ON APPEAL FROM THE HIGH COURT OF JUSTICE
KING’S BENCH DIVISION
MEDIA & COMMUNICATIONS LIST
Mr Justice Nicklin
Royal Courts of Justice
Strand, London, WC2A 2LL
Before :
LADY JUSTICE NICOLA DAVIES
LORD JUSTICE ARNOLD
and
LORD JUSTICE WARBY
Between :
(1) SIMON BLAKE (2) COLIN SEYMOUR (3) NICOLA THORP | Claimants/ Respondents |
- and – | |
LAURENCE FOX | Defendant/ Appellant |
Greg Callus (instructed by Gateley Legal) for the Appellant
Heather Rogers KC and Beth Grossman (instructed by Patron Law Limited) for the Respondents
Hearing date: 15 August 2023
Approved Judgment
This judgment was handed down remotely at 10.30am on 25 August 2023 by circulation to the parties or their representatives by e-mail and by release to the National Archives
.............................
LORD JUSTICE WARBY :
Introduction
On 4 October 2020 three individuals posted tweets about the defendant which used the word “racist”. He responded by tweeting about each of them using the word “paedophile”. They sued him for libel. He counterclaimed. There was a trial of preliminary issues before Nicklin J (“the judge”). This is an appeal against decisions he made on three of those issues: the natural and ordinary meanings of the tweets complained of; whether they were statements of fact or opinion; and, in the case of the defendant’s tweets, whether they were defamatory at common law.
The essential facts
The judge set these out clearly and succinctly in his earlier reserved judgment on mode of trial, [2022] EWHC 1124 (QB), [2022] 4 WLR 77, on which the following is based.
The first claimant (“Mr Blake”) was a trustee of Stonewall until 30 June 2021. The second claimant (“Mr Seymour”) is an entertainer who has appeared in various television programmes, including the first season of Ru Paul’s Drag Race UK. The third claimant (“Ms Thorp”) is an actor, television commentator and writer. She has appeared in the ITV drama Coronation Street, and other television programmes. The defendant (“Mr Fox”) is an actor, perhaps best known for his portrayal of the character “Hathaway” in the ITV drama Lewis between 2006-2015. Each of the parties is active on the social media platform Twitter.
On 1 October 2020, the supermarket chain Sainsbury’s Plc (“Sainsbury’s”)”.published two tweets on its Twitter account @sainsburys.
The first tweet, at 10.11, displayed a graphic “Celebrating Black HistoryMonth” with the words:
“We are Celebrating Black History Month this October. For more information visit [website link given]. #blackhistorymonth”
The hyperlink included in this tweet linked to a page on Sainsbury’swebsite which was headed: “Celebrating Black History Month”. Under a sub-heading, “What we have been doing to support our colleagues”, Sainsbury’s included: “Recently we provided our black colleagues with a safe space to gather in response to the Black Lives Matters movement” (“the Sainsbury’s Website BLM Statement”).
The second tweet, at 15.22, contained a graphic with the words:
“We are proud to celebrate Black History Month together with our Black colleagues, customers and communities and we will not tolerate racism.
We proudly represent and serve our diverse society and anyone who does not want to shop with an inclusive retailer is welcome to shop elsewhere.”
On 4 October 2020, Mr Fox published the following tweet from his Twitter account (@LozzaFox) (“Mr Fox’s Sainsbury’s tweet”):
“Dear @sainsburys
I won’t be shopping in your supermarket ever again whilst you promote racial segregation and discrimination.
I sincerely hope others join me. RT.
Further reading here [website link given]”
Mr Fox’s case is that the link given in his tweet was to the Sainsbury’s Website BLM Statement. Mr Fox’s tweet quote-tweeted the second Sainsbury’s tweet.
Later on 4 October 2020, each of the claimants posted a tweet.
At 16.45, Ms Thorp tweeted:
“Any company giving future employment to Laurence Fox, or providing him with a platform, does so with the complete knowledge that he is unequivocally, publicly and undeniably a racist. And they should probably re-read their own statements of ‘solidarity’ with the black community.”
at 17.11, Mr Blake quote-tweeted Mr Fox’s Sainsbury’s tweet and said:
“What a mess. What a racist twat.”
at 17.19, Mr Seymour quote-tweeted Mr Fox’s Sainsbury’s tweet and said:
“Imagine being this proud of being a racist! So cringe. Total snowflake behaviour.”
A little later on 4 October 2020, in response to the claimants’ tweets set out in the previous paragraph, Mr Fox posted three tweets in each of which he quote-tweeted the relevant tweet and added the word paedophile (“the paedophile tweets”). The words he used were:
at 17.29, in response to Mr Blake’s tweet:
“Pretty rich coming from a paedophile.”
