Eppinga v Kalil

1. [1]

   PAYNE JA: Ellie Rose Eppinga appeals from orders made in the District Court dismissing her cross-claim brought in defamation against the respondents, Dr Mahmoud Kalil and Ms Magda Kalil. The original claim, also in defamation, was brought against Ms Eppinga by Dr Kalil and Ms Kylie Eather and later discontinued. Ms Eather was not a party to the cross-claim or the present appeal.

2. [2]

   The primary judge found that the respondents had established a defence of common law qualified privilege, rejecting the appellant’s argument that the respondents’ publication was actuated by malice. The primary judge’s findings about malice are the focus of the appeal.

Relevant Facts

1. [3]

   The proceedings arose from an unpaid work experience placement Ms Eppinga undertook on 25 January 2019 at the Beacon Hill Veterinary Hospital (“the Vet Hospital”), owned and operated by the respondents. At the time of the work-experience placement, the appellant was enrolled in a veterinary nursing course at TAFE.

2. [4]

   Ms Eppinga’s placement was remarkably brief. She arrived at about 8 am and left no more than 45 minutes later. Soon after arriving, the appellant gave evidence that she became concerned about the Vet Hospital’s hygiene standards. She expressed those concerns to the respondents and there was an argument, before Ms Eppinga was told to leave the premises.

3. [5]

   Beyond these basic facts, what happened at the Vet Hospital during the appellant’s placement is a matter of dispute.

4. [6]

   Shortly after Ms Eppinga left the Vet Hospital, Dr Kalil, Ms Eather and Ms Eppinga each contacted the police with reports about events occurring during Ms Eppinga’s work placement.

5. [7]

   At some point prior to 1 pm on 25 January 2019, Ms Eppinga attended Dee Why police station where she alleged that she had been verbally abused at the Vet Hospital. She said she did not want the matter to be investigated.

6. [8]

   Also on 25 January 2019, Ms Eather contacted Northern Beaches police. Police officers visited the Vet Hospital. The COPS report of that visit and related investigations was marked as Ex C9. When the police arrived, Ms Eather gave them her version of the morning’s events, including the following:

7. [9]

   It was common ground that “been up”, in the second line above, was a typographical error. I accept, as the appellant urged, that the report should likely have read “picked up”.

8. [10]

   After hearing Ms Eather’s version of events, a police officer called Ms Eppinga. During that telephone call (which is summarised in Ex C9), Ms Eppinga denied “picking up medication”.

9. [11]

   On 26 January 2019, Dr Kalil visited Dee Why police station, where he made a witness statement about the previous day’s events and about a number of threatening phone calls he described receiving on 26 January 2019. That statement was marked Ex C10. In that statement, Dr Kalil relevantly said:

10. [12]

    On 25 January 2019, at 16:53, the appellant made the following post on her Facebook account:

11. [13]

    Later on 25 January 2019, appellant also started a petition, in substantially identical terms to her Facebook post, on the website change.org.au. Finally, she made a public post to the “Sydney Northern Beaches Buy, Swap & Sell” group on Facebook:

12. [14]

    On 3 February 2019, the following post appeared on the Beacon Hill Veterinary Hospital Facebook page:

13. [15]

    This post is the matter complained of by the appellant in her cross-claim.

Procedural background

1. [16]

   On 11 February 2019, Dr Kalil and Ms Eather, began proceedings in defamation against Ms Eppinga and another person, Lisa Milham. A pleading and costs issue came to this Court: Muriniti v Kalil [2022] NSWCA 109. The claim complained about Ms Eppinga’s 25 January 2019 Facebook post and her related online publications. Ms Milham’s involvement is not relevant to the present proceedings. On 4 December 2020, Dr Kalil and Ms Eather discontinued their claim against Ms Eppinga and Ms Milham.

2. [17]

   On 27 June 2019, the appellant filed a cross-claim complaining about the 3 February 2019 post on the Vet Hospital’s Facebook page. The defendants to that action were Dr Kalil and Ms Magda Kalil.

3. [18]

   On 10 March 2021, Ms Eppinga filed an amended statement of cross-claim. In it, the appellant alleged that the respondents’ 3 February 2019 Facebook post conveyed the following 14 imputations:

4. [19]

   Whilst the matter complained of referred to Ms Eppinga only as “Ellie”, the appellant pleaded that she was identified by two readers who accessed the matter complained of. The respondents conceded before the primary judge that that the matter complained of was published to the two named readers, and that Ms Eppinga could be identified given the extrinsic facts those two readers knew. The respondents admitted that imputations (g) and (i) were conveyed and defamatory, but denied that the remaining imputations were conveyed.

