Rachel Riley v Laura Murray

This is the judgment following the trial of the Claimant’s libel action. The claim concerns a Tweet posted by the Defendant on 3 March 2019. The judgment is divided into the following sections:

The Claimant is a television presenter, probably best known for her appearance on the Channel 4 programme, Countdown. The Claimant was a regular user of Twitter, the social media platform. In March 2019, she had some 625,000 followers. From September 2018, the Claimant had publicly spoken to condemn what she regarded as the fostering of anti-Semitism in the Labour Party under Jeremy Corbyn.

In March 2019, the Defendant was the Stakeholder Manager for the leader of the Labour Party, at that time, Jeremy Corbyn MP. The Defendant was also a user of Twitter. In March 2019, she had some 7,252 followers.

This libel action concerns a Tweet by the Defendant, but to understand the context it is necessary first to set out two earlier Tweets; one by Owen Jones, The Guardian journalist, and the other by the Claimant.

On 10 January 2019, Owen Jones posted the following message on Twitter referring to an incident in which an egg had been thrown at Nick Griffin, the former leader of the British National Party:

At around 15.30 on 3 March 2019, Jeremy Corbyn was assaulted with an egg whilst he was visiting the Finsbury Park Mosque. A member of Mr Corbyn’s team apparently detained the assailant, and he was subsequently arrested. Media reports of the incident started to appear from around 16.30 on 3 March 2019. At 17.25 The Guardian reported that a man had been arrested for hitting Mr Corbyn with an egg. A party source, quoted in the article, said that Mr Corbyn had not been injured and had continued his visit following the incident.

At 18.16, the Claimant posted the following Tweet (“the Good Advice Tweet”):

The Claimant was referring to the incident involving the assault on Mr Corbyn earlier that afternoon, but only a reader of the Good Advice Tweet who was aware of the assault on Mr Corbyn would have understood the reference from the Claimant’s use of the egg, and a rose, to depict the Labour Party.

One person who did understand this reference in the Good Advice Tweet was the Defendant. At 20:10, the Defendant posted the following Tweet in reply to the Good Advice Tweet:

The Defendant posted two further responses to other Tweets:

At 20.37, she replied to @Problemspartof:

At 20.40, she replied to @Problemspartof and @briantaylor56:

At 21.03, the Defendant posted a further Tweet (“the Defendant’s Tweet”):

The Defendant’s Tweet did not reply to, quote Tweet, or otherwise include (for example by screenshot) the Good Advice Tweet. That is going to be a point of some significance in this case. In her evidence at trial, the Defendant stated that the reason she had not quote-Tweeted or otherwise included the Good Advice Tweet was that she did not want to “drive additional traffic” to the Good Advice Tweet. Whatever the motivation, in practical terms, it meant that the Good Advice Tweet was not immediately available to anyone reading the Defendant’s Tweet. Unless the reader manually searched through the Claimant’s Tweets, s/he would be entirely dependent upon the Defendant’s description of the Good Advice Tweet to understand what the Claimant had said.

The Good Advice Tweet and the Defendant’s Tweet provoked a large number of responses (see further [36] below).

The Defendant relies upon direct responses to, and quote-Tweets of, the Good Advice Tweet to demonstrate, she contends, that a significant number of people interpreted the Good Advice Tweet to mean that Jeremy Corbyn deserved to be attacked because he was a Nazi. Mr McCormick QC identified these responses in two Appendices to the Defendant’s skeleton argument for trial. By way of example, and selected from the immediate responses to the Good Advice Tweet on 3 March 2019, the Defendant relies upon the following:

(@p_m_b): “When you have to resort to this, you’ve lost the debate. @RachelRileyRR condoning violence against the leader of the opposition, a sitting elected MP, because he’s on the left and she’s on the right. Awful. @channel4 should review this behaviour”.

(@pagster57): “Wow! Rachel Riley calling Jeremy Corbyn who’d just been attacked, a Nazi. Are you going to act in this @Channel4? This is dangerous”.

(@StopCityAirport): “This is abhorrent. The suggestion that Corbyn is a Nazi because an egg was thrown at him. This is Hate Speech. This is the subtle nuances that make evil become the norm. Awful. Shameful. Desperate.”

(@CallumMilburn24): “I agree with almost all of the constructive criticism surrounding the Labour Party and Anti-Semitism but this is outrageous. Insinuating Corbyn is a Nazi is losing touch with reality.”

(@RIGIDDOGMA): “Cheering on far-right violence is it?”

(@moroniscarrot): “Both calling the leader of her majesty’s opposition a nazi and celebrating a physical assault on him inspired by such rhetoric? That’s a really good look that. Very moral high ground.”

(@blazerunner): “Riley should face criminal charges for incitement. At the very least she should be cautioned by police concerning her attitude to the Leader of the Opposition”.

(@ClaireTromans): “Imagine being the kind of person to celebrate an almost 70-year-old man being hit in the head. Just imagine”.

(@a_nitak): “Revealing your true colours here Rachel. Condoning violence? Quite disgusting”.

(@TheEmanFifty): “A 69-year-old man gets punched and you condone such actions? Stay away from the crack pipe Riley and stick to the calculator”.

(@beverlydawnrose): “Corbyn’s fought for equality for all people his whole life and was physically assaulted today, while visiting a mosque. You’re applauding his assault & calling him a Nazi. What? You’re reported”.

On the other hand, the Claimant can point to responses to the Good Advice Tweet – usually responding to other comments criticising the Claimant – that show that the reader understood the message it conveyed differently:

(@iainl7): “Nobody’s gloating at Corbyn getting egged. It’s a bad thing that undermines the very serious claims against him. However, an awful lot of people are gloating at Jones’s fetish for political violence coming back to bit him”.

(@GeneralWoundwort): “I didn’t read it as that but as tongue-in-cheek highlighting [Owen Jones’] sanctimony over this. If you don’t want your preferred public figures egged, don’t justify egging public figures you dislike…”

(@DisraeliRascal): “She didn’t label him a nazi… what is it with twitter: she highlighted a clear hypocrisy”.

(@Cameleopardisuk): “The fact that you can’t see how this is a jab at the lefts (sic) hypocrisy its incredible. If its (sic) wrong to throw an egg at someone you agree with its wrong to do it to someone you disagree with – simple.”

(@LBuhmgravy): “How so? Rachel has just used Owen’s remarks about a similar situation! She isn’t throwing the egg or calling Corbyn a Nazi! Careful! With your 2+2 making 5 is it actually you likening Jezza to a Nazi… [and then in response to another reply that has been deleted] Is she? Or is she showing Owen Jones justifying violence from the left not so very long ago?”

(@lucid_leigh): “You’re being intentionally obtuse, her argument is that you shouldn’t throw eggs at anybody. Everyone is entitled to protection of the law and it’s not for Owen Jones to decide who should and shouldn’t be assaulted”.

(@DruPKok): “I think she’s pointing out the hypocrisy of Owen’s position rather than either calling Corbyn a Nazi or supporting egg throwing. Why are Corbyn supporters being wilfully blind [to] Owen Jones’ tweet…?”

(@JamesCleverly) (then Deputy Chairman of the Conservative Party): “Seems a lot of people can’t tell the difference between advocating eggs being thrown at people (which I am not) and me supporting RR in highlighting hypocrisy (which I am). Don’t throw eggs at people, even if you disagree with them (in case you’re not sure of my views on this).”

(@FoundLozzed): “For those who are so wound up, they ‘twang’ when they walk, this reads as ‘Owen Jones is a hypocrite’. Nothing more, nothing less…”

(@GracePetrie): “I don’t believe that Rachel Riley thinks Corbyn is a Nazi. I think the inference she has made is hyperbolic. People will take her at her word on both sides, awful things will be said about her by his defenders and him by hers, and things will get steadily, relentlessly worse”.

(@oddjob_sean1): “Hypocrisy. Dictionary result for hypocrisy. Noun: hypocrisy; plural noun: hypocrites. The practice of claiming to have higher standards or more noble beliefs than is the case. AKA Owen Jones tweet in January Rachel Riley pointed out”.

At 00.17 on 4 March 2019, the Claimant responded to the Defendant’s Tweet by ‘quote-Tweeting’ it:

At 07.38 on 4 March 2019, the Defendant replied to the Claimant’s Tweet:

The Defendant’s Tweet is no longer available. When the Defendant awoke on 4 March 2019, she found she had been sent many abusive messages in response to the Defendant’s Tweet. In her witness statement for trial, she summarised allegations calling her a “stupid little girl”, “racist fat filth”, “vile”, “lying idiot” and “lying bitch”. In response, the Defendant deactivated her Twitter account, at around 9am on 4 March 2019, and, on 15 March 2019 she deleted her Tweet. It is common ground that from the point at which the Defendant suspended her Twitter account, the Defendant’s Tweet would no longer have been available on Twitter (unless it was available as a screenshot posted by another user).