At 17.30, in response to Mr Seymour’s tweet:
“Says the paedophile.”
At 17.51, in response to Ms Thorp’s tweet:
“Hey @nicolathorp
Any company giving future employment to Nicola Thorpe (sic) or providing her with a platform does so with the complete knowledge that she is unequivocally, publicly and undeniably a paedophile.”
At 18.24 on 4 October 2020, Mr Fox posted the following tweet:
“Language is powerful. To accuse someone of racism without any evidence whatsoever to back up that accusation is a deep slander. It carries the same stigma and reputation destroying harm as accusing someone of paedophilia. Here endeth the lesson.”
At some point during the morning of 5 October 2020, Mr Fox deleted all the paedophile tweets.
The pleaded cases
It was the claimants who started the legal action, but for clarity and convenience I shall consider the publications chronologically, starting with the claimants’ own tweets.
Mr Fox’s pleaded case was that these all contained an allegation of fact that he was a racist. His defence and counterclaim (“DAC”) alleged that the tweets of Mr Blake and Mr Seymour “each meant and was understood to mean that the defendant was a racist” (paragraph 77) and that Ms Thorp’s tweet “meant that the defendant was unequivocally and undeniably a racist” (paragraph 78). Paragraph 79 of Mr Fox’s DAC said this:
“Although ‘racist’ is an ordinary English word requiring no definition, for the avoidance of any doubt it means someone who is hostile to people of different ethnicities, races or skin colours; and/or who believes that some racial or ethnic groups, or people with certain skin colours, are inferior to others; and/or who believes that people should be segregated based on their racial or ethnic origins or the colour of their skin …”
The claimants pleaded that their tweets were all expressions of opinion about what Mr Fox’s public statements had shown him to be. In their statement of case on the issues for resolution at the trial of preliminary issues they pleaded their cases on meaning in this way:
Mr Blake contended that the meaning of his tweet was that: “the defendant’s latest Tweet about Sainsbury’s was a ‘mess’ and showed that he was a ‘Racist twat’”.
Mr Seymour contended that his tweet meant that: “the defendant’s response to the action taken by Sainsbury’s was cringeworthy and showed him to be a racist”.
Ms Thorp argued that the meaning of her tweet was that: “the defendant’s public statements, including his response to Sainsbury’s, showed him to be unequivocally, publicly and undeniably a racist”.
In response to paragraph 79 of the DAC, the claimants’ primary case was that “… if and insofar as any of their tweets bore the meaning that the defendant was ‘a racist’ the natural and ordinary meaning would be in that form, using the word (‘racist’) without any further definition”. They too maintained that “‘racist’ is an ordinary English word that requires no further definition.”
The claimants went on to deny that their tweets bore any of the specific meanings set out in DAC 79. They said that this was a “reductive and inapposite” approach to the meaning of the term “racist”. Although the word “might encompass” the matters identified by Mr Fox that “is not the natural or most likely interpretation of the word today”. It “covers a broad spectrum of conduct (including speech) and attitudes in relation to racial or ethnic issues”. Examples were given, including failing to take racism seriously and mocking, or demonstrating insensitivity to, or showing disrespect for, the experience and concerns of black people and people of colour. The claimants suggested that if (contrary to their primary case) the term “racist” as used by them required any further definition, it meant that Mr Fox was a racist in one of these ways.
As for Mr Fox’s paedophile tweets:
Each of the claimants maintained that the relevant tweet contained an allegation of fact that he or she
“was a paedophile, who had a sexual interest in children, and had (or was likely to have) engaged in sexual acts with or involving children, such acts amounting to serious criminal offences.”
Mr Fox denied that the paedophile tweets were defamatory of any of the claimants. His pleaded case is that these tweets would not have been understood literally but rather as “tit-for-tat vulgar abuse”. He pleaded that all readers of each of his tweets would have been aware that it was made in direct response to an allegation of racism against him by the particular claimant; that there was no apparent cause or reason for that claimant to allege that he was a racist; and that he was retaliating by calling that claimant a “paedophile”. Mr Fox maintained that a reader of the paedophile tweets would have understood that he:-
“was making … the rhetorical point that it was wrong to throw around seriously defamatory allegations on Twitter without any factual foundation and that the Defendant was giving the Claimants a taste of their own medicine (accusing a serious but outlandish term which if a true allegation would not be made in these terms).”
Relevant legal principles
These are not controversial. For present purposes the essentials can be quite shortly stated.
In most defamation claims, the first key issue is the natural and ordinary meaning of the words complained of. This is defined as the single meaning the words would convey to the hypothetical ordinary reasonable reader. That meaning is to be determined objectively by reference to the words themselves. No other evidence is admissible. The author’s intention is irrelevant as is evidence about the meaning that readers actually took from the statement complained of. But the medium of expression and the context in which the words complained of appear are both important.
