Case Analysis: Article 10, the Reynolds Test and the Rule in the Duke of Brunswick’s Case: the decision in Times Newspaper Ltd v the United Kingdom

Rory Dunlop

This case analysis looks at the recent decision of the European of in Times Newspaper Ltd v the United Kingdom1. In this case, the Times sought to challenge two important aspects of the English of defamation: the Reynolds test for qualified privilege and the rule in the Duke of Brunswick’s case. The Court decided that the criticisms of the Reynolds test were manifestly ill-founded but left open the question of whether the rule in the Duke of Brunswick’s case was compatible with Article 10. This analysis examines the Court’s reasoning in relation to the Reynolds test and argues that the Court should find the rule in the Duke of Brunswick’s case to be incompatible with Article 10.

1. Background

Before looking at the case which is the subject of this analysis, it is necessary to look first at the two elements of the English law of defamation that the applicant intended to challenge.

(a) The Reynolds Test

The so-called Reynolds test emanates from the of the House of Lords in Reynolds v Times Newspapers2. Albert Reynolds was a former Irish Prime Minister accused in the British press of deliberately misleading the Irish Parliament. Reynolds sued for defamation and the defendants pleaded, inter alia, qualified privilege.

At the the returned a in the plaintiff's favour and he was awarded the sum of 1p by way of damages. The Court of Appeal set aside the jury's verdict and ordered a retrial. The court also ruled that the publication was not covered by qualified privilege. The defendants appealed, arguing that there should be a ‘generic’ qualified privilege encompassing the publication by a newspaper of political information.

The House of Lords rejected this argument. They considered that the proposed new category of qualified privilege would fail to provide adequate protection for reputation. In addition they held that it would be unsound in principle to distinguish political discussion from other matters of serious public concern.3 However, they did say that the defence of qualified privilege would be available in respect of political information when the defendant satisfied the old test, i.e.:

“where the person who makes a communication has an interest or a duty, legal, social, or moral, to make it to the person to whom it is made, and the person to whom it is so made has a corresponding interest or duty to receive it.”4

1 Application numbers 23676/03 & 3002/03, decision of 11 October 2005. 2 [2001] 2 AC 127 3 ibid. at 204 G-H 4 ibid. at 194 H, citing Adam v. Ward [1917] A.C. 309, 334

1

Lord Nicholls, giving the leading judgment, went on to provide a non-exhaustive list of the factors to take into account when deciding whether the defence should be available5:

“1. The seriousness of the allegation. The more serious the charge, the more the public is misinformed and the individual harmed, if the allegation is not true. 2. The nature of the information, and the extent to which the subject-matter is a matter of public concern. 3. The source of the information. Some informants have no direct knowledge of the events. Some have their own axes to grind, or are being paid for their stories. 4. The steps taken to verify the information. 5. The status of the information. The allegation may have already been the subject of an investigation which commands respect. 6. The urgency of the matter. News is often a perishable commodity. 7. Whether comment was sought from the plaintiff. He may have information others do not possess or have not disclosed. An approach to the plaintiff will not always be necessary. 8. Whether the article contained the gist of the plaintiff's side of the story. 9. The tone of the article. A newspaper can raise queries or call for an investigation. It need not adopt allegations as statements of fact. 10. The circumstances of the publication, including the timing.”

The application of these factors came to be known as the Reynolds test. This test was criticized by some commentators and .

Clayton and Tomlinson criticised it in ‘The Law of Human Rights’6 on two grounds. First, they said that it did not provide full recognition of the species of ‘qualified privilege’, developed by the Convention , described as a ‘safeguard to journalists’ when ‘matters of legitimate public concern’ are being discussed by the press. They cited Bergens Tidende v Norway 7 as an example of this species of qualified privilege.

Secondly, they argued that it left the law of defamation in a state of uncertainty in relation to media discussion of matters of public interest. They said the following8:

“In the absence of a developed body of , a responsible media organisation publishing material relating to matters of ‘public interest’ will be unable to determine whether a defence is available. It seems that the full background to the publication will have to be investigated in each case in order to satisfy the court that there is a ‘right to know’. It is arguable that this uncertainty, of itself, constitutes an unacceptable restriction on freedom of expression.”

