//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[2026] ACTCA 5
respondent. Appeal dismissed. The appellant is to pay the respondent's costs: [98].
The appellant complained that a customer had accused her of selling a fake Chanel bag and had posted about the dispute on Facebook. The central video did not name her. Although the video attracted online activity, the evidence did not show how many people understood it to be about her, or adequately connect alleged avoidance and a short-term business decline to reputational harm from the publications. The majority held that the serious harm threshold was more demanding than merely non-trivial harm and dismissed the appeal because the appellant had not proved the element. McWilliam J also agreed the appeal should be dismissed, while finding that the primary judge's reasoning included material errors.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
2,112 words · generated 07/10/2026
Supamas Supaphien operated a business providing a personal shopping service for luxury goods through Facebook. In September 2021, Natesai Chaiyabarn accused her at a Thai Buddhist temple of selling a counterfeit Chanel handbag. The next day, Chaiyabarn posted a video on Facebook describing her experience as a warning to others. The video did not identify Supaphien by name or nickname. It attracted 185 likes, 144 comments and 34 shares ([1], [18]-[20]).
3 cited instruments linked from this judgment.
Supaphien sued for defamation. Her pleaded case covered the initial oral accusation, the video, comments posted in response, and a screenshot allegedly sent through Facebook Messenger ([13]-[32]). The publications allegedly conveyed that she had sold counterfeit goods, defrauded a customer, or was dishonest. Some later comments and the screenshot allegedly identified her using her nickname, “Jing Jo”, and photographs ([21], [27], [31]-[32]).
The appeal was not a trial of whether the handbag was genuine or whether each pleaded imputation was conveyed. The issue had been brought forward for determination as a separate question: had publication of the pleaded defamatory matters caused, or was likely to cause, serious harm to Supaphien's reputation? The primary judge answered “No” and dismissed the proceedings ([8]-[9]). The Court of Appeal dismissed her appeal and ordered her to pay Chaiyabarn's costs ([98]).
The evidence did not include testimony from the people said to have identified Supaphien. The pleadings and particulars identified five such people, but none gave evidence. The appellant also did not tender the “subsequent comments” said to identify her ([33]-[34]). Her evidence included an assertion that 34 people avoided her and a table concerning business sales and gross profit. The judges considered those materials insufficiently detailed to establish that the publications had caused serious harm to her reputation ([93], [141]-[143]).
Section 122A(1) of the Civil Law (Wrongs) Act 2002 (ACT) makes serious harm to reputation an element of the cause of action. The Court explained that the plaintiff bears the onus of proving that element, and that the judge determines it under s 122A(3) ([4]-[5]). The legislation also permits the issue to be decided before trial. A determination that the element is not established may be made on the pleadings in the circumstances set out in s 122A(7), while a separate question may be determined on evidence ([37]-[40]).
The appellant argued that the primary judge had applied too high a threshold by saying serious harm sat above substantial harm but below grave harm. The majority accepted that there was a legitimate concern about substituting synonyms for statutory language, referring to Basten AJA's reservation in Rader. But it was not persuaded that using a spectrum was an error of principle. The majority rejected the argument that “serious” meant only more than trivial. It stated that “the distinction between trivial and serious is not binary” ([62], [73]). It therefore rejected the relevant ground, while leaving the central question as whether the evidence proved serious harm.
The principal evidentiary difficulty was identification. The video did not name Supaphien. The number of likes, comments and shares did not establish that the recipients understood the video to refer to her. As the majority put it, “It did no harm to her reputation if all that was passed along the grapevine was that an unnamed person sold a fake Chanel handbag” ([79]-[80]). In other words, circulation of an allegation about an unidentified seller did not, without more, prove reputational harm to this plaintiff.
The majority also considered the alleged avoidance and business downturn. It found that the appellant had not provided evidence adequately connecting those matters to the publications. The business table contained limited information and did not establish why sales changed, what the figures represented or whether the change followed the relevant posts in a way that could support the suggested inference ([93]). The alleged avoidance by 34 people lacked detail about what avoidance involved, when it occurred and why. That left an insufficient basis to conclude that the publications caused the change in how others regarded Supaphien ([141]-[143]).
McWilliam J agreed with the result but differed on the primary judge's reasoning. In that judge's view, the primary judge had made material errors by treating aspects of the evidence in isolation and by relying on observations about social media informality in assessing serious harm ([112], [122]-[125], [138]). McWilliam J considered that the Thai community and the social-media-based nature of the business were relevant context, not matters to be discounted merely because the community's size or the appellant's level of involvement was not quantified ([113]-[117]). Nevertheless, after independently assessing the evidence, McWilliam J concluded that it remained insufficient ([139]-[143]).
The judgment describes the pre-amendment common law position as including a presumption of damage where defamatory matter was published of and concerning the plaintiff, alongside a tendency-to-injure approach. The majority referred to Newman v Whittington for the historical account and explained that the presumption had posed difficulties in cases involving publications of limited seriousness or circulation ([86]). It considered that the serious harm element changed the plaintiff's task by requiring proof of harm rather than reliance on the former presumption ([87]-[88]).
The statutory amendment followed the Model Defamation Amendment Provisions 2020 and was modelled on s 1 of the Defamation Act 2013 (UK) ([3]). The judgment discusses the UK authorities Lachaux v Independent Print Ltd and the interpretation of the serious harm threshold there. The majority treated the statutory wording and the ACT legislative context as supporting a threshold higher than non-trivial harm ([64]-[73]).
The judgment does not establish a single, precise verbal definition of “serious”. The majority considered the Rader formulation permissible but did not say that every future court must use its spectrum. McWilliam J expressly preferred the approach of focusing on the ordinary statutory language, referring to MG v PJ and Mond v The Age Company Pty Limited ([107]-[110]). The shared practical point is that the court evaluates reputational impact on the evidence and in the whole circumstances, not merely the inherent gravity of the allegation.
The procedural provisions are also significant. Section 122A allows determination before trial or during trial, and s 122A(7) permits a determination on pleadings where the pleaded particulars are insufficient. Here, however, the application proceeded as a substantive separate question on evidence, rather than only as a pleading challenge ([8], [37]-[40]). The judgment's account therefore distinguishes a case that fails on the face of its pleaded particulars from one in which the plaintiff has had an opportunity to prove the element but has not done so.
At [73], the majority stated: “The distinction between trivial and serious is not binary.” In plain English, the statutory test is not satisfied simply by showing that harm was more than trivial. There is a meaningful threshold to cross, and the particular words used do not replace the need to assess the evidence.
At [79]-[80], the majority focused on the fact that the video did not identify Supaphien. The observation that it did no harm if the grapevine carried only a story about an unnamed person means that publication numbers cannot be considered independently of identification. The plaintiff must establish a connection between the publication and the plaintiff's reputation.
At [88], the majority said that the proceedings failed because Supaphien did not discharge her burden of proof on an element of the cause of action. This was not a finding that the allegations were true, or that the words could not be defamatory. It was a conclusion about the evidence supporting serious harm.
At [143], McWilliam J stated that the evidence, taken at its highest and with the available inferences, did not provide a sufficient basis to prove serious harm. That passage captures the concurrence's distinction between criticism of the primary judge's reasoning and the appellate court's assessment of the result. A flaw in the reasoning did not require allowing the appeal if the evidence still did not establish the element.
The decision is particularly relevant where a plaintiff relies on online publications that do not name them, or where identification depends on context, comments, nicknames or images. The judgment treats proof of who understood the publication to refer to the plaintiff as a central factual question. Engagement figures such as likes, comments and shares do not, by themselves, establish the number of people who identified the plaintiff ([19]-[20], [79]-[83]).
It is also relevant where serious harm is decided as a separate question before a full trial. The case illustrates that, where the issue is substantively determined on evidence, taking the plaintiff's case at its highest does not remove the need to prove the element. The court is not required to fill evidentiary gaps by speculation ([40], [44], [78]).
The judgment may also matter where alleged harm is supported by people avoiding the plaintiff or by a business downturn. It highlights the importance of evidence connecting those consequences to the publication, rather than merely showing that they occurred at some point afterwards. The Court did not hold that such evidence can never establish serious harm. It found the evidence in this case too imprecise to support the necessary inference ([93], [141]-[143]).
Finally, the case concerns publications within a community that is also the plaintiff's customer base. McWilliam J regarded that setting as relevant context, particularly because the appellant's business operated through social media and served members of the Australian Thai community ([115]-[117], [125]). The majority's result nevertheless turned on the proof available, including the lack of evidence establishing the scope of identification and the causal link to reputational impact.
The judgment cites decisions available to it, rather than reporting later treatment of Supaphien itself. Accordingly, it does not establish how courts treated this appeal after 20 February 2026.
On the meaning of “serious”, the majority considered Rader v Haines, where Brereton JA's spectrum formulation had been supported by Macfarlan JA, while Basten AJA expressed a reservation about synonyms and scales ([49]-[55]). The majority concluded that the spectrum approach was permissible, not that it was the only permissible formulation ([62]).
McWilliam J noted that the Queensland Court of Appeal in MG v PJ had adopted Basten AJA's preference for the plain and ordinary meaning of the statutory word, and referred to a similar approach in Mond ([107]-[110]). The reasons also refer to Peros v Nationwide News Pty Ltd (No 3) and Whittington v Newman on inferences and the need to connect a response such as avoidance to the defamatory publication ([134], [140], [142]). These references show that, by the time of this judgment, Australian decisions were discussing both the construction of “serious” and how serious harm may be proved by inference.
