved a claim against a solicitor for professional negligence. There was an underlying issue as to whether a claim that the plaintiff had enjoyed against a surgeon had become time-barred. The surgeon had operated on the plaintiff's knee and the operation had left her with foot drop. The trial judge held that she did not herself have knowledge that the surgeon had caused this outcome but should have sought medical advice that would have disclosed this fact. Thus she acquired constructive knowledge under section 14(3)(b). The Court of Appeal upheld both his decision and his reasoning. Hoffmann LJ said, at p 448:

"Section 14(1)(b) requires that one should look at the way the plaintiff puts his case, distil what he is complaining about and ask whether he had, in broad terms, knowledge of the facts on which that complaint is based. "

He went on to make a statement that, if not read with care, is capable of misleading. He said, at p 449

"The purpose of section 14(1), as Lord Donaldson MR pointed out in Halford v Brookes [1991] 3 All ER 559 at 573, [1991] 1 WLR 428 at 443, is to determine the moment at which the plaintiff knows enough to make it reasonable for him to begin to investigate whether or not he has a case against the defendant. He then has three years in which to conduct his inquiries and, if advised that he has a cause of action, prepare and issue his writ. Ordinarily it will suffice that he knows that the injury was caused by an act or omission of the defendant. But there may be cases in which his knowledge of what the defendant did or did not do is so vague and general that he cannot fairly be expected to know what he should investigate. He will also not have reached the starting point if, in an unusual case like Driscoll-Varley v Parkside Health Authority, he thinks he knows the acts and omissions he should investigate but in fact he is barking up the wrong tree."

The statement that time begins to run when the plaintiff knows enough to make it reasonable for him to begin to investigate whether or not he has a case against the defendant relates to an investigation whether, having regard to the knowledge of attributability that has been acquired, a case against the defendant exists. The passage should not be read as holding that time begins to run as soon as the claimant knows enough to make it reasonable to make a further investigation of the facts that are relevant to attributability.

119. I can now proceed to Spargo v North Essex District Health Authority [1997] PIQR P235 in which Brooke LJ set out the effect of a number of decisions, including Nash v Eli Lilly:

"(1) The knowledge required to satisfy section 14(1)(b) is a broad knowledge of the essence of the causally relevant act or omission to which the injury is attributable;

(2) 'Attributable' in this context means 'capable of being attributed to', in the sense of being a real possibility;

(3) A plaintiff has the requisite knowledge when she knows enough to make it reasonable for her to begin to investigate whether or not she has a case against the defendant. Another way of putting this is to say that she will have such knowledge if she so firmly believes that her condition is capable of being attributed to an act or omission which she can identify (in broad terms) that she goes to a solicitor to seek advice about making a claim for compensation;

(4) On the other hand she will not have the requisite knowledge if she thinks she knows the acts or omissions she should investigate but in fact is barking up the wrong tree: or if her knowledge of what the defendant did or did not do is so vague or general that she cannot fairly be expected to know what she should investigate; or if her state of mind is such that she thinks her condition is capable of being attributed to the act or omission alleged to constitute negligence, but she is not sure about this, and would need to check with an expert before she could be properly said to know that it was."

This summary has been treated as definitive – see McGee on Limitation Periods, 6^th ed (2010) at 8.026. The summary is a valiant attempt to summarise the previous jurisprudence, but is capable of confusing. It does not deal with constructive knowledge. Significantly it states that a firm belief in attributability can amount to knowledge.

120. Courts have had particular difficulty in interpreting Nash v Eli Lilly and Spargo in circumstance where it is reasonable to expect a potential plaintiff to seek expert medical advice on causation, whether or not he or she holds a firm belief that an injury has been caused by medical treatment or by conditions at work – see O'Driscoll v Dudley Health Authority [1998] Lloyd's Rep Med 210, Ali v Courtaulds Textiles Ltd (2000) 52 BMLR 129 and Sniezek v Bundy (Letchworth) Ltd [2000] PIQR P213. The facts of this last case might have been devised as an examination question on limitation. The claimant experienced a sensation of burning on his lips and throat and formed the firm view that this was the result of exposure to polymer at his workplace, a view that he never abandoned. He stopped work because of this in 1988 and sought legal and medical advice. He was seen by a number of experts, all of whom could find nothing wrong with him. One indeed concluded that his symptoms were psychosomatic. Ultimately, in 1994 a senior ENT Registrar advised that his symptoms might well be attributable to exposure to polymer. The trial judge took that as the date on which he acquired knowledge for the purpose of section 14. The Court of Appeal did not agree, holding that the claimant's firm belief in the face of expert advice to the contrary constituted knowledge for the purpose of section 14.
121. In Haward v Fawcetts [2006] UKHL 9; [2006] 1 WLR 682 the House of Lords considered knowledge for the purposes of sections 11 and 14 in the context of a claim for professional negligence in giving investment advice. Lord Nicholls of Birkenhead at para 9 approved the approach of Lord Donaldson MR in Halford v Brookes. He said

"Thus, as to the degree of certainty required, Lord Donaldson of Lymington MR gave valuable guidance in Halford v Brookes [1991] 1 WLR 428, 443. He noted that knowledge does not mean knowing for certain and beyond the possibility of contradiction. It means knowing with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking advice and collecting evidence: 'Suspicion, particularly if it is vague and unsupported will indeed not be enough but reasonable belief will normally suffice'. In other words, the claimant must know enough for it to be reasonable to begin to investigate further."

At para 11 he paraphrased the position by stating that time does not begin to run until the claimant knows that there is a real possibility that his damage was caused by the act or omission in question.

The decision of Foskett J in relation to knowledge

122. I propose to refer to only a few particularly relevant incidents of the hearing before Foskett J. Suspicion that veterans may have been injured by exposure to ionising radiation in the tests has a long history. This was explored in voluminous evidence. This focussed on events in the 1980s and thereafter. Of particular significance was the formation in May 1983 of the British Nuclear Test Veterans' Association ("BNTVA"). The objects of the BNTVA included

"the relief of persons suffering from disability attributed to the effects of exposure to radioactivity particularly dealing with nuclear weapons tests…To conduct or promote research into the causes, effects, and treatment of such disablement and to claim financial assistance, benefits and compensation as they may be entitled to." (My emphasis)

268 of the claimants have been members of the BNTVA and all five of the lead claimants selected by the MoD have been members.

123. Substantial evidence was also adduced before Foskett J of the progress of scientific research and opinion in relation to the effect of exposure to ionising radiation, with particular reference to the position of the veterans.
124. In 1985 a veteran called Melvyn Pearce commenced an action against the MoD in which he alleged that a lymphoma that he had developed in 1978 was caused by exposure to ionising radiation during nuclear testing at Christmas Island. After a successful excursion as far as the House of Lords on a preliminary point on Crown immunity ([1988] AC 755) Mr Pearce discontinued his action because those acting for him concluded that it would be impossible to prove that radiation had caused his cancer.
125. The Government commissioned the National Radiological Protection Board ("NRPB") to survey the possible effects of radiation on the servicemen who had participated in the tests. The NRPB published reports in 1988, 1993 and 2003. These showed no general increase in mortality on the part of veterans either generally or from cancers. The reports have been subject to expert criticism the methodology of which has itself been criticised by the MoD.
126. A number of other epidemiological surveys were carried out in relation to veterans who had taken part in the atomic testing. None of these led to the conclusion that veterans were suffering a disproportionate incidence of injuries.
127. I have referred at para 97 to the Rowland report which was published in 2008, but shown privately to the appellants in 2007. This dealt with assays on 50 New Zealand veterans who served on two ships that took part in some of the tests, which were compared with results of similar assays on 50 controls. Two of the assays showed no difference, but the third, which was an assay known as mFISH, indicated that aberrant changes to chromosomes had occurred in veterans with three times the frequency of similar changes in the controls. The report concluded that the likely cause of this was ionising radiation thus indicating that the veterans had incurred long term genetic damage as a result of their participation in the tests. The report emphasised, however at p 6, "that the current study makes no claims on the health status of the veterans".
128. Before Foskett J the MoD argued that there was no evidence that the veterans had been exposed to abnormal ionising radiation or that such radiation had caused their injuries, but that the belief that their injuries were attributable to radiation none the less amounted to knowledge of attributability for the purposes of sections 11 and 14. Counsel for the veterans' description of this argument was set out by Foskett J as follows at para 519

"Mr Browne has submitted that there is a logical tension in the defendant's case on limitation in that it is contended that each individual claimant should at the first sign of significant health effects have realised the link to the negligent acts or omissions concerning the tests such that the limitation clock started ticking, yet in the same breath it is contended that no reputable scientist then (or indeed now) could support any such link thus denying any chance of establishing liability. He has submitted that the defendant's argument seems to proceed on the basis that whatever the state of scientific knowledge might have been at any time as to whether any participant or veteran had been exposed to substantial ionising radiation and as to whether such exposure had the potential to cause the significant injury suffered, none the less if the lay participant or veteran suspected that he had been exposed to such radiation and suspected that there might be a link between such exposure and the relevant injury, then he should be fixed with actual knowledge to that effect."