The New Zealand Court of Appeal also argued, in the case of Lange v Atkinson9, that the Reynolds test would create excessive uncertainty. This case concerned an article by a political scientist accusing the former Prime Minister, David Lange, of hypocrisy. The defendant pleaded qualified privilege and was successful before the High Court. The Court of Appeal dismissed the appeal and, in the process, appeared to develop a ‘generic’ qualified privilege defence, for statements made about the actions and qualities of current, former and prospective members of

5 ibid. at 205 A-C 6 2000 edition at § 15.249. 7 Application number no. 26132/95, Judgment of 2 May 2000. Clayton & Tomlinson relied on paragraphs 53 and 60 in particular 8 2000 edition at § 15.249 9 [2000] 3 NZLR 385

2

Parliament, similar to the one which was later rejected by the House of Lords in Reynolds10. The Privy Council, on appeal, sent the case back to the New Zealand Court of Appeal to take into account the decision of the House of Lords in Reynolds11. The New Zealand Court of Appeal upheld their original judgment, with modifications, and declined to follow the House of Lords.12

They considered that a distinction needed to be maintained between the occasion of privilege and its abuse. The quoted a report carried out on libel and its chilling effect on the media.13 The report concluded that ‘uncertainty in both the principles of defamation law and their practical application induce great caution on the part of the media’. The Court added the following:14

“The blurring, perhaps even the removal, of the line between the occasion and its abuse in Lord Nicholls of Birkenhead's non-exhaustive list must add significantly to that uncertainty. In the absence of compelling justification that consequence appears undesirable.”

Later, in their conclusions, they said:15

“[T]he Reynolds decision appears to alter the structure of the law of qualified privilege in a way which adds to the uncertainty and chilling effect almost inevitably present in this area of law. We are not persuaded that in the New Zealand situation matters such as the steps taken to verify the information, the seeking of comment from the person defamed, and the status or source of the information, should fall within the ambit of the inquiry into whether the occasion is privileged. Traditionally such matters are not of concern to that question in the kind of setting presently under discussion. In particular, source and status may be relevant, but only in the area of reports of meetings and suchlike. For the reasons expressed in our earlier judgment, we do not consider it necessary, nor would it be in accord with principle, to import into this inquiry, for the limited purposes of the specific subject-matter now under discussion but not otherwise, a specific requirement of reasonableness.”

(b) The Rule in the Duke of Brunswick’s case

The Duke of Brunswick v Harmer16 is the case normally cited for the established common law rule that each publication of a defamation gives rise to a separate cause of action. The case itself concerned an article published in a magazine on 19 September 1830, when the limitation period for libel was six years. The article defamed the Duke of Brunswick. Seventeen years after its original publication, a servant of the Duke obtained a back number of the magazine from the office of the magazine’s publishers and from the British Museum. The Duke sued on those two publications. The defendant contended that the cause of action was time-barred, relying on the original publication date. The Court held that the delivery of a copy of the newspaper to the plaintiff’s agent constituted a separate publication in respect of which suit could be brought.

10 Lange v Atkinson [1998] 3 NZLR 424 at 468 11 Lange v Atkinson [2000] 1 NZLR 257, unreported in England 12 Lange v Atkinson [2000] 3 NZLR 385. See ‘Defamation Law in New Zealand ‘Refined’ and ‘Amplified’’ by W.R.Atkin, Common Law World Review 30.2 (237) June 2001 in which Mr Atkin argues convincingly that the later decision of the New Zealand Court of Appeal in fact bears more similarities to the judgment in Reynolds than to their original judgment. 13 Libel and the Media: the Chilling Effect [1997] by Eric Barendt, Laurence Lustgarten, Kenneth Norrie and Hugh Stephenson, referred to in Reynolds, op. cit. at p. 1032 14 Ibid. at § 24 15 ibid. at § 38 16 [1849] 14 QB 185

3

This rule was subsequently approved by the majority of the House of Lords in Berzovsky v Michaels17, where the issue was whether England was the appropriate forum for bringing an action in relation to a defamatory article published in an American business magazine of which only relatively few copies were ever distributed in England. In the course of their judgments, Lord Steyn, Lord Hoffmann and Lord Hope each observed18 that, according to the English common law, a separate cause of action arose in respect of each publication of the article.

