What happened
The case arose from the 19 September 2016 publication of Steve Cannane's book Fair Game: The Incredible Untold Story of Scientology in Australia by HarperCollins Publishers Australia Pty Ltd. Chapter 14, titled "Deep Sleep" (pp 176-204), focused on the use of deep sleep therapy (DST) and electroconvulsive therapy (ECT) at Chelmsford Private Hospital in Sydney during the 1960s and 1970s. The chapter named the late Dr John Herron (a psychiatrist) and Dr John Gill (a general practitioner who became de facto medical superintendent) as among "Bailey and his fellow Chelmsford doctors" who continued DST despite warnings, mounting deaths and professional rejection of the practice. It described DST as unproven and controversial, linked it to at least 24 deaths at the hospital and 24 suicides within a year of release, and portrayed the doctors as operating like a "secretive cult" where authority was rarely questioned, death certificates were falsified, families were lied to and denied visitation, and health funds were defrauded. The chapter credited Scientologists, particularly nurse Rosa Nicholson, with exposing the scandal, leading to the 1988-1990 Royal Commission conducted by Slattery AJ. That Commission made damning findings against the doctors, including that DST was experimental, lacked justification, caused deaths, and that records were falsified.
On 19 September 2017, Dr Herron and Dr Gill each commenced separate proceedings in the Federal Court seeking damages for defamation. Dr Gill gave an undertaking to pay any costs ordered against Dr Herron. The trial before Jagot J lasted approximately 10 weeks. The primary judge dismissed both proceedings with costs on 25 November 2020: Herron v HarperCollins Publishers Australia Pty Ltd [2020] FCA 805 and Herron v HarperCollins Publishers Australia Pty Ltd (No 3) [2020] FCA 1687. Her Honour found that seven of the pleaded imputations were conveyed about Dr Herron and one about Dr Gill, but the remaining contested imputations (E, G, H, J, K, L, M) were not conveyed. All conveyed imputations were held to be substantially true under s 25 of the Defamation Act 2005 (NSW). The publishers also succeeded on the statutory qualified privilege defence under s 30. Evidentiary rulings admitted Royal Commission material (including evidence given under compulsion) and reports from four deceased experts (Drs Gandevia, Greenaway, Hassall and Joseph) who had given evidence to the Royal Commission. The primary judge held that s 17(2) of the Royal Commissions Act 1923 (NSW) did not apply to prevent admissibility in federal court proceedings.
A joint notice of appeal was filed on 23 December 2020. Dr Herron died on 7 February 2021. The further amended notice of appeal raised numerous grounds, but the Full Court (Rares, Wigney and Lee JJ) found it unnecessary to deal with all because evidentiary errors (particularly the admission and use of the dead experts' reports and Royal Commission material without proper consideration of ss 135 and 136 of the Evidence Act 1995 (Cth)) infected the truth findings, making a retrial inevitable on that defence. The court held that all contested imputations were conveyed, s 17(2) of the Royal Commissions Act had no application in federal jurisdiction (being inconsistent with s 56 of the Evidence Act), the qualified privilege defence failed because the publishers made no attempt to obtain or publish the applicants' side of the story despite knowing they disputed the Royal Commission findings and that court proceedings had been stayed for delay, and s 10 of the Defamation Act did not prevent Dr Gill from challenging the costs orders in the Herron proceeding given his undertaking and the merger of the cause of action in the judgment. The appeal was allowed with costs. Orders dismissing both proceedings were set aside. Judgment was entered for Dr Gill on qualified privilege. He was released from his undertaking. The Gill proceeding was remitted for retrial on remaining issues (primarily substantial truth). The Herron proceeding abated under s 10.
The decision turned on close analysis of chapter 14 as a whole. The back cover and chapter itself presented the book as based on "years of interviews and meticulous research". The ordinary reasonable reader—a person reading a serious investigative work—would not parse the text with lawyerly precision but would read between the lines, understanding generalised references to "the doctors", "Chelmsford doctors" and "the surviving Chelmsford doctors" (after naming Dr Herron, Dr Gill and others at pp 178-179) as applying to all four named individuals. The narrative wove Mr Hart's case (treated by Dr Herron without consent, leading to a jury verdict against him for false imprisonment, assault and battery) with the broader scandal, Scientologists' role, the Royal Commission findings, and assertions that the doctors "continued to avoid accountability". This context conveyed the contested imputations.