129. Foskett J at para 520 said that he sympathised with the submission that there was a "logical tension" in this argument. That was why he would have liked to decide the limitation issue on the basis of his "preferred view", which he expounded at paras 514 to 518 as follows. Knowledge of attributability under section 14 meant that each veteran had to have knowledge of two matters. The first was that his injury was capable of being caused by abnormal radiation. The second was that he had been exposed to such radiation. This knowledge could only be obtained from scientific material. Insofar as the veterans had believed that they had been exposed to prompt radiation, there had been no foundation for that belief and it was unsound. They had been "barking up the wrong tree". Not until the Rowland report was there material upon which knowledge could be based both that the veterans could have been exposed to fall-out contamination well after the explosions and that this exposure was capable of causing chromosomal aberrations that evidenced the kind of mechanism that could have led to at least some of the injuries of which they complained.
130. In short, Foskett J favoured a test of knowledge of attributability that required belief to be reasonably founded on fact. He concluded, reluctantly, that the decided cases precluded the adoption of his preferred view. Instead his approach was to look in each case for the moment at which the veteran had manifested not merely suspicion but a firm belief that his illness was attributable to exposure to radiation. This was the moment at which the relevant knowledge of attributability was acquired.

The decision of the Court of Appeal in relation to knowledge

131. The Court of Appeal held that Foskett J had been right to reject the preferred view. After reference to authority, and citation of the passage from Haward v Fawcetts that I have set out at para 121 above, the Court summarised the position as follows at para 92:

"So, in a case where the claimant's state of mind is more accurately described as one of belief rather than knowledge, it seems to us that what matters is whether his state of belief is such as to make it reasonable to expect him to begin to investigate further. In general that assessment will have to be made by reference to the things that he has said and done. For example, if he says that, at such and such a time, he had a firm belief that his illness had been caused by radiation, it would obviously be reasonable to expect him to begin investigating. If he said that he had a firm belief that his illness could have been caused by radiation, that would also, we think, be enough."

132. The Court went on at para 93 to indicate a critical disagreement with the test of Lord Donaldson MR that Lord Nicholls had approved:

"We note that, in Halford, Lord Donaldson MR suggested that a belief would have to be reasonable before it could amount to knowledge. With great respect, we do not think that the belief needs to be objectively reasonable. We think that what matters is the claimant's subjective state of mind. If a claimant comes to believe that there is a causal connection between his condition and the matters complained of, it will matter not from where he has derived that belief, even it were from an incompetent expert adviser or from a newspaper article which was not based on sound research. If the belief were of such strength that it was reasonable to expect him to start investigating his claim, it would amount to knowledge within section 14."

133. The Court applied a test of subjective belief when considering the individual cases. By way of example I quote the following passage from para 222 in respect of Mrs Clark's state of mind in relation to the cause of her late husband's cancer:

"In so far as Mrs Clark's state of mind would more aptly be described in terms of belief rather than knowledge, she needed only to have a strong enough belief to make it reasonable to expect her to start making inquiries."

The application of this test led the Court to conclude that all 9 claimants had acquired knowledge of attributability more than three years before the commencement of proceedings.

Conclusions in respect of the first generic issue

134. Section 14 is about knowledge, actual and constructive, of facts. The object of the section is apparent from its terms. Time will begin to run when the claimant has, or ought reasonably to have, knowledge of the facts that make up the essential elements of his claim. These are:

i) the fact that he has sustained a significant injury;

ii) the identity of the defendant;

iii) the act or omission alleged to constitute negligence, nuisance or breach of duty;

iv) the fact that the injury is attributable to that act or omission;

v) if the act or omission was of a person other than the defendant, the identity of that person and the additional facts supporting the bringing of the action against the defendant.

It is unnecessary for the claimant to know that the alleged act or omission constituted negligence, nuisance or breach of duty. Knowledge is limited to fact, not law.

135. Section 14(3) defines constructive knowledge. It lays down a test of knowledge that the claimant ought reasonably to have been expected to acquire either from facts "observable or ascertainable by him" or facts ascertainable by him with the help of medical or other appropriate expert advice which it is reasonable for him to seek. Thus the claimant is expected, where it is reasonable to do so, to make inquiries in order to ascertain the relevant facts and, if he does not do so, will be deemed to have knowledge of the facts that those inquiries would have disclosed. The words that I have emphasised underline the fact that objective standards have to be applied. There has been some divergence of view as to whether, when applying the test of what is reasonable, allowance has to be made for particular characteristics of the claimant – contrast the views of Lord Hoffmann, Lord Walker and Lady Hale in Adams v Bracknell Forest Borough Council [2004] UKHL 29; [2005] 1 AC 76 at paras 44-46, 75-78 and 91 respectively. Lord Hoffmann disagreed with statements in earlier cases that it was necessary to have regard to the character and intelligence of the plaintiff when considering whether he had acted reasonably. He applied a test that was strictly objective. His approach was followed by the Court of Appeal in B v Nugent Care Society (Practice Note) [2009] EWCA Civ 827; [2010] 1 WLR 516, rightly in my view. If a claimant is suffering from any disadvantage in comparison to the reasonable man regard can be had to this when exercising discretion under section 33.
136. The wording and the scheme of section 14 does not permit the replacement of the test of actual or constructive knowledge of the specified facts with a test of subjective belief. There is good reason for this. It is not desirable that a plaintiff should commence an action on the basis of subjective belief that is not reasonably founded on a basis of fact. Nor would it be just to discriminate between two claimants who had identical knowledge of the material facts on the ground that one believed that they demonstrated attributability while the other formed the view that they did not.
137. Section 14 envisages different stages in the acquisition of knowledge and different degrees of knowledge. At each stage the claimant's state of mind has to be assessed according to an objective standard. In some circumstances the claimant will have the knowledge of attributability as a result of his own observation of the circumstances in which he has sustained his injury. The facts that he knows will leave him in no doubt as to what act or omission has caused his injury and who was responsible for it. In that situation he will normally have the knowledge that would lead a reasonable person to consult a solicitor with a view to making a claim for compensation. In other circumstances he will not know, from his own observation, all the relevant facts in relation to the cause of his injury. His limited knowledge may be such as would lead a reasonable person to make further investigations as to the facts. In that situation he will have imputed to him the knowledge that he would have acquired if he had made those investigations. The question will then be whether that knowledge would have led a reasonable person to consult a solicitor with a view to making a claim for compensation. At no stage, as a matter of law, is it relevant to consider the subjective belief of the claimant divorced from the facts that have led to that belief. When considering the claimant's state of mind the relevant question will not be what he believed, but what he reasonably believed. A reasonable belief will be based to some degree on known facts.
138. In practice the distinction between knowledge and belief is not one that will normally arise. The starting point will be a claimant who has commenced proceedings more than three years from the date when the cause of action arose. The claimant will normally have pleaded a viable case; the act or omission alleged will be one that was capable of having caused the injury in respect of which he claims. It will be common ground that he had the relevant knowledge of attributability at the time that he commenced the proceedings. The issue will be when he first acquired that knowledge.
139. The distinction between knowledge and belief is critical in the present case because it is common ground that when the nine lead claimants started these proceedings there were no known facts capable of supporting a belief that the veterans' injuries were attributable to exposure to ionising radiation. Insofar as veterans believed that their injuries were attributable to such exposure that belief was not reasonable. No individual claimant was in a position to know that his injury was attributable to exposure from his own observation or from facts that he was capable of ascertaining by himself. The reasonable course for any veteran who suspected that his injury might be attributable to exposure to radiation was to seek expert advice. Some did so. Those who did not were all in a similar position. They had constructive knowledge of the scientific data available to those from whom they should reasonably have sought advice.
140. Experts had carried out epidemiological surveys to see if these suggested that the veterans were suffering a disproportionate incidence of injuries compared to the rest of the population and concluded that they were not. The MoD denied that the veterans had been exposed to ionising radiation and there was no known reason to gainsay this. In short, there was no scientific evidence available that provided significant support to the belief that the veterans' injuries were attributable to exposure to ionising radiation.
141. Foskett J and the Court of Appeal fell into error in equating subjective belief with knowledge. In so far as there were statements in earlier decisions of the Court of Appeal to which I have referred in paras 112 to 121 above which lent support to that approach, those statements were unsound. I question both the reasoning and the conclusions about knowledge of the Court of Appeal in Sniezek v Bundy, Foskett J's preferred view was correct in principle. Belief in attributability had to be founded on known fact if it was to amount to "knowledge". It had to be "reasonable belief". The Court of Appeal was wrong to alter this test, which had been advanced originally by Lord Donaldson MR, so as to remove the requirement that the belief should be reasonable. The search for the moment when each of the lead veterans formed a subjective belief that it was possible that his injuries were attributable to exposure to radiation was misconceived.
142. The scientists were the people to whom it was reasonable for the veterans to look for advice but, at least before the publication of the Rowland report, such scientific data as was or became available did not support the theory that there was a serious possibility that the veterans' injuries were attributable to exposure to radiation. In their Amended Points of Claim on Limitation those veterans who commenced proceedings before the publication of the Rowland report allege that their knowledge of the existence of scientific evidence that demonstrated that their injuries were attributable to the acts or omissions of the defendants did not arise until after the commencement of proceedings. It is open to question whether the Rowland report demonstrates that the veterans' injuries are attributable to exposure to radiation, as to which see paras 155 to 157 below. What is not open to question is that, prior to the publication of that report, there was no evidence that demonstrated that the veterans' injuries were attributable either to prompt radiation or to fall-out radiation. In effect the proceedings were commenced on a speculative basis. The veterans further plead that the consequence of this is that their claims are not time barred. Whether that submission is correct is the second generic issue.