This rule has also been the subject of much criticism by judges and commentators alike. Even in 1887 Pollock suggested that the law went “wrong from the beginning in making the damage and not the insult the cause of action.”19 In 1953 Dean Prosser described the rule as possibly appropriate to ‘small communities and limited circulations’ but ‘potentially disastrous today’.20 Robertson and Nicol have described the rule as “absurd but venerated”.21

Judges too have noted the irony of applying such a rule in the modern age. Kirby J made the following wry remarks in Dow Jones & Co Inc. v Gutnick22:

“The idea that this Court should solve the present problem by reference to judicial remarks in England in a case, decided more than a hundred and fifty years ago, involving the conduct of a manservant of a Duke, despatched to procure a back issue of a newspaper of miniscule circulation, is not immediately appealing to me.”

The Gutnick case concerned whether or not a businessman mostly based in Australia should be allowed to sue the publisher of the Wall Street Journal in Australia for an article that it published in its newspaper and on its website. The defendant argued that this would create a chilling effect on the willingness of newspapers and writers to post their articles on the internet, because it would expose them to litigation all round the world, in where they could not be expected to know the relevant law. This argument failed in part because it was contrary to the established law of defamation, including the rule in the Duke of Brunswick’s case.23 However, Kirby J at least, made clear that this result was not wholly satisfactory but that he considered it would be necessary for any change in the law in this area to be made by Parliament.24

2. The Case: Times Newspaper Ltd v the United Kingdom

(a) The Facts

On 8 September and 14 October 1999 the Times published two articles in its newspapers and on its website (“the articles”), alleging that Mr Loutchansky, a Russian businessman, was being investigated for money laundering.

On 6 December 1999 Mr Loutchansky brought proceedings for libel (“the first action”) in respect of the articles against the Times and the journalists responsible. The defendants accepted that the articles were defamatory and did not seek to justify them. Instead they relied on the defence of

17 [2000] 1 WLR 1004 18 ibid. at pp. 1012D, 1024G and 1026H respectively 19 The Law of [1887] at 210 20 ‘Interstate Publication’ [1953] 51 Mich LR 959 21 Robertson & Nicol on Media Law, 4th Ed [2002] at 101 22 [2002] 210 CLR 575 at § 92 23 However, as is argued below, the rule in the Duke of Brunswick’s case can be abolished without affecting the result of this case. 24 [2002] 210 CLR 575 at §§ 164-6

4 qualified privilege, contending that the allegations were of such a kind and such seriousness that they had a duty to publish them and that the public had a corresponding right to know them.

While these proceedings were ongoing the articles remained on the Times’ website, where they were accessible to all internet users as part of the applicant’s archive of past issues. On 6 December 2000 Mr Loutchansky brought a second action for defamation in relation to the continuing internet publication of the articles (“the second action”). Initially, the defendants’ only defence was one of qualified privilege. The two actions were consolidated and set down for a split trial on issues of liability and then quantum.

The articles remained on the internet archive although, on 23 December 2000, the applicant added the following preface to the first article:

“This article is subject to High Court libel litigation between Grigori Loutchansky and Times Newspapers. It should not be reproduced or relied on without reference to Times Newspapers Legal Department.”

In or around March 2001 the defendants applied to re-amend their defence in order “to contend that as a matter of law the only actionable publication of a newspaper article on the internet is that which occurs when the article is first posted on the internet” and that, as a result, the second action was barred by section 4A of the Limitation Act 1980 (‘the single publication rule argument’).

On 19 March 2001 Gray J dismissed this application, relying in particular on the authority of the rule in the Duke of Brunswick’s case25.

On 27 April 2001, Gray J ruled that the defence of qualified privilege was not made out and gave judgment for Mr Loutchansky in relation to the first action, for damages to be assessed26. His judgment reviewed the case-law on qualified privilege and in particular the words of Lord Atkinson in Adam v Ward cited above27. However, he added his own gloss to that test28:

“I take that form of duty, albeit one not owed at law, to be a duty such that a publisher would be open to legitimate criticism if he failed to publish the information in question.”

He then went through each of the factors identified in the Reynolds test and concluded that neither article was protected by qualified privilege. He relied in particular on the cryptic nature and/or unreliability of the defendants’ sources, their failure to verify the story and their failure to take reasonable steps to contact Mr Loutchansky or present his side of the story.