The majority also considered Lachaux, the UK authorities, and the approach to proving actual impact. It treated those materials as informative because the Australian model provision was based on the UK provision, while deciding the ACT statutory question by reference to its text and context ([3], [63]-[73]).
The judgment leaves room for debate about how to describe the serious harm threshold. The majority held that the Rader spectrum was not an error of principle and rejected a merely non-trivial threshold. McWilliam J preferred avoiding additional judicial labels and relying on the ordinary statutory meaning ([62], [73], [107]-[110]). The judgment does not impose a precise universal verbal test beyond requiring proof of serious harm to reputation.
The evidence needed to prove identification by inference remains fact-specific. The majority observed that inference may be possible where the plaintiff is named, and McWilliam J stated that identification may be proved by inference in an appropriate case ([83], [140]). The decision does not specify a minimum number of witnesses or a fixed evidentiary formula.
The judgment also leaves open how courts should evaluate social-media reach within a specific community. McWilliam J considered that the community and business context mattered, while the majority concluded that the evidence did not show how many recipients identified the appellant. The case therefore does not make online engagement figures irrelevant; it demonstrates that those figures require evidence connecting circulation to identification and reputational impact ([79]-[83], [115]-[125], [137]-[143]).
Finally, the judgment notes that serious harm may be established by inference but does not define when the inference is sufficiently strong. The practical question remains whether the evidence, considered as a whole, permits a reliable conclusion that the publication caused or was likely to cause serious reputational harm. This is general legal information, not legal advice. Zoe can make mistakes. Before acting on this, relying on it, or using any draft, have a qualified Australian lawyer check it.
Source: the ACT Courts (www.courts.act.gov.au). Converted to plain text by ZoeLaw. Not the official version.
SUPREME COURT OF THE AUSTRALIAN CAPITAL TERRITORY COURT OF APPEAL
Case Title: Supaphien v Chaiyabarn Citation: [2026] ACTCA 5 Hearing Date: 14 May 2024 Decision Date: 20 February 2026 Before: McCallum CJ, Loukas-Karlsson and McWilliam JJ Decision: (1) The appeal is dismissed. (2) The appellant is to pay the respondent’s costs. Catchwords: APPEAL – CIVIL LAW – Defamation – requirement for plaintiff to establish serious harm – where serious harm element brought forward as a separate question – whether primary judge erred in construing “serious” as falling on a spectrum between “significant” and “grave” – model defamation amendment provisions – where alleged defamatory material did not identify the plaintiff by name – proof of publication to only five people – whether serious harm established Legislation Cited: Civil Law (Wrongs) Act 2002 (ACT) ss 122A, 124A, 124B, Ch 9 Civil Law (Wrongs) Amendment Act 2021 (ACT) Court Procedures Rules 2006 (ACT) rr 425, 1521 Defamation Act 2005 (NSW) s 10A Defamation Act 2005 (Vic) s 10A Defamation Act 2013 (UK) s 1 Cases Cited: Bazzi v Dutton [2022] FCAFC 84; 289 FCR 1 David William Pallas & Julie Ann Pallas as trustees for the Pallas Family Superannuation Fund v Lendlease Corporation Ltd [2024] NSWCA 83; 114 NSWLR 81 Lachaux v Independent Print Ltd [2015] EWHC 2242 (QB); [2016] QB 402 Lachaux v Independent Print Ltd [2020] AC 612 MG v PJ [2025] QCA 99 Mond v The Age Company Pty Limited [2025] FCA 442 Newman v Whittington [2022] NSWSC 249 Peros v Nationwide News Pty Ltd (No 3) [2024] QSC 192 Rader v Haines [2022] NSWCA 198 Selkirk v Hocking [2023] FCA 432 Supaphien v Chaiyabarn [2023] ACTSC 240 Whittington v Newman [2024] NSWCA 27 Texts Cited: ACT, Legislative Assembly, Hansard, 12 May 2021 nd David Rolph, Rolph on Defamation (Lawbook Co, 2 ed, 2024) Explanatory Statement, Civil Law (Wrongs) Amendment Bill 2021 (ACT) Parties: Supamas Supaphien (Appellant) Natesai Chaiyabarn (Respondent) Representation: Counsel C Erskine SC w JA Castelan (Appellant) M Richardson SC (Defendant) Solicitors Australian Law Partners (Appellant) Aulich Civil Law (Respondent) File Number: ACTCA 35 of 2023 Decision Under Appeal: Court: ACT Supreme Court Before: Curtin AJ Date of Decision: 1 September 2023 Case Title: Supaphien v Chaiyabarn Citation: [2023] ACTSC 240 Court File Number: SC 322 of 2022 McCALLUM CJ and LOUKAS-KARLSSON J: Introduction
Ms Supamas Supaphien conducted a business described as providing “a personal shopping service connecting consumers and sellers of luxury goods on the Facebook social media platform”. In September 2021, a customer, Ms Natesai Chaiyabarn, approached Ms Supaphien in the presence of two others at a Thai Buddhist temple and accused Ms Supaphien of selling her a fake Chanel handbag. The following day, Ms Chaiyabarn posted a video on Facebook which, without naming Ms Supaphien, told Ms Chaiyabarn’s story (of having purchased what she later came to believe was a fake Chanel handbag) as a cautionary tale. The video was “liked” 185 times, received 144 comments and was shared 34 times.
Ms Supaphien brought proceedings for defamation against Ms Chaiyabarn. The proceedings initially also included a second defendant, who had participated in the on-line conversation, but the claim against her was resolved.
Proceedings for defamation in the Territory are governed by Ch 9 of the Civil Law (Wrongs) Act 2002 (ACT). Chapter 9 was amended in 2021 to enact the Model Defamation Amendment Provisions 2020: see Explanatory Statement, Civil Law (Wrongs) Amendment Bill 2021 (ACT). One of the more significant reforms proposed by the Model Provisions was the introduction of a test of “serious harm” as an element of the cause of action for defamation. The serious harm threshold was modelled on s 1 of the Defamation Act 2013 (UK). It has been introduced in every jurisdiction in Australia apart from Western Australia. In the Territory, the serious harm provision is s 122A of the Civil Law (Wrongs) Act, which came into force on 1 July 2021: see Civil Law (Wrongs) Amendment Act 2021 (ACT).
Section 122A(1) provides: It is an element (the serious harm element) of a cause of action for defamation that the publication of defamatory matter about a person has caused, or is likely to cause, serious harm to the reputation of the person.
Section 122A(3) confers the task of determining “whether the serious harm element is established” on the judicial officer (as opposed to the jury). In characterising serious harm as an element of the cause of action, the legislators plainly intended that the onus of satisfying the judicial officer of that element falls to the plaintiff. The equivalent provision in NSW (s 10A of the Defamation Act 2005 (NSW)) has been interpreted in that way: see Newman v Whittington [2022] NSWSC 249 at [47] (Sackar J); applied in respect of the Victorian provision (also s 10A of the Defamation Act 2005 (Vic)) in Selkirk v Hocking [2023] FCA 432 at [5] (O’Callaghan J).
Ms Chaiyabarn applied to have the issue of serious harm determined before the trial, as allowed under s 122A(4). Importantly, the procedure she invoked in making that application was to bring forward a separate question, as allowed under r 1521 of the Court Procedures Rules 2006 (ACT). The application foreshadowed that the issue might be determined by reference only to the pleadings (“whether the plaintiff’s pleaded particulars are sufficient to establish the element of serious harm within the meaning of s 122A of the Act”).
Section 122A does contemplate that an application to have the issue of serious harm determined before trial can be determined on the pleadings, but only where the determination is that the element is not established. Thus, s 122A(7) provides “the judicial officer may determine the serious harm element is not established on the pleadings without the need for further evidence if satisfied that the pleaded particulars are insufficient to establish the element”. In such a case, the proper test would be akin to that which is applied in a strike-out application where the question is whether the pleading discloses a reasonable cause of action. Where a pleading fails that test, more than one remedy is available. The pleading can be struck out with leave to replead. Alternatively, the proceedings can be dismissed if, for example, it is not apparent that the deficiency can be rectified.
It was accordingly open to Ms Chaiyabarn to seek to have the issue determined by reference only to the pleadings. However, the application did not in fact proceed in that way. A timetable was put in place for evidence to be filed. The primary judge (Curtin AJ) was careful to ascertain at the outset of the hearing that the application before his Honour was not confined to an application in the nature of a strike-out application directed to the content of the pleadings. Ms Chaiyabarn did not abandon strike-out as an alternative position, but the relief sought included an order dismissing the proceedings. The primary judge specifically confirmed with both counsel that he was being asked to determine the issue of serious harm as a separate question and to do so having regard to the evidence each had put before the Court. The separate question was then framed in the judgment (with respect, appropriately) by reference to the language of the section, as follows: Has the plaintiff established that publication of the pleaded defamatory matters has caused, or is likely to cause, serious harm to the reputation of the plaintiff.
The primary judge determined that serious harm was not established. Accordingly, his Honour answered the separate question “No” and made orders dismissing the proceedings and requiring the plaintiff to pay the defendant’s costs: Supaphien v Chaiyabarn [2023] ACTSC 240 at [171].
Ms Supaphien appeals from those orders. We have determined that the appeal must be dismissed for the following reasons. For convenience, the parties are referred to throughout this judgment either by name or by their respective roles in the appeal; Ms Supaphien as the appellant and Ms Chaiyabarn as the respondent.
The appellant’s case
The respondent’s application was determined by reference to the appellant’s amended statement of claim filed on 12 December 2022. The causes of action pleaded in that document may be summarised as follows.