The second generic issue: the effect of starting proceedings for personal injury without reasonable grounds for belief that the injury was caused by breach of duty

143. What is the position where a claimant starts an action for personal injury against a defendant in circumstances where he has no reasonable grounds for believing that his injury is attributable to the act or omission that he alleges against the defendant? Such a situation was considered by the Court of Appeal in Whitfield v North Durham Health Authority [1995] 6 Med LR 32. The question arose in limitation proceedings of the effect of a writ that had been issued but not served. Waite LJ held that the issue of a writ was not determinative of knowledge under section 14. He observed that it might be the product of

"a generalised – though as yet unspecifically informed – sense of grievance (memorably rendered by Stanley Holloway as "Somebody's got to be summonsed")"

144. Where a personal injury action is commenced more than three years after the cause of action arose and the defendant raises a challenge on the ground that it is time-barred, the onus is on the claimant to prove that the action was started less than three years from the date on which he acquired knowledge, as defined by section 14 – see the comprehensive analysis of burden of proof in the context of limitation of Mance J in Crocker v British Coal Corporation (1995) 29 BMLR 159 at pp 169 -173. If the claimant's response to the limitation challenge is to allege that he started proceedings without knowledge that his injury was attributable to the act or omission that he alleges caused it the defendant is likely to contend that the action should not be permitted to proceed. There are three arguments that can be advanced for bringing such an action to an end, all three of which have been advanced in the present case:

i) The claimant is time-barred because the requirements of sections 11 and 14 are not satisfied.

ii) The claim should be struck out on the ground that it discloses no reasonable cause of action and is an abuse of process.

iii) Summary judgment should be given in favour of the defendant.

I shall consider each in turn.

Limitation

145. In the present case the MoD sought to show that the veterans' claims were time barred by showing that they had subjectively come to believe that their injuries were attributable to exposure to radiation more than three years before they commenced proceedings, notwithstanding that, according to the MoD, they had no reasonable grounds for that belief. For the reasons that I have given that approach was misconceived.
146. At one time I was attracted to the argument that a claimant who cannot point to any moment in time before commencement of proceedings when he acquired knowledge of attributability is not in a position to discharge the burden of proving that he commenced the proceedings within three years of acquiring knowledge as required by section 11. The MoD did not advance that argument and, on reflection, I believe that it would have been fallacious. Section 11 provides that an action may not be brought after "the expiration of the period" of three years that commences with "knowledge". If the claimant has not acquired "knowledge" before bringing the action, that period has not begun to run.
147. For these reasons I have concluded that on the unusual facts of this case the MoD was not in a position to raise a limitation defence to the veterans' claims. The Court of Appeal's finding that all the claims were out of time must accordingly be reversed.
148. The MoD had other options. It could simply have left the veterans to attempt to prove the exposure and causation that the MoD denied. In the case of the claim brought by Pearce they took no limitation defence and the claim was ultimately dropped because of the problem of proving causation. The Ministry might have taken the view that the same was likely to occur in the present case. It did not, however, take that course. Nor did it formally apply to strike out the proceedings or apply to the court to grant it summary judgment. None the less in the course of the limitation proceedings it sought to persuade the court to do one or the other of its own motion. I turn to consider whether the court should have done so.

Strike out

149. In this case the MoD invited Foskett J to strike out the lead claims pursuant to CPR r 3.4 on the ground that each claim was bound to fail as the claimant was not in a position to establish causation. He declined to do so on the grounds that it would not be right to prejudge the issue of causation. He was not persuaded that the veterans were bound to fail on that issue. The Court of Appeal held at paras 70-72 that the judge had reached the right answer, but that he should have based his decision on procedural grounds. The Court held at para 71 that CPR r 3.4(2) only permits the court to strike out proceedings where the terms of the pleading itself justify this course. That may be true, but CPR rr 3.1(1) and 3.4(5) preserve the inherent jurisdiction of the court to strike out proceedings on the ground of abuse of process and the Court of Appeal was wrong not to consider the overall merits of the veterans' position.
150. I am not persuaded that the conduct of the veterans in commencing proceedings before they had reasonable grounds to believe that their injuries were caused by exposure to radiation constituted an abuse of process. A case such as this poses special problems for the litigant. It is not uncommon for a number of people who suffer injury or disease to form the suspicion or even belief that this is attributable to the exposure of their bodies to some noxious substance or process. Initially there may be no significant scientific support for such suspicion or belief. The suspicion or belief that the MMR injection causes autism, or that the use of mobile phones causes brain tumours are, perhaps, examples of such suspicions or belief. The parents of victims of thalidomide, or those who suffered from taking Opren, may initially have had no sound basis for suspecting the cause of the conditions caused by those products. In such circumstances no individual victim can reasonably be expected to commence proceedings on suspicion. Nor can any individual reasonably be expected single-handed to obtain the necessary expert assistance to investigate whether his suspicions or belief are well founded. Group action is the sensible way forward. Once a group is formed the practical course for anyone who suspects or believes that he may be in the same position as the other members of the group is likely to be to join the group. In that case the knowledge of the group and those advising it will become the constructive knowledge of the individual.
151. Whether and when it will be reasonable for a group to commence legal proceedings will depend upon the particular circumstances. Normally investigation of the facts will precede the commencement of proceedings. The pre-action protocol for disease and illness claims may well be relevant. It is possible to conceive of circumstances where it may be reasonable to commence group proceedings even though investigations are ongoing and there is uncertainty as to attributability of a disease to a suspected cause. This might be a reasonable precautionary step in order to forestall the possibility of a limitation challenge such as the present or it might assist in obtaining funding.
152. In the present case problems of funding played a role in the initiation and pursuit of the litigation. Some funding appears to have been obtained from the Legal Services Commission to carry out investigations and proceedings were instituted with this assistance at the end of 2004. The Treasury Solicitor then agreed to an extension of the time for service of Particulars of Claim because of difficulties arising out of the funding certificate. The Legal Services Commission withdrew funding on 17 August 2005. This led to a hiatus in the proceedings. Ultimately a Conditional Fee Agreement backed by after-the-event insurance was obtained. Although this was limited to the limitation issue alone this may have had some influence upon the fact that master particulars of claim were settled and were served on 29 December 2006.
153. Having regard to this history I do not consider that the initiation of this group action, albeit that it was launched on a speculative basis, constituted an abuse of process and it would not have been right to strike it out on that basis. If the MoD wished to bring the proceedings to an end the appropriate course was to seek summary judgment pursuant to CPR r 24.2 on the ground that the claimants had no real prospects of succeeding on their claims. Although the MoD did not make a formal application to this end it did give the veterans notice that it would be contending at the limitation hearing that the claimants had no real prospects of success and invited Foskett J to strike the proceedings out on this basis.