Judges must seek to place themselves in the position of a reader who is neither avid for scandal nor unduly naïve. They should beware of over-elaborate analysis, especially when dealing with postings on social media such as Twitter, which are “in the nature of conversation rather than carefully chosen expression”. The meaning that an ordinary reasonable reader will receive from a tweet is likely to be “more impressionistic than, say, from a newspaper article” and “the essential message that is being conveyed by a tweet is likely to be absorbed quickly by the reader”. Judges should have regard to the impression the words make upon them. They can take judicial notice of particular characteristics of a given readership if these are matters of common knowledge but should beware of impressionistic assessments of those characteristics. The correct approach, and the established practice, for a judge deciding meaning at first instance is to read or watch the offending publication to capture an initial reaction before reading or hearing argument.
These points, and a fuller account of the principles governing the determination of meaning, are all to be found in Koutsogiannis v The Random House Group Ltd [2019] EWHC 48 (QB),[2020] 4 WLR 25 [12] (Nicklin J), Stocker v Stocker [2019] UKSC 17,[2020] AC 393 and Millett v Corbyn [2021] EWCA Civ 567, [2021] EMLR 19 [8]-[9]. The words I have quoted are from Stocker [43]-[44] (Lord Kerr, with whom the other Justices agreed).
The second issue in this appeal is one that often arises in a defamation claim: whether the words complained of are an allegation of fact or a statement of opinion. Section 3 of the Defamation Act 2013 provides for a defence of “honest opinion” which is relatively generous. But the first condition for the availability of this defence is that the statement was one of opinion: see s 3(2) of the 2013 Act. A statement will only be defensible under s 3, therefore, if it is recognisable as a comment or opinion as distinct from an imputation of fact. If it is not, the defendant will need to prove that it is substantially true (s 2 of the 2013 Act) or that it was a reasonable publication on a matter of public interest (s 4 of the Act).
Opinion is synonymous with “comment”. It is something which is or can reasonably be inferred to be a deduction, inference, conclusion, criticism, remark, observation or the like. As with meaning, the court deciding whether a statement is one of fact or opinion looks only at the words complained of and their immediate context, and the ultimate question for the court is the objective question of “how the words would strike the ordinary reasonable reader”. This question may be considered after the meaning has been decided, or at the same time, or in the reverse order, which is common practice.
This is a highly fact-sensitive process that focuses on the particular statement at issue. One factor for consideration is whether the statement contains any indication of the basis on which it is made. At common law a statement that contains no indication of or reference to any supporting facts is liable to be treated as a statement of fact. The second condition for the statutory defence of honest opinion is “that the statement complained of indicated whether in general or specific terms the basis of the opinion”: s 3(3) of the 2013 Act. Beyond these extreme cases, “[t]he more clearly a statement indicates that it is based on some extraneous material, the more likely it is to strike the reader as an expression of opinion”.
Again, the principles are set out more fully in the judgment of Nicklin J in Koutsogiannis at [16]-[17] and Millett v Corbyn [13], [19]. Reference may also be made to Telnikoff v Matusevich [1992] AC 343, 352 (which deals with context), Joseph v Spiller [2010] UKSC 53, [2011] 1 AC 852 [88]-[89] (Lord Phillips), the judgment of Sharp LJ in Butt v Secretary of State for the Home Department [2019] EWCA Civ 933, [2019] EMLR 23 at [32]-[33] and [39] (which is the source of the quotation at [23] above), and my judgment in Triplark Ltd v Northwood Hall (Freehold) Ltd [2019] EWHC 3494 (QB) [17] (from which the quotation at the end of [24] above is taken).
The third relevant aspect of the law is the common law on what is defamatory. In short, a statement is defamatory if it (a) attributes to the claimant behaviour or views that are “contrary to common shared views of our society” and (b) would tend to have a “substantially adverse effect” on the way that people would treat the claimant: see Millett v Corbyn at [9].
A fourth relevant aspect of defamation law is the principle that “mere vulgar abuse” is not actionable. The law is summarised in Gatley on Libel and Slander 13th ed at para 3-037:
“Insults or abuse which convey no defamatory imputation are not actionable as defamation. Even if the words, taken literally and out of context, might be defamatory, the circumstances in which they are uttered may make it plain to the hearers that they cannot regard it as reflecting on the claimant’s character so as to affect his reputation because they are spoken in the ‘heat of passion, or accompanied by a number of non-actionable, but scurrilous epithets, e.g. a blackguard, rascal, scoundrel, villain, etc.’ for the ‘manner in which the words were pronounced may explain the meaning of the words.’”
This can be seen as a logical consequence of the law’s concentration on the impact a statement would have on the ordinary reasonable reader and the way they would treat the claimant, and a reflection of the importance attributed to context and medium.