At 10.12 on 4 March 2019, the Claimant re-Tweeted a post by Stephen Pollard, the then editor of the Jewish Chronicle. Mr Pollard had posted a screenshot of the Defendant’s Tweet with the message:

At 16.44 on 4 March 2019, the Jewish Chronicle Tweeted that the Claimant had instructed solicitors to pursue a libel claim against “Corbyn staffer Laura Murray”. In her evidence, the Claimant denied that she had spoken to anyone at the Jewish Chronicle or announced that she was intending to sue the Defendant. She stated that she thought that she had contacted her solicitor to seek advice about the Defendant’s Tweet, but had taken no steps towards proceedings at that stage. The Jewish Chronicle must have obtained the story from another source.

At 16.48, the Claimant re-Tweeted a post which suggested that the Defendant was moving to the complaints team in the Labour Party “to help them clear anti-Semitism cases”, and added the comment:

The deletion of the Defendant’s Tweet means that the Twitter analytics data, which would have given reliable evidence as to the extent of direct publication of the Defendant’s Tweet, is no longer available. It also effectively deleted the replies and quote Tweets that other Twitter users had contributed following publication of the Defendant’s Tweet. (Footnote: ¹) Some replies have nevertheless been found and I will identify some of these later in the judgment.

There has been some criticism of the Defendant for the deletion of the Defendant’s Tweet, including a substantial section in Mr Bennett QC’s written trial submissions. I do not consider that this criticism is justified. Although there was some suggestion, on Twitter, that the Claimant was going to sue the Defendant for libel (see [20] above), a letter of claim was not sent until 28 March 2019. It specifically requested the deletion of the Defendant’s Tweet, and did not require that the Defendant should preserve any documents or more specifically the Twitter analytics data. As a matter of fact, the Defendant’s Tweet had already been deleted by this stage, and with it the analytics data. Mr Bennett QC suggested that, as the analytics data had been destroyed, the Court ought not to “feel constrained in reaching conclusions as to the scale [of publication] by reason of the absence of the [data]” and that the Court ought to draw inferences about the scale of publication.

As I have said, I do not consider that the Defendant is open to any legitimate criticism for the loss of the Twitter analytics data following her deletion of the Tweet. No letter of claim had been received by this point and proceedings had not been commenced. Whilst, in the absence of the analytics data, the Court can draw sensible inferences from evidence it accepts as reliable, Mr Bennett QC appeared to be urging some form of rough-and-ready judicial guesswork as a penalty for the Defendant having caused the loss of the analytics data. I will approach the task of assessing the scale of publication by assessing the evidence I have.

The Claim was commenced on 31 May 2019. As is now common in defamation proceedings, a direction was made for the trial of preliminary issues of (1) the natural and ordinary meaning of the Defendant’s Tweet; (2) whether it conveyed a statement of fact or expression of opinion; and (3) whether the meaning found was defamatory of the Claimant. In a judgment handed down on 24 April 2020 ([2020] EMLR 20), I determined the preliminary issues as follows [25]:

In reaching the conclusion that meaning (2) was a statement of fact, I rejected the Defendant’s contention that the ordinary reasonable reader would have understood the Defendant’s Tweet, in this respect, to be an expression of opinion: her summary or description of the Good Advice Tweet (see [22] and [27]). An application for permission to appeal the determination of the preliminary issues was refused by the Court of Appeal on 27 August 2020. For the purposes of this judgment, and as will become apparent, the finding that the meaning (2) was a statement of fact is of some importance.

In this judgment, I will refer to the meaning in paragraph (2) as “the Factual Allegation” and the meaning in paragraph (3) as “the Opinion”.

On 27 April 2020, the Claimant served Amended Particulars of Claim reflecting the Court’s decision on the preliminary issues.

On 26 May 2020, the Defendant filed her Defence. In summary:

it was denied that publication of the Defendant’s Tweet had caused or was likely to cause serious harm to the Claimant’s reputation, as required by s.1 Defamation Act 2013;

meanings (1) and (2), found to be allegations of fact, were substantially true, and the Defendant had a defence under s.2 Defamation Act 2013;

meaning (3), found to be an expression of opinion, was the Defendant’s honest opinion, and the Defendant had a defence under s.3 Defamation Act 2013;

the Defendant’s Tweet was a publication on a matter of public interest and was protected under s.4 Defamation Act 2013; and

the Defendant’s Tweet was protected by qualified privilege being a reply to the attack made in the Good Advice Tweet.

The Claimant filed a Reply on 24 June 2020. In summary:

in response to the denial that the Claimant could demonstrate serious harm to reputation as a result of publication of the Defendant’s Tweet, in a Schedule to the Reply, the Claimant identified responses to the Defendant’s Tweet that had been posted on Twitter and Facebook.

in answer to the defence of truth, the Claimant stated:

in respect of the honest opinion defence, that (a) the defence could not succeed unless the defence of truth in respect of meaning (2) was made out; and (b) the Defendant did not hold the Opinion when the Defendant’s Tweet was published;

in relation to the public interest defence, that (a) the statement complained of was not and did not form part of a statement on a matter of public interest; and (b) any belief that publication of the Defendant’s Tweet was in the public interest was not reasonable; and

in respect of the reply to attack qualified privilege defence, the defence was not available because the Defendant was under no proper duty, whether as a response to the Good Advice Tweet or otherwise, to publish the Defendant’s Tweet.

The issues I must resolve are as follows:

has the Claimant demonstrated that publication of the Defendant’s Tweet has caused or is likely to cause serious harm to her reputation as required by s.1 Defamation Act 2013? If so,

has the Defendant demonstrated that the Factual Allegation is substantially true under s.2 Defamation Act 2013;

has the Defendant demonstrated that the Opinion is protected as honest opinion under s.3 Defamation Act 2013;

has the Defendant demonstrated that the Defendant’s Tweet was a publication on a matter of public interest under s.4 Defamation Act 2013; and

if the Claimant succeeds on liability, what remedies should she be granted.

Although there are many issues to be resolved in this case, few of them depend upon resolution of any dispute of fact or conflict in the evidence between the witnesses who were called to give evidence at the trial. The Claimant and Defendant gave evidence, and both were cross-examined. Some of the cross-examination of the Claimant concerned matters that have very little to do with the issues I have to decide. There was no suggestion put to either witness that she was not telling the truth in her evidence. I am satisfied that both witnesses were truthful, and both were doing their best to assist the Court in their evidence. Largely, the key evidence is contained in documents. Insofar as I need to resolve any disputes of fact, my conclusions are stated below.

Section 1(1) Defamation Act 2013 provides:

Following the Supreme Court decision in Lachaux -v- Independent Print Ltd [2020] AC 612, I take the summary of the relevant principles from Turley -v- Unite the Union [2019] EWHC 3547 (QB):

The investigation of the actual impact of the Defendant’s Tweet is somewhat hampered by the absence of analytics data from Twitter as to the extent of publication of the Defendant’s Tweet (and posted responses/replies to it). This means that evidence that could have been important on the issue of harm to the Claimant’s reputation is not available.

Nevertheless, the Claimant has identified and relied upon various pieces of evidence, including posts on social media responding to the Defendant’s Tweet, in support of her case on serious harm to reputation:

There are at least two screen shots of the Defendant’s Tweet which show that it had been re-Tweeted at least 1,585 times, liked by 4,932 people and provoked 736 responses. It is not known when these screen shots were taken, but it must have been at some point prior to the deletion of the Defendant’s Tweet on 15 March 2019 (see [18] above).

The evidence as to the period immediately following its posting at 21.03 demonstrates that, by 21.20, the Defendant’s Tweet had been re-Tweeted 91 times and received 208 likes. Another screenshot demonstrates that, in the period of about an hour after the Defendant’s Tweet, there had been 363 re-Tweets, 981 likes and 50 responses. By just after midnight, the Defendant’s Tweet had received 94 responses, 661 re-Tweets and 1,764 likes.

Responses to the Defendant’s Tweet, in the 2 hours after it was posted, included the following:

(from @SouthwoldL10): “Why are the police not questioning this has been?”;

(from @MrsDPTrellis): “Isn’t that criminal hate speech?”;

(from @gangleri2000): “Rachel Riley should be given an ASBO”;

(from @beverlydawnrose): “Today I unblocked Rachel Riley briefly just to report her and would ask as many people as possible to do the same. Applauding a physical attack on Jeremy Corbyn, a lifelong campaigner for equality, should be enough to get her off twitter – I would hope”;

(from @JackRussellsMom): “Rachel Riley needs to be arrested for some of the hate she’s tweeted. She’s disgusting”;

(from @blazerunner): “Riley should face criminal charges for incitement. At the very least she should be cautioned by police concerning her attitude to the Leader of the Opposition”; and

(from DaEqualityBear): “Wower’s (sic) I really disagree with Corbyn, I also think anyone making comments like that Rachel really are dangerous, it’s almost incitement, and it’s dangerous”;

The Defendant’s Tweet was also posted by users of Facebook:

On 3 March 2019, at 22.16, Ian Humphries posted the Defendant’s Tweet with the comment: “Disgusting from Rachel Riley” (“the Ian Humphries Post”). Replies to this post on Facebook included: “There should be an outcry for her to be sacked from Countdown… She should be sacked for ‘bringing Countdown into disrepute’”; “I hope this has been reported”; and “She should be arrested for inciting racial hatred”.

On 3 March 2019, at 22.32, Dilys Hadley shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “I hope that her words were reported to the police as hate crime. And that everyone boycotts her and her programme”.