Summary judgment

154. The Court of Appeal stated at paras 5 and 70 that the judge declined to give the MoD summary judgment pursuant to CPR r 24(2) although I do not believe that the judge, having declined to strike out the proceedings, gave separate consideration to summary judgment. At all events the Court of Appeal itself considered whether it should grant the defendant summary judgment on the ground that the claims had no real prospect of success. The Court ruled at para 75 that it would not be right to do, basing its ruling on purely procedural grounds:

"We are of the view that we should refuse summary judgment on purely procedural grounds. We recognise that the claimants had been informally put on notice that causation would be raised in an application for summary judgment. Further, we acknowledge that, even without such informal notice, it behoved the claimants to prepare themselves to show the general merits of their claims in case the judge had to consider whether to exercise his discretion under section 33. But notwithstanding those two factors, we consider that we should not grant summary judgment in the absence of a formal application. The claimants should have been left in no doubt that they faced summary judgment if they could not show an arguable case on causation. It was simply not appropriate in a case of this importance and complexity to place on the judge the decision as to whether or not to exercise the jurisdiction under Part 24 of his own motion. Thus, because of the lack of formal notice, we consider that it would not be fair to give summary judgment against the claimants under this rule."

155. The Court of Appeal considered the problems that the veterans would experience in establishing causation when it came to consider of the exercise of discretion under section 33. They concluded, at para 156, that the veterans' cases on causation "faced very great difficulties" which were "much more serious than they appeared to Foskett J." I would endorse that conclusion. The current difficulties facing the veterans in relation to causation appear to me to be very great indeed. The Rowland report assists them a little but it does not have the significance that Mr Dingemans has sought to attach to it.
156. The Rowland report shows that many of the New Zealand veterans had a raised incidence of chromosome translocation that suggested exposure to abnormal, albeit low level, fall-out radiation. But this was not true of all of the veterans assayed. The assays of some showed no abnormalities. This is no more than one would expect. Exposure to fall-out radiation results from inhalation or ingestion of fall-out. It may result from swallowing sea water while swimming or eating contaminated fish. Thus it can vary from one man to the next. The most that can be deduced from the Rowland report is that it is probable that individual veterans were exposed to low level fall-out. There is currently no evidence that there is any correlation between the raised incidence of chromosome translocation of individual New Zealand veterans and the incidence of cancer or any of the other conditions of which the claimant veterans complain. Nor is not suggested that the aberrant chromosomes identified by the mFISH assay could themselves have had a mechanistic link in the contraction of cancer, although there is an established mechanistic link between some chromosome aberrations and cancer. The Rowland report results simply constitute a biomarker suggesting exposure to radiation.
157. The most that the veterans as a group are currently in a position to establish is that there is a possibility that some of them were exposed to a raised, albeit low level, of fall-out radiation and that this may have increased the risk of contracting some at least of the injuries in respect of which they claim. This falls well short of establishing causation according to established principles of English law. Foskett J was prepared to contemplate the possibility that the Supreme Court would extend the principle in Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22; [2003] 1 AC 32 so as to equate causing an increase of risk with causing injury. The Court of Appeal at para 154 held that there was no foreseeable possibility of this. In the light of the observations of this Court in Sienkiewicz v Greif [2011] UKSC 10; [2011] 2 AC 229 the Court of Appeal was plainly correct.
158. For these reasons I do not believe that the veterans' claims have a reasonable prospect of success. Despite this conclusion I have decided that the Court of Appeal was right not to grant the MoD summary judgment. There are two interrelated reasons for this. The first is that I agree with the Court of Appeal that it would be unjust to enter summary judgment against the 9 lead claimants when the MoD made no formal application for summary judgment. While the veterans' allegation that they had been ambushed by the MoD was unjustified, they could properly complain that they had not been given adequate warning that they would have to resist summary judgment.
159. The other reason is that these are lead cases in group litigation. The object of selecting lead claimants for the purpose of trying preliminary issues is that the decisions of the court in the lead cases will be determinative, or treated as determinative, in the other cases that raise the same issues. It does not seem to me fair to those other claimants to expect them to accept a ruling that the claims have no reasonable prospects of success when that issue was not fairly and squarely on the table when the arrangements were made for the hearing of the lead cases.

Section 33

160. As I have concluded that none of the claims is out of time, the question of exercising discretion under section 33 does not arise. Had it arisen I might not have reacted to it in the same way as the Court of Appeal. The 1011 claimants must include a significant number in respect of whom there is no limitation defence. These will include those whose injuries were diagnosed, or appreciated to be significant, less than three years before proceedings were commenced. One of the principal reasons for limitation of actions is to protect defendants from being vexed by stale claims. Where group claims are proceeding in any event, this is not such a significant consideration. It may be unjust to preclude some claimants from participating in the litigation on the ground that they did not bring their claims soon enough. This is a factor that the Court of Appeal does not appear to have taken into account. It might have led me to differ from the Court of Appeal's conclusion. As it is, on my view of the case no issue arises under section 33.
161. For the reasons that I have given I would allow all 9 appeals and reinstate the actions of the lead claimants.