The judgment
The judge referred back to his mode of trial judgment for details of the tweets complained of and the circumstances in which they came to be published. He summarised the pleaded cases and the procedural background to the trial. He set out the applicable legal principles in a way that has attracted no criticism, and then outlined the competing submissions of the parties before giving his decision.
The judge held that the single, natural and ordinary meaning of each of the claimants’ tweets about Mr Fox was “that the defendant was a racist” and that this “was an expression of opinion, and obviously so”. In doing so he accepted the submission of Ms Rogers KC for the claimants, that “there are some words that almost always signify that they represent the person’s opinion” and that “racist” was one of them. He said that
“’Racist’ is quintessentially one of those words. It almost invites the question from someone who hears the allegation: ‘why did you say that?’ It is very different from the allegation that somebody is a paedophile.”
The judge found support for his conclusion that the tweets of Mr Blake and Mr Seymour were statements of opinion in the fact that each had quoted Mr Fox’s Sainsbury’s tweet which the judge said “would appear to the ordinary reasonable reader to be the basis of a comment that the defendant was a racist”. In those circumstances the judge held that the condition laid down by s 3(3) of the 2013 Act was satisfied.
Ms Thorp’s tweet was different from those of Mr Blake and Mr Seymour: she did not quote-tweet Mr Fox’s Sainsbury’s tweet and she stated that Mr Fox was “unequivocally, publicly and undeniably” a racist. The judge rejected a submission that these features of Ms Thorp’s tweet meant that it amounted to an allegation of fact. He was satisfied that it would be understood to be an expression of opinion. The adjectives used by Ms Thorp were “simply” the “forceful expression of her opinion” and “rhetoric” which did not “convert” the expression of an opinion into an allegation of fact. But the judge held that Ms Thorp’s tweet did not indicate in general or specific terms the basis of her opinion, so she had “not established the second condition of the defence of honest opinion under s 3(3).”
The judge held that the claimants’ statements that Mr Fox is a racist were defamatory at common law. They do not challenge that conclusion.
As for the paedophile tweets, the judge held that their meaning was “that each of the claimants was a paedophile, someone who had a sexual interest in children and who had or was likely to have engaged in sexual acts with or involving children, such acts amounting to serious criminal offences” and that this was an allegation of fact. In other words, he accepted the claimants’ case.
The judge rejected Mr Fox’s argument that the ordinary reasonable reader would have understood him to be making a rhetorical comment, saying:
“That is one meaning that some readers may have thought his Tweets meant. It is not the natural and ordinary meaning. It is an extrapolation from the primary and obvious meaning of the words. It can only be arrived at after some interpretation. … Such an interpretation would only emerge after some analysis. For many readers it is likely to be arrived at only if they had someone prompting them to consider whether the Tweet had a second theoretically or logically deducible meaning beyond its plain meaning. …. The Defendant may have intended to convey this second meaning, but his intention is irrelevant to the objective single meaning of the Tweet.”
The judge also rejected Mr Fox’s argument that “an allegation that someone is a paedophile is mere abuse”. He set out the passage from Gatley that I have quoted above and said (at [53]) that “In my judgment, an allegation that a person is a paedophile does not qualify in the sense of being mere abuse as indicated in that passage. There was nothing in [the paedophile tweets] to indicate the word was not to be given its clear meaning.”
Having reached these conclusions the judge inevitably rejected Mr Fox’s argument that his tweets were not defamatory.
The appeal
Mr Fox appeals with the permission of Nicola Davies LJ on four grounds which his Counsel, Mr Callus, has developed in argument before us.
The first ground of appeal is that the judge’s decision on the fact/opinion issue was wrong in relation to each of the six tweets complained of. Mr Callus argues that the claimants’ tweets were all unequivocal statements of fact not opinion; and Mr Fox’s tweets were not allegations of fact but rhetorical statements of opinion “to the effect that the allegation of racism that had been levelled against him in the tweet [he] quoted was as baseless and abusive as an allegation that that person (making that allegation of racism) was a paedophile”.
The judge is said to have made three errors of law: (1) addressing the tweets in groups or batches, according to their common terms, rather than individually, setting out and considering the detail of the specific words used in each tweet and their particular context; (2) accepting the claimants’ argument that to say someone is “racist” will generally be an expression of opinion, when there can be no presumption about the meaning of any given statement, all decisions on that issue being inescapably fact-sensitive; (3) adopting an unduly condescending view of Twitter users, who are “reasonably sophisticated” and know that tweeters often need to encapsulate their arguments in pithy, creative and rhetorical ways.
