APage_img1117-12-470x312It is still very early days for section 1(1) of the Defamation Act 2013, which is likely to be a source of uncertainty for practitioners while decisions clarifying its application are awaited from specialist defamation judges and the appeal courts. This may well be having a chilling effect on potential claimants owing to the costs risks attaching to the testing of new legal boundaries.

Whilst the Act identifies in ss 2(4), 3(8) and 4(6) specific parts of the common law that have been abolished, section 1(1) is in potential conflict with established common law principles not overtly abolished.

The section could have been interpreted consistently with these principles, but for the addition at a very late stage of the Bill through Parliament of section 1(2), which has rendered the proper construction of section 1(1) less obvious.

There is also some room for confusion with concepts and terminology adopted in defamation that may have contributed to uncertainty over the application of section 1(1).

Section 1(1) provides:

A statement is not defamatory unless its publication has caused or is likely to cause serious harm to the reputation of the claimant.

This sub-section has the purpose, according to Explanatory Note 11, of building upon Thornton v Telegraph Media Group ([2011] 1 WLR 1985)(“Thornton“) and the “threshold of seriousness” to raise the bar for bringing a defamation claim. In principle, that was a straightforward aim and easy to grasp. It was also the general understanding of its purpose amongst practitioners during the Bill’s passage through Parliament.

At common law, in the case of defamation actionable per se, words were actionable if they had the capacity to cause some, albeit modest, harm to the claimant’s reputation. In Thornton Tugendhat J identified the need to define the word “defamatory” to include a minimum threshold of seriousness to exclude trivial claims, which, as the Court of Appeal recognised in Jameel v Dow Jones [2005] QB 946, might constitute an interference with freedom of expression. However, it has been widely assumed by commentators that, in order to meet the new threshold test, a claimant now has to, or may have to, adduce evidence of actual serious harm or its probability.

Early support for this assumption was provided by Bean J in Cooke & Midland Heart Ltd v MGN [2014] EMLR 31 at [43]:

I do not accept that in every case evidence will be required to satisfy the serious harm test. Some statements are so obviously likely to cause serious harm to a person’s reputation that this likelihood can be inferred. If a national newspaper with a large circulation wrongly accuses someone of being a terrorist or a paedophile, then in either case (putting to one side for the moment the question of a prompt and prominent apology) the likelihood of serious harm to reputation is plain, even if the individual’s family and friends knew the allegation to be untrue. In such a case the matter would be taken no further by requiring the claimant to incur the expense of commissioning an opinion poll survey, or to produce a selection of comments from the blogosphere which might in any event be unrepresentative of the population of “right thinking people” generally. But I do not consider that the Article in the present case, with the meaning relating to the claimants which I have held it to have, comes anywhere near that type of case.1

There has been some indication in decisions since Cooke that the courts may adopt a less demanding approach than suggested by Bean J’s examples. See HH Judge Parkes QC in Donovan v Gibbons [2014] EWHC 3406 (QB) at [6] and Warby J in Ames v Spamhaus Project Ltd [2015] EWHC 127 (QB) at [55]:

“….there may be circumstances in which one would naturally expect to see tangible evidence that a statement had caused harm to reputation, but as practitioners in this field are well aware, it is generally impractical for a claimant to seek out witnesses to say that they read the words complained of and thought the worse of the claimant. I note from Cooke [42(f)] that the defendant’s submission on that case did not go so far as to say that this could never be done; rather, it was submitted that the court should be “wary” of attempts to rely on inference.

However, prior to the Act, no such evidence was required or, more significantly, permitted at the stage of determining whether words were defamatory in their natural and ordinary meaning. If Bean J and those who anticipated his approach are correct, section 1(1) has not merely raised the bar, it has altered the nature of the cause of action for libel and some slanders, so that they are no longer actionable per se.

At common law, in the case of defamation actionable per se, the “harm” that gives rise to the cause of action is the publication to a third person of defamatory words about the claimant. The test of defamatory is whether the words have a tendency to cause harm to the claimant’s reputation. The cause of action arises (or not) at the moment of publication, that is to say, as soon as words with that tendency are published to the person by whom they are read or heard; and not later. The claimant need not wait for some concrete evidence that the publication of the words has in fact provoked an adverse reaction to him before issuing his Claim Form.