As already noted, the appellant conducted the business of providing a personal shopping service for luxury items on Facebook. She pleads that she is known in the Thai community in Australia by the nickname “Jing Jo”, a Thai phrase meaning kangaroo, and that she uses that nickname to identify herself on Facebook and other social media platforms.
First matter complained of: slander
The first defamatory matter complained of is an oral statement alleged to have been made at a Thai Buddhist temple in the Territory. The respondent is alleged to have approached the appellant and said to her words in the Thai language which translate to the following words in the English language: You sold me a fake Chanel handbag. I’m going to return it and I want my money back.
It is alleged that the statement was made in the presence of two others, Vira Vumbaca and Nucy Phitsinee Wilkinson. Ms Wilkinson became the second defendant. The pleading does not specify whether those two other people understand the Thai language but it may be inferred that at least Ms Wilkinson does as she later posted comments on Facebook in Thai.
The appellant responded: The bag is not fake, I bought it from the Chanel shop.
The pleading does not specify whether those words were spoken in English or whether that is also a translation from the Thai language. The appellant immediately called the Chanel store in Melbourne. The sales assistant spoke directly to the respondent and said, “the handbag is not a fake, Sue purchased it from this store”. It may be inferred that the reference to “Sue” is a reference to the appellant.
The appellant alleges that the oral statement at the temple (the first matter complained of) conveyed the following defamatory imputations: Imputation A: The Plaintiff had sold a handbag to the First Defendant which she had passed off as a genuine Chanel handbag when it was a fake or counterfeit product; Imputation B: The Plaintiff had defrauded the First Defendant. Second matter complained of: the video clip
The second defamatory matter complained of is the video clip allegedly uploaded by the respondent on her Facebook page. The video was not in evidence before the primary judge or this Court. A translation of its content as provided in the pleadings is as follows: One person sells brand names in Canberra. I bought myself a birthday present. I have decided to buy from this seller because she looks high profile, high class, reliable, speaks well, and so I decided to buy her. At that time I didn’t think the product is fake. After seven months past, the quality of the bag has changed. The shape has changed. My bag started to show damage because it’s fake. The product started to change. The colour started to come off the chain. The colour started to come off. The quality was poor. I didn’t think Chanel would make such a bad quality product. I sent my bag to a reliable company and they determined that the bag is counterfeit. The reason why I haven’t gone to check with Chanel is because they don’t confirm whether is the product genuine or fake. So I went back to talk to the seller, the seller would not return my money or take responsibility. She accused us of swapping the bag with a fake one. So, I want to come out and tell everyone that before thinking of buying things for over eight thousand dollars please check the seller’s history first. I don’t want this kind of thing to happen to other people again. At this time, I won’t provide further information because I have already reported it to the police, lawyers and other government agencies and I will continue to take legal action. If there is any progress I will update.
Fourth matter complained of: exchange between Wilkinson and Gaylard
The fourth matter consists of an exchange of comments uploaded in the same forum between Ms Wilkinson and a Facebook user named Chaisuda Gaylard. The respondent is alleged to be jointly and severally liable for those comments because she uploaded the video, received notifications when comments were posted and did not take down the comments.
The comments exchanged were as follows: Gaylard: Too bad. Heard of her story when she lived in Melbourne and moved to Canberra but I don’t know if it’s the same person? Wilkinson: Chaisuda Gaylard need to be exposed… Gaylard: Nucy Wilkinson Is it better to wait for karma to return back to her? Wilkinson: Chaisuda Gaylard She does something wrong again and again. She doesn’t care… (Emphasis in original.)
The appellant alleges that the comments conveyed the following imputations by way of false innuendo: Imputation F: The Plaintiff has sold many counterfeit handbags; Imputation G: The Plaintiff does not care that she is defrauding people. Fifth matter complained of: screenshot sent via Messenger
Finally, the appellant complained of a defamatory screenshot posted by the respondent. A Facebook user named Piyachay Srijan asked the respondent for information about the identity of the person who had sold her the fake Chanel bag. In response, it is alleged that the respondent sent a screenshot via the Facebook Messenger application with two photographs of the appellant. The screenshot included the appellant’s Facebook profile name, which was also her nickname “Jing Jo”.
The imputations allegedly conveyed by the screenshot are: Imputation H: the Plaintiff is the person who sold a counterfeit handbag to the First Defendant; Imputation I: the Plaintiff is the person who has been identified in the Defamatory Video and the First Defamatory Comment and the Second Defamatory Comment as being dishonest. Scope of publication and harm alleged
The pleading accordingly names only four people alleged to have identified the appellant from the matters complained of: Vira Vumbaca, Ms Wilkinson (the former second defendant), Chaisuda Gaylard and Piyachay Srijan. Vira Vumbaca and Ms Wilkinson heard the original slander and then also heard the conversation with the Chanel representative who supported the appellant’s contention that the handbag was not fake. In further particulars filed on 12 December 2022, the appellant identified one further person alleged to have identified her from Ms Wilkinson’s comment. In total, only five people are specified as having identified the appellant from the matters complained of.
Ms Supaphien did not call evidence from any of those people or any other person who viewed the video and understood it to refer to her. The “subsequent comments” referred to in paragraph 13(e) of the pleading (set out above) were not tendered. Without such evidence, the important issue of the scope of publication (beyond the five people named) was left to inference based on vague, high-level particulars.
The nature of the application before the primary judge
As already noted, the task of determining whether serious harm is established is conferred on the judge, even if the proceedings are being heard or to be heard with a jury. Section 122A allows considerable flexibility as to how and when the judge is to make that determination. The issue of serious harm can be determined “at any time before the trial for the proceedings commences or during the trial”: s 122A(4). A determination that the serious harm element is not established can be made on the pleadings: s 122A(7). In other words, the judge can dismiss a claim at any stage on the basis that the pleaded case, if proved, could not establish the serious harm element. Any such determination would, as a matter of fairness if not practical necessity, require the judge to take the pleading at its highest (as if proved).
Alternatively, the matter can be determined on evidence, whether before the trial or at any stage during the trial. Where a party makes an application to have the matter determined before the trial, subs (5) requires the judicial officer to determine it “as soon as practicable before the trial commences” unless satisfied of special circumstances justifying its postponement. Subsection (6) provides guidance as to the kind of “special circumstances” that might justify postponement of the issue once raised by a party. That provision states that, in deciding whether there are special circumstances, the judge may take into account the cost implications for the parties; the resources available to the court at the time and the extent to which establishing the serious harm element is linked to other issues for determination during the trial for the proceedings. In that way, the section seeks to strike a fair balance between proportionality and the interest of the claimant in having a full opportunity to put forward their case.
The clear purpose of those provisions is to provide a fair, flexible mechanism for protecting the parties from being put to the expense of a full hearing, only to have the claimant fail on the element of serious harm. The vice addressed by those provisions was the increasing burden on the courts and the community of defamation cases arising from publications of limited circulation in which the costs of the trial and the burden on the courts were vastly disproportionate to the interests at stake.
In the present case, as already explained, the issue was brought forward as a separate question. The primary judge recorded at [3] of his judgment that the appellant did not oppose that course. Nor did she attempt to establish special circumstances warranting the postponement of the issue until the trial. This meant that each party was expected to put forward their case on the issue of harm and that the judge’s determination as to whether serious harm was established would bind the parties as on a final hearing. The application was not confined to one that could be determined on the pleadings, as in the case of a determination under s 122A(7) or, for example, a strike-out application under r 425 of the Court Procedures Rules.
Grounds of appeal
Application of the wrong test
The first ground addressed in the appellant’s written submissions is the alleged application of the wrong test for “serious harm”. This topic was said to include grounds 1, 2, 3 and 9. However, it is ground 2 that specifies the error complained of, as follows: The learned trial judge erred in concluding that serious harm for the purposes of section 122A of the [Civil Law (Wrongs) Act] meant more than substantial harm (Reasons [114]).
The primary judge set out the principles for judging serious harm at [98]-[121] of the judgment. The discussion begins at [98] with the unexceptionable proposition that s 122A is concerned with harm to reputation (as distinct from hurt to feelings). His Honour noted at [103] that the plaintiff bears the onus of proving the serious harm element, citing the decision of Sackar J in Newman referred to above. He further noted at [103] and [104] that the plaintiff must prove serious harm as a fact in every case, citing Newman at [69], and that this must be proved by evidence of the actual impact of the publication on the person’s reputation, citing Newman at [65], [68].
The passage impugned by ground 2 is at [114], where his Honour said: ‘Serious’ sits on the spectrum above ‘substantial’ but below ‘grave’. ‘Serious’ means ‘significant’ or ‘worrying’, not ‘slight’ or ‘negligible’. It means more than merely ‘substantial’: Rader at [27].
The reference to Rader is a reference to the decision of the New South Wales Court of Appeal in Rader v Haines [2022] NSWCA 198. The primary judge can hardly be criticised for simply reciting a statement of an intermediate Court of Appeal. However, the appellant argues that the passage cited is wrong and should not be followed.
The plaintiff in Rader sued his former parents-in-law over an email they had sent to his parents, who were in England. The headnote to the judgment describes the email as “ostensibly concerning the education of their mutual grandchildren, but also canvassing the ongoing acrimony between the appellant and his ex-wife”. The primary judge (Gibson DCJ) dismissed the case on two grounds including failure to establish the element of serious harm. Mr Rader appealed.