Primary judgment

1. [20]

   The matter was heard over four days in February and April 2023. On 21 April 2023, Weber SC DCJ delivered judgment for the respondents. His Honour found that all but imputations (b), (d), (h) and (j) were conveyed and defamatory.

2. [21]

   His Honour also found that the scope of the matter’s publication was “extremely limited”: PJ [22]. The evidence led by the appellant established publication to only two persons, and in the context of a website like Facebook, more extensive publication was not presumed. The limited scope of publication, his Honour found, was relevant to the assessment of damages.

3. [22]

   The principal issue was the respondents’ pleaded defence of qualified privilege. The basis for the defence was that the matter of complained of was a “reply to an attack” and therefore published on a privileged occasion. The pleaded “attack” was the appellant’s 25 January 2019 Facebook post and her change.org petition. The appellant conceded that the matter complained of was published on a privileged occasion and was sufficiently connected to that occasion. The only issue was malice.

4. [23]

   For the principles on malice, the primary judge cited Roberts v Bass (2002) 212 CLR 1. No complaint was made about his Honour’s description of the governing law.

5. [24]

   The appellant’s case on malice before the primary judge focused on Dr Kalil’s state of mind when publishing the matter complained of. (Ms Kalil’s state of mind appears not to have been in issue and no reference was made to her or her state of mind on appeal). The appellant argued that Dr Kalil acted with malice because he had no actual belief that Ms Eppinga was trying to steal medication from the Vet Hospital. Dr Kalil did not give evidence. The primary judge found he would only draw an inference of the kind drawn in Jones v Dunkel (1959) 101 CLR 298 if an evidentiary onus had shifted to the respondents. His Honour found the evidentiary onus had not shifted. That was because of two aspects of the evidence.

6. [25]

   The first relevant aspect of the evidence was the witnesses statement Dr Kalil gave to the police on 26 January 2019, Ex C10. This statement was tendered, without qualification, by the appellant. In this statement, Dr Kalil stated that he saw Ms Eppinga “holding medication” and believed that she was about to steal it. As part of the statement, Dr Kalil agreed that he would be prepared to repeat its contents as evidence in court, and would be liable to prosecution if what he had said was wilfully false.

7. [26]

   The second relevant aspect of the evidence was the testimony of Ms Eather and Ms Eppinga. Ms Eather’s evidence was that, on 25 January 2019, she had entered the Vet Hospital’s “treatment room” with Dr Kalil and found Ms Eppinga with a drug in her hand: PJ [50]. The primary judge found Ms Eather to be a “straight-forward and reliable” witness of credit. By contrast, the primary judge was unimpressed by Ms Eppinga, who he found was defensive, reluctant to make concessions and “missed no opportunity to advance her case by advocacy from the witness box”.

8. [27]

   His Honour drew a specific adverse inference from Ms Eppinga’s evidence about her knowledge of Dr Kalil’s Muslim faith. That evidence was relevant to the allegation, in the matter complained of, that Ms Eppinga had used “a range of vile and racist profanity” when confronted. The evidence was also relevant in light of Dr Kalil’s allegation, made in Ex C9, that during a telephone call on 26 January 2019 Ms Eppinga had said to him “I know you are Muslim, and the police are on your side”. In both evidence in chief and cross-examination, Ms Eppinga denied any knowledge of Dr Kalil’s faith. However, that evidence was contradicted by online communications Ms Eppinga made in January 2019:

   1. (1)

      In a Facebook post, Ms Eppinga said “I arrived at 8 am the day before yesterday … and some strange Muslim chanting music was bellowing through the clinic”.

   2. (2)

      In a Facebook message to her cousin, she then said: “The police are on the fkn Muslims side it baffles me!”. Ms Eppinga accepted that this was a reference to Dr Kali.

9. [28]

   When confronted, in cross-examination, with the second of these communications, Ms Eppinga reluctantly conceded that, on 25 January 2019, she knew Dr Kalil was a Muslim. The primary judge referred unfavourably to other aspects Ms Eppinga’s evidence, and contemporaneous documents fortified his view that she was not a witness of credit. For example, Ms Eppinga’s oral evidence was that she had been physically assaulted at the Vet Hospital. However, according to Ex C9, the COPS report made on 25 January 2019, Ms Eppinga’s contemporaneous complaint was of verbal abuse only.