It is at this first stage, of defining the cause of action, where use of language may have sown confusion. The “harm” or the harmful event is the publication of the words, not any consequences of publication. Although the common law looks exclusively to the tendency of the words to have an adverse effect upon the claimant’s reputation, some of the expressions used have not been entirely apt to make this clear. In Thornton, Tugendhat J, after reviewing the principal authorities on the test of defamatory, articulated what has become known as the “threshold of seriousness” in these words at [96]: “the publication of which [the claimant] complains may be defamatory of him because it affects in an adverse manner the attitude of other people towards him, or has a tendency to do so”. The “or” is unhelpful because it suggests two distinct possibilities when in reality there is only one, because it is the tendency of words to have an adverse impact that is the test of defamatory. The test is whether words have been published (or are threatened to be published) which would tend to affect in an adverse manner the attitude of other people towards the claimant. That is what gives rise to the cause of action at common law.

Another area of possible confusion is that section 1(1) uses the expression “or is likely to cause serious harm to the reputation of the claimant”. At common law, the words “tendency”, “likely” and “calculated to” are used interchangeably in defining the test of defamatory (see, for instance, Tugendhat J in [29] on p 1994B and D of Thornton). It is also clear from [93] – [94] of Thornton that Tugendhat J was adopting the traditional approach of treating “tendency” and “likelihood” as interchangeable terms and that unless words had a tendency or likelihood to have adverse effects upon the reputation of the claimant, it would be difficult to justify why there should be a presumption of damage.

So it may be that section 1(1) was drafted to mirror Tugendhat J’s formulation of the test of defamatory at [96] of Thornton, adjusted only so as to qualify the test by the addition of the word ‘serious’. This would have achieved the aim identified in Explanatory Note 11 whilst preserving the common law approach in all other respects.

As to some extent follows from the nature of the cause of action, if the words are defamatory, damage is presumed, there being no requirement to prove that harm was actually suffered; although evidence of actual harm is admissible at the assessment of damages stage, at which point it may have considerable impact.

The presumption of damage is irrebuttable, so that a statement is actionable as defamatory and damage will be presumed (however modest) even if the publisher can show that the words were not believed by, or the claimant was unknown to, the publishees. In Jameel v Dow Jones at [32] –[41] a challenge to the presumption of damage as incompatible with article 10 of the European Convention on Human Rights was rejected by the Court of Appeal. However the court identified the need to redress the operation of the common law rules in those rare cases where a claimant brings an action for defamation in circumstances where his reputation has suffered no or minimal actual damage. However, it was said that the way to do so was not to make radical change to the common law but for the courts to put a halt to actions as an abuse of process if they were not serving the legitimate purpose of protecting the claimant’s reputation, such as in Jameel, where publication was only minimal and all publishees had been identified.

In the straightforward case of natural and ordinary meaning, the test of defamatory at common law is purely objective. The starting point is the meaning of the words. The court applies the single meaning rule to select, sometimes from a spectrum of possible meanings, the single ‘right’ meaning. That meaning is decided according to established principles of construction modelled around the concept of the hypothetical reasonable reader, who is taken to be representative of those who would read the words, not any actual reader of the words. For the purpose of this exercise, no evidence of the meaning in which the words were in fact understood is admissible; such evidence is irrelevant – see, for instance, Goddard LJ in Hough v London Express Newspapers Ltd [1940] 2 KB 507 at 515:

In the case of words defamatory in their ordinary sense the plaintiff has to prove no more than that they were published: he cannot call witnesses to prove what they understood by the words…  The only question is, might reasonable people understand them in a defamatory sense.

There is very good reason for the common law approach of excluding evidence of how individual readers understood or responded to the words. For instance, powerful evidence of serious harm might include any of the following (if clearly referable to the publication): being spat at or called names in the street; the receipt of hate mail; a fire bomb through the letter-box; an abusive twitter storm. But how would such evidence assist the court to choose between competing meanings of the words applying the test of whether the claimant’s reputation would suffer serious harm in the eyes of the hypothetical reasonable reader? Reasonable readers do not exhibit such extreme behaviours; nor does the hypothetical reasonable reader usually resort to sacking employees who are defamed without first inquiring into the truth of the words in accordance with fair employment procedures, with the result that at the stage of assessment of damages such evidence of harm may be challenged as too remote. However, the admission of such evidence at the threshold stage, let alone a requirement for it to be adduced in most cases at that point, would surely alter fundamentally the traditional objective approach to the test of defamatory.

The principle that the cause of action arises at the moment of publication is also evident in claims based on a legal innuendo. Where only readers with knowledge of particular facts would understand the words in a defamatory sense, the extrinsic facts must have been within the knowledge of the publishees at the moment of publication; see Lord Denning in Grappelli v Block [1981] 1 WLR 922 at 825:

I would go by the principle, which is well-established, that in defamation — be it libel or slander — the cause of action is the publication of defamatory words of and concerning the plaintiff. The cause of action arises when those words are published to the person by whom they are read or heard. The cause of action arises then: and not later.