On the evidentiary issues, the Full Court emphasised that the primary judge's construction of s 17(2) of the Royal Commissions Act was erroneous. That provision provides that answers or documents produced under compulsion to a Royal Commission "shall not ... be admissible in evidence against that person in any civil or criminal proceedings". The primary judge had read it as applying only where the evidence was deployed to establish civil or criminal liability against the witness. Rares J (with whom Wigney J agreed) held that the natural and ordinary meaning prohibits admissibility against the witness in any civil or criminal proceeding. The purpose is to encourage complete candour by giving "complete immunity" (Colonial Sugar Refining Co Ltd v Attorney-General (Cth) (1912) 15 CLR 182 at 219 per Isaacs J; Giannarelli v The Queen (1983) 154 CLR 212 at 220 per Gibbs CJ). A construction limited to proceedings where the witness is a defendant would undermine that purpose, as the witness could never be certain of protection when giving evidence. However, s 17(2) has no direct application in federal jurisdiction because it purports to regulate the exercise of federal jurisdiction and is inconsistent with s 56(1) of the Evidence Act ("evidence that is relevant in a proceeding is admissible in the proceeding"). It is not picked up by s 79 of the Judiciary Act (Rizeq v Western Australia (2017) 262 CLR 1; Masson v Parsons (2019) 266 CLR 554). Constitutional arguments under the Melbourne Corporation principle and s 118 of the Constitution were rejected; the Commonwealth's power under s 51(xxxix) to regulate federal jurisdiction includes the power to enact laws that do not pick up inconsistent State provisions. Lee J preferred not to decide the precise construction of s 17(2) given the inconsistency with s 56 but agreed the primary judge erred in not considering discretionary exclusion or limitation under ss 135 and 136 of the Evidence Act. The fact that evidence was given under compulsion, with the expectation of s 17(2) protection, was relevant to the danger of unfair prejudice.
The dead experts' reports were prepared for the Royal Commission in 1989-1990 and admitted under s 63 of the Evidence Act (first-hand hearsay exception for civil proceedings where the maker is unavailable). The primary judge held they fell within s 79 as expert opinion. The Full Court disagreed. The reports did not adequately disclose the facts, assumptions or reasoning process. For example, Dr Gandevia's statistical analysis of death certificates compared crude death rates at Chelmsford with New South Wales general population rates without age-standardisation or full underlying data; some death certificates and medical records were not in evidence. Dr Greenaway, Dr Hassall and Dr Joseph relied on incomplete records, some of which were unavailable at trial. None of the reports complied with the Federal Court Rules 2011 (Cth) r 23.13 or the Harmonised Expert Witness Code of Conduct. Critically, the opinions could not be tested by cross-examination. The inability to cross-examine went not only to weight but to probative value. The primary judge erred in not excluding the reports under s 135, as low probative value (undisclosed assumptions, no literature review in some cases, reliance on untested Royal Commission material) was substantially outweighed by the danger of unfair prejudice. Lee J noted the "decision making process too intricate to unpack and put back together" where other evidence was "infected by inadmissible evidence". The truth findings, which relied heavily on these reports to establish that DST caused deaths and constituted malpractice, could not stand. A retrial on truth was required.
The qualified privilege defence failed independently. Section 30(1)(c) required the publishers to prove their conduct in publishing was reasonable in the circumstances. The primary judge held it was reasonable for Mr Cannane to rely on the Royal Commission report, which had "stood unchallenged and undisturbed for decades", and that seeking the applicants' response would have been a "waste of time" given Slattery AJ's findings on their credibility. The Full Court disagreed. Chapter 14 was not a fair report of the Royal Commission; it added new material from Mr Hart's interview, unpublished manuscript and 60 Minutes program. It asserted that after the Royal Commission the "Chelmsford doctors continued to avoid accountability for their actions" (p 190). The publishers knew the applicants were alive, strongly disputed the findings, and that the New South Wales Court of Appeal in Herron v McGregor (1986) 6 NSWLR 246 and Gill v Walton (1991) 25 NSWLR 190 had permanently stayed disciplinary proceedings because departmental delay had caused irreparable prejudice to their defence. The High Court refused special leave. These curial findings were not "mere legal red tape" or "exculpatory" but showed that the doctors could not fairly contest the allegations. Mr Cannane's "zero confidence" that they would tell the truth did not make it unnecessary to seek a response; he could have published their side (or the court findings) and expressed scepticism. The back cover claimed the book was based on "meticulous research". HarperCollins led no evidence of its own state of mind or inquiries. In these circumstances the conduct was not reasonable. Rares J (with Wigney J agreeing) held Dr Gill entitled to judgment on the defence. Lee J agreed.
Dr Herron's death raised s 10 of the Defamation Act, which provides that a person (including a personal representative) cannot assert, continue or enforce a cause of action for defamation in relation to the publication of defamatory matter about a deceased person. The publishers argued Dr Gill could not challenge the dismissal of the Herron proceeding or the costs orders. The Full Court rejected this. Dr Herron's cause of action had merged in the judgment (Blair v Curran (1939) 62 CLR 464). The appeal concerned Dr Gill's independent interest under his undertaking to pay costs. An appeal is not the assertion of the underlying defamation cause (Ryan v Davies Bros Ltd (1921) 29 CLR 527). Dr Gill was granted leave nunc pro tunc to appeal the Herron proceeding. He was released from the undertaking. The Herron proceeding abated.
The disposition reflected these conclusions. The appeal was allowed with costs. The primary judge's orders were set aside. Judgment was entered for Dr Gill on qualified privilege. The Gill proceeding was remitted for retrial on remaining issues (primarily substantial truth, with the qualified privilege defence resolved in his favour). Dr Gill was released from his undertaking. The parties were directed to file short submissions on the form of orders.
