Identification: a person does not have to be named to be defamed

A defamation claim fails unless the plaintiff can show that the publication was about them. Lawyers call this the identification element, and the traditional phrase is that the matter must have been published “of and concerning the plaintiff”. The question is not whether the publisher meant the plaintiff. It is whether the audience would reasonably have understood the publication to refer to the plaintiff. That objective test produces some surprising results. A person who was never named can be identified, a publisher can be liable to someone they have never heard of, and an insult aimed at a whole category of people may identify nobody at all. Section references are to the Defamation Act 2005 (NSW) unless otherwise stated.

The test for identification

The NSW Court of Appeal has summarised the element in these terms: to succeed, a plaintiff must prove not only that the defendant published the material and that it is defamatory, but also that the material was published of and concerning the plaintiff.¹ Identification can be proved in a number of ways. It is usually clear because the plaintiff is named and the publisher intended to refer to them, but neither being named nor the publisher’s intention is necessary for identification.²

Where the plaintiff is not named, Australian courts apply the test stated by Isaacs J in the High Court in 1918: are the words such as reasonably in the circumstances would lead persons acquainted with the plaintiff to believe that he or she was the person referred to?³ The general public need not be able to work out who was meant. It is enough that people who knew the plaintiff, reading the publication reasonably, would have understood it to be about them.

Identification is an objective question about the understanding of the audience, not a question about the publisher’s state of mind. Dixon J explained in Lee v Wilson & Mackinnon that liability depends upon mere communication of the defamatory matter to a third person, and that the communication may be quite unintentional.⁴ The same case adopted the proposition that liability for libel does not depend on the intention of the defamer but on the fact of defamation.⁵

Named plaintiffs and shared names

Where the plaintiff is named in the publication and the matter is defamatory on its face, identification is normally straightforward. The plaintiff proves publication and does not need to prove that any reader had prior knowledge of them.⁶ Complications arise where more than one person answers to the name.

Lee v Wilson & Mackinnon is the classic Australian example. A newspaper report referred to “Detective Lee” in connection with corruption allegations. Two police officers named Lee, neither of whom was the person meant, sued. The High Court held that each could recover. Dixon J reasoned that a description which on its face designates one person only may nevertheless be a libel of two or more, if it is capable of denoting each of them and is reasonably understood by one group of readers to refer to one, and by another group to refer to another.⁷ The Court of Appeal in Zoef v Nationwide News Pty Ltd treated it as common ground that where a publication includes a name, each person of that name may sue, regardless of the publisher’s intention to refer to a particular person, where in context the words are reasonably understood to refer to each of them.⁸

In Zoef, an article reported that a Sutherland Shire tailor named Tony Zoef had been charged with weapons offences. The person charged was the tailor’s son, who lived at the same address and shared the name. The article included the son’s age and an indistinct photograph of him, but the headline and opening paragraph pointed to the father’s name, profession and locality. The Court of Appeal upheld the finding that the father was identified. The threshold for a reader to reasonably identify the plaintiff is not high. The ordinary reasonable reader draws conclusions from general impressions, does not go back and read the article again, and is influenced by headlines and prominent features rather than a detail such as an age buried in a later paragraph.⁹ The reader is taken to be aware that there may be more than one person of the same name in a district, but that did not make it unreasonable to conclude that this article referred to the well known local tailor.¹⁰ The father recovered $150,000.¹¹

Identification without a name

A plaintiff who is not named can be identified by any combination of details that points to them: a description of their role or appearance, a photograph, a nickname, a job title, a business name, a location, or a reference to events associated with them. The law distinguishes between two situations. In the first, the ordinary reasonable reader would know from the publication itself to whom it refers. In the second, the reference would be apparent only to people who had knowledge of special circumstances outside the publication.¹²

The second situation is common and is handled through what are called extrinsic facts. If the plaintiff’s identity would be known only to readers with knowledge of particular facts, the plaintiff must prove that the matter was published to people who had that knowledge and who believed that the plaintiff was referred to.¹³ The plaintiff must also show that those readers were ordinary sensible readers who could reasonably have reached that conclusion.¹⁴ The extrinsic facts relied on must be true. A plaintiff cannot build identification on readers’ mistaken beliefs about the surrounding facts, although readers who knew true facts may still have erroneously but reasonably understood the article to refer to the plaintiff.¹⁵

Pedavoli v Fairfax Media Publications Pty Ltd illustrates the risk. The Sydney Morning Herald reported that a teacher at a named school had resigned amid allegations of sexual misconduct with students. The teacher was not named, but the article described her as being in her late twenties and said she was believed to teach drama and English. Those details did not match the teacher who had actually resigned, but they matched Ms Pedavoli, who had done nothing of the kind, and pointed unequivocally to her.¹⁶ McCallum J found that she was identified even though she was not named, and awarded $350,000.¹⁷

Pedavoli also addressed an important timing issue. The defendants argued that the claim was limited to readers who, at the moment of publication, already knew the facts that made the description fit the plaintiff. The starting principle, which the court accepted, is that where a plaintiff relies on readers’ knowledge of extrinsic facts, that knowledge must exist at the time of publication to that reader.¹⁸ But the court adopted the reasoning of Eady J in an English decision that a publisher who gives tantalising detail and implicitly invites readers to find out who is meant can be liable to those who do so. The article offered enough detail to encourage readers in the school community to identify the teacher, and the publisher was taken to have invited that course.¹⁹