[Counsel for the Plaintiff] urged us to say that in slander it may be different. He suggests that the cause of action there does not arise until there is damage — like actions in negligence and the like. I prefer to go by the principle that in defamation a cause of action arises (and a writ can be issued) as soon as the words are published to a person then knowing all the material facts. If there are extrinsic facts, he must know them then — at the time of publication. That is when a cause of action arises. It cannot be made into a cause of action by reason of facts subsequently coming to the knowledge of the reader or hearer.

Consistently with the common law approach, the words in section 1(1) “has caused” could be construed as referring to actions for defamation commenced post-publication, where the serious harm (i.e. the act of publication of the words complained of) has occurred by the time of the Claim Form. The words “or is likely to cause” could refer to actions commenced pre-publication, where, if not restrained by interim or final injunction, there is a threat of publication or further publication likely to cause serious harm. This interpretation would preserve the distinction between harm to reputation, as an ingredient in the cause of action, and actual damage, evidence of which is relevant to the court’s assessment of damages if the claim succeeds.

However, in Cooke, Bean J viewed section 1(1) differently, but also as problematic, at [31]:

The words [in section 1(1)] “has caused” involve looking backwards in time, the words “or is likely to cause” involve looking forwards. The Act does not make clear the moment which marks the dividing line between past and future. It cannot be the moment of publication, since at that moment no harm “has been caused”. The two logical possibilities seem to be the date of issue of the claim and the date of the trial (or of the trial of the preliminary issue of serious harm). Either of these has the curious effect that whether a statement is held to have been defamatory on the day it was published might depend respectively on the timing of the issue of proceedings, or the timing of the trial.” (italics added)

However, if the common law approach is preserved there was no need for the Act to identify the moment which marks the dividing line between past and future. The dividing line is already clear and it is “the moment of publication” because that is when the harmful event occurs (or ‘seriously harmful’ under section 1(1)).

So does section 1(1) do away with the hypothetical reasonable reader and override the purely objective approach to what is defamatory; and is the distinction now lost between the concept of harm as an ingredient of the cause of action and actual damage as the focus of the damages assessment?

That the requirement of serious harm as the test of defamatory was, in the original form of section 1, to bear its traditional common law meaning of tendency to (serious) harm rather than requiring evidence of actual damage is reinforced by the clause having been drafted to read “a statement is not defamatory unless its publication has caused or is likely to cause serious harm”. Save for the word “serious”, this is quite close to the common law formulation of Tugendhat J in Thornton at [96] set out above.

However, if the purpose of section 1(1), when first drafted to stand alone, had been to make actual damage a requirement of the cause of action and to use “likely” in a sense different from “tendency to”, this would have pointed to the clause being worded “a statement is not actionable unless its publication has caused or is likely to cause serious harm”.

Explanatory Note 10 arguably detracts from a construction of section 1(1) as a strengthened mirror of the common law as stated in Thornton because the highlighted sentence below from the Note is otiose if the common law is preserved but qualified by the requirement of “serious”:

10. [Section 1(1)] provides that a statement is not defamatory unless its publication has caused or is likely to cause serious harm to the reputation of the claimant. The provision extends to situations where publication is likely to cause serious harm in order to cover situations where the harm has not yet occurred at the time the action for defamation is commenced… .

This sentence is, however, not entirely incompatible with the construction suggested above in which section 1(1) is distinguishing between actions commenced post-publication and actions commenced pre-publication. However, it is impossible to reconcile the preservation of the common law objective approach to defamatory meaning and the presumption of damage once section 1(1) is read with the later added sub-section (2), which says:

For the purposes of this section, harm to the reputation of a body that trades for profit is not “serious harm” unless it has caused or is likely to cause the body serious financial loss.

A body is very unlikely to have sustained serious financial loss at the moment of publication (subject to an instant drop in the share price of a quoted company being admissible), which is the point that troubled Bean J about section 1(1). Sub-section (2) is not compatible with a construction in which “serious harm” means serious harm in the abstract, as in a tendency to cause serious harm. It must be imposing a requirement of actual serious harm or its probability, which is a test that plainly would call for evidence. So whilst it leaves libel no longer actionable per se in the case of bodies trading for profit, it is not easy to see how it can be treated as exclusively applicable to bodies trading for profit. Inevitably, it must also colour the interpretation of section 1(1), making it difficult to see how libel can any longer be regarded as actionable per se in any circumstance.

Adrienne Page QC is a leading specialist silk whose practice covers defamation, malicious falsehood, privacy, confidence, harassment by speech and media law issues.

This paper was originally published on the 5RB website and is reproduced with permission and thanks