LADY HALE

162. Limitation of actions is a creature of statute, not of the common law. Until the Limitation Act 1623, there were no limitation periods for non-land-related claims. When introduced, they were and remain a procedural, not a substantive, bar to the claimant's action. If the defendant does not plead limitation, the cause of action subsists and the court must try the claim. In an age of private claims against private defendants, it may even have been regarded as ungentlemanly to raise a limitation defence, but we can hardly blame the Government or a liability insurer for seeking to protect the interests of the taxpayer or its policy holders and shareholders by doing so in a case such as this.
163. The current law of limitation is complicated and incoherent. This is, as the Law Commission pointed out in 1998, largely because it "has been subjected to a wide range of ad hoc reforms, following the recommendations of reform bodies charged with recommending reforms of particular pockets of law". The Commission went on to comment that "the traditional approach of limitation periods running from accrual of a cause of action has led to problems, which the Legislature has tried to solve, either by moving to a discoverability starting date (as in the Latent Damage Act 1986) or by relying on a judicial discretion to disapply the limitation period (as in the Defamation Act 1996), or by using both approaches (as in the regime for personal injuries)" (Consultation Paper No 151, Limitation of Actions, para 1.21).
164. In other words, this is a field in which statute has intervened for policy reasons. But in policy terms the current regime for personal injury claims, combining discoverability with discretion, might be thought to have the worst of all possible worlds. From the defendant's point of view, one aim of limitation periods is to ensure that a fair trial will still be possible because the evidence will not have been lost or deteriorated. Another aim is that the defendant will not be harassed with stale claims and he (or his insurer) can treat matters as closed after a certain length of time. From the state's point of view, there is also an interest both in fair trials and in an end to litigation. From the claimant's point of view, there may be some interest in being encouraged to get on with it while the evidence is still fresh, but in general the claimant will want as long as possible in which to "recognise and consider their cause of action, to take legal advice on their case, and to attempt to negotiate a settlement with defendants" (para 1.36). In policy terms, the crucial question is whether a fair trial is still possible in the individual case, coupled with the ability to write off claims after a period of time.
165. Where a cause of action depends upon damage resulting from the defendant's tort, a limitation period based upon the accrual of the cause of action may have nothing to do with whether a fair trial will still be possible or with the interests of the defendant in not being harassed by stale claims. Mr Sinfield's illness was not diagnosed until many years after the exposure which is alleged to have caused it. The action was brought within the three years after the diagnosis. The defendant will have to live with the evidentiary and other consequences of that. Even if the illness had occurred earlier, and thus the cause of action had arisen earlier, it was not discovered until the diagnosis. When the Limitation Act 1963 responded to Cartledge v E Jopling and Sons Ltd [1963] AC 758, by introducing discoverability into the personal injury limitation regime, it was prioritising the interests of the claimant in being compensated for his injury over the interests of both the defendant and the state. But given the length of time which some illnesses take to develop after exposure to the causative agent, perhaps it did not seem such a very radical step.
166. By contrast, the introduction of the power to disapply the limitation period in personal injury cases in the Limitation Act 1975 was a radical step. But it was a step more closely linked to the policy aims underlying the limitation legislation. It enables the court to ask whether the defendant deserves to enjoy the windfall of a limitation defence, or the claimant to lose the benefit of a claim, by reference to the crucial questions of whether a fair trial of the action will still be possible and whether there is a good reason for the delay in bringing the claim. In policy terms, the world would be a more sensible and predictable place if we had only the discretion provided by section 33 of the Limitation Act 1980, without the discoverability provisions in sections 11 and 14. It might be better still if the cause of action accrued at the date of the wrongful act or omission rather than at the date of damage. Hence, in policy terms, it is understandable that the Law Commission welcomed the approach to discoverability in personal injury actions which has been developed in the Court of Appeal. In policy terms, shifting the burden to the discretion in section 33 is preferable to postponing the date when time begins to run.
167. The Court of Appeal's approach, as we know, culminated in the view that a strong but completely irrational belief that an injury was attributable to the act or omission of the defendant equated with "knowledge" of that "fact" for the purpose of section 14(1) of the 1980 Act. Once there was power to disapply the limitation period in cases where a fair trial would still be possible, justice could be done to both parties irrespective of the date of knowledge. The realist in me is not surprised that the Court of Appeal applied a subjective test to the date of knowledge. The court could then get on with weighing the competing interests under section 33. I remember doing just that in Roberts v Winbow [1999] Lloyd's Rep Med 31.
168. However, that is not what the statute provides. Like it or not, time does not begin to run until the claimant has "knowledge" of the essential "facts". We have been focussing in this case on knowledge "(b) that the injury was attributable in whole or in part to the act or omission which is alleged to constitute negligence, nuisance or breach of duty". The Court of Appeal has reached the position that a sufficiently strong subjective belief to send the claimant to a solicitor to investigate making a claim is "knowledge" of attributability for this purpose, even though there is no reasonable basis in evidence or objective fact for that belief. This leads to unedifying inquiries, such as those which took place in these cases, into the strength of the various claimants' belief, however unreasonable – inquiries made on behalf of a defendant who has always maintained that there is no reasonable basis for their beliefs (and thus contributing to the strong sense of injustice they feel). On the Court of Appeal authorities, a claimant who strongly believed, on no reasonable ground whatsoever, that his illness was caused by exposure to radiation "has knowledge of the fact that" his injury is attributable to that exposure, whereas a claimant who strongly believed that it was not, on the reasonable ground that those in a position to know the truth denied it, has no such knowledge. The strength of a claimant's subjective belief is not a sensible basis for deciding who does, and who does not, have an absolute right to pursue his action.
169. I suspect that that point would never have been reached had the law confined itself to knowledge of the fact of injury. That was the problem in Cartledge v Jopling. Diagnosis is a relatively clear cut question. You do or do not have the disease. You do or do not know, on the basis of a reliable diagnosis, that you do or do not have the disease. The hypochondriac who believes, on no ground whatsoever, that he has a disease cannot be said to have knowledge of the fact that he has an injury, let alone that it is significant. If this had been the only context in which knowledge had come to be explored, I do not think that the Court of Appeal would have found itself in the place where it eventually arrived.
170. But most of the cases are about knowledge of attributability. Even on the basis that attributable means "capable of being attributed to", it is difficult to see how an unreasonable belief in attributability can amount to "knowledge of the fact" of attributability. It clearly does not. It amounts to an unreasonable belief. Only when there is some reasonable basis in evidence or objective fact for that belief can it be turned into something approaching knowledge. We do not need in this case to debate the point at which belief can turn into knowledge because we can confidently state that there was no such basis for the veterans' beliefs, however strongly held, until at the earliest the publication of the Rowland report.
171. It is my belief, therefore, that the Court of Appeal, bit by bit in the way that Lord Kerr and Lord Phillips have described, turned a very sensible policy approach to how the law should approach limitation in personal injury cases into a construction of section 14 of the Limitation Act 1980 which cannot be justified by the words used.
172. Some of these claimants did have a strongly held belief that their illnesses were caused by exposure to radiation. That depended on (a) proof of their exposure and (b) proof that their exposure caused their illnesses. There was no evidence to get them to point (a) until the Rowland report, but that only supplied the possibility. Each of the claimants would have to be assessed to see whether the same chromosomal effects were found in them. There is still no evidence to supply a causal connection between that exposure and the claimants' various illnesses.
173. In agreement with Lord Phillips and Lord Kerr, upon whose analysis of the law I cannot improve, I would hold that none of these actions is statute barred and the discretion under section 33 does not arise. As we are in a minority, it is not necessary for me to choose between them, but logically I prefer Lord Phillips' view. Rowlands has got the claimants further than they were before, but it has not supplied the basis for a belief in causation as well as exposure. So the claimants still do not have the knowledge required for time to begin to run.
174. There is a very good case for the law being different. But I do not think that we should translate our view of what a sensible law of limitation would say into our view of what it does say. Knowledge and belief are different concepts and there is no reason to believe that Parliament intended to equate the two.
175. I would therefore allow this appeal and make the order which Lord Phillips proposes.

LORD KERR

The genesis of the statutory provisions

176. In the Report of the Committee on Limitation of Actions in Cases of Personal Injury (1962) (Cmnd 1829) (prompted by the decision in Cartledge v E Jopling & Sons Ltd [1963] AC 758 and leading to the enactment of the Limitation Act 1963) at para 8 the following appears:

"[The claimant] may, for example, obviously be suffering from a particular disease without being able to attribute it to the conditions under which he has been working, either because there is no sufficiently widespread knowledge of the causal connection between the processes on which he is, or formerly was, engaged and the disease in question, or because he has no reason to suppose that these processes in fact expose, or exposed, him to some noxious substance."

177. The problem in the state of the law which the Committee had identified was therefore firmly linked to an absence of knowledge of the fact that an employee's disease had been caused by work processes. And this was reflected in the terms of the 1963 Act. By section 1(1) of that Act, section 2(1) of the Limitation Act 1939, which had imposed a time limit of three years for bringing certain actions, was to no longer afford a defence where the requirements of section 1(3) were fulfilled. Section 1(3) provided:

"The requirements of this subsection are fulfilled in relation to a cause of action if it is proved that the material facts relating to that cause of action were or included facts of a decisive character which were at all times outside the knowledge (actual or constructive) of the plaintiff until a date which –

(a) either was after the end of the three-year period relating to that cause of action or was not earlier than twelve months before the end of that period, and

(b) in either case, was a date not earlier than twelve months before the date on which the action was brought" (emphasis added).