10. [29]

    The primary judge concluded that Ms Eather’s evidence was to be preferred to Ms Eppinga’s where it conflicted. The primary judge found that Ms Eppinga had a drug in her hand when Ms Eather and Dr Kalil entered the treatment room. Along with Ex C10, this evidence was sufficient to convince the primary judge not to draw a Jones v Dunkel inference from Dr Kalil’s failure to give evidence. In those circumstances, the primary judge found that the appellant failed to discharge her onus to prove malice. Ms Eppinga’s cross-claim therefore failed.

Grounds of appeal

1. [30]

   The appellant advanced 32 separate grounds of appeal.

2. [31]

   In oral submissions, the appellant indicated that the bias ground (ground 32) was no longer pressed. The remaining grounds can be conveniently grouped into four issues. The first, qualified privilege, has three sub-issues:

   1. (1)

      Qualified Privilege (grounds 1-3, 6-22, 24-25, 27-30):

   2. (2)

      Failure to give adequate reasons (grounds 23 and 26);

   3. (3)

      Imputations (ground 31);

   4. (4)

      Damages – extent of publication (grounds 4 and 5).

3. [32]

   While there was a dispute in the written submissions about whether or not this appeal required leave, the respondents were content to proceed on the basis that leave would be granted. I am content to proceed on the basis of the joint submission that leave should be granted but wish to record that, even making every assumption favourable to the appellant, I would not have awarded damages in excess of $100,000. Leave was required and should have been sought.

Issue 1: Qualified Privilege

1. [33]

   At common law, the defence of qualified privilege requires, first, identification of a “privileged occasion”: LVMH Watch & Jewellery Australia Pty Ltd v Michael Lassanah & Ors [2011] NSWCA 370 at [91]-[92] per Bergin CJ in Eq, Giles and Campbell JJA agreeing. Secondly, the defamatory statements must be “sufficiently connected” to the privileged occasion: Bashford v Information Australia (Newsletters) Pty Ltd (2004) 218 CLR 366; [2004] HCA 5 at [27]-[30] per Gleeson CJ, Hayne and Heydon JJ; at [135]-[136] per Gummow J; at [191]-[192] per Kirby J; KSMC Holdings Pty Ltd t/as Hubba Bubba Childcare on Haig v Bowden [2020] NSWCA 28 at [49] per Payne JA.

2. [34]

   A person who responds to a public attack on their reputation can avail themselves of the privileged occasion that attack generates. The publisher subject to a prior attack has a defence of qualified privilege, provided the response is “sufficiently connected” to the occasion of privilege: Adam v Ward [1917] AC 309; Loveday v Sun Newspapers Ltd (1938) 59 CLR 503; Hubba Bubba at [49].

3. [35]

   In this case, Mr Muriniti, who represented the appellant at trial and in this Court, specifically accepted that as it responded to Ms Eppinga’s published attacks, the matter complained of was published on an occasion of qualified privilege. Mr Muriniti also conceded that each of the defamatory statements alleged were sufficiently connected to the privileged occasion and were relevant to that occasion.

4. [36]

   Qualified privilege is, however, defeated by a finding that the publisher, in making the defamatory statement, was actuated by malice. As I said in Hubba Bubba:

5. [37]

   The principal issue in this appeal was the primary judge’s finding that Ms Eppinga had not proved the respondents were actuated by malice in publishing the matter complained of. The appellant’s challenges on this issue were directed to the primary judge’s assessment of the evidence. There was no challenge to the principles of law he applied.

6. [38]

   An important part of the primary judge’s decision on malice was his Honour’s observation of the two principal witnesses in the witness box, being the appellant and Ms Eather, and his assessment of their credit. Where their evidence conflicted, the primary judge preferred the evidence of Ms Eather to the evidence to Ms Eppinga, including on the issue of what occurred in the treatment room.

7. [39]

   The appellant challenged the primary judge’s credit findings. The appellant submitted that Ms Eather should have been assessed as a witness of limited credit because:

   1. (1)

      she was “combative, non-responsive, argumentative” and “an advocate for Dr Kalil”.

   2. (2)

      during oral testimony, Ms Eather allegedly gave two conflicting accounts of whether or not she saw Ms Eppinga “going through drawers”. This inconsistency, the appellant submitted, indicated a deliberate change of story.