The defendants appealed both rulings. They argued, firstly, that the rule in the Duke of Brunswick’s case breached Article 10, pointing out that, as a result of the rule, newspapers which maintained internet archives were vulnerable to claims in defamation for years and even decades after the initial hard copy and internet publication. They said that this would inevitably have a chilling effect on the willingness of newspapers to provide internet archives and thus would limit their freedom of expression. In its judgment dismissing the appeal, the Court of Appeal stated29:

25 Loutchansky v Times Newspapers and others [2001] EMLR 36 26 Loutchansky v Times Newspapers and others [2001] EMLR 38 27 See footnote 4 above. 28 [2001] EMLR 38 at § 18 29 Loutchansky v Times Newspapers and others [2002] 2 WLR 640 at § 74.

5

“We do not accept that the rule in the Duke of Brunswick imposes a restriction on the readiness to maintain and provide access to archives that amounts to a disproportionate restriction on freedom of expression. We accept that the maintenance of archives, whether in hard copy or on the internet, has a social utility, but consider that the maintenance of archives is a comparatively insignificant aspect of freedom of expression. Archive material is stale news and its publication cannot rank in importance with the dissemination of contemporary material nor do we believe that the law of defamation need inhibit the responsible maintenance of archives. Where it is known that archive material is or may be defamatory, the attachment of an appropriate notice warning against treating it as the truth will normally remove any sting from the material.”

Secondly, in relation to the ruling of 27 April 2001, the defendants argued that Gray J had applied the wrong test for qualified privilege. With some reluctance the Court of Appeal agreed, holding that Gray J’s gloss on the traditional test for qualified privilege was too stringent. They added30:

“There will undoubtedly be occasions when one newspaper would decide to publish and quite properly so, yet a second newspaper, no less properly, would delay or abstain from publication. Not all journalists can be or should be expected to reach an identical view in every case. Responsible journalism will in certain circumstances permit equally of publication or of non- publication.”

The first action was remitted back to Gray J to re-examine his findings of fact with the correct standard in mind. On 26 November 2002, Gray J gave judgment for Mr Loutchansky again, holding that neither of the articles was protected by qualified privilege31. He went through each of the ten Reynolds factors in detail and reached the same conclusion as in his first judgment for very similar reasons.

(b) The Complaints

The Times brought two applications to the European Court of Human Rights, which were subsequently joined. The main complaint in the first application32 was that the Reynolds test failed to satisfy the foreseeability requirement of Article 10 because it was too vague and unpredictable. The Times argued that the factors in the Reynolds test conflated the public’s right to know with the conduct of the publisher (in particular the first and third factors concerning “the seriousness of the allegation” and “the source of the allegation”). It claimed that this led to legal uncertainty and it relied on the criticisms made by Clayton and Tomlinson and by the New Zealand Court of Appeal in Lange v Atkinson.

In the second application33, the Times complained that the application by the Court of Appeal of the rule in the Duke of Brunswick’s case breached its rights under Article 10. It stressed, in particular, the chilling effect of the said rule. They argued that it was a disincentive for newspapers to maintain internet archives because it leaves them vulnerable for an indefinite period in the future to an indefinite number of claims against which they may well, because of the passage of time and the fading of memories, be unable to defend themselves.

30 Ibid. at § 49 31 [2002] EWHC 2490 (QB) 32 No. 23676/03 33 No. 3002/03

6

(c) The Decision

(i) The first application

The Court looked first of all at the object of the Reynolds test. It noted that the Reynolds test was intended to exempt newspapers from their ordinary obligation to verify defamatory factual statements in certain specific circumstances: namely where, taking into account all the relevant circumstances, the newspaper acted in accordance with the standards of “responsible journalism”.

Secondly, the Court recalled its own case law on when journalists should be exempted from their ordinary obligation to verify defamatory factual statements. It noted that, in that case-law, it had taken into account numerous factors, including factors very similar to the impugned factors in Reynolds: for example the nature and degree of the defamation34; the authority of the source35, whether the newspaper had conducted a reasonable amount of research before publication36, whether the newspaper presented the story in a reasonably balanced manner37 and whether the newspaper gave the person defamed the opportunity to defend themselves38.

Thirdly, the Court observed that there had been a reasonable amount of case-law on the application of the Reynolds test which demonstrated certain principles, such as how important it is for the defence of qualified privilege that the source of the information be reliable39, that the person defamed be given the opportunity to defend himself40 and that the story be balanced41.