The appeal was heard by Macfarlan JA, Brereton JA and Basten AJA. The main judgment was written by Brereton JA, with whom Macfarlan JA agreed without qualification: at [1]. Basten AJA also agreed but with a qualification as to the very part of the judgment cited by Curtin JA in the present case. The passage in question concerns the meaning of the term “serious”. Brereton JA said at [27]: Courts appear so far to have avoided endeavouring to explain the word ‘serious’. Indeed, it has been said that ‘serious’ is an ordinary word in common usage. But in my view there is utility in giving some further explanation of its content. It is used in the sense third mentioned in the Oxford Dictionary definition, namely ‘significant or worrying because of possible danger or risk; not slight or negligible’, for which the example given is ‘she escaped serious injury’, and the synonyms include ‘severe’ and ‘grave’. In Lachaux, whereas the Court of Appeal had considered that the new statutory test was the same as the common law ‘tendency to cause substantial harm’ test, in Thornton v Telegraph Media Group, albeit raised to the level of ‘serious harm’, the Supreme Court confirmed that it ‘raises the threshold of seriousness above that envisaged in Jameel (Yousef) and Thornton’. At least, this shows that ‘serious’ involves more than merely ‘substantial’. In Monroe v Hopkins, Warby J concluded that ‘whilst the claimant may not have proved that her reputation suffered gravely, I am satisfied that she has established that the publications complained of caused serious harm to her reputation and met the threshold set by s 1 of the 2013 Act’. In my opinion, ‘serious’ harm sits on the spectrum above ‘substantial’ but below ‘grave’. Importantly, there can be harm which, though substantial, does not reach the level of serious harm. (Footnotes omitted.)
Failure to consider the context of the publications and the grapevine effect
The next topic addressed in the appellant’s written submissions is the trial judge’s alleged failure properly to take account of the context of the publications and the grapevine effect. This topic is directed to grounds 4, 5 and 8 in the Notice of Appeal.
Ground 4 alleges numerous errors on the part of the primary judge including his Honour’s remarks that the ordinary reasonable reader of social media reads in a somewhat informal manner and that social media is a casual medium in which the reader “reads and passes on”; his observation that there was no evidence that the people to whom the various comments were published were known to the appellant; his observation that there was no evidence that the 34 people the appellant claims shunned and avoided her did so because of the publications; the absence of any evidence from the appellant as to publication and downloading or any platform of facts from which an inference of downloading could be drawn; the limited scope of the publications and downloading; his Honour’s view that, in the world of social media, the publications were not extensively available and that the extent of publication was limited; his Honour’s view that the harm alleged could only be past harm, given the transitory nature of material posted on social media; and the appellant’s failure to demonstrate that the decline in profit of her business was not due to other possible causes than the defamatory publications.
Ground 8 alleges that, “considering the extent of the publication of the pleaded defamatory matters”, the primary judge erred in failing to take account of the grapevine effect.
These grounds proceed on a misapprehension of the process adopted by the primary judge, the correctness of which the parties confirmed at the outset of the hearing before his Honour. In her written submissions, the appellant states, “it was agreed that the appellant’s evidence and pleadings were to be taken at their highest. Accordingly, there was no need for the appellant to have adduced evidence that specifically supported the allegations contained in her pleadings”. However, as already explained, the parties asked the primary judge to determine the separate harm element as a separate question in the proceedings. The fact that, for the purpose of considering whether the proceeding was liable to be dismissed in limine, the pleadings were to be taken at their highest did not obviate the need for proof of the serious harm element of the cause of action.
The submissions in support of this aspect of the appeal also fail to grapple with a fundamental difficulty faced by the appellant, that the principal defamatory publication (the video clip) did not name her. The repeated reference in the submissions to the fact that the video was “liked” 185 times, received 144 comments and was shared 34 times overlooked this fundamental hurdle. Not only was the primary judge not required to assume or infer that all of those people understood the video clip to refer to the appellant, on the evidence before his Honour, it was not open to draw any such inference. The evidence was virtually silent on an issue as to which the appellant bore the onus of proof.
Irrelevant considerations
The final topic addressed in the submissions is ground 7, which contends that the primary judge erred by taking certain “irrelevant considerations” into account. This ground of appeal must also be rejected. The first “irrelevant consideration” specified in ground 7 is the primary judge’s finding that the ordinary reasonable reader of social media reads in an informal manner, even though it was admitted that the matters complained of conveyed the imputations pleaded by the appellant.
The appellant submitted that the qualities attributed to the ordinary reasonable reader of social media are only relevant to the task of determining whether or not an imputation is conveyed. It was further submitted that in any event an allegation of fraud in an online communication is not something that a person would skim over and ignore.
It may be accepted that the informal nature of social media is relevant to the question of defamatory meaning: Bazzi v Dutton [2022] FCAFC 84; 289 FCR 1. However, as submitted on behalf of the respondent, it is not irrelevant to the task of determining serious harm. As the point was put by senior counsel for the respondent, “it’s hard to see how such an obviously true observation can be appropriate and orthodox over here and then heretical when you move to the next step”.
The remaining matters relied upon in support of ground 7 were observations by the primary charge about matters as to which there was no evidence. It is wrong to characterise the primary judge’s consideration of the absence of evidence as an irrelevant consideration. The whole point of the separate question was to determine a matter as to which the appellant bore the onus of proof. The absence of evidence on the matters specified by the primary judge was highly relevant. The fact that some of those matters may have been difficult to prove did not obviate the need to prove them.
Orders
For those reasons, none of the grounds of appeal is made out. Accordingly, the orders we propose are: (1) Dismiss the appeal. (2) Appellant to pay the respondent’s costs. McWILLIAM J:
I have had the benefit of reading the reasons of McCallum CJ and Loukas-Karlsson J in draft form. I agree that the appeal should be dismissed, because fundamentally the evidence in this case was insufficient to establish that the publications in question caused an impact to the appellant’s reputation that constituted serious harm. The comprehensive nature of their Honours’ reasons, including the discussion of the relevant legislative provision, its historical context and purpose of the amendment that introduced the element of serious harm, permit these reasons for arriving at the same result to be condensed.
The Court’s task
Starting with the task on appellate review, as a finding of serious harm is an evaluative and not a discretionary finding, the correctness standard applies. As stated in MG v PJ [2025] QCA 99 (MG) at [67]: A consideration of the evidence, including available inferences, and the whole of the circumstances either establishes the serious harm element or it does not.
This was a case where the appellant’s evidence was by way of affidavit and there was no cross-examination. The appellate court is in as good a position as the primary judge to draw inferences from the evidence before the court. To succeed, the appellant had to establish not just error in the primary judge’s reasoning, but that the finding itself as to serious harm was incorrect. That was recognised in the appellant’s grounds of appeal, which challenged the ultimate conclusion as well as the reasoning process.
Complaints on appeal
The complaints on appeal were directed to a failure to apply the correct test, which was said to be a matter that permeated the reasons, as well as specific aspects of the reasoning that, taken together, were also broadly indicative of a failure to properly approach the question of serious harm. These specific aspects were: the failure to take proper account of the context of the publications and the social media medium, a failure to take proper account of the uncontradicted evidence and the taking into account of a number of irrelevant considerations.
For efficiency, I have dealt with the various errors under the one umbrella in explaining why I would find in the appellant’s favour in terms of error in the reasoning process on the question of whether serious harm to the appellant’s reputation had been established. Ultimately though, establishing error does not change the result because I consider that the primary judge’s conclusion – that the element of serious harm had not been established on the evidence – was correct.
The “serious harm” element
The appellant’s complaint that the primary judge adopted and applied an improper test for what constituted serious harm derived from his Honour stating at [114] of the primary judgment: ‘Serious’ sits on the spectrum above ‘substantial’ but below ‘grave’. ‘Serious’ means ‘significant’ or ‘worrying’, not ‘slight’ or ‘negligible’. It means more than merely ‘substantial’: Rader at [27].
As explained in the reasons of McCallum CJ and Loukas-Karlsson J, that was a reference to a passage from Rader v Haines [2022] NSWCA 198 (Rader) where Brereton JA (Macfarlan JA agreeing) discussed the dictionary definition of “serious”, and synonyms for the word, ultimately concluding at [27] that “serious” harm sits on the spectrum above “substantial” but below “grave”. The full extract is set out in the reasons of McCallum CJ and Loukas-Karlsson J.
In Rader, Basten AJA dealt with the question of what constitutes “serious harm” at [90]- [91]: 90. I agree with the orders proposed by Brereton JA. With one minor reservation, I also agree with his reasons. 91. The reservation relates to the exegesis on the meaning of “serious” in s 1 of the Defamation Act 2013 (UK). There is a risk in seeking synonyms, which may later be treated as valid replacements for the ordinary English word adopted by the Parliament. There is also a risk in seeking to place the term on a scale, between other terms of equal imprecision. The critical concept is “serious harm to the reputation of the claimant”; it is that to which the court is required to attend by reference to the evidence of a range of matters. Analysis of individual component words is apt to distract from that inherently impressionistic exercise.
Subsequently the Queensland Court of Appeal has adopted the approach of Basten AJA in MG, with Flanagan JA stating at [83] (citations omitted): The approach of Basten AJA has received judicial support in subsequent decisions. I respectfully agree with his Honour. The plain and ordinary meaning of the word “serious” should be adopted as a matter of course. This Court need not advance any alternative nomenclatures when construing this term.
Brown JA and Davis J agreed with his Honour.