Why the court decided this way
The Full Court decided the appeal in Dr Gill's favour on the contested imputations, evidentiary errors, qualified privilege and s 10 issues because the primary judge's reasoning could not be reconciled with established principles of defamation law, evidence and federal jurisdiction.
On meaning, the court applied the orthodox test that the ordinary reasonable reader is not a lawyer but a person of ordinary intelligence reading casually, prone to loose thinking, reading between the lines and drawing derogatory implications more freely than a lawyer (Lewis v Daily Telegraph Ltd [1964] AC 234 at 258 per Lord Reid, 277 per Lord Devlin; Mirror Newspapers Ltd v World Hosts Pty Ltd (1979) 141 CLR 632; Reader's Digest Services Pty Ltd v Lamb (1982) 150 CLR 500 at 505-506 per Brennan J; Trkulja v Google LLC (2018) 263 CLR 149 at 160-161 [32]; Favell v Queensland Newspapers Pty Ltd (2005) 221 ALR 186 at 190 [10]-[12]). Chapter 14 began by naming Dr Bailey, Dr Herron, Dr Gill and Dr Gardiner as "fellow Chelmsford doctors" not deterred by warnings or the "death toll mounting before their eyes" (pp 177-178). It then referred to "the doctors and psychiatrists" operating like a "secretive cult" where "death certificates were falsified", "family members were regularly denied visitation rights and routinely lied to about how seriously ill their loved ones were", and "doctors defrauding the patient's health funds" (p 179). Later references to "the Chelmsford doctors" continuing to avoid accountability (p 190) and the Royal Commission's findings about Bailey falsifying "as many as 17 death certificates" (p 192) were read as illustrations, not qualifications, of the earlier generalised charges. The reader would not parse the text proleptically or assume only Dr Bailey was responsible; the "smoke" of generalised misconduct implied the "fire" of involvement by all named doctors (Lewis at 285 per Lord Devlin). The book's serious tone, claim of meticulous research, and focus on exposing "atrocities" and "horror show of sustained medical malpractice and abuse" encouraged the reader to associate all four doctors with the catalogue of misconduct. The primary judge's finding that Dr Gill was not a psychiatrist (and therefore not included in imputations referring to "a psychiatrist") was inconsistent with her own later acceptance (at [859]) that there was no material difference between "psychiatrist" and "doctor" in context, and with the reader's inference that any doctor administering DST/ECT in a psychiatric hospital must be a psychiatrist. The contested imputations were therefore conveyed.
The evidentiary errors were fundamental. The dead experts' reports (Drs Gandevia, Greenaway, Hassall and Joseph) were prepared for the Royal Commission in 1989-1990. They contained opinions on causes of death, whether DST contributed, and statistical mortality. None complied with the Federal Court Rules 2011 (Cth) r 23.13 or the Harmonised Expert Witness Code of Conduct. Facts, assumptions and reasoning were not adequately disclosed. Some underlying death certificates, medical records and other material were not in evidence. The opinions could not be tested by cross-examination at the Royal Commission or at trial. The primary judge admitted them under s 63 of the Evidence Act (first-hand hearsay where maker unavailable) and held they satisfied s 79. The Full Court held they did not. Section 79 requires the opinion to be wholly or substantially based on specialised knowledge; that cannot be assessed if the facts, assumptions and reasoning are not exposed (Makita (Australia) Pty Ltd v Sprowles (2001) 52 NSWLR 705 at 743-744 [85] per Heydon JA; Dasreef Pty Ltd v Hawchar (2011) 243 CLR 588 at 604 [37] per French CJ, Gummow, Hayne, Crennan, Kiefel and Bell JJ; HG v The Queen (1999) 197 CLR 414 at 428 [41] per Gleeson CJ). The reports were therefore inadmissible under s 76 (opinion rule). Even if admissible, they should have been excluded under s 135. Probative value was low given the deficiencies, age, lack of cross-examination and unavailability of supporting material. The danger of unfair prejudice was high: the applicants could not test the opinions, yet the reports were used to establish that DST caused deaths and constituted malpractice. The primary judge's statement that there was "no valid reason to discount [the dead experts'] opinions merely because they were not available for cross-examination" (J [458]) reversed the proper approach. Authority requires that evidence not tested by cross-examination be treated with "considerable reserve" (Clyne v Law Society of New South Wales (NSWCA, 4 September 1987, unreported) per Mahoney JA; Amalgamated Television Services Pty Ltd v Marsden [2002] NSWCA 419 at [188]-[191]). Lee J described the decision-making process as "too intricate to unpack and put back together" where other evidence was "infected by inadmissible evidence". The truth findings, which relied on these reports to establish causation of deaths and malpractice, could not stand. A retrial was required.