On 3 March 2019, at 22.54, Mark Strawbridge posted the Defendant’s Tweet without comment. Replies to this post on Facebook included: “So now they are encouraging violence against Corbyn. She should lose her fucking job”; and “More than her fucking job, her liberty too she has saught (sic) to incite violence…”. The responses, however, show that not all readers responded in this way. The first comment inquired: “Did she really say that?”, and later a reader posted: “As vile as Ms Riley is, this is not what she said. She screenshot an Owen Jones tweet from January about eggs being thrown at someone who behaves like a Nazi and her comment was ‘good advice’”.

On 3 March 2019, at 23.02, Leah Levane posted the Defendant’s Tweet, with the comment: “This is appalling. Jeremy Corbyn gets attacked while visiting his local Mosque (it’s the Annual Open Mosque Day in the UK) and this is the response of those who oppose him! I’m never glad when anyone gets assaulted…”.

On 3 March 2019, at 23.34, Jacqueline Walker shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “I had heard the name, but I don’t watch TV so didn’t really know much about her. After Googling her, I am appalled, she’s a big name on BBC. The BBC should sack her, they are fast enough at getting rid of other celebrities who step out of line, and rightly so. The public licence fee pays her wages, if I had a license (sic) I’d be straight onto the BBC about this” (another poster points out that the Claimant in fact appears on Channel Four); “She should be charge with incitement. She is radicalising thugs like the one that attacked the leader of Her Majesty’s Opposition with his fist. The thug that attacked the Labour leader should have the years ahead in jail to consider his actions, but so too should those who incited his… violence”; “Isn’t that incitement to violence?”; “Time for her to be charged with inciting hatred and unrest”; “She should be reported to the police under the prevention of harassment act”; “She should be locked up and the key should be thrown away”; “She should be sacked from her job, horrible little mare!”; and “Stick to shitty Countdown Rachel Riley and take your vile comments with you”.

On 4 March 2019, at 00.21, John Clements shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “She should stick to maths. I for one will never watch either version of Countdown again until she is sacked”; “This bitch needs a visit from the police”; “She should be sacked by channel four”; and “String her up”.

On 4 March 2019, at 01.15, Liz Roberts shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “Riley needs to be sacked from Countdown… That’s not appropriate behaviour. It really shows what kind of woman she really is”.

On 4 March 2019, at 06.42, Janet Field posted the Defendant’s Tweet with the comment: “Disgusting inciteful threatening behaviour from Rachel Riley. Needs police intervention” (“the Janet Field Post”). Replies to this post on Facebook included: “Inciting violence. What a wicked idiot”, “Hope the Police are looking into her”; “Should be sacked immediately”; “Well I hope she’s going to be interviewed by the police for inciting violence”; and “This is surely a tweet to (sic) far for Rachel Riley. Get rid now”.

On 4 March 2019, at 07.33, Kevin Fuller posted the Defendant’s Tweet with the comment: “She is effectively inciting violence using hate speech about our Jezza.” Replies to this post on Facebook included: “Jeremy should file a complaint to the police about her this is inciting violence against him,… what a nasty person she is”; “She might be good at numbers but that’s the extent of her talents… inciting violence is not a good look and she should apologise!!”; “She should be sacked from her position, absolutely vile.”

On 4 March 2019, at 10.05, Mick Shaw shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “Instigating criminal behaviour unlawful please sue”; and “I think she should be taken to court what a vile creature she [has] turned out to be”.

On 4 March 2019, at 10.41, Carolyn Marsden shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “She’s not right in the head!!!”; “Is that not inciting violence and therefore an offence?”; “But [they] won’t she will be on all over the media and they will make excuses for her vile statement and nothing will come from it because they take care of their own”; “Her comments need reporting, They are hate speech!!!”; and “She’s clearly not stupid, which is even more worrying. I could be more forgiving of a genuine idiot, who lacked the intellectual equipment to tell right from wrong”.

On 4 Mar 2019, at 10.38 and 11.09, Jane Wilson shared the Ian Humphries Post on Facebook without comment. Replies to this post on Facebook included: “Inciting violence me thinks!”; “I will find the tweet and report it, if it is still there”; “She really is a piece of work… vile woman and by this tweet is encouraging hatred and violence… she should be sacked”; “Just goes to prove behind a doe-eye sweet smile lies an evil piece of work… it always comes out in the end”; “I saw it, it was disgusting. She has also been reported to Channel 4!”; “Should she keep her job on countdown???”; and “… to think the silly mare even attending University”. Again, at least one reader took a different view, posting: “As much as I despise Rachel Riley, it was actually a retweet from Owen Jones… She didn’t call Jeremy a Nazi. As I said, I don’t like the woman but saying she called Jeremy a Nazi is not true”.

On 4 March 2019, at 15.04, Gigi Camille shared the Ian Humphries Post and commented: “Then she should be prosecuted for incitement”. Replies to this post on Facebook included: “She should be sacked from her job on Countdown”.

On 4 March 2019, at 15.19, Carole Hope shared the Janet Field Post on Facebook without comment. Replies to this post on Facebook included: “Oh she’s not stupid at all and she knows exactly what she is doing and I’m really concerned that channel 4 actually employ someone with such right wind dangerous views”; and “She’s inciting hate, is that no a criminal offence or maybe it’s only if its (sic) against a tory”.

Assessment of whether the Defendant’s Tweet caused serious harm to the Claimant’s reputation is potentially complicated, evidentially, by two particular matters:

Shortly after the Good Advice Tweet appeared, on 3 March 2019, Owen Jones posted on Twitter;

at 19.03, he quote-Tweeted the Good Advice Tweet and made a comment directed at the Claimant:

at 19.08, he Tweeted:

at 19.11, he Tweeted what appears to be a screenshot of the Good Advice Tweet with a response, “Shame it wasn’t a brick”, with the comment:

The Claimant posted a Tweet, at just after midnight on 4 March 2019, denouncing the Defendant’s Tweet as an “appalling distortion” (see [16] above).

Owen Jones has over a million followers on Twitter. As the Claimant stated in her evidence, the effect of Mr Jones’ Tweets, in response to the Good Advice Tweet, was to cause something of a ‘pile on’ (the social media phenomenon of someone highlighting a post or individual in circumstances where that person’s followers are likely then to post criticism of the post/individual). At the relevant time, in matters concerning the Labour Party and particularly support for Mr Corbyn, Mr Jones and the Claimant were likely to be perceived as coming from opposing camps.

In support of the Claimant’s case that the Defendant’s Tweet has caused serious harm to her reputation, Mr Bennett QC relies upon three matters: (1) the seriousness of the defamatory imputations; (2) the scale of publication; and (3) actual evidence of harm to the Claimant’s reputation.

In respect of the seriousness of the allegations, Mr Bennett submits that the allegation that someone has stated publicly that a politician deserved to be violently attacked was very serious. At the time of the Defendant’s Tweet, the murder of the MP Jo Cox, in June 2016, would have very much been a matter of real public concern. Mr Bennett QC submits that the allegation would have carried weight because of the Defendant’s position in the Labour Party. The evidence of the extent of publication on Twitter alone (as demonstrated by the responses, likes and re-Tweets – see [36(i)] above) demonstrates a solid basis on which the Court can infer serious harm to reputation. But there is also the evidence that, shortly after publication, the Defendant’s Tweet is then shared on Facebook, a separate platform. Mr Bennett submits that the reactions of individuals to the Defendant’s Tweet, captured in real time, on Twitter and Facebook, is clear evidence of serious reputational harm and these responses will represent only a fraction of those who would have concluded the same about the Claimant but did not publicly voice their thoughts.

The Defendant accepts that the natural and ordinary meaning of the Defendant’s Tweet, found by the Court, contains serious allegations against the Claimant. But she contends that the Claimant cannot rely upon evidence of reputational harm caused by her decision to ‘quote-Tweet’ (and thereby republish) the Defendant’s Tweet (see [16] above). The Defendant contends that at least some of the responses, likes and re-Tweets of the Defendant’s Tweet will have been caused by the publication of the Claimant’s Tweet at 00.17 on 4 March 2019.

I have deliberately limited the evidence I have set out above, relevant to the issue of serious harm, to restrict it to evidence which I am satisfied can be directly linked to the Defendant’s Tweet. Such is the nature of social media, that very quickly discussion can fragment, with users commenting upon the contributions of others. In part, this is simply the modern version of the well-recognised ‘percolation’ or ‘grapevine’ effect in the spread of defamatory allegations, and some of this is properly to be regarded as the harm to reputation arising from the original publication. Nevertheless, the Court must take care to ensure that the relevant harm to reputation has been caused by the publication complained of.

It is not necessary for me to resolve the Defendant’s argument that she is not responsible for further publication of the Defendant’s Tweet caused by the Claimant’s decision to quote-Tweet it (at 00.17 on 4 March 2019), because I am quite satisfied that the evidence demonstrates that serious harm to the Claimant’s reputation had been caused prior to this point. As s.1 is a threshold issue, once it has been surmounted by a claimant, it is not necessary to consider by what margin. This is a matter, if it is reached, that would be relevant to damages.