In support of his second point, Mr Callus relies on certain phrases and passages in the judgment, including the judge’s use of the word “convert”. In connection with his third point, Mr Callus submits that the judge paid insufficient attention to the context, which will have made the rhetorical character of Mr Fox’s tweets obvious to ordinary reasonable readers. He places particular emphasis on Mr Fox’s response to Ms Thorp, which not only quote-tweeted what she had said but also replied in identical terms, except that “paedophile” was used in place of “racist”. This, it is submitted, was manifestly a case of sarcastic mimicry.
The second ground of appeal is closely linked to this last point. The contention is that the judge erred in his determination of the meaning of the paedophile tweets and whether or not these were defamatory at common law. Mr Callus argues that “[I]f the defendant’s tweets are recognisably rhetorical comment to the ordinary and reasonable reader, then they are not defamatory at common law either.” At worst, he says, those tweets suggested that the claimants had made an absurd and baseless allegation of racism, which is an imputation of which none of them seeks to complain.
The third ground of appeal is that the judge erred in declining to define the meaning of “racist”. Mr Callus argues that a definition is required to resolve the pleaded issues on meaning and define the parameters of the case. It was thus an essential part of the judge’s task at the trial, as he is said to have acknowledged this in the course of argument. Mr Callus submits that ground three must succeed if (as he maintains) the allegation is one of fact, because the judge’s ultimate decision not to define the term “racist” was exclusively based on his conclusion that it was a statement of opinion. The judge had accepted that the position would be different if it was an allegation of fact.
Mr Callus further submits that a definition is a vital step in managing the case fairly and expeditiously. He gives an illustration of the consequences of not defining racism. Mr Blake and Mr Seymour have pleaded the defence of honest opinion. A key issue is whether the third condition is met, namely that “an honest person could have held the opinion on the basis of … any fact which existed at the time the statement complained of was published” (s 3(4)(a) of the 2013 Act). That issue, says Mr Callus, can only be decided if the court knows what “the opinion” was.
Fourth and finally, it is argued that the judge erred in finding a meaning which went beyond those pleaded by the claimants. Their case was, as I have explained, that their tweets meant that Mr Fox’s public statements had “showed him to be” or “showed that he was” a racist. And Mr Blake and Mr Seymour relied for this purpose only on Mr Fox’s Sainsbury’s tweet. On that footing, it is said, the issues on honest opinion would or should be limited and relatively straightforward to resolve. However, the judge did not accept the claimants’ case on meaning. He found a meaning that is considerably wider. Mr Callus accepts that a judge is entitled to depart from the parties’ pleaded meanings but submits that in this case the judge gave no reason, and had no good reason, to do so.
Again, Mr Callus adds that the judge’s approach has undesirable practical consequences. The claimants have treated the judge’s determination as giving them the freedom – which they have taken - to rely in support of the opinion on a wide range of other, extraneous matters none of which was mentioned or alluded to in the tweets complained of. As matters stand, there is liable to be argument about whether this is a legitimate course to take. At common law, defendants did not have that freedom. Read literally, the Objective Honesty Condition would allow reliance on any fact provided only that it existed at the time of publication. In Riley v Murray [2022] EWCA Civ 1146, [2023] EMLR 3 at [50]-[58] I expressed my doubts that Parliament intended to change the law in this respect, and Arnold and Dingemans LJJ agreed; but we did not need to decide the point, which remains open.
On behalf of the claimants Ms Rogers KC and Ms Grossman resist the appeal on all four grounds, submitting that we should not interfere with the judge’s decisions, as they did not involve any legal error and were well within the range of reasonable decisions open to him. There is no cross-appeal nor any respondent’s notice.
Discussion
I shall again take matters in chronological order, dealing first with the meaning of the claimants’ tweets and whether they were statements of fact or opinion.
These are issues of fact. The court of first instance decides them by assessing the impact the statement complained of would have on the ordinary reasonable reader, applying the well-established principles I have outlined above. An appeal court will be slow to interfere with such a decision.
In Stocker the Supreme Court held that on an appeal against a judicial determination of meaning the court should exercise “disciplined restraint”. It should not interfere just because it would prefer a different meaning or conclusion within the reasonably available range. The court identified a range of reasons for such appellate self-discipline, including but not limited to the advantages a judge at first instance may have over one hearing an appeal: see [58]-[59].
Although the question of whether a statement is one of fact or opinion is a binary one the approach is the same. We have recently held that the Stocker test applies on an appeal against the binary question of whether the ordinary reasonable reader would or would not understand a statement to refer to the claimant: Dyson Technology Ltd v Channel Four Ltd [2023] EWCA Civ 884 [36]. In Millett v Corbyn this court observed that “we do not second-guess” decisions on whether a statement is one of fact or opinion which involve the application of accepted principles to the undisputed facts of the case; in the absence of legal error an appeal will only succeed if the court is satisfied that, allowing for the advantages available to the first instance court, the finding was wrong: [21], [36], [37].