Similar errors affected the Royal Commission material. The primary judge admitted it on the basis that s 17(2) of the Royal Commissions Act applied only where evidence was used to establish liability against the witness. Rares J (with Wigney J agreeing) held this construction was wrong. The natural and ordinary meaning prohibits admissibility "against that person in any civil or criminal proceedings". The purpose is to encourage candour by giving "complete immunity" from subsequent use (Colonial Sugar Refining at 219 per Isaacs J; Giannarelli at 220 per Gibbs CJ; Sorby v The Commonwealth (1983) 152 CLR 281 at 310 per Mason, Wilson and Deane JJ). A construction limited to proceedings where the witness is a defendant would leave the witness uncertain of protection when giving evidence. However, s 17(2) has no direct operation in federal jurisdiction. It purports to regulate the exercise of federal jurisdiction and is directly inconsistent with s 56(1) of the Evidence Act ("evidence that is relevant in a proceeding is admissible in the proceeding"). It is not picked up by s 79 of the Judiciary Act (Rizeq; Masson v Parsons). Lee J agreed that s 17(2) had no application but considered the primary judge's construction erroneous and that the existence of s 17(2) (and the witnesses' expectation of protection at the time they gave evidence) was relevant to the exercise of discretion under ss 135 and 136. The primary judge did not consider those discretions. Given the volume of material and its permeation through the trial (including credit findings), the Full Court could not undertake the discretionary exercise on appeal. The matter required remittal.
The qualified privilege defence failed because the publishers' conduct was not reasonable. Section 30(3)(h) required consideration of whether the matter published contained the substance of the person's side of the story and, if not, whether a reasonable attempt was made to obtain and publish a response. Mr Cannane made no attempt to contact Dr Herron or Dr Gill despite knowing they were alive and disputed the Royal Commission findings. He omitted reference to the Court of Appeal decisions in Herron v McGregor (1986) 6 NSWLR 246 and Gill v Walton (1991) 25 NSWLR 190 permanently staying disciplinary proceedings on the ground that departmental delay had caused irreparable prejudice to the doctors' defence. The High Court refused special leave. These were not "mere legal red tape" or "exculpatory"; they showed the doctors could not fairly contest the allegations. The book's assertion that the doctors "continued to avoid accountability" (p 190) could only be based on subsequent events, not the Royal Commission. The back cover claimed "meticulous research". HarperCollins led no evidence of its own state of mind. In these circumstances the conduct was unreasonable. The primary judge's emphasis on the Royal Commission's "unchallenged" findings overlooked that the doctors had been prevented from challenging them by the stays. The failure to include the court findings or seek a response was not justified by Mr Cannane's "zero confidence" that the doctors would be truthful; he could have published their side and expressed scepticism. The defence therefore failed. Dr Gill was entitled to judgment on it.
Section 10 of the Defamation Act did not prevent Dr Gill from challenging the costs orders in the Herron proceeding. Dr Herron's cause of action had merged in the judgment (Blair v Curran (1939) 62 CLR 464). The appeal concerned Dr Gill's independent interest under his undertaking. An appeal is not the assertion of the underlying defamation cause (Ryan v Davies Bros Ltd (1921) 29 CLR 527 at 532-534). Dr Gill was granted leave nunc pro tunc. He was released from the undertaking. The Herron proceeding abated.
The overall disposition reflected these conclusions. The appeal was allowed because the evidentiary errors infected the truth findings, making a retrial necessary on that defence (with qualified privilege resolved in Dr Gill's favour). The court directed the parties to file submissions on the form of orders.
Before and after state of the law
Before this decision, the law on imputation conveyance was well settled: the ordinary reasonable reader test required the publication to be read as a whole, with readers drawing implications more freely than lawyers, especially derogatory ones from imprecise or generalised language (Lewis v Daily Telegraph Ltd [1964] AC 234; Mirror Newspapers Ltd v World Hosts Pty Ltd (1979) 141 CLR 632; Reader's Digest Services Pty Ltd v Lamb (1982) 150 CLR 500; Trkulja v Google LLC (2018) 263 CLR 149; Favell v Queensland Newspapers Pty Ltd (2005) 221 ALR 186). The decision applies those principles to a book chapter using collective nouns after naming individuals, confirming that a reader would not read down generalised allegations to apply only to the most prominent figure (Dr Bailey) but would attribute them to all named doctors. It reinforces that the test is "one in generosity not parsimony" (Trkulja at 160 [32]).
On the interaction between State laws and federal jurisdiction, the law was settled by Rizeq v Western Australia (2017) 262 CLR 1 and Masson v Parsons (2019) 266 CLR 554: State laws cannot regulate the exercise of federal jurisdiction; s 79 of the Judiciary Act picks up and applies State laws as Commonwealth laws only to fill gaps, and only if the Constitution or a Commonwealth law does not "otherwise provide". The decision applies that framework to s 17(2) of the Royal Commissions Act, holding it inconsistent with s 56(1) of the Evidence Act and therefore not picked up. It confirms that the test for "otherwise provide" in s 79 is equated with inconsistency under s 109 of the Constitution (Masson at 579-580 [43]). Constitutional arguments under the Melbourne Corporation principle (Melbourne Corporation v Commonwealth (1947) 74 CLR 31) and s 118 were rejected; the Commonwealth's power under s 51(xxxix) to regulate federal jurisdiction includes the power not to pick up inconsistent State provisions. The decision does not change the law but applies it to Royal Commission evidence in federal court defamation proceedings.