Finally, the Court held that, on the facts of the particular case as found by the domestic court (i.e. that the Times had published defamatory statements of fact, based on sources that were either unreliable or cryptic without contacting the person being defamed or attempting to put their side of the story) the Times should have realised that it might be considered to have failed to live up to the standard of “responsible journalism”, as described in Reynolds and that, as a result, it would not be able to benefit from the defence of qualified privilege.

(ii) The second application

The Court did not make any decision in relation to the second application. It has communicated that complaint to the Government for their response.

3. Analysis

(i) The Reynolds Test & Article 10

It is submitted that the ECHR was correct to reject the criticisms of the Reynolds test. The decision provides a fairly comprehensive answer to the first criticism made by Clayton and Tomlinson, that the Reynolds test did not give full recognition to the Strasbourg case law. The

34 Bladet Tromsø and Stensaas v. Norway no. 21980/93, § 66, ECHR 1999-III 35 ibid. 36 Prager and Oberschlick v. Austria, judgment of 26 April 1995, Series A no. 313, § 37 37 Bergens Tidende and Others v. Norway, no. 26132/95, § 57, ECHR 2000-IV 38 ibid., § 58 39 GKR Karate v Yorkshire Post, [2000] 1 WLR 2571 40 Gilbert v MGN [2000] EMLR 680 41 Al Fagih v HH Saudi Research & Marketing UK Ltd [2002] EMLR 13

7

Court emphasised the similarities between the Reynolds test and the Strasbourg jurisprudence. It would be surprising if it were otherwise, given that Lord Nicholls formulated the Reynolds test immediately after a thorough examination of the Strasbourg jurisprudence42. Indeed, it could be argued that the Reynolds test is a helpful clarification of the Strasbourg jurisprudence on when defendants should be excused from their ordinary obligation to verify defamatory statements of fact.

The second criticism, that the Reynolds test creates excessive uncertainty, is equally weak. The Times’ submission (echoing the New Zealand Court of Appeal in Lange v Atkinson43) that there needs to be a distinction between the public’s right to know and the conduct of the publisher is plausible at first sight. However, it does not repay further analysis. As the Court pointed out, it is important to consider the context. Qualified privilege is an exceptional defence, necessary only when the defendant is unable to prove the truth of the allegations that he has made. The purpose of this defence is to ensure that in certain situations, where it is in the public interest for allegations to be published even though that they cannot be proved to be true, the publisher should be protected.44

The crucial question, therefore, in any case of qualified privilege (as recognised both by the Court and by the House of Lords in Reynolds45) is whether or not it was in the public interest for the allegation to have been published, notwithstanding that it could not be proved to be true. Central to that question is the conduct of the publisher and the factors identified in the Reynolds test and the Strasbourg jurisprudence: it is unlikely to be in the public interest for a journalist to publish a defamatory statement of fact that he cannot prove about a public figure, without his having made, for example, reasonable efforts to check the veracity of the allegation and/or to present it fairly. If it were otherwise, the law would allow for unchecked harassment of politicians and public figures by the media (which is often controlled by owners with their own agenda); it would probably discourage many eligible people from going into public office; and it might well undermine any respect that the public may have for the media. The truth could quickly be buried under a mass of lies and slander. Malice, being notoriously difficult to prove46, would not be enough to protect against this.

Take an example: a newspaper publishes a story alleging that a well-known politician, X, is a paedophile. This allegation comes from a single, very unreliable source. The newspaper makes no attempt to check up on the source, or to put the allegations to the politician. Instead it leads with a blazing headline saying in emotive terms that X is a paedophile. Even though the story is never proven to be true, it ruins X’s career. X cannot prove malice because the newspaper refuses to reveal its source. Can it really be said, to borrow the Times’ words, that the public had a “right to know” this allegation?47

Even the New Zealand Court of Appeal recognised, in their second judgment in Lange v Atkinson48, that the reasonableness of the defendants’ conduct needs to be taken into account

42 op. cit. 203H – 204F 43 see footnote 12 above. 44 When “the common convenience and welfare of society” call for frank communication on questions of fact, to quote Parke B in Toogood v Spyring [1834] 1 CM & R 181 at 193 45 Reynolds op. cit. at 193H – 194A citing Toogood v Spyring and Davies v Snead [1870] LR 5. 46 Ibid. at 201G 47 Arguably, it is unhelpful, in this context, to talk of the “right to know”. Knowledge implies truth and, by definition, the allegations that require a defence of qualified privilege have not been proven to be true. 48 Lange v Atkinson [2000] 3 NZLR 385. See by W.R.Atkin, Common Law World Review 30.2 (237) June 2001

8 when assessing whether or not a defence of qualified privilege should be available. The court achieved this by broadening the test for malice to include consideration of factors like the steps taken to verify the information and whether the claimant’s side of the story was sought or published. These were the factors from the Reynolds test which it criticised, on the grounds that they should not “fall within the ambit of the inquiry into whether the occasion is privileged”.