The publisher’s intention is irrelevant

It follows from the objective test that a publisher’s honest and reasonable belief that they were writing about someone else, or about nobody in particular, is no answer to the identification element. In Lee v Wilson & Mackinnon the newspaper intended to refer to a real “Detective Lee” and was still liable to two other officers of that name.⁷ In Pedavoli, although the error was the careless inclusion of incorrect information rather than deliberate, the publisher was liable just as if the identification had been intentional.¹⁶

Intention and care are nonetheless relevant to defences and to damages. A publisher who promptly corrects and apologises may be able to rely on the offer to make amends procedure, which was the central issue in both Zoef and Pedavoli. In each case the offer was held to fall short and the publisher went on to pay damages.²⁰

Groups and classes

A statement about a class of people raises a distinct problem. If a publication says that all lawyers are thieves, no particular lawyer can sue, unless there is something to point to that individual. The High Court recognised in David Syme & Co v Canavan that a reference to a class of persons of which the plaintiff is a member can be a sufficient reference to the plaintiff if readers would reasonably believe the plaintiff was referred to, but that this is a fact the plaintiff must prove.²¹ The House of Lords in Knupffer v London Express Newspaper Ltd later settled the approach that Australian courts continue to apply.²¹ Lord Atkin said that it is a mistake to lay down a rule about libel on a class and then qualify it with exceptions. The only relevant rule is that the words must be understood to be published of and concerning the plaintiff. It is irrelevant that the words were published of two or more persons, or that those persons were called by some generic or class name, if the words would reasonably be understood as published of each of them.²²

Two features of the case law matter in practice. First, the size and definition of the group matter. Lord Porter observed that the size of the class, the generality of the charge and the extravagance of the accusation may all be taken into account, though none is conclusive.²³ A defamatory statement about a body of trustees or directors, or the tenants of a particular building, may be understood as published of each member.²⁴ A statement about a large or indeterminate number of people described by a general name usually is not. In Knupffer itself, an attack on a political movement with members spread across several countries did not identify the head of its British branch, even though witnesses said their minds went to him when they read it.²⁵

Second, an accusation against one unidentified member of a small group may reflect on every member. In Christiansen v Fairfax Media Publications Pty Ltd, Nicholas J held that an article alleging that one of three dismissed employees was the culprit was capable of casting a cloud of suspicion over each of them, and so was capable of conveying that there were grounds to suspect the plaintiff. The smallness of the group provided the basis for that imputation.²⁶ Whether such a meaning is actually conveyed remains a question of impression in each case.²⁷

Businesses and trading names

Statements about a business often identify the people behind it. A sole trader who trades under a business name is identified by a reference to the business. A statement that a small company is dishonest may be reasonably understood as a statement about its directors or owners, particularly where the business is closely identified with one or two people in its community, on the same reasoning that applies to trustees and directors.²⁴ In Zoef, the plaintiff was identified by the combination of his name, his trade and the locality in which he practised it.²⁸

This point has added significance because of the restrictions on corporate plaintiffs. A corporation can only sue for defamation if it is an excluded corporation, that is, a not for profit body or a corporation with fewer than 10 employees that is not an associated entity of another corporation, and in either case not a public body.²⁹ Where the company itself cannot sue, the Act expressly preserves any cause of action of an individual associated with the corporation in relation to defamatory matter about the individual, even if the same matter also defames the corporation.³⁰ Whether the individual is in fact identified is then decided on the ordinary principles described above.

Proving identification

Identification involves a two stage process. The first question, whether the matter is capable of referring to the plaintiff, is a question of law for the judge. The second, whether the plaintiff was in fact identified, is a question of fact for the jury, or for the judge sitting alone.³¹ If the evidence is not capable of establishing identification, the case does not go to the tribunal of fact.³²

Where the plaintiff is named and the matter is defamatory on its face, little evidence is needed beyond proof of publication.⁶ Where the plaintiff is not named, the usual evidence comes from witnesses who read or heard the publication and understood it to refer to the plaintiff, together with proof of the extrinsic facts they knew. In NSW, the rules of court require a plaintiff who is not named to give particulars of identification, including the identity, by name and address or by class, of the people to whom the identifying facts were known.³³

The number of readers who could identify the plaintiff also affects other elements of the claim, such as serious harm. Since 1 July 2021 in most jurisdictions, a plaintiff must prove that the publication has caused, or is likely to cause, serious harm to their reputation.³⁴ An unnamed plaintiff who can be identified only by a small circle of readers may face a more difficult task on serious harm than a plaintiff named to the world at large, though as Pedavoli shows, a small readership can still be the one whose opinion matters most to the plaintiff’s reputation. Damages in that case were assessed on the basis that the article was not a publication to the world at large, yet still approached the statutory cap.³⁵ Western Australia has not enacted the serious harm element, and the common law presumption of damage continues to apply there.

Getting advice

A person who has not been named but who is recognisably the subject of a publication may have a claim, but will need evidence of who understood the publication to refer to them and what those readers knew. A publisher who has described someone inaccurately, or used a name shared with another person, may be liable to a plaintiff they never had in mind, and the early steps taken in response to a complaint can affect the defences available. Harris Defamation Lawyers acts for plaintiffs and defendants across Australia and can assess whether the identification element is likely to be made out.