On expert evidence, the law required opinions to be wholly or substantially based on specialised knowledge with disclosed facts, assumptions and reasoning (Makita (Australia) Pty Ltd v Sprowles (2001) 52 NSWLR 705; Dasreef Pty Ltd v Hawchar (2011) 243 CLR 588; HG v The Queen (1999) 197 CLR 414). The decision applies that to reports prepared for a Royal Commission decades earlier, holding they were inadmissible under s 79 because the basis was not exposed and could not be tested. It reinforces that inability to cross-examine affects both probative value and the danger of unfair prejudice under s 135. The decision does not change the law but illustrates its strict application where expert reports are old, prepared for a different purpose, and cannot be tested.
On qualified privilege, the law required the defendant to prove conduct was reasonable in all the circumstances, with s 30(3)(h) directing attention to whether the substance of the person's side of the story was published or a reasonable attempt made to obtain and publish a response (Lange v Australian Broadcasting Corporation (1997) 189 CLR 520 at 574; Rogers v Nationwide News Pty Ltd (2003) 216 CLR 327 at 339-340 [30]-[32] per Gleeson CJ and Gummow J). The decision applies that to a publisher relying on a Royal Commission report without seeking the subjects' response or including court stays of proceedings for delay. It confirms that a publisher's "zero confidence" in the truthfulness of a response does not make it unnecessary to seek one; the response (or the court findings) could be published with scepticism. The decision does not change the law but illustrates that reliance on an inquiry's findings is not a shield if the publisher goes beyond fair report and omits material that would give a balanced picture.
On s 10 of the Defamation Act, the law reflected the common law rule that a personal action dies with the person (Woolworths Ltd v Crotty (1942) 66 CLR 603). The decision confirms that where a cause of action has merged in a judgment, an appeal challenging a costs order (particularly where the appellant has an independent interest via undertaking) does not involve asserting the defamation cause itself (Ryan v Davies Bros Ltd (1921) 29 CLR 527). The decision does not change the law but applies it to a situation where one appellant dies during an appeal and the surviving appellant has funded the litigation and given an undertaking.
After the decision, the law is clarified in several respects. First, in federal court defamation proceedings, s 17(2) of the Royal Commissions Act has no direct operation; Royal Commission evidence is admissible under the Evidence Act but may be excluded or limited under ss 135 and 136 if the circumstances of compulsion and the expectation of protection create unfair prejudice. Second, publishers relying on Royal Commission findings for serious allegations must still satisfy s 30(3)(h) by seeking a response or including material that would give the subjects' side (including subsequent court findings of prejudice). Third, old expert reports prepared for inquiries must still satisfy s 79; if they do not disclose facts, assumptions and reasoning, or cannot be tested, they are likely to be excluded under s 135. Fourth, the ordinary reasonable reader test applies robustly to books using collective nouns after naming individuals. The decision will require publishers and practitioners to be more cautious when relying on historical inquiries without balancing material, and to ensure expert evidence complies with modern standards even if prepared long ago. It reinforces the high bar for truth and qualified privilege defences in serious defamation cases involving historical events.
Key passages with plain-English translation
Paragraph [30] (Rares J, with Wigney and Lee JJ agreeing on meaning): "The test for determining what an ordinary reasonable reader of a publication would have understood it to convey adapts to the nature of the particular publication. Here, the reader is a person who has decided to read a serious paperback work about, as the front cover states, 'The incredible untold story of scientology in Australia', by a well-known investigative journalist."
Plain-English translation: When deciding what a publication means, judges must imagine a typical reader of that type of material. For a serious investigative book, the reader is someone who buys it expecting careful research, not a lawyer dissecting every word.
Paragraph [32] (Rares J): "The ordinary reasonable reader would draw the implication that whenever chapter 14 used the word 'doctors' it referred to all of the four doctors whom it named on pp 178-179 as 'Bailey and his fellow Chelmsford doctors, John Herron, John Gill and Ian Gardiner'."
Plain-English translation: Once the book names four doctors together and then talks about 'the doctors' doing bad things, the average reader will assume all four are included, not just the most famous one.
Paragraph [37] (Rares J): "The reader is allowed to draw implications much more freely than a lawyer, especially when the implications are derogatory: Trkulja 263 CLR at 160-161 [32]."
Plain-English translation: Ordinary readers don't read like lawyers. They jump to negative conclusions more easily, especially if the words suggest wrongdoing. Courts must allow that.
Paragraph [113] (Rares J): "If s 17(2) of the NSW Act were construed by having regard to whether a witness before a Royal Commission were a party to a proceeding on a particular side of the record, the legislative purpose of ensuring that the evidence that he or she gave to the Commission was complete and truthful would be defeated."
Plain-English translation: If the protection against using Royal Commission evidence only applied when the person is a defendant, witnesses would never know if they were safe to speak fully. That would defeat the whole point of the law, which is to get the truth by promising protection.
Paragraph [125] (Rares J): "The tender of evidence or a document, that a witness gave or produced under compulsion to a Royal Commission, to found a defence of fair report, qualified privilege or honest opinion is not a tender against a witness. Rather, it is a tender against a party in order to prove as a fact that something was said or produced to the Royal Commission which the defendant or respondent reported as a fact that had occurred or used as material on which he, she or it expressed an opinion."