In effect, therefore, the test for qualified privilege in Lange is the Reynolds test split into two stages. Whatever other advantages this may have, it adds nothing to the certainty or forseeability of the test for qualified privilege.49 The reality is that some uncertainty is simply inevitable in this area50: a difficult balance needs to be struck between the need for the press to be free and the need for the press to be responsible. As the Court pointed out, will inevitably have to exercise a certain amount of discretion, after having taken into account all the relevant factors. By listing the most important of those factors, the Reynolds test has helped to limit that discretion.

(ii) The rule in the Duke of Brunswick’s case

The Times’ criticisms of the rule in the Duke of Brunswick’s are much stronger. As the Times pointed out, this rule creates a powerful disincentive for newspapers to maintain internet archives because it leaves them vulnerable for an indefinite period in the future to an indefinite number of claims against which they may well, because of the passage of time and the fading of memories, be unable to defend themselves.51

It is true, of course, that this argument has less weight in the Times’ case than it might do in others. The Times itself cannot really complain of having problems with the passage of time and the fading of memories because the second action was brought only just over a year after the original publication at a time when the Times had already been obliged to prepare its defence to the first action.

Nonetheless, the Court can, and it is submitted should, look beyond the facts of the present case to the “chilling effect” of the rule. This has often been the approach of the Court when considering proportionality in the context of Article 10 in the past. Compare for example Cumpana and Mazare v Romania, where the Court considered that a sentence of imprisonment awarded to journalists for a defamatory article was a breach of their rights under Article 10, even though they never served the sentence52. It seems clear that the Court was concerned by the effect that the sentence might have in the future (deterring the media from exercising its right to freedom of expression), rather than the hardship it caused to the individual applicants in the past53. After all, the applicants never served their sentence and the Court accepted that it was not disproportionate for them to have received some sort of criminal conviction.

49 W.R.Atkin argues convincingly, in ‘Defamation Law in New Zealand ‘Refined’ and ‘Amplified’’, cited above, that the Lange formulation has significant disadvantages: in particular, it will be very difficult for the claimant to prove that the defendant behaved unreasonably because they will not have access to the defendant’s sources. He concludes “New Zealand might have been wiser simply to have followed Reynolds.” 50 The split decisions of the House of Lords in Reynolds and the Grand Chamber in Bladet Tromso are just two notable examples of the possibility for disagreement between judges. 51 See Chase Securities Corpn v Donaldson (1945) 325 US 304, 314 and Gregoire v G P Putnam’s Sons (1948) 81 NE 2d 45, 48-49, where the US courts relied on very similar reasons for applying a single publication rule. 52 Cumpana & Mazare v Romania, Application no. 33348/96, § 116, ECHR 2004-... 53 Compare Nikula v. Finland, no. 31611/96, § 54, ECHR 2002-II and Steur v. the Netherlands, no. 39657/98, § 29, ECHR 2003-XI where the Court found that warnings imposed on advocates violated their right to freedom of expression, even though they were purely informal and did not impose any sanction.

9

In any case, the Times did suffer some hardship as a result of the rule in the Duke of Brunswick’s case: it had to incur the costs of defending two actions in relation to the same articles (as opposed to just one).

It is submitted that the Court of Appeal’s reasons for rejecting this argument were weak. It relied on two reasons in particular: first, that internet archives are a “comparatively insignificant aspect of freedom of expression” because the material contained within them is “stale news”; secondly, that the law of defamation need not inhibit the maintenance of archives because “where it is known that archive material is or may be defamatory, the attachment of an appropriate notice warning against treating it as the truth will normally remove any sting from the material.”

As to the first argument, it is submitted that the Court of Appeal grossly underestimated the value of internet archives. These archives provide more than mere “stale news” - they are important sources for education and historical research, particularly as they are readily accessible to the public (more accessible than traditional library archives) and, for the most part, free.