Similarly, in Mond v The Age Company Pty Limited [2025] FCA 442 (Mond), Wheelahan J stated at [373]: The term “serious harm” in s 10A(1) of the Defamation Act is not defined. It is an open-textured phrase where no precise or universal test has been authoritatively stated at appellate level in relation to the Australian legislation. In Rader v Haines [2022] NSWCA 198 (Rader) at [27] Brereton JA, with whom Macfarlan JA agreed, stated in the context of s 1 of the United Kingdom Act that “serious” sits on a spectrum that is above “substantial”, but below “grave”. However, at [91] Basten AJA stated that there was a risk in seeking synonyms of “serious” and in seeking to place the term on a scale between other terms of equal imprecision. In Peros at [95], Applegarth J agreed with Basten AJA. I also agree with Basten AJA, noting the accepted dangers of substituting judicial language for the text of legislation: see Shi v Migration Agents Registration Authority [2008] HCA 31; (2008) 235 CLR 286 at [92] (Hayne and Heydon JJ).
The primary judge erred in the approach taken
The reasoning of the primary judge demonstrates that his Honour did have an overall appreciation of the evaluative task or the critical concept. His Honour stated at [107]: The requirement for serious harm to reputation is concerned with actual or likely reputational damage – that is, the impact of the imputation, in all the circumstances, on the [appellant’s] reputation – arising from a combination of the inherent tendency of the words and their actual impact on those to whom they were communicated: Rader at [28].
However, the actual consideration of the evidence in this case was not consistent with a reasoning process that accorded with the approach or test his Honour had set out. As will be explained, at times, it is apparent that the primary judge misdirected attention to determining where isolated facts fell on a spectrum of harm (applying an incorrect test), rather than assessing the causal impact of the imputations to the appellant’s reputation in the whole of the circumstances, and whether that impact was one of serious harm. There were also instances where the primary judge embarked upon considerations that would have been irrelevant if his Honour had been applying the correct approach. These difficulties in the primary judge’s reasoning were not isolated, or so few or inconsequential, as to be immaterial overall to the primary judge’s evaluation. That is why I have concluded that the primary judge erred in the reasoning process.
The appellant’s membership of the Australian Thai community
One of the circumstances of this case involved a cultural element, in that the appellant was a member of the Australian Thai community. In that regard, the primary judge stated at [153]: The [appellant] gave evidence she was a member of the Australian Thai community, but there was no evidence as to how large that community was (at the relevant time) nor the [appellant]’s degree of involvement with that community.
The appellant submitted that these considerations were irrelevant, and I agree that the size of the Australian Thai community or how involved the appellant was in that community were not really to the appellant’s point. This was the appellant’s community, whatever its size and whatever opportunities for social connection or involvement it provided.
The appellant’s evidence was that she was the owner and operator of a social media-based luxury goods business since May 2020. The customers of her business were members of the Australian Thai community. Her reputation was vital to the success of her business. The Facebook posts in question involved a post from a community member on her Facebook page, and republications of the material to subsets of that community – “Thais in Canberra” and “Na Rak Na Jab Tong” Thai community Facebook groups.
As I read the import of that evidence, it was that this community was where her reputation was vulnerable to harm. This was the appellant’s realistic target market for selling expensive handbags and why her reputation in that community for honesty as a seller of authentic luxury products mattered.
Further, the existence of the community was but one fact in a range of matters. Directing attention (and criticism) to a lack of evidence about the size of the Australian Thai community itself and the degree of the appellant’s involvement in it suggests that the primary judge did not consider or appreciate the evidence as part of the whole circumstances from which to evaluate the impact to reputation and whether it was serious.
Social media nature of the publication
Another indicator that the primary judge diverted from the proper evaluative approach or did not apply the correct test is seen from the primary judge’s consideration of how social media communications will be read by the ordinary reasonable reader and how the material would be interpreted in a casual medium.
The relevant part of the reasons is at [119]: Social media publications will be read by the ordinary reasonable reader in a somewhat informal manner: Randell v McLachlain at [18]. The search for how an ordinary reasonable reader would interpret a tweet or Facebook post should reflect the circumstance that: social media is a casual medium; posts are in the nature of conversation rather than carefully chosen expression; and these mediums are pre-eminently ones in which the reader reads and passes on: Stocker v Stocker [2019] UKSC 17; [2020] AC 593 per Lord Kerr JSC at [43], cited with approval in Bazzi v Dutton [2022] FCAFC 84; 289 FCR 1 per Rares and Rangiah JJ at [29].
This idea is repeated at [151]: In terms of harm to the [appellant]’s reputation, all the publications other than the First Defamatory Statement were made on social media. As has been said (see [119] above), the ordinary reasonable reader of social media reads in a somewhat informal manner and social media is a casual medium in which the reader reads and passes on.
The appellant submitted that the primary judge erred because he failed to appreciate that online publication can be serious and can have serious consequences. It was submitted that his Honour appeared to underestimate the impact that social media can have, especially within a particular community.
I accept that submission. If his Honour was by that reasoning intending to convey that social media itself is a casual forum, such that the impact could not be “serious” because the reader merely “reads and passes on”, that was an erroneous statement. Otherwise, it is unclear why this formed part of the primary judge’s reasons on this separate question. The fact that social media publications were read in an informal manner was not a matter relevant to the serious harm evaluation purportedly being undertaken at that point in the primary judge’s reasons.
The respondent submitted that these were unexceptionable statements, and I agree they may well be, if they were viewed as stand-alone propositions of law dealing with what imputations were conveyed or the natural and ordinary meaning, taking account of how they might be read. But the difficulty is that on the separate question, the imputations themselves were accepted by the respondent to be serious. The critical distinction was discussed in Rader in relation to cognate legislation at [19]: Section 1 of the UK Act says that before a statement is defamatory it must have caused or be likely to cause “serious harm to the reputation of the claimant.” This requirement concerns the seriousness of the harm to the reputation, not of the imputation … the application of the test [for serious harm] is to be determined by reference to the actual facts about the impact of the imputation, and not just the meaning of the words; although the meaning is necessarily part of the impact, because all else being equal a grave imputation will cause more harm than a trivial one, it is only part of the equation, because the gravest imputation will cause no harm if it has no impact. …
The primary judge’s discussion of identification
The primary judge dealt with identification of the appellant at [152]: There is no evidence that the people to whom the various comments were published were known to the [appellant], or at least known so well that they would not have been disposed to believe the imputations.
The respondent accepted that the legal terminology here was wrong but submitted that this should be taken as meaning there was no evidence of identification. I do not accept that submission because the language appeared (at least to me) to be deliberate, in that it echoes what the primary judge had earlier stated as an applicable principle at [110]: A grave imputation may not result in serious harm, for example where the publication is to a small number of persons well acquainted with the plaintiff who are not disposed to believe it, and any impact of the imputation on the plaintiff’s reputation is transitory or ephemeral: Rader at [28]. Serious harm need not be permanent harm: Rader at [32].
Identification was an issue about which there was little and ultimately insufficient evidence, as McCallum CJ and Loukas-Karlsson J have explained. The point is simply that I think the primary judge intended to refer back to the principle his Honour had earlier stated, and do not accept this should be read as addressing the evidence as to identification that did exist, such as it was, particularly where no reference was made to the appellant’s evidence as to the use of the nickname or the reference to kangaroo. In any event, in my view, that aspect of the reasons did not assist in demonstrating an application of the proper reasoning process or test.
The primary judge’s treatment of the evidence that 34 people avoided the appellant
A separate demonstration that the adoption of a spectrum of harm distracted his Honour from the proper approach to the task is seen in an example his Honour used when discussing the evidence that 34 people avoided the appellant following the publications. His Honour made the point at [154] that no evidence was given of any specific instance of avoidance, or the level of interaction the appellant had enjoyed with the 34 people prior to the publications, compared with afterwards. The primary judge stated at [155]: For example, if the [appellant] only had rare, social media interactions with those 34 people before the publications, the fact that she had none subsequently (if that were the fact) would not support a claim that any harm was “serious”, but would rather suggest that the harm was minor, slight or inconsiderable.
The appellant was not cross-examined. That 34 people avoided the appellant following the publications was not challenged. What inference may be drawn from that fact may be debateable, and I will return to that below. Here, it may fairly be accepted that what his Honour was endeavouring to explain was the point that a level of detail was missing in the evidence. The circumstances of that number of people shunning or avoiding the appellant post publication were important, because such information assists the court in working out whether the inference to be drawn from the fact that such avoidance occurred was attributable to a group cancel culture that had arisen as a result of the publications, or whether there was some other cause.
Why the example is significant, though, is because of the reasoning process articulated in it. The example demonstrates: (a) the piecemeal approach to serious harm being undertaken, and (b) a focus on the personal impact to the appellant by reference to how much contact she had previously had with the members of the Thai community that were said to have avoided her and whether that contact was via a social media interaction.
As an exposition of the primary judge’s understanding of the task, it indicates a flawed reasoning process that had wandered away from considering the facts presented and the inferences to be drawn from them as part of an overall matrix of circumstances from which to assess the impact upon a person’s reputation.
Factual error and the grapevine effect
The primary judge stated at [156]: Further, no evidence was given as to whether any of those 34 people were any that had liked, commented, or shared one or more of the publications. That is, there is no evidence that any of those 34 people avoided the [appellant] because of the publications.