Plain-English translation: You can use Royal Commission evidence to prove that someone said something (for example, to defend a fair report claim), but you cannot use it to prove that what they said was true if that would harm the person who said it.
Paragraph [138] (Rares J): "The proper construction of s 17(2) is the same as Brennan J held for s 6DD (154 CLR at 228), namely: The occasion for its application is the tendering against a witness (i.e., a witness who has given evidence before a Commonwealth commission) of evidence of what the witness stated or disclosed in the course of giving his evidence before the commission. The section requires the court to refuse admission in evidence of any such statement or disclosure."
Plain-English translation: The law means exactly what it says: you cannot use what someone was forced to say at a Royal Commission against them in court. The court must refuse to let it in.
Paragraph [177] (Rares J, citing Lange at 574): "Whether the making of a publication was reasonable must depend upon all the circumstances of the case. But, as a general rule, a defendant's conduct in publishing material giving rise to a defamatory imputation will not be reasonable unless the defendant had reasonable grounds for believing that the imputation was true, took proper steps, so far as they were reasonably open, to verify the accuracy of the material and did not believe the imputation to be untrue. Furthermore, the defendant's conduct will not be reasonable unless the defendant has sought a response from the person defamed and published the response made (if any) except in cases where the seeking or publication of a response was not practicable or it was unnecessary to give the plaintiff an opportunity to respond."
Plain-English translation: To get qualified privilege, the publisher must act reasonably. That almost always means checking the facts, believing the story is true, and giving the person criticised a chance to reply—unless it's impossible or pointless. Just relying on an old inquiry isn't enough if you could have asked for the other side.
Paragraph [266] (Lee J): "As for whether the reports were admissible pursuant to s 79 of the Evidence Act, the deficiencies in the reports were manifest and by no means merely technical or formal. The facts or assumptions upon which the doctors based their opinions were, for the most part, not disclosed or made sufficiently clear in the reports. Nor was the reasoning process which lay behind the opinions exposed, or at least adequately exposed. In some cases the documents or other material relied upon by the doctors was unclear, unavailable or not in evidence. It was, in all the circumstances, not possible to assess or determine the extent to which the opinions expressed by the doctors were based on their specialised knowledge, as required by s 79 of the Evidence Act."
Plain-English translation: The dead experts' reports were so unclear about what facts they relied on and how they reached their conclusions that the court could not tell if the opinions were properly based on their expertise. That is the basic test under s 79. The reports therefore could not be admitted as expert opinion.
Paragraph [268] (Lee J): "As for the prospect of discretionary exclusion under s 135 of the Evidence Act, the problems arising from the deficiencies in the reports were compounded by the fact that the opinions expressed by the doctors were not tested by cross-examination at the Royal Commission and were not able to be tested at the trial in this proceeding. The inability of Mr Herron and Dr Gill to cross-examine the doctors went not only to the danger of unfair prejudice, but also to the probative value of the evidence. The primary judge erred in not having regard to the fact that the opinions of the doctors had not been, and were unable to be, tested when assessing the probative value of the evidence for the purposes of s 135 of the Evidence Act."
Plain-English translation: Not being able to cross-examine the experts made the evidence both less valuable and more unfair. The trial judge should have taken that into account when deciding whether to exclude it under s 135. She didn't, and that was an error.
Paragraph [299] (Lee J): "It is ultimately not possible to excise the evidence of the 'dead experts' from the primary judge's reasoning process in respect of the substantial truth of the imputations. It cannot be concluded that her Honour would necessarily have made the same findings even without the evidence of the 'dead experts', or that the substantial truth of the imputations had been proved on the basis of the other evidence."
Plain-English translation: The trial judge's decision that the allegations were true was so mixed up with the dead experts' evidence that we cannot separate it. We cannot say she would have reached the same conclusion without it. A new trial is needed.
Paragraph [187] (Rares J): "Mr Cannane made no mention in chapter 14 of those significant, consistent findings of the Courts from 1986 to 1993, or of the impact of the significant prejudice that the delay between events over a decade earlier would occasion Mr Herron and Dr Gill in dealing with allegations in, or the findings of, the Royal Commission. The primary judge erred in finding that Mr Cannane's reasons for failing to approach Mr Herron or Dr Gill for comment were reasonable. His conduct was unreasonable particularly because he knew each of them was alive and strongly disputed the findings of the Royal Commission. Yet, he made no reference to those matters or the findings of the Courts about the significant prejudice to which they had been exposed since 1985, by delay of others in bringing proceedings against them, well before the Royal Commission, so as to require these proceedings to be stayed."
Plain-English translation: The book left out important court decisions that said the doctors could not fairly defend themselves because of long delays by the authorities. That was not reasonable. The publisher knew the doctors disputed the Royal Commission and were still alive. He should have included their side or the court findings. Leaving them out made the publication unfair.