Was the judge wrong to find that the claimants’ tweets were all statements of opinion (ground one)?
I do not think so. In my view the suggestion that he took a sweeping approach, treating these individual cases collectively as a group or batch, is misplaced. He was very familiar with this case. His mode of trial judgment set out the detail of the tweets complained of and the sequence of relevant events in the way that I have adopted in this judgment. It was not necessary for him to set this all out again to ensure that he applied his mind to the specifics or to demonstrate that he had done so.
I am not persuaded, either, that the judge made the second legal error attributed to him by Mr Callus. His self-direction on the law (at [28]-[31]) is not criticised and in my view is unimpeachable. It includes reminders that the process is “highly fact-sensitive”, that the court should not adopt “prescriptive” or “bright-line” rules, and that the answer is likely to turn on the particular context in which the words complained of appeared. These are supported by extensive citation of authority, several of the decisions quoted being judgments of Nicklin J himself. I do not think we can fairly read the turns of phrase later in the judgment that are relied on by Mr Callus as signalling a departure from these sound principles.
I do not believe the judge adopted or applied any mistaken presumption about the nature of the word “racist”. In my opinion he was simply noting, as part of his assessment of the statements complained of, that in practice the word “racist” tends to be used in an evaluative way. In doing so he was bringing to bear his own experience of the use of English. That is an inescapable part of any decision on the natural and ordinary meaning of words. It is consistent with the principle that judges should “have regard to” the impression which the offending statement makes on them. A judge commits no error of law in doing this so long as he keeps well in mind that the essential touchstone is the reaction of the ordinary reader of the particular publication and adheres to the other principles I have mentioned. In my judgment, the judge’s approach was a proper one.
Were his conclusions nonetheless wrong?I think not. In my judgment the judge was clearly right to find that the tweets of Mr Blake and Mr Seymour were statements of opinion. Comments or opinions can take many forms but these were classic instances of the genre. By quote-tweeting Mr Fox’s Sainsbury’s tweet they set out the facts in clear and unequivocal terms (“see what Laurence Fox has said on Twitter”). The body of the tweet then made various observations about those facts (“mess”, “racist”, “twat”, “proud”, “cringe”, “snowflake behaviour”). “Twat” was mere vulgar abuse. Some of the other words were not defamatory, or not seriously so (“mess”, “cringe”, “snowflake”). “Racist” was used by Mr Blake as an adjective and by Mr Seymour as a noun. But in each case the word in its context was clearly an evaluative statement about Mr Fox’s behaviour.
When it comes to Ms Thorp’s tweet the answer is not so obvious. For the reasons urged on the judge by Mr Callus this tweet falls into a different category and it certainly could be analysed as a statement of fact, or one that falls to be treated as such at common law. But that was not the judge’s impression. One of the reasons for appellate restraint in this field, as in others, is that the appeal court lacks the advantages enjoyed by the court of first instance. In this context one advantage is the ability to approach these issues with a mind relatively free of the baggage of argument and previous decisions. In the judge’s position I might have reached the same conclusion. In my judgment we would be wrong to interfere. I note, incidentally, that even if this was a statement of fact it would make little difference to the shape of the case. The failure to satisfy s 3(3) of the 2013 Act means that Ms Thorp cannot plead the defence of honest opinion. She can only succeed by proving that the allegation was substantially true. We are told that she has in fact pleaded that defence.
Was the judge wrong to decline to define the term “racist” (ground three)?
Given my conclusions on ground one some of the points raised by Mr Callus fall away. The question can be narrowed down to this: having found that the words complained of contained a statement of opinion that Mr Fox “was a racist” was the judge wrong not to define what “racist” meant in this context? My answer is that his decision was not wrong but entirely legitimate.
As Ms Rogers says, it is odd for Mr Fox to complain about the judge’s decision on this point when in substance the judge accepted Mr Fox’s pleaded case on the meaning of the statements complained of, including his primary case (with which the claimants agree) that there is no need for any definition of “racist”. The only difference between the meanings found by the judge and those complained of by Mr Fox in paragraphs 77 and 78 of the DAC is that the judge did not incorporate the words “unequivocally and undeniably” that featured in Mr Fox’s version of the meaning of Ms Thorp’s tweet.
The judge rejected Mr Fox’s case that the allegation of racism was one of fact, but I do not think that can create or increase a need to define the term. Mr Callus has not suggested as much. His main argument is that it does not matter for this purpose whether the statement is one of fact or comment; either way, the court is obliged to provide a definition because the parties have pleaded rival contentions on the meaning of the term. I do not think that can be right. On each side these are fall-back arguments that need consideration only if the court rejects the pleader’s primary case. The judge accepted the primary case of both parties that no definition was required.