Taking the snapshot at midnight following the Defendant’s Tweet, it had received 94 responses, 661 re-Tweets and 1,764 likes. Although not in the league of mainstream media publications, that is evidence of significant publication. It is a reasonably safe assumption that most of those who re-Tweet or like a Tweet agree with its message. As the essential message of the Defendant’s Tweet was the Opinion that the Claimant was “as dangerous as she is stupid”, agreement with this strongly suggests that those who liked or re-Tweeted the Defendant’s Tweet accepted, as true, the Factual Allegation. In my judgment, on its own this evidence would provide a solid basis on which to infer serious harm to reputation, but the Claimant’s evidence goes further and demonstrates actual instances of reputational harm.

The evidence of responses to the Defendant’s Tweet, on both Twitter and Facebook, is set out in [36(iii) and (iv)] above. This is clear, contemporaneous, evidence of the impact of the defamatory sting of the Defendant’s Tweet. The evidence of the Defendant’s Tweet being posted on Facebook shows how effectively, and very quickly, defamatory publications can be republished and the harm to reputation spread. In just over an hour following publication of the Defendant’s Tweet, the Ian Humphries Post published the Defendant’s Tweet on Facebook. Within 24 hours, the Ian Humphries Post was shared, and the Defendant’s Tweet further republished, by nine other Facebook users. At each stage, the evidence supports the conclusion that serious reputational harm was being caused to the Claimant’s reputation. The value of this evidence is not only its cogency on the issue of reputational harm, but also because it is unconnected with the Claimant’s quote-Tweet of the Defendant’s Tweet at 00.17 on 4 March 2019 (see [16] above).

In Turley, I referred to similar evidence produced by the Claimant in the following terms [114(ii)]:

In my judgment, and evaluating the totality of the evidence, the Claimant has demonstrated, as a fact, that the harm caused to her reputation by publication of the Defendant’s Tweet (including the identified republication for which the Defendant is responsible) was serious. The Claimant has satisfied the requirements of s.1 Defamation Act 2013.

The Defendant has advanced a defence of truth in respect of the Factual Allegation (see Section C: [25]-[27] above). In other words, she contends that the allegation that the Claimant had publicly stated in a tweet that Jeremy Corbyn deserved to be violently attacked was (and is), as a matter of fact, substantially true. When considering this defence, it is important to note that the Defendant’s argument that this was an expression of her opinion, was previously rejected by the Court, for the reasons explained in the judgment dated 24 April 2020.

The Defendant’s truth defence, pleaded in the Defence and advanced at trial, is:

that the Good Advice Tweet: “(a) meant; (b) was capable of meaning; (c) was capable of being understood to mean; and/or (d) would have been understood to mean by some or all the people who read it, that Mr Corbyn deserved to be violently attacked” and that the meaning “substantially conveyed” by the Good Advice Tweet was that Mr Corbyn deserved to be violently attacked; and/or

that it was reasonable for a reader of the Good Advice Tweet to understand and/or interpret it to mean that the Claimant was stating that Mr Corbyn deserved to be violently attacked.

Section 2(1) Defamation Act 2013 provides:

Again, there is little dispute as to the key principles that apply to a defence of truth under s.2. Mr McCormick QC referred to my summary of the law from Turley [125]:

There is, however, a substantial dispute as to one legal principle: the proper approach to determining the meaning of the Good Advice Tweet and whether the ‘single meaning rule’ should apply.

Unusually for a defamation claim, the basic facts relevant to the defence of truth are not disputed. What the Claimant said in the Good Advice Tweet is a matter of record. On the issue of truth, both what the Claimant intended to convey in the Good Advice Tweet, and what the Defendant understood it to mean, are not relevant. Although the Claimant was asked some questions on this issue during her cross-examination, the Defendant does not advance any case, under her truth defence, as to the intended meaning of the Good Advice Tweet.

Mr Bennett QC, for the Claimant submits that, in the determination of the Defendant’s truth defence, the meaning of the Good Advice Tweet falls to be construed, objectively, in accordance with the established rules of determining meaning in defamation claims (see e.g. Koutsogiannis -v- The Random House Group Limited [2020] 4 WLR 25 [11]-[17]). He contends that application of those principles would lead the Court to attribute a single meaning to the Good Advice Tweet and that the Court should then determine the defence of truth by reference to the single meaning found. He relies upon the endorsement of the single-meaning rule in Lord Neuberger’s judgment in Oriental Daily Publisher Ltd -v- Ming Pao Holdings Ltd [2013] EMLR 7 [141]-[142].

Mr McCormick QC argues that this would be the wrong approach. He submits that the Court has the task of determining whether the Defendant’s description of what was published by the Claimant in the Good Advice Tweet is substantially true. He relies upon the evidence in responses to the Good Advice Tweet as demonstrating that a significant number of people did interpret the Claimant’s words in substantially the same way that the Defendant did and expressed in the Defendant’s Tweet.

Mr McCormick relies upon criticism of the single meaning rule as artificial (see e.g. Bonnick -v- Morris [2003] 1 AC 300 [20]-[21] per Lord Nicholls and Stocker -v- Stocker [2020] AC 593 [33]-[34] per Lord Reed). Further he argues that the Court of Appeal, in Ajinomoto Sweeteners Europe SAS -v- Asda Stores Ltd [2011] QB, refused to extend the rule to the tort of malicious falsehood, with Rimer LJ noting that “if the single meaning rule did not exist, I doubt if any modern court would invent it, either for defamation or for any other tort”. Subsequently, and following Ajinomoto, the Court of Appeal has held that, for the purposes of malicious falsehood, the question is whether “a substantial number of persons would reasonably have understood the words to have such a meaning”: Cruddas -v- Calvert [2014] EMLR 5 [30] (see also Tinkler -v- Ferguson [2019] EWCA Civ 819 [29]).

Mr McCormick contends that the approach to determining meaning in malicious falsehood claims is now very similar to the test that used to be applied in defamation claims as to the capacity of words to bear a particular meaning, described by Sedley LJ in Berezovsky -v- Forbes Inc [2001] EMLR 45 [16] as a determination: “not what the words mean but what a jury could sensibly think that they meant. Such an exercise is an exercise is generosity, not parsimony”. Relying upon observations in McAlpine -v- Bercow [2013] EWHC 1342 (QB) [7] per Tugendhat J; Monroe -v- Hopkins [2017] 4 WLR 68 [35] per Warby J; and Stocker -v- Stocker [43]-[46] per Lord Kerr, Mr McCormick submits that posts on a medium like Twitter are “particularly susceptible to disagreement between reasonable people as to their meaning”. Fundamentally, he submits that extending the application of the single meaning rule to the determination of a defence of truth would be to use it for a purpose for which it is not suitable and doing so would unduly limit freedom of expression.

The key authority upon which Mr McCormick places substantial reliance is the decision of Haddon-Cave J in Begg -v- BBC [2016] EWHC 2688 (QB).

Begg was a libel claim that arose from an edition of the Sunday Politics television programme broadcast by the BBC. The words complained of are set out in the judgment: [3]. The BBC advanced a defence of justification (truth). At trial, the Judge determined the natural and ordinary meaning of the words complained of in the programme was [73]:

The BBC’s defence of justification included reliance upon various speeches, given by the Claimant. An issue for determination was whether these speeches substantially proved true the meaning that the Claimant had promoted and encouraged religious violence and, more generally, whether he was an extremist Islamic speaker who espoused extremist Islamic positions. The Judge explained his approach to determining the meaning of the speeches relied upon by the BBC as follows (emphasis added with underlining):

The Judge then considered each of the speeches ([134]-[334]), making findings in respect of the meaning of each speech, before finding that BBC’s defence of justification succeeded “overwhelmingly”: [365]-[373].

Based on this authority (particularly the underlined section above), Mr McCormick submits that the Court can, and should, consider whether a section of those who read the Good Advice Tweet would reasonably have understood the words to convey the meaning that the Claimant had publicly stated in a tweet that Jeremy Corbyn deserved to be violently attacked. He argues that, for this purpose, the Court can, and should, consider the evidence of how people did understand the Good Advice Tweet. Mr McCormick acknowledges that the evidence that the Court has – in the form of responses to the Good Advice Tweet – could not be called a “representative sample”, but he suggests that it is reasonable evidence of the range of interpretations that readers put on the Good Advice Tweet. Relying on the principles from the law of passing off, he contends that such evidence is admissible and that, although it is ultimately a matter for the Court’s assessment, the Court can take into account evidence of how words have been understood by “rational readers” (Footnote: ³): see Clark -v- Associated Newspapers Ltd [1998] 1 WLR 1558, 1568C-E per Lightman J.

The starting point, for the assessment of any defence of truth, is the imputation that the defendant seeks to prove as substantially true. This is the defamatory imputation that the Court has found the publication to bear. Together with the extent of publication, it is the measure of the objective harm to the Claimant’s reputation caused by publication of the imputation. A successful defence of truth means that the Court is satisfied, on evidence, that this objective harm to reputation is substantially justified.

In this case, the Court has ruled that the imputation, for these purposes, is the Factual Allegation:

It is important to note that this imputation has three key elements: (1) a public statement in a tweet; (2) that Jeremy Corbyn deserved to be attacked; (3) violently.