What fact patterns trigger this precedent
This decision will be triggered in future defamation cases involving publications that rely on the findings of public inquiries (Royal Commissions, coronial inquests, parliamentary committees) to make serious allegations against named individuals. It applies whenever the publication uses generalised language after naming people ("the doctors", "the surviving Chelmsford doctors") in a context that invites the reader to attribute collective misconduct to all of them. It is particularly relevant where the publisher has not sought or published the subjects' response, or has omitted material court findings that cast the inquiry's conclusions in a different light (for example, stays of proceedings for delay causing prejudice).
The evidentiary principles apply to any federal court proceeding where a party seeks to tender:
- Reports or opinions prepared for an earlier inquiry that do not disclose the facts, assumptions and reasoning process in a way that allows the court to assess whether the opinion is based on specialised knowledge (s 79 Evidence Act).
- Evidence of statements or documents produced under compulsion to a Royal Commission or similar body. While admissible under the Evidence Act, the circumstances of compulsion (and any statutory promise of protection) are relevant to whether the evidence should be excluded or its use limited under ss 135 or 136.
- Hearsay evidence of unavailable witnesses (including dead experts) where the notice under s 67 does not sufficiently identify the specific representations relied upon, or where the evidence cannot be tested.
The decision on qualified privilege is triggered where a publisher relies on an official report but goes beyond fair report by adding its own narrative, assertions of continued "avoidance of accountability", or failure to include subsequent curial findings that the subjects could not fairly defend themselves. It is especially relevant where the publisher knows the subjects are alive, dispute the findings, and could have been asked for a response.
The s 10 analysis applies where one party to a defamation appeal dies and the surviving party has an independent interest (for example, a costs undertaking) in challenging the judgment. It confirms that once a cause of action merges in a judgment, an appeal against costs orders is not the assertion of the defamation cause itself.
Practitioners should note the emphasis on the "prime duty" of an expert to expose the facts and reasoning (Makita; Dasreef; HG v The Queen). Old reports prepared for inquiries will rarely satisfy that duty. Publishers and their lawyers must ensure that any expert evidence relied upon for a truth defence complies with modern standards or risk discretionary exclusion. When relying on Royal Commission material, consider whether s 17(2) of the Royal Commissions Act (or analogues) creates a basis for limiting use under s 136 even if the provision does not directly apply in federal jurisdiction.
How later courts have treated it
As the decision is recent (29 April 2022), there are no reported cases that have yet cited or applied it. However, its principles are consistent with, and reinforce, existing High Court and appellate authority.
The approach to imputation conveyance builds on Trkulja v Google LLC (2018) 263 CLR 149 and Favell v Queensland Newspapers Pty Ltd (2005) 221 ALR 186. Later courts are likely to apply it to other publications that use collective nouns or generalised allegations after naming individuals (for example, reports of inquiries into institutional abuse, corporate misconduct or professional negligence). It confirms that the "generosity" of the ordinary reasonable reader test is not limited to newspapers or social media but applies to serious books.
The analysis of s 17(2) of the Royal Commissions Act and its interaction with the Evidence Act in federal jurisdiction follows directly from Rizeq v Western Australia (2017) 262 CLR 1 and Masson v Parsons (2019) 266 CLR 554. It is likely to be treated as authoritative on the point that State laws purporting to regulate admissibility cannot apply of their own force in federal court. The rejection of the Melbourne Corporation and s 118 arguments reinforces the exclusivity of Ch III. Future cases involving compelled evidence from State inquiries in federal proceedings will cite the decision for the proposition that s 56(1) of the Evidence Act prevails, but that the circumstances of compulsion remain relevant to ss 135 and 136.
The treatment of the dead experts' reports applies Dasreef Pty Ltd v Hawchar (2011) 243 CLR 588, Makita (Australia) Pty Ltd v Sprowles (2001) 52 NSWLR 705 and HG v The Queen (1999) 197 CLR 414. It illustrates that the "factual basis rules" are not a "counsel of perfection" only for modern reports; they apply with equal force to old reports prepared for inquiries. The emphasis on the danger of unfair prejudice from inability to cross-examine aligns with Clyne v Law Society of New South Wales (NSWCA, 4 September 1987, unreported) and Amalgamated Television Services Pty Ltd v Marsden [2002] NSWCA 419. Later courts are likely to treat the decision as confirming that expert reports from unavailable witnesses prepared without modern safeguards will usually be excluded under s 135 in serious cases where the opinions go to central issues such as causation of death or professional negligence.
The qualified privilege reasoning applies Lange v Australian Broadcasting Corporation (1997) 189 CLR 520 at 574 and Rogers v Nationwide News Pty Ltd (2003) 216 CLR 327 at 339-340 [30]-[32]. It is likely to be cited for the proposition that a publisher's "zero confidence" in the truthfulness of a potential response does not make it unnecessary to seek one; the response (or court findings of prejudice) can be published with appropriate scepticism. The decision's treatment of the omission of court stays of proceedings for delay (Herron v McGregor (1986) 6 NSWLR 246; Gill v Walton (1991) 25 NSWLR 190; Walton v Gardiner (1993) 177 CLR 378) is likely to be applied in cases where publishers rely on inquiries but omit subsequent curial findings that the subjects could not fairly defend themselves. It reinforces that reliance on an inquiry's "unchallenged" findings is not a shield if the publisher goes beyond fair report.