178. The "material facts" expression in this subsection was explained in section 7(3) of the Act:

"… any reference to the material facts relating to a cause of action is a reference to any one or more of the following, that is to say –

(a) the fact that personal injuries resulted from the negligence, nuisance or breach of duty constituting that cause of action;

(b) the nature or extent of the personal injuries resulting from that negligence, nuisance or breach of duty;

(c) the fact that the personal injuries so resulting were attributable to that negligence, nuisance or breach of duty, or the extent to which any of those personal injuries were so attributable."

179. For time to begin to run, therefore, the claimant had to know as a fact that the personal injuries resulted from negligence etc and that they were attributable to that default. As Lord Walker has observed (in para 29 of his judgment) these twin requirements were replicated in the Limitation Act 1975 (which is now consolidated as part of the Limitation Act 1980).
180. The 1975 Act followed on the Law Reform Committee's Interim Report on Limitation of Actions in Personal Injury Claims of May 1974 (Cmnd 5630). In para 49 of the Report, the committee confronted directly the question of how the date of knowledge should be determined. It outlined various tests that had been considered:

"It has not been suggested to us, and in our view could not reasonably be suggested, that the plaintiff's date of knowledge should arrive until he has knowledge (actual or constructive) both of his injured condition and of its having been caused by an act or omission of the defendant. In our view, the crucial question to be answered is whether the date of knowledge should arrive-

(1) on the plaintiff's acquiring knowledge of those facts; or

(2) on his acquiring knowledge of these facts and also that he has a worthwhile cause of action against the defendant; or

(3) at some intermediate point between these states of knowledge, as for example on his becoming aware, in the words of Lord Pearson, [in Smith v Central Asbestos Co Ltd [1973] AC 518] "(as a matter of fact in the same manner as a jury would decide) that the defendants were at fault and that his injuries were attributable to their fault"."

181. The Committee chose the first of these alternatives. It is significant that the way that it expressed its conclusion was that the plaintiff's knowledge should arrive only when he had actually or constructively acquired knowledge of two matters which it described as "facts". The first was his injured condition and the second that the condition had been caused by an act or omission of the defendant. In my view, the characterisation of the second of these as a fact clearly indicates that it was contemplated that there would have to be some factual foundation for it. I am reinforced in that view by the later passage in the Law Reform Committee's report at para 54 where it said that the only possible intermediate solution was that adopted by Lord Pearson in Smith's case (set out in sub-para (3) of para 49 of the report and quoted above).
182. Although the committee rejected this formulation it did so because the concept of fault could not be satisfactorily defined and because it contained a considerable subjective element. It did not suggest that the knowledge of the plaintiff should be other than knowledge of a particular set of facts. Nothing in the report suggests that the committee considered that mere belief in a state of affairs would be sufficient. Indeed, in para 55, it put its conclusion in this way: that the plaintiff should have knowledge, actual or constructive, both of his injured condition and of its having been caused by acts or omissions of the defendant. This seems plainly to point to the requirement that there be an objective basis for the knowledge of facts which the plaintiff had to have.
183. In light of this, it is unsurprising that the prefatory words of section 14(1) of the 1980 Act are that the date of knowledge is the date on which the plaintiff first had knowledge of the facts which are outlined in the sub-paragraphs which follow. The facts contained in those sub-paragraphs of which the plaintiff is required to have knowledge are (a) that the injury was significant; (b) that it was attributable to the act or omission of the defendant; (c) the identity of the defendant; and (d) if it is alleged that the act or omission was that of a person other than the defendant, the identity of that person and the additional facts supporting the bringing of an action against the defendant. In relation to the attributability element of the plaintiff's knowledge, what was stipulated was that the plaintiff was required to know as a fact that the injury was attributable to (in the sense of being capable of being the cause of) the act or omission.
184. From a purely textual analysis of the statute, therefore, it is impossible to suggest that what it intended to convey was that the test should be that the plaintiff would have statutory knowledge when he believed (or even firmly believed) that the injury was attributable to the defendant's act or omission. The natural meaning of the language used was that the plaintiff needed to know it rather than to believe it and that he needed to know it as a fact. Indeed, the phrase "knowledge of facts" permeates the section. Section 14(1)(d), for instance, provides that, if it is alleged that the act or omission was that of a person other than the defendant, time would begin to run only when the plaintiff had knowledge of the identity of that person and the additional facts supporting the bringing of an action against the defendant.
185. The word 'knowledge' appears nine times in the subsections which precede section 14(3) and in all but one of these, knowledge is associated with facts. Knowledge of facts is also the same formula used in section 14(1A). It is interesting to note, therefore, that knowledge of the facts is required for a wide spectrum of circumstances such as: (i) the identity of the defendant; (ii) that the injury was significant; and (iii) the further unspecified facts that are needed to support the bringing of an action against the defendant when it is alleged that the act or omission was that of a person other than the defendant.
186. The direct association between knowledge and facts continues in section 14(3) which provides:

"For the purposes of this section a person's knowledge includes knowledge which he might reasonably have been expected to acquire –

(a) from facts observable or ascertainable by him; or

(b) from facts ascertainable by him with the help of medical or other appropriate expert advice which it is reasonable for him to seek;

but a person shall not be fixed under this subsection with knowledge of a fact ascertainable only with the help of expert advice so long as he has taken all reasonable steps to obtain (and, where appropriate, to act on) that advice."

Subjective belief versus knowledge of facts

187. If, as the respondent contends, the rubric 'knowledge of the facts' is to be interpreted as firm belief in a state of affairs, it surely must have that meaning for each of the circumstances in which it is employed. Therefore, if the respondent is right, the claimant need only have a belief in the identity of the defendant rather than knowing who he is; he need only believe that his injury was significant rather than knowing it to be such; and need only believe the other facts that support the bringing of an action against the defendant, where it is alleged that the act or omission was that of a person other than the defendant.
188. If the respondent's argument was correct, there is therefore a wide and surprising range of matters and circumstances that a claimant needs only to believe in rather than to know, for statutory knowledge to arise. If that interpretation were to be accepted, the repeated use in the section of the word knowledge (which is in no sense a natural synonym of belief) is mystifying. It would have been a simple task to convey that a claimant need only believe in these matters if that had been the intention of Parliament. But it is, to my mind, inherently unlikely that Parliament could ever have intended that claimants should be encouraged to commence litigation when all that they had to go on was a belief, however strongly held.
189. It is possible to imagine states of mind that might be required of a claimant as points on a spectrum, with, at one end, simple belief with no factual foundation whatever and, at the other, objectively verifiable certainty. On the belief side of this spectrum, various alternative formulations of states of mind can be envisaged. Belief in a state of affairs which, if true, would constitute the requisite knowledge; or reasonable grounds for believing certain facts; or a real possibility that the facts which the claimant believes are correct; or credible information supporting the correctness of the facts. To interpret section 14 so as to fasten on any one of those formulations would require, at least, a determinedly purposive approach. But the respondent in this case would have us go further than any of these and construe the section as requiring no more than mere belief. Simple unvarnished belief is qualified only (in the respondent's submission) by the requirement that it be belief of such a quality as would prompt a sensible person to begin investigations into the viability of proceedings. This is, in essence, a purely subjective state of mind. The second element, that a sensible person holding such a belief would be moved to investigate whether to bring proceedings, does not sound at all on the source or accuracy of the belief. So an unreasoning belief, provided it was sufficiently firmly held, would qualify as knowledge activating the beginning of the limitation period. That precisely constructed and specifically defined set of circumstances – that an individual should have a belief that was personal to him or her and that this would act as the catalyst for an inquiry whether to launch proceedings – seems to me to be impossibly and impermissibly incompatible with the natural meaning of section 14.
190. It is not clear what becomes of the limitation period as a consequence of the progress (or lack of it) of the inquiries, if the respondent's argument is accepted. On its analysis, when the moment arrives that a claimant's belief in a set of affairs hardens into a conviction that inquiries should be made, time begins to run. But what happens if the inquiries are initially unproductive and the claimant's belief falters? Mr Gibson QC submitted that what mattered was the claimant's subjective state of mind and his degree of certainty. He said that statutory knowledge equates with firmness of belief. He also argued that belief was transformed to the status of fact by its inclusion in a statement of claim. But what happens if the claimant loses belief after the statement of claim has been issued? What if the firmness of his purpose, the strength of his belief crumbles? Is the limitation period to be suspended? If subjective belief of sufficient resolution is the catalyst for time starting to run, why should not a failure of belief stop time running? And how is the question of whether subjective belief is present to be judged? Is this inevitably linked to the issue of proceedings? What if a claimant testifies that he or she did not believe that there was a connection between the exposure and the onset of disease but was advised to issue proceedings in the hope that evidence could be obtained to forge such a link, are they to be fixed with a belief that, in truth and in fact, they did not hold? These considerations clearly point, in my opinion, to the unworkability of a system so directly linked to and uniquely dependent on a claimant's subjective belief.
191. Other anomalies arise, if subjective belief is substituted for knowledge based on objective fact. If two individuals are exposed to the same noxious materials and both develop disease in consequence, but medical science has not evolved to the point where either could be fixed with constructive knowledge that this was due to the exposure, time begins to run against the claimant who has a firm belief in the cause of his illness sufficient to lead him to consult a solicitor but not against his fellow employee who does not share that strength of belief. Such a situation defies logic as well as legal principle.