In the concerns notice dated 28 September 2021, three of the 34 people named by the appellant had commented on the video posted in a manner that suggested they had probably identified the appellant (for example, comments about the seller moving from Melbourne to Canberra and questioning if that was the same person, to which the respondent replied “need to be exposed” and “she does something wrong again and again, she doesn’t care”, and a separate comment by a different person among the 34 named that “you are big fan of her”). I acknowledge the respondent’s point that not all the matters in the concerns notice found their way into the pleaded case and the evidence was slim at best (discussed below), but I have traversed this as concerning because the comment itself overlooks that serious harm to reputation may be inferred: Whittington v Newman [2024] NSWCA 27 at [44] and the cases there-cited. Again, by itself, it is not fatal to the reasoning, and the primary judge had earlier noted (at [118]) that an inference of download can properly be drawn from a platform of facts. However, in combination with the other difficulties in the reasons, it does give rise to uncertainty whether this is another example of the primary judge misapplying the correct test or taking the wrong approach.
A separate troubling aspect of the primary judge’s reasoning appears at [158], where his Honour stated: There was no evidence…led by the [appellant] as to publication, downloading, nor any platform of facts from which an inference of downloading could be drawn. That is, there were allegations made of likes etc, but no evidence was presented to prove those facts.
That was problematic as a matter of fact, in that there was such evidence, as accepted by the respondent during the hearing. However, passing over those factual matters, of greater significance is the reasoning that then followed at [159]-[162]: 159. Further, and importantly, in terms of duration of harm, there was no evidence as to the dates of the likes, comments and shares. 160. However, even if evidence was led proving those matters as pleaded, I would not have found serious harm. The publications and downloading were limited. The First Defamatory Comment was published to two people (one of whom was the second defendant) and occurred contemporaneously with the [appellant]’s denial and contemporaneously with someone from the Chanel store in Melbourne saying, in the presence of the two other people, that the bag was genuine. 161. The Defamatory Video received 185 likes, 144 comments and 34 shares which, in social media terms, is not a significant number. The other publications were published or downloaded no more than those numbers suggest. In the “read and pass on” world of social media, the publications were not extensively available and therefore, in my view, the extent of publication was limited. 162. The harm alleged could only, realistically speaking, be past harm given the transitory nature of material posted on social media.
Serious harm to the appellant’s reputation was not established on the evidence
Putting the primary judge’s erroneous approach and reasoning to one side, the evidence was nevertheless insufficient to establish that what was published caused or was likely to cause serious harm to the appellant’s reputation.
I would not go as far as to say that the introduction of the serious harm test necessarily imposed a burden on a plaintiff to find witnesses who will say they saw the matter complained of and understood it to refer to the plaintiff. Serious harm may be established by fact or as a matter of inference, and such an expectation should be tempered by the reality that a plaintiff may struggle to identify or produce such evidence: MG at [91], citing Peros v Nationwide News Pty Ltd (No 3) [2024] QSC 192 (Peros) at [86]. Identification too may be proved by inference in an appropriate case. However, proof of the element was required and on a separate question being substantively and not summarily determined, it was not enough to simply plead the case and rely on the gravity of the imputations, really without much more.
The evidence as to financial impact on the appellant’s business was scant and imprecise, as the respondent submitted. It did not permit any inference that a downturn in business over one month was a result of damage to reputation. As the primary judge recorded, the financial impact evidence was led not to prove any pecuniary loss but only in support of a conclusion that the appellant was shunned and avoided. Even to that extent, it should have been well within the appellant’s abilities to set out what those figures represented (for example, the product that was sold or how many sales were made in a month), and possibly whether the sales were to members of the Australian Thai community. At the very least, identifying whether the limited sales in September occurred before or after the publications on the Facebook pages in September 2021 may have assisted in justifying the inference the appellant sought to draw from that financial evidence. The appellant’s own evidence was that she ceased her business “after September” in 2021, through a loss of confidence, which makes it difficult to prove that, to the extent any inference of impact by way of a reduction in sales even for one month was available, such impact was any more than a passing dip, let alone that it was attributable to the impact of the publications.
Similarly, returning to the unchallenged evidence that 34 people avoided the appellant following the publications, no details were provided as to the timing and conduct that constituted avoidance. The lack of detail other than the names of the people left the appellant with little justification for the inference she sought to draw as to the reason for their avoidance being impact to her reputation. Given the lack of direct evidence from those persons linking the publications (whether personally viewed or learned about through the grapevine) to their estimation of the appellant, and the difficulties with the financial evidence explained above, the avoidance evidence became critical. As stated by Applegarth J in Peros at [80] (cited in MG at [91]): …Care is required in such cases that the abuse, criticism or other response (such as being avoided) is caused by the publication of the defamatory matter, and not by other publications. …
Conclusion
Supaphien
Chaiyabarn [2026] ACTCA 5
(2008) 235 CLR 286
Appeal dismissed. The appellant is to pay the respondent's costs: [98].
It may be noted that those words do not identify the appellant by her name, or indeed by any nickname. Accordingly, apart from being confined to people who understand the Thai language, Ms Supaphien’s cause of action is further confined to publication to persons who understood the video to refer to her. The pleading alleges that the video received 185 likes, 144 comments and 34 shares of that post. It is then alleged that the video “was downloaded and comprehended by, and thereby published to, at least 185 people on the Facebook social media platform”. While it may be open to infer that the 185 people who “liked” the video understand the Thai language, it cannot be inferred or assumed that all 185, or indeed any number of them, understood the video to refer to the appellant.
The pleading further alleges that the video was republished by the respondent on two other Facebook groups, where it received a total of 31 further comments, and further republished by being shared by 37 members of one of those Facebook groups. Again, it cannot be inferred or assumed that all or any number of the people involved in those communications understood the video to refer to the appellant. That was a matter as to which she bore the onus of proof.
The pleading does specify particulars of identification but they are sparse and vague. The pleading alleges that the video identified the appellant because: 1. the Second Defendant and Vira Vumbaca were each tagged as recipients of the Defamatory Video on Facebook and they knew that the First Defendant was accusing the Plaintiff of selling a fake or counterfeit handbag to the First Defendant; and 2. subsequent comments made in response to, and thereafter published with, the defamatory video identified the Plaintiff: a. by her nickname “Jing Jo”; b. by photographs of the Plaintiff using a kangaroo (known as Jing Jo in the Thai language) to partially obscure the Plaintiff’s face.
Senior counsel for the respondent accordingly submitted that, while it was accepted that the primary judge would take the pleading at its highest, doing so left the question of the scope of publication (beyond four people named in the pleading) entirely open. Providing adequate particulars of the scope of publication was essential in a case in which the principal matter complained of (the video clip) was a foreign language publication that did not refer to Ms Supaphien by name or nickname.
The appellant alleges that the video clip conveyed the following defamatory imputations: Imputation C: The Plaintiff had sold a handbag to the First Defendant which she had passed off as a genuine Chanel handbag when it was proven to have been a fake or counterfeit product. Imputation D: The Plaintiff had defrauded the First Defendant. Third matter complained of: Ms Wilkinson’s comment
The third matter complained of is a comment in the Thai language uploaded in response to the video clip. The author of that comment was the second defendant in the proceedings. Ms Supaphien contends that Ms Chaiyabarn is jointly and severally liable as publisher of that comment.
The comment, as translated from the Thai language, is as follows: Nat Natesai I’m not stupid but I believe people because I think they will be like us!! If we’re a seller, whether it’s real or fake, first-hand, or second-hand, we would be honest. I wouldn’t have to wait for the customer to come and tell me later, and then one probably wouldn’t be the first. There are no secrets in the world, especially the world of lies and deceit!! Pity those who stand on lies and try to deceive them until they think that what they have done is true. It’s something that people all over the world to. [C]an even fool herself. (Emphasis in original.)
The appellant alleges that the comment conveyed the following defamatory imputation: Imputation E: The Plaintiff is a dishonest person.
The appellant alleges that the comment was “of and concerning” her because subsequent (unspecified) comments made in response to the comment identified her by her nickname Jing Jo and by photographs of her using a kangaroo to partially obscure her face.
The pleading did not address the element of serious harm except by reference to the imputations. In the case of each matter complained of, after the paragraph specifying the imputations alleged to arise from the matter complained of, the following formula was adopted: The meanings set out above were defamatory of the plaintiff and were likely to cause the people to whom they were published to shun and avoid the plaintiff, to lower their estimation of her character and to avoid her business in the future and this amounts to serious harm. (Emphasis in original.)
The words “and this amounts to serious harm” are underlined because they were added in an amendment. The original statement of claim appears to have been drafted in ignorance of the introduction of the serious harm element, which came into force over a year before the proceedings were commenced. The pleading makes no attempt to particularise any harm in fact suffered by the appellant as a result of the publication of any of the matters complained of. That was standard in defamation pleadings before the introduction of the serious harm element. Since the introduction of that element, a failure properly to plead that element renders the claim liable to be dismissed.
The evidence relied upon by Ms Chaiyabarn in support of the separate question was an affidavit affirmed by her solicitor annexing an affidavit Ms Supaphien had sworn in response to a similar application filed by the second defendant before the claim against her was resolved. The solicitor’s affidavit also annexed the concerns notice that had been served by Ms Supaphien before the proceedings were commenced (as required by s 124A of the Civil Law (Wrongs) Act) and correspondence the parties had exchanged about particulars. It may be noted that the concerns notices did not include any complaint about the first matter complained of (the slander), which is fatal to that part of the claim: s 124B of the Act. However, senior counsel for the respondent very fairly noted that this was not taken as a point in the proceedings before the primary judge. Accordingly, the fairer approach is to consider the appeal on the premise that the slander forms part of the claim.
Ms Supaphien relied on an affidavit in which she confirmed the contents of her previous affidavit as establishing the harm she claimed to have suffered as a consequence of all of the publications.