The s 10 analysis applies Ryan v Davies Bros Ltd (1921) 29 CLR 527 and confirms that merger of a cause of action in a judgment takes the matter outside s 10. It is likely to be cited in any defamation appeal where one party dies and the survivor has an independent interest in costs or other orders.
Overall, the decision is likely to be treated as an important application of settled principles rather than a radical change. It will be cited for the strict approach to expert evidence from unavailable witnesses, the need to consider discretionary exclusion of compelled inquiry material even when s 17(2) does not directly apply, and the high bar for reasonableness under s 30 where a publisher does not seek a response or include material that would give the subjects' side. Practitioners will need to be more cautious when advising on the admissibility of old inquiry reports and the reasonableness of publishing serious allegations without giving the subjects an opportunity to reply.
Still-open questions
The decision leaves several questions for future cases.
First, the precise construction of s 17(2) of the Royal Commissions Act (and analogues such as s 6DD of the Royal Commissions Act 1902 (Cth)) remains open in State courts. Rares J (with Wigney J agreeing) held that the provision prohibits admissibility against the witness in any civil or criminal proceeding. Lee J preferred not to decide the point given the inconsistency with s 56 of the Evidence Act in federal jurisdiction. In a State court exercising State jurisdiction, a different construction might be adopted. The Court of Appeal's decision in Feldman v Nationwide News Pty Ltd (2020) 103 NSWLR 307 (which the primary judge followed) held that the provision applies only where the evidence is used to establish liability against the witness. That reasoning was not overruled and remains binding in New South Wales courts exercising State jurisdiction. Future cases will need to decide whether the Full Court's analysis is preferred.
Second, the interaction between s 17(2) (or analogues) and the fair report defence under s 29 of the Defamation Act 2005 (NSW) remains open in some respects. The decision confirms that the provision does not prevent evidence being tendered to prove that a statement was made in a Royal Commission for the purpose of a fair report defence. However, the precise limits of that exception, and whether s 17(2) could still operate to prevent use of the content to prove truth in a fair report case, were not fully explored. The implied freedom of political communication was held to operate to allow fair reports of Royal Commission proceedings, but the boundaries of that protection (for example, where the report goes beyond what was actually said) were not defined.
Third, the decision does not resolve the precise weight to be given to Royal Commission findings in a subsequent defamation trial. The court accepted that a lay publisher such as Mr Cannane was entitled to treat the Royal Commission as an authoritative source but held that he went further by making his own value judgments and assertions of continued "avoidance of accountability". Future cases will need to decide how far a publisher can rely on an inquiry's findings without independent verification, particularly where the subjects have not had an opportunity to test them in a curial setting.
Fourth, the decision leaves open the precise approach to discretionary exclusion under s 135 or limitation under s 136 of Royal Commission material in cases where s 17(2) does not directly apply. The Full Court held that the primary judge erred in not considering those discretions but could not undertake the exercise itself given the volume of material. Future cases will need to develop a principled framework for when the circumstances of compulsion (and the witness's expectation of protection) create a danger of unfair prejudice that substantially outweighs probative value.
Fifth, the decision does not address the position where a publisher seeks to tender only part of a dead expert's report or limits its use to a non-hearsay purpose (for example, to prove the fact that the opinion was given to the Royal Commission). The court proceeded on the basis that the reports were tendered for the truth of their contents. Future cases may explore whether more limited tender, coupled with a s 136 limitation, could render such evidence admissible.
Sixth, the precise limits of the "factual basis rules" for expert reports prepared for inquiries remain open. The decision applies Makita and Dasreef strictly but does not decide whether the Red Bull approach (that the rules are a "counsel of perfection" and often go to weight rather than admissibility: Sydneywide Distributors Pty Ltd v Red Bull Australia Pty Ltd (2002) 234 FCR 549 at 552 [7] per Branson J, 576 [87] per Weinberg and Dowsett JJ) applies to reports prepared for non-curial purposes. The court noted the reports were fundamentally different from those prepared for litigation but did not decide whether a less strict approach might apply in some cases.
Seventh, the decision does not resolve the position where one appellant dies and the estate is insolvent but a third party (not the surviving appellant) seeks to continue the appeal. The court granted Dr Gill leave nunc pro tunc because of his undertaking. The position of a non-party with no direct interest remains open.
Finally, the decision leaves open the precise approach to costs in a retrial where some defences have been resolved in the plaintiff's favour. The court directed the parties to file submissions on the form of orders. Future cases will need to decide how costs of the first trial should be dealt with where the plaintiff succeeds on some issues but a retrial is ordered on others.
These open questions will require careful consideration in future proceedings. The decision provides a strong foundation for arguing that old inquiry reports and compelled evidence must be approached with caution, that publishers cannot hide behind inquiries without giving the subjects a fair opportunity to respond, and that the rules of evidence apply with full force even to material prepared decades earlier for a different purpose. It is likely to lead to more cautious pleading of truth and qualified privilege defences and greater attention to the admissibility and weight of historical expert and inquiry material.