In any event, Ms Rogers is surely right to emphasise that the judge’s task was to decide the natural and ordinary meaning of the particular words complained of, and whether a term needs definition as part of that process is a fact-sensitive question that turns on the words themselves. The meaning conveyed by a statement can be more or less clear, defined or specific. It depends on the impact the words would have, read in their context, and not on the parties’ pleaded cases. I can see that it might be necessary to incorporate a definition or limitation on the meaning of a term if the judge is satisfied that the statement complained of used that term in some limited or specialised sense rather than its ordinary meaning. But neither side submitted that was the position and the judge evidently did not consider that it was. He concluded that the opinion stated by the claimants was simply that Mr Fox “is a racist”. All of this was open to the judge.
In argument, the judge made clear that he thought it would be wrong to go further, if the statement was one of opinion. As I read his observations what he meant was that it would be wrong for him to narrow down the ordinary meaning by supplying a definition which the claimants had not themselves provided, either expressly or by implication, in the statements complained of. I would agree with that. I rather think that the same reasoning would apply if the statement was one of fact but it is unnecessary to decide the point.
For these reasons, the fact that the parties had set out very different views about the ordinary meaning of the word “racist” did not impose a duty on the judge to resolve that aspect of the dispute at this stage of the action. The consequences will have to be worked out as part of the management of the case. I am not persuaded that any of them are intractable. It is not obvious to me why the form of the meaning identified by the judge should give rise to difficulty in applying s 3(4)(a). But whatever the consequences may be they cannot determine the natural and ordinary meaning of the claimants’ words.
Was the judge wrong to find that the opinion expressed by the claimants’ tweets was not the limited one which the claimants themselves had pleaded as the meaning of those tweets (ground four)?
Again, this is an odd point for Mr Fox to take given his own pleaded case. And again, I do not consider the judge was in error. It is rightly accepted that the court is not bound by the pleadings of the parties but may find a meaning different from those advanced by either side. That is axiomatic. The boundaries of that principle are the subject of some dispute but need not detain us. In this case, the judge evidently concluded that the meaning of the claimants’ tweets was not as limited as they sought to contend. The claimants were not just saying that Mr Fox’s Sainsbury’s tweet showed him to be a racist (or in Ms Thorp’s case, that this and other public statements had shown this). They were making the broader assertion that he was, generally, “a racist”. It was clearly open to the judge to reach that conclusion. It was the meaning advanced by Mr Fox and the claimants did not even argue that it was an impossible interpretation of what they had said.
Mr Callus’s complaint of lack of reasons lacks substance. The principles and practice in this area mean that judges are, quite rightly, not so much reasoning their way to a conclusion as gaining an impression of the meaning of a statement, then seeking to analyse and explain why they have responded as they have. It is well-established that the judge should avoid over-elaborate analysis and unduly detailed reasons: Koutsogiannis [12(iv), (v)]. In this case the judge made clear that he had that in mind. He referred to the undisputed facts, identified the applicable legal principles, summarised the essential features of the competing arguments and spelled out the precise meaning which he had drawn from each of the statements complained of. I would infer that he accepted at least some of the argument for Mr Fox on the issue of meaning. But in any event, his reasoning was sufficient.
I can see the force of Mr Callus’s point that the further conduct of this case might have been simpler if the judge had accepted the claimants’ narrow version of the opinion they stated in their tweets. It may be that the parties will feel the need to litigate the issue of law that we left unresolved in Riley v Murray. Again, though, we cannot know, and these are anyway case management issues that may arise further down the line. They clearly cannot be drivers of the court’s decision on the issue of meaning. I think Ms Rogers is right to warn of the dangers of judges being sidetracked by such irrelevant considerations when deciding that important but relatively straightforward issue.
Was the judge wrong to decide that the paedophile tweets were statements of fact with the meaning he identified (grounds one and two)?
I have rolled this up into a single question because that is the substance of the matter. In relation to each of the paedophile tweets Mr Fox’s case stands or falls by the proposition that the judge was wrong to reject his contention that the ordinary reasonable reader would have understood the tweet to be a rhetorical way of rebutting the charge of racism, rather than an allegation that the claimant in question was a paedophile.
I do not agree that the judge erred in law by taking a condescending view of the attributes of the ordinary reader. That argument seems to me to be based on a misinterpretation of what the judge said about the “rhetorical” meaning relied on being one that emerged only after some interpretation or analysis on the part of the reader. The judge was not saying in those passages that ordinary reasonable users of Twitter, or those who read Mr Fox’s tweets, are incapable of undertaking the process of pausing to consider and analyse the tweets, or unable to see rhetoric and sarcasm when it presents itself. The judge had reminded himself of the need to avoid impressionistic assumptions about a given readership. In these passages he was doing no more than applying to the facts of this case some recognised principles, including the proposition that judges should take account of the fact that social media are not generally provided for mature reflection by their readers but for consumption at some speed.