The authorities

192. The obvious starting point is that chosen by Lord Phillips (in para 112 of his judgment) of Davis v Ministry of Defence (unreported). In that case May LJ decried what he considered was the wrongful assimilation by the trial judge of firm belief and knowledge. The facts of the case required the question of belief versus knowledge to be addressed directly. The appellant had always believed that his dermatitis had been caused by working conditions. But he was advised that there was not sufficient evidence of this. His belief was therefore not supported by objective evidence. As Lord Wilson has pointed out (in para 9 of his judgment), in Halford v Brookes [1991] 1 WLR 428, 443F, Lord Donaldson MR characterised Davis as an exceptional case. Indeed, at 442H he suggested that the facts of Davis were highly unusual. But one wonders, in the light of contemporary experience, if the case was quite as unusual as it has been portrayed. Developing medical science about the aetiology of various conditions, particularly perhaps in the field of asbestos related disease, has shown its remarkable tendency to catch up with and provide support for the firmly held beliefs of workers that the condition from which they suffered was caused by their working conditions. In any event, the important thing, as it seems to me, is not whether Davis can be dismissed as an exceptional case but whether the reasoning that led to its outcome is sound and can be applied to the present appeal. In my judgment the reasoning is indeed sound and is directly relevant to the problem that this case poses.
193. Lord Wilson considers it to be "heretical" to allow a claimant to escape what he describes as the conventional requirement to assert his cause of action for personal injuries within three years of its accrual (para 6 of his judgment) "by establishing that, even after his claim was brought, he remained in a state of ignorance entirely inconsistent with it". One can see the initial attraction of that argument. But it is necessary, I believe, to take a step back. The important question on which to focus is when the action accrues. Like it or not, the legislature has decreed that this occurs when a claimant is in a mental state which amounts to knowledge of facts that are, among other things, capable of showing that the personal injury which is the subject of his claim was due to the default of the defendant. True it may be, as Lord Wilson says, that the concepts of belief and knowledge are inherently subjective but that does not mean that they are interchangeable. I know something to be true because I have a factual foundation for my knowledge of its truth. I may believe something to be true without any basis in practical reality whatever. And simply because I assert the truth of a particular proposition, I cannot be taken to know (as opposed to believe) it to be so.
194. I am of the opinion, therefore, that the Court of Appeal in Davis was right to recognise the clear distinction between belief (even firm belief) and knowledge. The latter concept is inexorably tied to an objective assessment of what is known rather than what is taken on faith or impression.
195. The approach of May LJ in Davis was approved by Slade LJ in Wilkinson v Ancliff (BLT) Ltd [1986] 1 WLR 1352 but the retreat from its clear demarcation of knowledge of facts from firm belief may be supposed to have begun with Halford. The facts of the latter case have been set out by Lord Phillips at para 114 of his judgment and need not be repeated. At 434B Russell LJ said:

"Once the plaintiff in this case realised that her daughter's death was capable of being attributed to the activities of the defendants or one or other of them, time began to run and, subject to the provisions of section 33 of the Act, she had three years thereafter in which to issue her proceedings."

196. It should be noted that, although the distinction between knowledge and belief might have begun to be blurred by this case, Russell LJ's formulation does not equate knowledge with belief. "Realised" is the word employed rather than "knew". Realisation is, of course, a concept which could span both knowledge and belief although it might be considered to be more comfortably accommodated in the former. In any event, Lord Donaldson MR addressed the issue in somewhat different terms. At 443E-F he said:

"The word [knowledge] has to be construed in the context of the purpose of the section, which is to determine a period of time within which a plaintiff can be required to start any proceedings. In this context 'knowledge' clearly does not mean 'know for certain and beyond possibility of contradiction.' It does, however, mean 'know with sufficient confidence to justify embarking on the preliminaries to the issue of a writ, such as submitting a claim to the proposed defendant, taking legal and other advice and collecting evidence.' Suspicion, particularly if it is vague and unsupported, will indeed not be enough, but reasonable belief will normally suffice."

197. It is important to observe that the Master of the Rolls said that it was necessary that the plaintiff should know with sufficient confidence that a state of affairs existed that justified embarking on the preliminary steps that would lead to the issue of proceedings. And, of course, in this case, Mrs Halford was deemed to know that her daughter had been killed by one or other of the defendants. It was not a question of her merely believing that one or other had committed the crime. But, importantly, Lord Donaldson MR in this passage described the quality of knowledge required as that which was sufficient to prompt inquiries to be made and in the same passage suggested that reasonable belief would normally be sufficient. In my opinion, these two concepts have been conflated in later decisions, so that belief (as opposed to knowledge) sufficient to prompt investigations has become the yardstick. I doubt if this was Lord Donaldson MR's intention. He did not elaborate on what was meant by 'reasonable belief' in this context but, juxtaposed as this was with the statement that a plaintiff should know with sufficient confidence that further inquiries were justified, one can only suppose that he considered that the belief would have to have a sufficiently secure factual foundation in order to activate the limitation period.
198. In Nash v Eli Lilly & Co [1993] 1 WLR 782, Purchas LJ disavowed an attempt to define knowledge – see 792C. But in the passage of his judgment at 796 F-H quoted by Lord Phillips at para 117 above he addressed the question of whether firm belief could be equated with knowledge. In the first part of this passage, Purchas LJ appears to strongly reject this notion for he said:

"… a firm belief held by the plaintiff that his injury was attributable to the act or omission of the defendant, but in respect of which he thought it necessary to obtain reassurance or confirmation from experts, medical or legal, or others, would not be regarded as knowledge until the result of his inquiries was known to him or, if he delayed in obtaining that confirmation, until the time at which it was reasonable for him to have got it."

199. It can be deduced from this passage that the firm belief, in order to be transformed to a condition of knowledge, required to be bolstered by "reassurance or confirmation from experts" – in other words, they would need to confirm that there was a sound basis in fact for holding the belief. So far, so unexceptionable. But the final sentence of the passage, as Lord Phillips has pointed out, led to the later misapprehension that belief, provided it was of sufficient firmness, could, without more, be equated with knowledge. There Purchas LJ, in what seems to be a complete reversal of the statement quoted above, said:

"If the plaintiff held a firm belief which was of sufficient certainty to justify the taking of the preliminary steps for proceedings by obtaining advice about making a claim for compensation, then such belief is knowledge and the limitation period would begin to run."

200. In an earlier passage at 792C-D the same approach can be detected:

"In applying the section to the facts of these cases, we shall proceed on the basis that knowledge is a condition of mind which imports a degree of certainty and that the degree of certainty which is appropriate for this purpose is that which, for the particular plaintiff, may reasonably be regarded as sufficient to justify embarking upon the preliminaries to the making of a claim for compensation such as the taking of legal or other advice."