   3. (3)

      there were said to be inconsistencies between Ms Eather’s evidence and what were said to be “incontrovertible facts”. In particular, the appellant focused on the evidence that Ms Eather had instructed Ms Eppinga to clean the treatment room, and the evidence about the physical position of the medicine Ms Eppinga was said to have picked up. It was submitted that Ms Eather’s evidence was fabricated to present a version of events favourable to Dr Kalil.

   4. (4)

      Ms Eather’s conduct in the defamation claim she and Dr Kalil brought against Ms Eppinga, including her role in the personal costs order made against the appellant’s lawyers, was said to reflect unfavourably on her credit.

8. [40]

   The appellant also submitted, without elaboration, that Ms Eppinga should have been accepted as a witness of credit.

9. [41]

   The appellant asks this Court to interfere on appeal with findings of fact made by the primary judge based upon the credit of the principal witnesses. The principles governing this exercise were explained in Fox v Percy (2003) 214 CLR 118; [2003] HCA 22, and authoritatively and more recently restated by Bell, Gageler, Nettle and Edelman JJ in Lee v Lee (2019) 266 CLR 129; [2019] HCA 28:

10. [42]

    In the present appeal, the findings of fact under challenge are based on credit findings “likely to have been affected by impressions about the credibility and reliability of witnesses formed by the trial judge as a result of seeing and hearing [the witnesses] give their evidence”. To set those findings of fact aside, this Court must be persuaded the findings were “glaringly improbably” or “contrary to compelling inferences”. I am not so persuaded.

11. [43]

    I have concluded that it was well open to his Honour to form the view that he did of the credit of both Ms Eppinga and Ms Eather. That view was formed based on a number of matters, including the manner in which both witnesses gave evidence. The primary judge found that Ms Eppinga was combative and defensive; Ms Eather was measured and prepared to make appropriate concessions. These were essentially demeanour-based credit findings.

12. [44]

    Ms Eather’s evidence was said to contain two inconsistent accounts of whether she saw Ms Eppinga “going through drawers” in the treatment room. In cross-examination Ms Eather said:

13. [45]

    Mr Muriniti then asked whether Ms Eather mentioned Ms Eppinga going through the drawers when she spoke to police on 25 January 2019:

14. [46]

    I reject the submission that Ms Eather’s accounts were inconsistent. Ms Eather’s position, from the outset, was that she found Ms Eppinga going through drawers in the treatment room. She did not observe Ms Eppinga doing so, but inferred that she was because a drawer near Ms Eppinga was open which previously had been shut. Ms Eather was clear in distinguishing between what she had seen and what she had inferred. Her answer “I did not. I had nothing to do with the document whatsoever” was, as Mr Muriniti conceded in this Court, Ms Eather’s explanation that she had no direct role in preparing the COPS report in Ex C9, and that she was not responsible for how the police had recorded her explanation of events.

15. [47]

    Nor do I accept Mr Muriniti’s submission Ms Eather’s evidence is suspect because Ms Eather had not at any earlier point explained, to the police or in communications with the veterinary board, the basis for her inference. Mr Muriniti’s cross-examination was the first occasion that it became relevant for Ms Eather to explain the inference she drew.

16. [48]

    I also reject the appellant’s submission that facts found based on Ms Eather’s evidence were glaringly improbable or inconsistent with other allegedly “incontrovertible facts”. The first alleged inconsistency was in Ms Eather’s account of whether she had instructed Ms Eppinga to clean the treatment room. Ex C9, the COPS record, suggested she had. In oral testimony, Ms Eather could not recall whether she had given that instruction:

17. [49]

    This evidence is consistent with a witness doing her best to recall events some years ago. This evidence does not adversely affect her credit. Ms Eather conceded that the contemporaneous COPS entry said what it did, but maintained she had no independent memory, as she sat in the witness box four years later, of instructing Ms Eppinga to clean the treatment room. Mr Muriniti suggested that it was implausible Ms Eather had forgotten her instructions to Ms Eppinga. I do not regard the evidence as in the slightest implausible, much less do I accept that Ms Eather’s evidence on this point was fabricated. No inconsistency between Ms Eather’s evidence and incontrovertible facts was shown.