Mr Jones, in his original Tweet had stated that if a person did not want eggs thrown at him/her, then “don’t be a Nazi”. The Claimant had responded in the Good Advice Tweet: “Good advice”. The Defendant’s Tweet added the specific elements of violence and that Mr Corbyn “deserved” to be attacked, which are both reflected in the meaning found by the Court.

The parties’ arguments on the Defendant’s truth defence have focused on whether it is appropriate to apply the single meaning rule to the interpretation of the Good Advice Tweet. I have reached the conclusion that, at least on the facts of this case, Mr Bennett’s argument that the single meaning rule should be applied to the Good Advice Tweet, must be rejected, but that Mr McCormick’s submissions must also be rejected.

The starting point is that, notwithstanding criticisms of its artificiality, the single meaning is rightly applied to the objective natural and ordinary meaning of a publication so as to provide a “practical and workable” solution to ascertaining (a) the objective harm done to reputation; and (b) the proper parameters of the defence of truth: see discussion in Berezovsky -v- Forbes [12]-[14].

However, the single meaning rule is not applied (fully or at all) in other areas of defamation where to do so would not be appropriate.

The single meaning rule was not applied directly when considering a Reynolds privilege defence: Bonnick -v- Morris. That is because the issues raised under that defence were different. In a Reynolds defence, the primary consideration was the state of mind of the defendant publisher and the meaning that s/he intended (or at least recognised) would be conveyed by a publication. Depending upon the facts of the case, the objective meaning of the publication might be relevant to the responsibility of the journalist: see discussion in Lord Nicolls’ judgment in Bonnick -v- Morris [24]-[25] and Eady J’s remarks to similar effect in Jameel -v- Wall Street Journal Europe SPRL [2004] EMLR 11 [70].

For similar reasons, the single meaning rule does not apply to any assessment of meaning in the context of a plea of malice: Loveless -v- Earl [1999] EMLR 530, 538-539 per Hirst LJ:

Properly understood, Begg is an example of the Court not applying the single meaning rule, in the context of an assessment of a defence of truth, because it was not appropriate to do so on the facts of that case. As is clear from Haddon-Cave J’s judgment, the question to be resolved at the heart of the BBC’s truth defence was whether the claimant’s speeches, relied upon by the defendant, had promoted and encouraged religious violence.

I do not accept Mr McCormick’s submission that the underlined passage from [61] sets out a general rule applicable to all cases where the meaning of a publication is relevant to a plea of truth. The Judge’s modification of the Jeynes’ principles in [62] demonstrates that his approach was case specific. Whether a finding that at least a section of the publishees reasonably understood the relevant publication to convey a particular meaning will be sufficient to prove the substantial truth of an imputation will be case specific and will depend on the parameters of the relevant truth defence.

The fundamental issue that I must resolve is whether the Defendant has shown that this defamatory imputation found by the Court in the Factual Allegation is substantially true.

It might be thought that the defence of truth has an unpromising start. The Good Advice Tweet, taken at face value, plainly does not state that Jeremy Corbyn deserved to be violently attacked. Whether it does so depends on whether such a meaning is inferred or detected as having been implied (a point that the Defendant’s contemporaneous Tweets demonstrate she clearly appreciated – see [10] above), but it is clear from its terms that the Good Advice Tweet was not meant to be read simply at face-value.

The evidence of reactions to the Good Advice Tweet shows that, whilst some people appeared to understand the Good Advice Tweet in the way described in the Defendant’s Tweet, other people did not. To that extent, the evidence supports the conclusion (which is obvious anyway) that the Good Advice Tweet was open to more than one interpretation and was therefore ambiguous. The Good Advice Tweet could be (and was) read as suggesting that there was an element of hypocrisy/inconsistency in Owen Jones’ original Tweet: that whether the throwing of eggs at politicians was acceptable depended on whether the target’s views were regarded as objectionable. Alternatively, the Good Advice Tweet could be (and was) read in a similar way to the Defendant’s interpretation; that Jeremy Corbyn was a Nazi and he too deserved to be similarly attacked.

The important point is that, in her choice of words, and particularly the decision not to include the Good Advice Tweet in the post, the Defendant’s Tweet removed that important ambiguity. Instead, the Defendant pronounced what the Claimant had said in the Good Advice Tweet as a matter of fact. That decision led to the Defendant’s Tweet being published (and republished) to people who were therefore unaware that what they were being told was only one interpretation of what the Claimant had said in the Good Advice Tweet. As the evidence from the responses to the Defendant’s Tweet demonstrates, most if not all of the people who had read it, took it at face value; that the Claimant had said what she was described by the Defendant as having said.

As I have sought to demonstrate above, the authorities establish that whether the single meaning rule should be applied, very much depends on why the meaning of the words is relevant to the issue to be resolved. Here, it is not appropriate to impose an artificial single meaning on the Good Advice Tweet. To do so, would stifle the very important fact that it was ambiguous. I therefore reject Mr Bennett’s submission that the single meaning rule should be applied.

Nevertheless, the Defendant’s defence of truth fails. What the Defendant stated, as a matter of fact, in the Defendant’s Tweet is not substantially true; it was at best half the story, presented to readers of the Defendant’s Tweet as if it was the full story. Critically, it took away the important fact that what the Good Advice Tweet said was a matter of interpretation or opinion, upon which reasonable views could differ, and replaced it with the Defendant’s unequivocal statement of what it meant as a matter of fact. In doing so, the Defendant’s Tweet was a misrepresentation of what the Claimant had said in the Good Advice Tweet.

The position in which the Defendant finds herself could easily have been avoided. If she had said, in the Defendant’s Tweet, for example, that the Claimant had posted a Tweet which was capable of suggesting, or implied, that Jeremy Corbyn deserved to be violently attacked then she may well have had a viable defence of truth (or honest opinion). But she did not do this. She took upon herself the burden of describing, as a matter of fact, what the Claimant had said and failed because she removed the element of ambiguity. Worse, she added the two elements that the Claimant had stated that Mr Corbyn “deserved to be violently attacked”. By doing so, the Defendant put forward the very worst construction that could be put upon the Good Advice Tweet and stated, as a fact, that this was what the Claimant had said.

These are not trivial differences, or ones that could be excused as small errors of detail, or exaggeration, within the permitted parameters of a defence of truth. There is a significant and material difference, not least in terms of likely harm to reputation, between offering an interpretation of what someone has said, and pronouncing unequivocally the interpretation, as a review of the responses to the Defendant’s Tweet more than demonstrates (see [36(iii)-(iv)] above). If anyone is guilty here of misapplying a single meaning, it is the Defendant.

I accept that Twitter is a fast-moving medium, and the law recognises that this must be reflected in the Court’s assessment of the objective meaning of Tweets. But, beyond that, defamation law makes little accommodation for those who post on Twitter impetuously. For them, a modern reworking of a familiar maxim would be: Tweet in haste, repent at your leisure (as most defamation cases involving Twitter appear to bear out). The Defendant has failed to prove the truth of the Factual Allegation.

The Court determined, as a preliminary issue, that the second defamatory element of the Defendant’s Tweet was the Opinion:

In respect of this part of the defamatory imputation, the Defendant has relied upon a defence of honest opinion under s.3 Defamation Act 2013.

In her Defence, the honest opinion defence is advanced in the following terms:

In her Reply, the Claimant contended:

the honest opinion defence could not succeed if the truth defence to the Factual Allegation failed; it was the express premise for the Opinion; and

the Defendant did not hold the Opinion.

Section 3(1) Defamation Act 2013 provides (so far as is material):

The Defendant has not made any submissions on the principles of law that apply. The Claimant referred me to some passages from Duncan & Neill on Defamation and authorities cited there (5th edition, 2020).

There is no dispute between the parties that s.3(2) is satisfied. The Court determined this at the trial of the preliminary issues.

As to s.3(3), Mr Bennett QC, for the Claimant, submits that the basis of the opinion was not properly indicated in the Defendant’s Tweet. The Defendant cannot rely on the Good Advice Tweet (in paragraph 14.2 of the Defence) as that was only available elsewhere. The reader of the Defendant’s Tweet only had the Defendant’s summary of what the Claimant was alleged to have said. The Defendant’s reliance on her “understanding” of the Good Advice Tweet (paragraph 14.3 of the Defence) is self-serving and impermissible. Mr Bennett submits that an opinion itself cannot itself be based upon an opinion held by the Defendant.

Mr McCormick QC contends that the second condition is met. The Defendant’s Tweet did indicate in general terms the basis of the opinion.

As to the final condition, under s.3(4)(a), Mr Bennett QC accepts that a defendant does not have to demonstrate reasonableness. The subsection reflects the old common law test of “whether any man, however prejudiced and obstinate, could honestly hold the view expressed by the defendant”: Telnikoff -v- Matusevitch [1992] 2 AC 343, 354g per Lord Keith. However, where the factual premise for the opinion stated in the words complained of is not demonstrated to be true, then an honest person could not express the opinion sought to be defended.

Mr McCormick argues that the issue is not determined by the fate of the truth defence to the Factual Allegation. He submits that the third condition requires proof of facts on which an honest person could have held the opinion expressed. All that the Defendant needs to establish is that an honest person could have expressed the Opinion based on the terms of the Good Advice Tweet. The reaction to the Good Advice Tweet more than demonstrates that an honest person could have expressed the Opinion.