As already noted, the pleading specifies only four persons (including the former second defendant) who were said to have received one or other of the matters complained of and to have identified Ms Supaphien as the person referred to. One further person is identified in the particulars. None of them gave evidence, nor was there any evidence from the appellant as to any relationship or connection between her and any of them.
Ms Supaphien was not cross-examined. The primary judge recorded that the parties had agreed in those circumstances that his Honour would “take the plaintiff’s case at its highest”: at [9] of the judgment. However, that did not require the judge to fill gaps in the pleadings and the evidence with speculation or unavailable inference. In determining the separate question, the task was to determine whether serious harm was established.
The reason the judgment refers to the UK provision (“s 1 of the 2013 Act”) is that, as the email was sent to and received in England, the cause of action was governed by the UK Act. We have already noted that the Australian model provision was modelled on the UK provision.
Basten AJA’s agreement in the result in Rader was unsurprising. Brereton JA had explained at [63] that “the publication was limited to the appellant’s parents, who of all the people in the world best knew him and were best positioned to judge for themselves whether there was substance to the imputations, and whose relationship with him was such as to withstand even the most serious allegations”. The only point on which Basten AJA did not agree with the Brereton JA’s reasons was in holding “one minor reservation” concerning his Honour’s approach to construing the serious harm provision. Basten AJA explained this reservation at [91]: The reservation relates to the exegesis on the meaning of ‘serious’ in s 1 of the Defamation Act 2013 (UK). There is a risk in seeking synonyms, which may later be treated as valid replacements for the ordinary English word adopted by the Parliament. There is also a risk in seeking to place the term on a scale, between other terms of equal imprecision. The critical concept is ‘serious harm to the reputation of the claimant’; it is that to which the court is required to attend by reference to the evidence of a range of matters. Analysis of individual component words is apt to distract from that inherently impressionistic exercise.
The appellant’s argument in support of ground 2 is that Basten AJA’s approach (there is a risk in seeking synonyms; attention should be confined to the words in the statute) is right and Brereton JA’s approach, with which Macfarlan JA agreed (“’serious’ harm sits on the spectrum above ‘substantial’ but below ‘grave’”), is wrong. Thus, it was argued that Curtin AJ erred because he applied the formulation stated by Brereton JA rather than adopting Basten AJA’s preferred approach of confining attention to the words used, without resort to synonyms or a scale of seriousness.
Because Brereton JA’s judgment enjoyed the unqualified agreement of Macfarlan JA, counsel for the appellant spent considerable time in oral submissions attempting to persuade this Court not to follow that decision (while the decision in Rader was concerned with s 1 of the UK Act, it was assumed that the decision would have the same standing as a decision on the model serious harm provision, given that the model provisions were modelled on the UK section). Subject to what follows, for reasons of comity, this Court would ordinarily follow such a decision, being a decision of the intermediate appellate court of another jurisdiction concerning the same provision. It was noted in that context that the decision in Rader has been followed in other Australian decisions. As McWilliam J notes, other decisions have expressed a preference for Basten AJA’s approach.
Counsel for the appellant submitted that, while historically the test for not following a decision of an intermediate appellate court of another jurisdiction on the same provision has been formulated in terms of whether the decision is “plainly wrong”, the preferable formulation is that adopted in the recent decision of the New South Wales Court of Appeal in David William Pallas & Julie Ann Pallas as trustees for the Pallas Family Superannuation Fund v Lendlease Corporation Ltd [2024] NSWCA 83; 114 NSWLR 81, namely, whether there are “compelling reasons” to depart from a decision in relation to which considerations of judicial comity in a federal judicature are engaged: [19]-[23] (Bell CJ); [138] (Gleeson JA); [139]-[140] (Leeming JA); [160] (Stern JA); but see Ward P at [127].
We respectfully agree.
The appellant submitted that there are two compelling reasons for not following the principle stated by Brereton JA in Rader (that “serious” harm sits on the spectrum above “substantial” but below “grave”).
First, she submitted that the notion of a spectrum is not supported in the extrinsic materials. It was submitted that, to the contrary, the debate in the ACT Legislative Assembly seems to have assumed that serious means “not trivial”.
It may be accepted that the presentation speeches to which we were taken by the appellant explained the introduction of the serious harm test by reference to “the increasing use of defamation law for trivial and vexatious matters”: ACT, Legislative Assembly, Hansard, 12 May 2021 at 1354 (Mr Rattenbury, Attorney-General). However, there is nothing to suggest the legislature took the distinction between a trivial matter and a matter in which serious harm is established to be a binary classification. If that had been the intention, the Model Provisions could easily have framed the threshold as a requirement to prove that the publication of defamatory matters has caused, or is likely to cause, “non-trivial harm” or “harm that is more than trivial” to the person’s reputation. As acknowledged by counsel for the appellant during oral argument, it is possible that, in deciding to remove a defence (the defence of triviality) and reverse an onus by creating a new element of the cause of action, the legislature intended at the same time to make the new test more stringent.
As to the submission that there is nothing in the debate in the Legislative Assembly that supports the notion of a spectrum, that also may be accepted. However, the suggestion that this omission is significant misconceives the different roles of the legislature and the judiciary. In debating whether to pass a law, Parliament does not seek to lecture the judiciary about how to interpret it. The task of statutory interpretation is for the judiciary.
There is, with respect, much to be said for Basten AJA’s preference for focusing on the word in the section rather than seeking to expand upon it with synonyms, let alone placing it within a scale of intensification where, as Basten AJA observed, it sits “between other terms of equal imprecision”: at [91]. However, we are not persuaded that the “place on a scale” approach amounts to an error of principle or demonstrates the application of a wrong test. Rather, it reflects a choice between permissible approaches to the task of statutory construction. The critical question is whether Curtin AJ’s approach put the bar for establishing “serious harm” too high. That question is considered below in the discussion of the other grounds of appeal.
The second compelling reason argued for not adopting the construction of “serious harm” applied in Rader was that the UK authorities “also do not support the notion of a spectrum”. It was submitted that there is in the English jurisprudence no decision where the idea of a spectrum has been adopted. At the same time, however, senior counsel for the appellant acknowledged that the decision of Warby J in Lachaux v Independent Print Ltd [2015] EWHC 2242 (QB); [2016] QB 402 “held that serious might mean more than substantial”.
The passage in question is at [45], where Warby J said: In my judgment this approach leads to the clear conclusion that in enacting s 1(1) Parliament intended to do more than just raise the threshold for defamation from a tendency to cause ‘substantial’ to ‘serious’ reputational harm. The intention was that claimants should have to go beyond showing a tendency to harm reputation. It is now necessary to prove as a fact on the balance of probabilities that serious reputational harm has been caused by, or is likely to result in future from, the publication complained of.
The publishers appealed from Warby J’s decision on the grounds that the judge had erroneously construed and applied the serious harm test. Mr Lachaux sought to uphold the decision on the ground that the judge could have taken a shorter route to reach the conclusion of serious harm. The Court of Appeal construed the section differently, holding that the introduction of the serious harm element did not affect the common law presumption of damage, but nonetheless upheld the judge’s holding of serious harm. The publishers appealed to the Supreme Court. Lord Sumption, with whom all other members of the Court agreed, rejected the Court of Appeal’s conclusion that the presumption of damage has survived the introduction of the serious harm test: Lachaux v Independent Print Ltd [2020] AC 612. His Lordship said at [12]: Although the Act must be construed as a whole, the issue must turn primarily on the language of section 1. This shows, very clearly to my mind, that it not only raises the threshold of seriousness above that envisaged in Jameel and Thornton but requires its application to be determined by reference to the actual facts about its impact and not just to the meaning of the words.
Lord Sumption concluded at [20] that “Warby J’s analysis of the law was coherent and correct, for substantially the reasons which he gave”.
The appellant submitted that Australian authorities, notably including the decision of the New South Wales Court of Appeal in Rader, have seized on Lord Sumption’s statement out of context to support the proposition that the introduction of the serious harm test in Britain raised the threshold. The submission suggested that it was wrong to draw that principle from Lachaux because that case was decided in a jurisdiction where harm is presumed and it is not necessary to prove actual harm. But the position at common law was relevantly the same in Australia before the adoption of the Model Provisions. A helpful and comprehensive discussion of the common law developments in both jurisdictions before the introduction of the statutory element of serious harm may be nd found in Dr David Rolph’s text, Rolph on Defamation (Lawbook Co, 2 ed, 2024) from [9.20].
In summary, the appellant’s argument was that there are compelling reasons for not following the decision of the majority in Rader to “superimpose a spectrum on the meaning of serious”. It was submitted that there is nothing in the Parliamentary debates to support the proposition that “serious” sits on a spectrum above “substantially” and below “grave”. Secondly, it was submitted that the decision in Rader, being a decision interpreting the English provision, relied on English precedent and that there is no support in the English authorities for putting “serious” on a spectrum. The significance of the error, so the argument went, is that the Rader test overstates the requirements of the serious harm element in Australia.
Having comprehensively criticised the construction that places “serious” on a scale between “significant” and “grave”, the appellant instead contended that the proper construction of the word “serious” should be ascertained by reference to its definition in the Macquarie dictionary. As with the scale posited in Rader, the dictionary offers several meanings of greater or lesser intensity. The appellant relied on the third meaning, “being in earnest; not trifling”. In making that submission, the appellant placed the word “serious” on a scale of meanings of various intensity.
Upon analysis, the transposition of the definition “being in earnest; not trifling” for “serious” into s 122A does not make sense. “Being in earnest” and, as a phrase that means the same thing, “not trifling” refer to a state of mind. That dictionary definition provides no guidance as to the degree of harm that is required by s 122A(1), which is clearly an objective test.