The second argument is equally unconvincing. It is not at all clear that newspapers will be able to protect themselves from litigation by attaching an appropriate notice warning to everything that either is or may be defamatory. It is hard to see how a purely general warning could be said to remove the sting and, given that the number of articles which might be considered defamatory is likely to very large, it will be almost impossible to attach an individual warning notice to each of them. Besides, even if such warning notices do give protection in some cases, even the Court of Appeal seems to accept that they will not give protection in all cases. Consequently, the chilling effect would still exist.

The final point to note is that the rule in the Duke of Brunswick’s case is not reasonably necessary to protect the reputation or rights of others. It could easily be replaced by a single publication rule whereby causes of action in defamation are deemed to accrue at the first moment of publication. If this rule were adopted, defamed persons would still be able to bring actions to maintain and protect their reputation. The only difference would be that, because of section 4A of the Limitation Act 1980 they would have had to bring such an action within one year of the initial publication. The strictness of this rule would be mitigated by the fact that section 32A of the same Act provides exceptions, so that, for example, if a person did not find out that he had been defamed until some time after the initial publication, the limitation period for him to bring a cause of action could be extended.

Such a single publication rule would not have to be identical to the rule in much of the United States. There the single publication rule was extended so that the choice of law to be applied came to be largely affected54, perhaps even determined55 by the proposition that only one action could be brought in respect of the alleged defamation, and that the place of publication was where the person publishing the words had acted. As the majority of the Australian High Court pointed out in Gutnick, there are two separate issues which should not be conflated: the date of accrual of the cause of action and the law to be applied to any action56.

54 See, for example, Hartmann v Time Inc 166 F 2d 127 (3rd Cir 1948); Dale System v Time inc. 116 F Supp 527 at 529-30 (DC Conn 1953). 55 Restatement of Conflict of , 2d, [1971] § 150. See also Zuck v Interstate Publishing Corp 317 F 2d 727 at 734 (2nd Circ 1963). 56 Op. cit. at § 35. See also ‘The Single Publication Rule in Libel: A Fiction Misapplied’, Harvard Law Review Vol. 62 [1949] 1041

10

The present approach of the English courts as to choice of law in internet defamation is in line with that of Gutnick57. However, as Kirby J pointed out, that approach is not wholly satisfactory and the Court in Strasbourg may well have to consider a complaint in relation to it. For example, a US citizen successfully sued in an English court for a remark which he made in the United States but which was reported in England might have a reasonable argument that this breached his right to freedom of expression.58 He could argue that the interference was not ‘prescribed by law’ in that, as a US citizen (where the law of defamation is quite different59), he would not have had “an indication that [was] adequate in the circumstances of the legal rules applicable.”60

However, the Court does not need to decide this issue in the present case. Instead, it can just find that the application of the rule in the Duke of Brunswick’s case violated the Times’ rights under Article 10 because it penalised the Times for maintaining internet archives, because it deters the Times from maintaining such archives in the future, and because it is not reasonably necessary for the protection of the reputation and rights of others.

57 See e.g. Lewis & ors v King [2004] EWCA Civ 1329 58 Similar cases have been brought, for example Lewis & ors v King, cited above, and Richardson v Schwarzenegger & ors [2004] EWHC 2422. The claimant in Lewis & ors v King was Don King, the US boxing promoter, who was suing Judd Burstein for remarks he made about Mr King on a couple of boxing websites. The High Court and Court of Appeal held that England was an appropriate forum for the case. The claimant in Richardson v Schwarzenegger is a British television host who accused Arnold Schwarzenegger of touching her breast. In the course of his election campaign, Mr Schwarzenegger’s publicist alleged that it was Ms Richardson who had behaved provocatively and that Ms Richardson had concocted her story. This allegation was reported in the Los Angeles Times. Hard copies of this paper were published in England and the article was also posted on the internet. The High Court refused to set aside the Master’s order giving permission to serve the second defendant outside the . 59 The US law of defamation is far less favourable to claimants. See New York Times v Sullivan 376 US 254. 60 See Sunday Times v the United Kingdom (no. 1), judgment of 26 April 1979, Series A no. 30, p. 31, § 49, where this is said to be one of the two requirements that flow from the expression ‘prescribed by law’.

11