It is common ground that s.3(2) is met. In my judgment, so too is s.3(3). The Defendant’s Tweet does indicate, clearly, the basis of the expressed opinion. The issue (at this stage) is not whether the factual premise is right, but whether it was sufficiently indicated. It was.

The real battleground is s.3(4)(a). The authors of Duncan & Neill, in paragraph 13.24, observe the following as to s.3(4)(a) (with footnotes omitted):

In the footnotes, the authors refer to the following paragraphs from my decision in Morgan -v- Associated Newspapers Ltd [2018] EWHC 3960 (QB):

The authors of the Blackstone’s Guide referred to, in paragraph 4.50, state:

Certainly, at common law, on this aspect of the defence of honest opinion, the position was well settled. The opinion had to be based either on facts which were proved true or protected by privilege. If the facts upon which the opinion purported to be founded were not proved to be true, or published on a privileged occasion, the defence of fair comment was not available: Tse Wai Chun -v- Cheng [2001] EMLR 31 [18] per Lord Nicholls. As authority for this proposition, Lord Nicholls relied upon London Artists Ltd -v- Littler Grade Organisation Ltd [1969] 2 QB 375 in which Edmund Davies LJ had stated (at 395c-e):

The requirement, under common law, that a defendant must prove the facts that underpinned an expressed opinion has been accepted by the ECtHR not to be incompatible with Article 10: Alithia Publishing Co Ltd -v- Cyprus (App. No. 17550/03) [69].

My observations in Morgan were obiter; the defence was struck out for other reasons. Here the point calls for decision. If s.3(4)(a) permitted a defendant to succeed with a defence of honest opinion where the expressed premise of a stated opinion was false, then that would represent a fundamental (and radical) departure from the settled position at common law. The facts of this case demonstrate this clearly. The Defendant’s Tweet contained a statement of fact – the Factual Statement – which was the express basis for the stated Opinion. The meaning found by the Court made this clear, by the inclusion of the words “by so doing…” I have found that the Defendant has failed to establish that the factual premise for her opinion – the Factual Statement – was true. At this point, the Defendant’s honest opinion defence simply collapses, for the reasons explained by Edmund Davies LJ in London Artists -v- Littler.

In my judgment, the section quoted from Blackstone’s Guide correctly summarises the position. The subsection has not revolutionised the defence of honest opinion. In line with development of the common law, s.3(4)(a) permits a degree of latitude in the proof of facts upon which an honest person could have held the expressed opinion (and it was this latitude which meant that s.6 Defamation Act 1952 could be repealed without replacement, see paragraph 28 of the Explanatory Note). It does not provide an escape route for defendants who have expressed an opinion upon stated facts they cannot prove to be true.

For these reasons, the Defendant’s honest opinion defence fails.

In light of this, it is not necessary for me to consider the Claimant’s plea that the Defendant did not actually hold the Opinion. However, as the point was not abandoned at trial, in fairness to the Defendant, I should shortly state my conclusion on this point. An allegation under s.3(5) that the Defendant did not hold the opinion is tantamount to a plea of malice, reflecting the position as it was at common law. Malice is a serious allegation. It requires a clear pleading and cogent evidence. The Claimant has advanced neither.

The Defendant was not effectively challenged on her evidence that she did hold the Opinion. Frankly, Mr Bennett QC lacked an evidential basis upon which to do so, and quite properly he did not try. In my judgment, the Defendant did honestly hold the Opinion. However, the basis on which she expressed that opinion was flawed and, for the reasons I have explained, her honest opinion defence has failed. This is no reflection upon the Defendant’s honesty. I have already stated that I am satisfied that the evidence of both the Claimant and Defendant was truthful (see [32] above).

Section 4 of the Defamation Act 2013 provides:

The key authorities on the proper interpretation of s.4 are Economou -v- de Freitas [2019] EMLR 7; Doyle -v- Smith [2019] EMLR 15; Turley -v- Unite the Union [2019] EWHC 3547 (QB); Hijazi -v- Yaxley-Lennon [2021] EMLR 7; and the important Supreme Court decision in Serafin -v- Malkiewicz [2020] 1 WLR 2455. Since the trial in this claim, on 1 July 2021, I handed down judgment in the case of Lachaux -v- Independent Print Ltd & Another [2022] EMLR 2. As the decision applies the existing legal principles, I did not seek any further submissions from the parties, and neither party has sought an opportunity to make any further submissions on the Lachaux decision.

There are three issues for the court to determine under s.4(1):

was the statement complained of, or did it form part of, a statement on a matter of public interest? If so,

did the defendant believe that publishing the statement complained of was in the public interest? If so,

was that belief reasonable?

Economou [87].

The terms of s.4(2) make clear that an assessment of the public interest defence under s.4 requires a consideration of all the circumstances. The defence is not restricted to conventional journalists or publishers but is available “to anyone who published material of public interest in any medium”: Economou [80]. In Economou [110], Sharp LJ explained:

Providing they are not treated as any sort of ‘checklist’, the now familiar ten illustrative factors, identified by Lord Nicholls in Reynolds -v- Times Newspapers Ltd [2001] 2 AC 127, 205, remain potentially relevant when assessing whether a defendant’s belief that publication was in the public interest was objectively reasonable. In Serafin, Lord Wilson traced the legislative history of s.4 through the post-Reynolds authorities in [57] to [59], and observed in [60]:

Mr McCormick has relied upon Warby J’s observations in Barron -v- Vines [2015] EWHC 1161 (QB) as to the importance of not allowing “the law of defamation to stifle political debate”: [45]; and that it is a matter of “high importance to afford political speech protection from the chilling effects which the law of defamation can have”: [59].

Based on the authorities of Flood, Serafin and Turley, Mr McCormick submits that whether a statement was on a matter of public interest is to be interpreted broadly, excluding only matters which are “personal and private”. This is an objective assessment. The Defendant’s Tweet was, he argues, “self-evidently on a matter of public interest”.

Mr Bennett contends that the Defendant’s Tweet was not a meaningful contribution to public discourse and “there is not a public interest in one member (from a particular political faction) of Twitter telling other members of Twitter that her opinion is that the Claimant (who is perceived to be from another faction) is dangerous etc.” He submits that the Defendant’s Tweet was published as part of a ‘pile on’ against the Claimant started by Owen Jones’ Tweet, at 19.03, stating that the Claimant was “in the absolute gutter” (see [37(i)(a)] above).

The Defendant’s evidence, at trial, was that she read the Good Advice Tweet and thought that the Claimant was “endorsing and encouraging” the attack on Jeremy Corbyn; that this was “reckless and irresponsible”, “deliberately provocative” and “sent a dangerous message to the wider world and was a dangerous and stupid thing to do”. Mr McCormick points to the Defendant’s Tweets, in the period 20.10 to 20.40 (see [9]-[10] above), as providing contemporaneous support of what she thought about the Good Advice Tweet. The Defendant monitored reaction to the Good Advice Tweet during the evening and saw several replies which made points similar to the ones she later expressed in the Defendant’s Tweet. Her evidence at trial was that she believed that her Tweet was in the public interest.

In his written submissions before trial, Mr Bennett did not appear to attack the Defendant’s claim that she believed the publication of the Defendant’s Tweet was in the public interest. He reserved most of his fire for the final issue: whether the belief was reasonable. Similarly, when he cross-examined the Defendant, Mr Bennett did not challenge the Defendant on her belief that posting the Defendant’s Tweet was in the public interest.

As with most s.4 defences, this is where the main battleground lies.

In her Defence, the Defendant relies on the following particular matters:

that she reasonably believed that the Good Advice Tweet:

meant and/or would be understood by some (if not most) readers of the tweet to mean that the Claimant was stating that Mr Corbyn deserved to be violently attacked (because he is a Nazi);

publicly encouraged violent attacks against Mr Corbyn, who had already been a target for death threats; and

demonstrated that the Claimant was playing a dangerous and unhealthy role in public life;

that she had made clear that the Defendant’s Tweet was in response and/or was a reaction to a Tweet posted by the Claimant following the attack on Jeremy Corbyn that same day, and that therefore readers could, if they wished, easily identify and review the Good Advice Tweet;

that the Good Advice Tweet conveyed her interpretation of the meaning and effect of the Good Advice Tweet, “rather than being a quotation or verbatim or literal account of the words used by the Claimant” and that her interpretation “is supported by the fluid, hyper-impressionistic way in which readers consume Twitter”; and

that she reasonably believed that it was important that she respond to the Good Advice Tweet by pointing out what she believed to be its lack of justification and dangerous implications.

In her Reply, the Claimant largely contended that the matters relied upon by the Defendant were irrelevant. As to a positive case in rebuttal, the Claimant contended:

In his submissions, Mr McCormick argues that there is no substance to the complaint that the Defendant did not contact the Claimant before publishing the Defendant’s Tweet. First, this is not to be regarded as a requirement of the s.4 defence: Serafin [76]. Second, the Defendant did send a response to the Good Advice Tweet at 20.10 (see [9] above), to which the Claimant did not respond, which substantially contained the same allegation as later conveyed in the Defendant’s Tweet. Third, this was not a case in which it was necessary to attempt to obtain a comment from the subject of the proposed publication. The basic facts – the attack on Jeremy Corbyn and the Good Advice Tweet – were not capable of any dispute.