The respondent relied on the remarks of Warby J in Lachaux at [29] where his Honour said: A requirement that ‘serious’ harm to reputation has been or is likely to be caused is clearly more demanding than the common law test identified in Thornton’s case: a tendency ‘substantially’ to affect in an adverse way people’s attitudes to the claimant. Parliament has left it to the judges to apply the serious harm requirement, ‘serious’ being an ordinary word in common usage. (Footnotes omitted.)
The respondent did not seek to defend Brereton JA’s spectrum but submitted that, in introducing the serious harm element, the ACT Legislature clearly intended to impose a more demanding test than “not trivial”. It was submitted that to construe “serious” as simply meaning more than trivial is “utterly inconsistent with any kind of ordinary meaning of that word”. We respectfully agree.
We are of the view that the appellant’s submission that s 122A was introduced only to address the question of trivial or vexatious claims and not to raise the threshold of seriousness above the trivial must be rejected. There is no warrant in the text of the section for such a construction. The distinction between trivial and serious is not binary. Having regard to the context in which the Model Provisions were introduced, it is clear that the legislation intended to introduce a higher threshold than non-trivial harm.
We would reject ground 2.
The appellant’s reliance on the grapevine effect is similarly misconceived. It did no harm to her reputation if all that was passed along the grapevine was that an unnamed person sold a fake Chanel handbag.
In oral submissions, the appellant submitted that, no matter what test is applied for serious harm, the evidence relied upon by her was sufficient to establish that she suffered or was likely to suffer serious harm to her reputation. She identified five matters leading to this conclusion: (a) the gravity of the imputations; (b) the nature of the recipients. It was submitted that the publications were made within the appellant’s community, the Thai community in Canberra and Melbourne where “gossip sticks and travels fast”, especially on social media; (c) the fact that the defamatory video was shared to 2 other Facebook pages and received further comments on those pages and was shared again 37 times from those pages; (d) the fact that the plaintiff was shunned by 34 people whom she named; and (e) the alleged decrease in the appellant’s business turnover.
The appellant accepted that the evidence did not identify the people who “liked” the video but submitted the Court could infer that it would be more than the number of likes. There is no basis for drawing any such inference.
The appellant submitted that the Court should have regard to the difficulties a plaintiff has in finding witnesses who will say they saw the matter complained of and understood it to refer to the plaintiff. However, that is the very burden imposed, deliberately so, by the introduction of the serious harm test. We accept that, in a case in which the plaintiff was named, the serious harm test could conceivably be satisfied by drawing an inference but the position is different in a case in which the plaintiff is not named.
As submitted by the respondent, the appellant elected to go into evidence on two issues, the financial loss and the shunning and avoiding by the 34 persons named in the concerns notice. The respondent submitted that the evidence on those issues was so poor and imprecise that the trial judge was entitled to take very little notice of it.
That submission must be accepted. The appellant’s case overlooked the nature of the determination of a separate question and seemingly also overlooked the proposition that the introduction of the serious harm element imposed a burden of proof on the plaintiff to prove harm to reputation in place of the common law presumption of damage.
The historical context in which the serious harm element imposing that burden of proof came to be introduced in Australia is explained in the judgment of Sackar J in Newman. In addition to providing a helpful exposition of relevant developments in the law, his Honour explains how those principles worked in practice, reflecting his Honour’s lengthy experience at the defamation bar in New South Wales. His Honour explains that, at common law, there was a presumption that the plaintiff had suffered damage because of the publication of defamatory matter and an “associated rule that the cause of action is made out if the statement complained of is inherently injurious or that it has a tendency to injure the claimant’s reputation” (referred to in Newman at [62]). These aspects of the cause of action posed a significant challenge to defendants in cases lacking in gravity or seriousness. Triviality was a defence but, as noted in Newman at [45], it faced “perhaps the inevitable hurdle of the presumption of damage which works in the plaintiff’s favour whenever it was proved that a defamatory imputation was published of and concerning the plaintiff” (citations omitted).
The introduction of the serious harm element had the effect of abolishing the presumption of damage, instead requiring proof of damage by the plaintiff. In Newman at [70], Sackar J noted the kind of evidence that had been called to prove serious harm in the English case of Lachaux: the plaintiff and the other witnesses had given evidence. The Court also received print runs of the readership of the publications and of the user numbers of online publications. The scale of the publication was taken into account, as was the fact at least one person in the UK knew of the publication and there was a likelihood there were others.
None of that kind of evidence was adduced here. Ultimately, the proceedings were dismissed because the appellant simply failed to discharge her onus of proof on an element of the cause of action. In the circumstances of the present case, the accepted seriousness of the imputations was not enough. The onus was on the plaintiff to prove actual facts about the impact of the imputations in circumstances where she was not named in the publications. Failure to take proper account of the evidence and pleadings
The next ground of appeal must fail for essentially the same reasons. It is ground 6 in the Notice of Appeal, which contends: Notwithstanding that it was agreed that the plaintiff’s case on her pleadings and the evidence would be taken at its highest and there was no cross examination of the plaintiff, the learned trial judge erred by failing to properly take into consideration the plaintiff’s uncontradicted evidence…
The Notice of Appeal proceeds to list twelve aspects of the evidence. At the hearing of the appeal, the appellant rested on her written submissions on this issue, which were sparse. She submitted that, in light of the matters listed, the evidence as to which was uncontradicted: the inescapable conclusion is that the appellant suffered serious harm to her reputation as a result of the publications or was likely to suffer serious harm. The primary judge erred in failing to reach that conclusion. The conclusion by the primary judge was glaringly improbable or contrary to compelling inferences.
The respondent relied in this context on the decision of the Court of Appeal in Rader at [67], where the Court said: In concluding that, despite the seriousness of the imputations, a discord in the appellant’s relationship with his parents for some weeks did not amount to ‘serious harm’, the trial judge was applying her vast experience in this field to a task which Lord Sumption said: …was to evaluate the material before [the Court], and arrive at a conclusion on an issue on which precision will rarely be possible. A concurrent assessment of the facts was made by the Court of Appeal. Findings of this kind would only rarely be disturbed by this court, in the absence of some error of principle potentially critical to the outcome. (Footnotes omitted.)
As explained in the discussion of the previous topic, the fundamental hurdle for the appellant was her failure to prove any facts from which any inference could be drawn as to the number of people who downloaded the video and understood it to refer to her. On the pleadings, with the addition of the particulars, that number was five and the sharing of the video played itself out over a matter of days.
Furthermore, as submitted on behalf of the respondent, the supposed evidence of the downturn in profits of the business was hopeless. The appellant’s affidavit included a table purportedly showing her business’s financial position leading up to the publication of the matters complained of and her business’s financial position during the month in which the matters were published. The legal status of the business was not explained. The table showed only “sales” and “gross profit” over four months including the month during which the matters complained of were published. There was no explanation of the number of sales, expenses or any other matter from which the gross profit each month would be compared with other months. The evidence was unchallenged but it proved almost nothing.
I respectfully agree with what has been set out in MG (being the more recent decision of an intermediate appellate court) and in Mond and would similarly take the approach discussed by Basten AJA in considering whether there was serious harm to the reputation of the appellant.
Indeed, the primary judge had even included part of that passage earlier in his Honour’s reasons (at [105]). Here, the seriousness of the imputation was not disputed. That is, it was accepted that the imputations that the appellant sold fake handbags and defrauded customers were serious. The primary judge was required to evaluate whether serious harm to reputation had occurred, or was likely to occur, on that basis and the fact that his Honour was citing inapt legal propositions when evaluating the seriousness of harm to reputation is a further indicator of a misapplied or incorrect approach.
Contrary to the primary judge’s statements above, the fact that the publications occurred on social media was a matter of some significance in circumstances where the appellant’s entire business was social media-based and the posts occurred on Facebook forums within the very community from which her potential customers were drawn, as evidenced by some of the comments posted in response to the publication.
The concerns I have about that passage are: (a) First, as the appellant submitted, it suggests that the primary judge misunderstood the nature of the evidence, in that the number of likes, comments and shares does not necessarily equate to the number of times a post is viewed. (b) Second, the primary judge’s statement that he would not have found serious harm because the extent of publication was limited, and that the number of people liking, commenting or sharing the post “in social media terms, is not a significant number”, stripped the extent of publication from its context, which was publication in a specific community which in turn had particular significance for the appellant’s reputation. Contrary to the primary judge’s view, the fact that the post generated that level of activity in those specified Facebook communities over such a short space of time (publication and republications occurred on 11-13 September 2021 and the concerns notice setting out the publications and comments was sent on 21 September 2021) indicates a greater impact than that stated by the primary judge. (c) Third, the statement that he “would not have found serious harm” because the publications and downloading were limited again reveals that the primary judge was directing attention to individual aspects of the evidence and where they sat on a spectrum of serious harm, this time with regard to the extent of publication.
The above matters cover the material parts of the primary judge’s reasons addressing harm to the appellant’s reputation. The remainder of the reasons dealt with the decline in profit of the appellant’s business. Taking the errors in the reasoning together, it cannot be said that any error in approach was limited to an isolated or insignificant aspect of the reasoning. I therefore accept that the primary judge erred in incorrectly applying the proper test or approach to the separate question.
Taking the evidence at its highest, including the whole context and inferences to be drawn from it, the evidentiary material that was led did not provide a sufficient basis to prove an impact on the appellant’s reputation that amounted to serious harm.