I do accept that this proposition should be applied with some care. As Mr Callus points out, there is a distinction to be drawn between the characteristics of the medium and those of the readership. It does not follow from the fact that tweets are generally read swiftly that their readers are careless, superficial or unsophisticated. Account should also be taken of the nature of the particular message. Tweets vary in their length and form. Not all are concise and conversational. A body of extraneous text can be included in a screen shot. And it is well-known that some information is posted on Twitter via threads, composed of multiple individual posts, which can be serious contributions to knowledge about topics of political or social importance.
Those points have little significance when it comes to what Mr Fox tweeted in response to Mr Blake and Mr Seymour. Those were short and pithy tweets of between three and six words. They followed swiftly after the tweets to which they responded. They do not give the appearance of being carefully considered or crafted. They are straightforward assertions. The one striking word was “paedophile”. The reader trying to understand what Mr Fox was getting at was given very little else to work with. The only relevant context (on the judge’s findings) was that which would have been apparent to all readers. In substance that was no more than the quote-tweet. On the face of it, the allegation was the one complained of.
That is a serious allegation. It has no apparent connection with the statement quote-tweeted by Mr Fox. That statement was clearly an attack on him. The reader would probably have understood that Mr Fox was seeking to counter the charge that he was a racist. But it by no means follows that it would be obvious to the reader that what he was trying to do was to make the somewhat complex rhetorical point that has now been identified. It is common experience that people accused of wrongdoing sometimes lash out in response by denouncing their accusers, in all seriousness, for some similar or other misconduct. The question for the judge, of course, was what the reasonable reader would probably make of these two tweets. I can see nothing wrong with his answer to that question in relation to either of them or with the reasoning he gave for providing that answer.
Mr Fox says that he did not intend to allege that any of the claimants was in fact a paedophile. But I do not think he can complain of being misunderstood on these occasions. The constraints of Twitter gave him plenty of room to say more than he did in these tweets. There is a good deal of force in Ms Rogers’ submission to us: if Mr Fox had wanted to say “I am no more a racist than you are a paedophile” he could have done so.
All of this said, and despite the self-denying principles I have outlined above, I take a different view when it comes to the judge’s conclusions about Mr Fox’s response to the tweet of Ms Thorp. Here I think there may be something in the submission that the judge’s conclusions were affected by considering the paedophile tweets as a package, or at any rate without sufficiently concentrating on their particularities. Mr Callus candidly conceded that this is how both parties framed their arguments at first instance. At all events I have been persuaded that the judge was wrong to find that the natural and ordinary meaning of this tweet was that Ms Thorp was a paedophile.
The exchange between Ms Thorp and Mr Fox was significantly different from the others. Ms Thorp’s tweet was different in the ways I have outlined above. Mr Fox’s response did of course include the word “paedophile” but I am satisfied that the ordinary reasonable reader of that tweet would not have taken the word literally. This tweet was sent an hour after Ms Thorp had posted hers. It was self-evidently a considered response. Mr Fox adopted the precise wording of the tweet to which he was responding, and which he quoted, with the notable exception of substituting the word “paedophile” for “racist”. No reasonable reader could have failed to see this or to discern that it was deliberate. The repetition of Ms Thorp’s very words quite plainly and obviously involved an element of mimicry. The significance of this was either obvious or the reader was prompted to think about what Mr Fox was getting at. Either way, the conclusion would, in my opinion, have been that Mr Fox was not using the word “paedophile” literally, to accuse Ms Thorp of being a paedophile; he was using that word rhetorically as a way of expressing his strong objection to being called a racist. Used in that way it was not defamatory.
In my judgment, the ordinary reader would have understood Mr Fox to be sending a message on the lines encapsulated in the submission of Ms Rogers that I have quoted above. To distil the meaning with more precision, I would put it this way: the single, natural and ordinary meaning of the tweet complained of by Ms Thorp is that she “had posted a tweet making an allegation of racism against Mr Fox which was outrageous and untrue.”
Conclusions
For the reasons I have given I would dismiss Mr Fox’s appeal against the judge’s findings in respect of the claimants’ tweets. I would also dismiss Mr Fox’s appeal against the judge’s findings in respect of his own tweets in response to those of Mr Blake and Mr Seymour. I would however allow Mr Fox’s appeal against the judge’s findings about his response to Ms Thorp’s tweet. It would seem to follow that her claim should be dismissed.
LORD JUSTICE ARNOLD:
I agree.
LADY JUSTICE NICOLA DAVIES:
I also agree.