201. What, with respect, these passages neglect to address is the source of the certainty. One can express oneself certain in a particular belief but knowledge depends on factual information rather than simple belief, however fervently held. Where the decision in Nash fell into error was in concentrating on the plaintiff's state of mind and the degree of certainty with which the belief that constituted that state of mind was held rather than making the all important link between knowledge and facts sufficient to support it.
202. Essentially the same confusion of knowledge with belief is apparent in Spargo v North Essex District Health Authority [1997] PIQR P225 where Brooke LJ said at P242:

"A plaintiff has the requisite knowledge when she knows enough to make it reasonable for her to begin to investigate whether or not she has a case against the defendant. Another way of putting this is to say that she will have such knowledge if she so firmly believes that her condition is capable of being attributed to an act or omission which she can identify (in broad terms) that she goes to a solicitor to seek advice about making a claim for compensation"

203. In the first sentence of this passage, the plaintiff is treated as having the necessary knowledge when she knows enough. In the second sentence firm belief is said to be "another way of putting" the need to know. I must respectfully disagree. Knowing is not believing. To know something to be true is different from believing it to be so.
204. The confusion begun by Nash between knowledge and belief was continued in Sniezek v Bundy (Letchworth) Ltd [2000] PIQR P213 CA. In that case Simon Brown LJ recognised that the difficulty in conceptualising what is meant by "knowledge" in section 14 generally arose in relation to "knowledge of the fact of attributability" (P233). And at P234 he observed that, as had been held in Spargo, "a real possibility of establishing causation constitutes attributability". In an earlier passage at P223-P234 he said this:

"… it seems to me that the real contrast being struck in Nash v. Eli Lilly is between on the one hand the mere believer whose situation is described in the first passage in the judgment, and on the other hand the firm believer sufficiently certain of his case to have clearly in mind (although always, of course, subject to the taking of appropriate advice and the preparation of evidence) the making of a compensation claim."

205. I do not have difficulty with the proposition that knowledge that there is a real possibility that the condition was caused by the act or default of the defendant constitutes statutory knowledge of attributability. And one can understand why it might be considered that there is a small step between knowing of a real possibility and firmly believing that there is a connection between the injury and the default of the defendant. But there is a significant difference between the two. Knowledge of a real possibility that the act or omission of the defendant caused the injury involves some evaluation of the factual foundation for the claim. It is not essential to the holding of a firm belief that a similar examination be conducted.
206. The focus of the debate should be on what the claimant knew as a fact – or, at least, on what he knew was a possible fact – and not on what he believed. This is, I think, well captured by Lord Mance in Haward v Fawcetts [2006] UKHL 9, [2006] 1 WLR 682 at para 128 where he said: "the question is when Mr Haward actually knew both enough of the acts or omissions now alleged to constitute negligence and that the loss suffered was capable of being attributable thereto to make it reasonable for him to begin to investigate whether or not the claimants had a claim" (emphasis added). The claimant needs to know that the acts or omissions possibly caused the injury. He does not need to know that they constitute actionable fault. But he needs to know – at least – that there is a reasonable possibility that those acts or omissions were responsible for his injury. As Lord Nicholls put it in para 11 of Haward, "time does not begin to run against a claimant until he knows there is a real possibility his damage was caused by the act or omission in question" (emphasis added).

The effect of the issue of proceedings

207. Lord Wilson considers that it is "a legal impossibility" for a claimant to assert that he did not have knowledge of attributability for the purpose of section 14 (1) after he has issued his claim. My disagreement with this view is inevitable in light of the conclusion that I have reached as to the proper interpretation to be given to knowledge for the purposes of section 14. The plain fact is that a claimant need only verify that he believes the facts stated in his claim form to be true. The terms in which this requirement is couched reflect, in my opinion, the prosaic truth that the significant averments in claim forms consist of assertions that are, by definition, likely to be controversial. Claimants cannot be supposed to "know" that what is asserted is true in any conventional sense. It is only if one gives 'knowledge' a specialised meaning (which abandons its normal connotation for a concept that is entirely different) for the purposes of section 14 that it becomes possible to say that when a claimant says that he believes something to be true, he is in effect to be taken as saying that he knows it to be such.

The proper test

208. Foskett J formulated his "preferred view" of the elements of the knowledge that a claimant is required to have in para 514 of his judgment. According to this test they required to know:

i) That the injury of which he complains was capable of being caused by the act or omission of the defendant; and

ii) That there is credible evidence that the act or omission alleged had occurred.

209. Lord Phillips considered that this preferred view (which, in light of the authorities, Foskett J felt unable to follow) was correct in principle and, with respect, so do I but with one minor qualification, which may amount to no more than a slight reworking of the same test. In Haward and Spargo it was held that the claimants were required to know that there was a real possibility of the act or omission having caused the damage. This reflected the circumstance that "attributable" in the context of section 14 has been construed as meaning "capable of being attributed". If a condition is capable of being attributed to exposure to a noxious agent, it follows that there is a real possibility that it was so attributable if the condition has developed and harmful exposure is established. I would therefore express the test compendiously as follows: the claimant must know from objectively verifiable facts that there was a real possibility that the injury suffered was due to the act or omission complained of.

Applying the test to these cases

210. The Ministry of Defence has consistently denied that the veterans were exposed to ionising radiation that was in any way capable of causing injury. That remains their position. Although many of the veterans have for many years believed (with varying levels of conviction and passion) that the conditions that they have suffered were caused by exposure to this radiation, until they were made aware – or ought to have become aware - of the findings of the Rowland study they could not have known from objectively verifiable facts that there was a possible connection between their exposure and the various conditions from which they have suffered. As Foskett J said, before the Rowland report, "all the evidence raised nothing more than a suspicion of exposure to excess ionising radiation with no clear link to the conditions of which complaint is made" – para 517 of his judgment.
211. In my judgment, the Rowland report has supplied the necessary objective factual foundation for knowledge on the part of the veterans that there is a real possibility of a connection between their exposure and the conditions that they have suffered. Lord Phillips has said that the most that the veterans as a group can show, on the basis of this report, is that "there is a possibility that some of them were exposed to a raised, albeit low, level of fall-out radiation and that this may have increased the risk of contracting some at least of the injuries in respect of which they claim" (para 157 of his judgment). But this is enough, in my opinion, to meet the test of knowledge from objectively verifiable facts that there is a possible connection between tortious exposure and injury. I would therefore hold that time has begun to run from the date that they became aware of or ought to have been aware of the contents of the Rowland report.

Striking out and summary judgment

212. For the reasons given by Lord Phillips, with which I agree, Foskett J was right not to accede to the application to strike out these proceedings and to refuse the defendant's application for summary judgment. I consider that the judge was fully justified in his view that the veterans were not bound to fail on the issue of causation. True it is that the evidence on this issue does not look especially promising for them at present. But, as Lord Phillips has illustrated, there are examples in the past of group litigation where the signs cannot have been propitious when litigation was commenced and, as he has said, depending on the particular circumstances of the case, it may well be prudent to issue proceedings, even though investigations are continuing.
213. It may well be correct, as Lord Phillips has suggested in para 154, that the judge did not give separate consideration to the question of summary judgment, having decided to refuse the strike out application but, if this is the position, it is hardly surprising. Once he had decided that the veterans were not bound to fail, it was not open to him to accede to the application for summary judgment against them.
214. Moreover, at a more general level, I would question the propriety of giving summary judgment on a preliminary hearing on whether an action is statute-barred, particularly where detailed medical evidence is required in order to address the question of causation. Although in the present case there was an extensive review of the evidence, its presentation was principally geared to the examination of its relevance to the date of the claimants' knowledge for the purposes of section 14. As Foskett J said (at para 5 of his judgment) "the merits of the individual claims [did] not arise for consideration". A confident judgment on the myriad issues that the question of causation in a complicated case such as the present raises will normally only be possible when there has been a full ventilation and testing of the experts' evidence on both sides.

Conclusion

215. I would allow the appeals and make the order that Lord Phillips has proposed.