More broadly, the Defendant relied upon the evidence from several Tweets that demonstrated that other people had reached a similar conclusion as to the meaning of the Good Advice Tweet. The Claimant had been criticised by others, principally Owen Jones, but she had not, until the morning of 4 March 2019, put forward any explanation for what she had meant by the Good Advice Tweet if it was not the interpretation that the Defendant and others had put upon it.

As to criticism that the Defendant should have quote-Tweeted, or otherwise have included, the Good Advice Tweet in the Defendant’s Tweet, Mr McCormick submits that the Defendant’s reason for not doing so – to avoid driving further traffic to the Good Advice Tweet – was not unreasonable and was an exercise of “editorial judgment”, to which the Court should afford respect.

In his written submissions, Mr McCormick’s final paragraph submitted:

The Defendant has demonstrated this element of the s.4 defence. The Claimant’s Good Advice Tweet was published on Twitter, on a public platform, and was intended by the Claimant to be read both by her large number of followers and more widely. The Claimant was well aware that her words, in the Good Advice Tweet, were likely to provoke public comment and engagement. Comment upon, or response to, the Good Advice Tweet was a matter of public interest. It was certainly not a “personal and private” matter.

I am also satisfied that the Defendant has demonstrated this element of the s.4 defence; indeed, her evidence on this point was not challenged by the Claimant.

In my judgment the Defendant has failed to demonstrate that her belief was reasonable.

My conclusion in the meaning judgment was [29]:

The most significant factor, in my assessment, is the failure of the Defendant to include the Good Advice Tweet in the Defendant’s Tweet. I accept the Defendant’s evidence that she did not do so because she did not want to drive further traffic to it, but this cannot be a good enough reason for depriving readers of the Defendant’s Tweet of accurate information about, and the proper context of, the Good Advice Tweet. In Turley, I observed that “it can never be in the public interest for a journalist to misrepresent in an article the information or evidence that s/he has obtained”: [153]. The Defendant is not a journalist, but this fundamental principle applies equally to her or anyone else seeking to avail themselves of a s.4 defence in answer to a defamation claim. A person who misrepresents the material they have is likely to find it difficult to establish that s/he reasonably believed that the resulting inaccurate publication was in the public interest.

That is the position here. Essentially, in the Defendant’s Tweet, the Defendant misrepresented what the Claimant had said in the Good Advice Tweet. I do not accept that the decision not to include the Good Advice Tweet, or accurately to describe what it said, was an exercise in editorial judgment that can excuse this critical failure. The effect was not trivial, it was significant. It prevented readers of the Defendant’s Tweet understanding what the Claimant had actually said and that what the Defendant was offering was her interpretation of it. As the reactions to the Good Advice Tweet demonstrate, some readers might nevertheless have shared the Defendant’s view, but critically, others would not have done.

I do not accept that any real weight can be attached to what is contended to be the “fluid, hyper-impressionistic way in which readers consume Twitter”. If that is advanced as an explanation for the Defendant’s failure to appreciate that the Good Advice Tweet was capable of another interpretation, then that only serves to emphasise the importance of not depriving your readers of the source material you are interpreting. The contemporaneous Tweets of the Defendant (see [10] above) demonstrate an awareness that she was offering an interpretation of the Good Advice Tweet. Within 30 minutes, the Defendant’s Tweet had converted that into the only interpretation.

I broadly accept Mr McCormick’s submissions that this was not a case in which the failure to put the allegation to the subject of the publication would have been particularly significant. Certainly, had the Defendant included the Good Advice Tweet in the Defendant’s Tweet, her s.4 defence would not have failed for want of putting the allegation to the Claimant. But then, had she taken this step, she would almost certainly have succeeded with her honest opinion defence and therefore had no need to rely on a s.4 defence.

As I have already noted (see [12] above), the failure to include the Good Advice Tweet, or otherwise indicate clearly that the Defendant’s Tweet was her deduction, conclusion or inference as to what the Claimant had said in the Good Advice Tweet, has had serious implications for each of the defences relied upon by the Defendant. I reject the Defendant’s submission that my rejection of each of the defences represents an unjustifiable restriction of her freedom of expression. English law provides several safeguards for those who publish defamatory statements. But, consistent with the proper respect for Article 8, which includes protection of reputation, freedom of expression is not an unqualified right. The English law of defamation seeks to strike a fair balance between these competing rights. In addition to the threshold requirements a claimant must surmount (including s.1 Defamation Act 2013), several substantive defences are available.

Warby J’s observations in Barron -v- Vines are a valuable and timely reminder of the importance of political speech, which will always be accorded proper weight, but it is not a trump card where other rights are engaged. Proper respect for freedom of expression under the defamation law in England & Wales is achieved by the established safeguards and available defences, seen as a whole package. In Barron -v- Collins [2015] EWHC 1125 (QB) Warby J explained [54]:

It is not a failure of the law of defamation if a person, by her actions or omissions, deprives herself of these defences. Fundamentally, for the reasons I have explained in this judgment, the Defendant misrepresented what the Claimant had said. That misrepresentation, which could easily have been avoided if the Defendant had included the Good Advice Tweet in her post, has been fundamental in my rejection of each of the substantive defences advanced by the Defendant.

It need not have been so. The evidence in this case provides hundreds of examples of people who were able to – and did – criticise the Claimant for posting the Good Advice Tweet; and, in turn, evidence of others criticising them for failing to understand that the Good Advice Tweet was pointing out hypocrisy. Each reader was free to make up his/her own mind about what the Claimant had said or meant in the Good Advice Tweet and to express their honest opinion on it. Providing it was clearly expressed as opinion, and they were honestly expressing their views, none of those individuals had reason to fear that they might be liable under the law of defamation. The Defendant has fallen foul of the law because, by not setting out the Good Advice Tweet, she instead took on the burden of describing, as a fact, what the Claimant had said in it. In doing so, she materially misrepresented what the Good Advice Tweet said, and, in the process, she defamed the Claimant and left herself with no viable defence.

In her Claim Form and Particulars of Claim, the Claimant originally sought remedies of damages and an injunction. Yet, even by the time of the issue of the Claim Form, on 31 May 2019, the Defendant had already deleted the Tweet and there is no evidence to suggest that, unless an injunction is granted, she threatens to publish the same or any similar allegations. Realistically, in light of this, Mr Bennett QC rightly did not pursue a claim for an injunction at the trial. That leaves the remedy of damages.

In her Particulars of Claim, the Claimant relied on several matters in aggravation of damages:

that the Claimant had campaigned actively against anti-Semitism in the Labour Party;

the Defendant bore animus towards the Claimant because she believed that the Claimant’s campaigning was causing harm to the Labour Party and, particularly, Jeremy Corbyn; and

the Defendant knew that the Claimant had not acted in the manner alleged in the Tweet and she posted the Defendant’s Tweet to undermine the Claimant and limit the damage she was causing to the Labour Party and Mr Corbyn.

In her Defence, the Defendant denied this alleged ‘improper motive’ and denied that she held any animus towards the Claimant. The Defendant also set out matters upon which she wished to rely in mitigation of damages, including:

the “overwhelming majority” of people to whom the Tweet was published will have read it through direct republication by the Claimant (particularly the Tweet at 00.17 on 4 March 2019 – see [16] above and the re-Tweet of Stephen Pollard’s post – see [19] above);

the limited time that the Defendant’s Tweet was published before it was removed from Twitter and subsequently deleted;

on 31 March 2019, the Claimant re-Tweeted a “false, defamatory, and highly damaging allegation” about the Defendant from the anonymous Twitter account “GnasherJew” to the effect that the Defendant had concluded, in her role in the Labour Party, that certain actions of people in the party were not anti-Semitic (“the GnasherJew Tweet”).

Neither party has identified the general legal principles that inform the assessment of an award of damages in a defamation case. This may be because the basic principle is uncontroversial. A successful libel claimant is entitled to recover, as general compensatory damages, such sum as will compensate him or her for the wrong he has suffered. I can take Warby J’s summary of the approach from Barron -v- Vines [2016] EWHC 1226 (QB):

Mr McCormick QC relies on the principle, from Pamplin -v- Express Newspaper Ltd [1988] 1 WLR 116, 120, that damages may be reduced (even to a nominal sum) taking into account any partial justification. Relying upon Flymenow Ltd -v- Quick Air Jet Charter Gmbh [2016] EWHC 3197 (QB), he also contends that culpable conduct of the claimant which provokes publication of the words complained of may also have the effect of reducing damages (again even to a nominal amount).

Counsel have also referred to the Court of Appeal’s decision in Campbell -v- News Group Newspaper Ltd [2002] EMLR 43. This was the defendant’s appeal against a jury’s award of £350,000 damages in a libel action. The claimant appeared in the Court of Appeal in person. As noted in the headnote, the claimant had been involved in serious misconduct up to and during the trial, including procuring his son and his son’s partner to make false statements, making a serious attack on the honesty of an innocent third party and seeking to implicate two solicitors in the most serious professional misconduct. The appeal was allowed and an award of £30,000 was substituted for the jury’s award. The judgment in the Court of Appeal was given by Schiemann LJ and included the following paragraphs on the issue of the Claimant’s conduct and mitigation of damages:

I have set out above (see [123] above) my assessment of the seriousness of the allegation. It is not suggested that it is anywhere close to the gravest allegations that can be made.

Following the deletion of the Defendant’s Tweet and the loss of the Twitter analytics data, reliable evidence of the extent of direct publication of the Defendant’s Tweet is not available. Nevertheless, there is reliable evidence that establishes the minimum number of publishees was at least 4,932 (the number of likes captured on a screenshot of the Defendant’s Tweet – see [36(i)] above). To that must be added:

the further direct republication of the Defendant’s Tweet both on Twitter (by at least 1,585 re-Tweets) and on Facebook (see [36(iv)] above) for which the Defendant is responsible as representing the natural, probable and foreseeable consequence of the original publication; and

the likely grapevine/percolation effect, the capacity for which is significantly increased on social media due to the ease with which posts can be quickly and easily disseminated.

Excluding republication of the Tweet caused by the Claimant’s own re-Tweets of it, and doing the best I can on the evidence I have, in my judgment it is likely that the Defendant’s Tweet was published to between 10,000 and 15,000 people. That is of an equivalent scale to the likely readership of the print edition of a local regional newspaper.

I have been satisfied, on the evidence, that the Claimant has demonstrated that the publication of the Defendant’s Tweet has caused serious harm to the reputation of the Claimant (see Section G: [33]-[47] above). But this is a threshold issue, and I must consider also consider any particular evidence of reputational harm.

The Claimant gave some evidence of adverse consequences which she attributed to the Defendant’s Tweet. She complained that she had received a large amount of abuse, which was of a level of seriousness that she was visited at home by police officers. The Claimant’s evidence was that the abuse and threatening messages that she received were all-consuming and that she found it difficult to concentrate on her work and led to sleepless nights. The Claimant also became the target of a concerted campaign to get her dismissed from her job with Channel 4 on the grounds that she had advocated violence. The Claimant had an “uncomfortable” meeting, at her instigation, with the Head of Daytime Programming at Channel 4 on 7 March 2019, to discuss complaints received about the Good Advice Tweet. The Claimant states in her evidence that, although Channel 4 accepted her explanation that the Defendant had misrepresented what she had said in the Good Advice Tweet, she nevertheless felt vulnerable.

As the Claimant recognises, however, it is very difficult to demonstrate that these consequences were caused by the Defendant’s Tweet, as opposed to other factors. First, and perhaps of most significance, criticism of the Claimant for the Good Advice Tweet had been led by Owen Jones (see [36(i)] above) before the Defendant’s Tweet was posted, and he had vastly more followers than the Defendant. Not all of the blame (or even most of it) for the way in which the Good Advice Tweet was interpreted can be laid at the feet of the Defendant. The social media ‘pile on’ was largely caused by Owen Jones’ posts; the Defendant had a very minor role by comparison.

Mr McCormick’s main submission, on the issue of damages is that the posting of the Good Advice Tweet was “profoundly irresponsible”. Relying principally on Flyemenow, he contends that the Claimant’s conduct “is such that only contemptuous damage could ever be appropriate”.

Mr Bennett contended that the Claimant is entitled to substantial damages arising from the seriousness of the allegation and the extent of publication. He submits that the Defendant cannot rely upon the re-Tweet by the Claimant of the GnasherJew Tweet (see [134(iii)] above) because she could have brought a Part 20 claim for defamation if she contended that she had been defamed by the Claimant.

This case is unusual. It turns, largely, on two Tweets: the Good Advice Tweet and the Defendant’s Tweet. I have found that the publication of the Defendant’s Tweet has caused serious harm to the Claimant’s reputation, and I have rejected the Defendant’s defences. Having established the elements of the tort of libel, the Claimant is therefore entitled to a sum in damages.

I reject the Claimant’s case that the Defendant was motivated by any improper purpose in posting the Defendant’s Tweet. I am satisfied that the Defendant acted honestly (see also [102] above). She made a mistake in the Defendant’s Tweet by not including the Good Advice Tweet.

I reject the submission that the GnasherJew Tweet has any mitigating effect on damages. Even after cross-examination of the Claimant on this point, I am left uncertain as to what, precisely, is being alleged as the act of misconduct. There is force in Mr Bennett QC’s submission that, if the GnasherJew Tweet is alleged to be false and defamatory, the proper way of having that issue resolved was to bring a counterclaim in respect of it. Pleading it as an alleged act of misconduct, relevant to damages, risks the introduction of satellite disputes of fact which the law relating to evidence admissible in mitigation of damages strives to keep within clear and proper bounds. Whatever wrongdoing is alleged from the GnasherJew Tweet, it is a long way from the sort of the sort of “discreditable conduct” of a party during the proceedings which would justify a reduction of damages under the principles in Campbell.

Mr McCormick QC relied upon the Flymenow case. In that case, the claimant brought a libel claim over a notice published by the defendant that suggested that the claimant was insolvent, in the sense that it was unable to pay its bills when due. Following a trial, the defendant’s defences, including a defence of truth, were dismissed. The unusual feature of the case was the fact that the claimant had been responsible for giving the defendant the impression that it was insolvent by failing to pay monies owed to it. Warby J found that the claimant “deliberately failed to discharge it debts knowing that they were due and payable [and]… repeatedly lied to [the defendant] in order to fob him off”: [94]. Although the Judge rejected the defendant’s defences, he awarded only £10 damages. He explained his reasons as follows:

In Joseph -v- Spiller, Tugendhat J found for the claimant in a libel action but awarded only nominal damages, of 1 pence, because the claimant had fabricated part of the claim for special damages.

The full passage from Gatley, referred to by Warby J, under the heading “Claimant’s own conduct: relevance in assessment of damages”, is as follows (with footnotes omitted):

The authority cited for the final sentence was Joseph -v- Spiller, and for the preceding sentence the decision of Morland J in Godfrey -v- Demon Internet Ltd (No.2) (unreported, 23 April 1999). That was a judgment on an application for permission to amend to add particulars of alleged conduct of the claimant – consisting of his postings online – in mitigation of any damages. The Judge noted the paucity of authority on the issue of a claimant’s conduct being relied upon in mitigation of damages, but identified dicta of Lord Radcliffe in Dingle -v- Associated Newspapers Ltd [1964] AC 371, 395 and Lord Hailsham LC in Broome -v- Cassell & Co [1972] AC 1027, 1071 where the latter said:

Based on these authorities, the Judge granted permission for the amendments and observed:

Godfrey -v- Demon was considered by the Court of Appeal in Burstein -v- Times Newspaper Ltd [2001] 1 WLR 579. May LJ said this about the decision:

Although posting the Good Advice Tweet could not be described as “bad conduct” of the Claimant, it properly falls to be characterised as provocative, even mischievous. It was calculated to provoke a reaction and it did. As the Claimant would readily have appreciated, the words she used in the Good Advice Tweet invited and required interpretation; read only in their literal sense they were meaningless. As explained in more detail above, there were two obvious meanings: the hypocrisy meaning or the meaning that suggested that Jeremy Corbyn deserved to be egged because of his political views. I am quite satisfied, on the evidence, that the Claimant was aware that the Good Advice Tweet was capable of being read in both senses. She may have intended the first, but she was certainly not blind to the second. It is telling that, in her re-Tweet at 16.48 on 4 March 2019, the Claimant suggested that she had been called “dangerous and stupid for talking about antisemitism yesterday”. That was a reference back to the Defendant’s Tweet and the Good Advice Tweet. But the Good Advice Tweet could only sensibly be regarded as concerning anti-Semitism if it was construed in the meaning of Jeremy Corbyn deserving to be egged because of his political views.

What impact should this aspect have on the damages award? This case has none of the additional features identified by Warby J in Flymenow [125] and it is not a case of the Claimant having abused the process of the Court or fabricated evidence, as in Joseph -v- Spiller. It would not be right to award the Claimant only nominal damages.

In my judgment, however, there is a clear element of provocation in the Good Advice Tweet, in the sense that the Claimant must have readily appreciated that the meaning of the Good Advice Tweet was ambiguous and could be read as suggesting, at least, that Jeremy Corbyn deserved to be egged because of his political views. In the context of her own high-profile campaign against anti-Semitism in the Labour Party, the risk of the Good Advice Tweet being read in that way was obvious. In that respect, the Claimant can hardly be surprised – and she can hardly complain – that the Good Advice Tweet provoked the reaction it did, including the Defendant’s Tweet. Those are matters which are properly to be taken into account when fixing the appropriate award of damages, whether under the ‘provocation’ principle or, if necessary, under Burstein as being matters directly relevant to the contextual background in which the Defendant’s Tweet came to be published.

This is not a case in which the damages award has an important role to play in vindicating the Claimant’s reputation. The Good Advice Tweet was ambiguous. The Defendant’s Tweet misrepresented it. This judgment, and the reporting of it, rather than reports of a simple figure of compensation, will make clear the vindication to which the Claimant is entitled.

Reflecting those factors, together with the seriousness of the allegation and the extent of its publication, I consider that the appropriate sum to award in damages is £10,000.