18. [50]

    The second alleged inconsistency with “incontrovertible facts” was between Ms Eather’s evidence at the trial and Ex C9’s account of whether the medication was in or on top of a medicine cabinet or on the treatment table. The relevant section of Ex C9, the COPS entry, is extracted above at [8]. There, Ms Eather is said to have observed Ms Eppinga picking up medication, before Dr Kalil told Ms Eppinga to “put it back on the table”. In cross-examination, Ms Eather said she observed Ms Eppinga near a medicine cabinet. Then came the following exchange:

19. [51]

    At its highest, this cross-examination established a lapse in memory, a possibility that Ms Eather candidly acknowledged. No inconsistency between the evidence and an “incontrovertible fact” existed.

20. [52]

    I have concluded that there were no glaring improbabilities in Ms Eather’s evidence nor were facts found on the basis of that evidence inconsistent with any incontrovertible facts. The primary judge enjoyed the advantage of seeing Ms Eather give evidence and consider her demeanour while giving that evidence. The primary judge was entitled to conclude that Ms Eather was a witness of credit. Nothing in the record of that evidence persuades me that this finding should be set aside.

21. [53]

    There was one remaining attack on Ms Eather’s credibility. In the primary claim, a personal costs order against Ms Eppinga’s lawyers made by a District Court judge was set aside on appeal by this Court: Muriniti v Kalil [2022] NSWCA 109. In that decision, Brereton JA set out nine reasons why the application for a personal costs order might be viewed as an abuse of process, but, at [102], stopped short of deciding it was. His Honour was highly critical of the solicitors who brought the application, Goldsmiths, who did not appear in this Court in the present case. Brereton JA was not, however, critical of their clients, Ms Eather or Dr Kalil. Nonetheless, the appellant described Ms Eather as “proactive in what was the commission of an abuse of process”. The submission appeared to be that, because Goldsmiths were acting on Ms Eather’s instructions, she was responsible for making an application for an unmeritorious and unsuccessful personal costs order. Ms Eather’s alleged involvement in “the commission of an abuse of process”, the appellant said, gave her a motive to lie in the present proceedings.

22. [54]

    I reject the appellant’s submission. Even if the appellant was right that Brereton JA’s findings reflected in some way unfavourably on the non-legally trained Ms Eather’s conduct in instructing her lawyers to seek the personal costs order, it simply does not follow that Ms Eather had a motive to lie in the present proceedings.

23. [55]

    There is, in any event an issue in considering this submission. Section 91(1) of the Evidence Act 1995 (NSW) provides:

24. [56]

    It may be, as Mr Muriniti submitted, that the personal costs order appeal was part of the same “proceedings” as this cross-claim. There may be some support for this view in King v Muriniti [2018] NSWCA 98. There Mr Muriniti represented a client whose four proceedings were dismissed by the Court of Appeal. The client could not meet the costs order against her, and so the defendants sought personal costs from Mr Muriniti under s 99 of the Civil Procedure Act 2005 (NSW). The defendants said that findings made in the course of the now-dismissed proceedings were relevant to the personal costs application. The question was whether s 91 applied. In deciding that it did not, Basten JA explained:

25. [57]

    His Honour then held that when the Court exercised its personal costs power under s 99 of the Civil Procedure Act, it was doing so within the underlying substantive proceedings: at [37]. Similarly, Emmett AJA said it was “artificial” to treat the costs application as separate proceedings simply because Mr Muriniti was not formally a party to the earlier proceedings: at [97].

26. [58]

    In the present case, it is ultimately unnecessary to determine whether evidence of Brereton JA’s findings is admissible despite s 91. Even if that evidence is admitted, the findings are neither relevant nor probative to the question of Ms Eather’s credit. This difficulty is insurmountable. Brereton JA made no finding about Ms Eather’s role in the costs application. It was never put to Ms Eather that she was responsible for that application. No doubt Goldsmiths lawyers sought Ms Eather and Dr Kalil’s instructions before making the costs application. But I also infer that Goldsmiths advised their clients, both lay people with no legal experience, that a costs application was a sensible forensic tactic. I am not persuaded that Ms Eather acted with an improper purpose in giving those instructions, no doubt on her lawyer’s advice. Even if I were so persuaded, it would not follow that Ms Eather’s credit should be reassessed. As I have explained, making every assumption favourable to Ms Eppinga, Ms Eather’s conduct in the costs application provides no support for an inference that she had a motive to lie in the present proceedings. Tellingly, before the primary judge it was not put to Ms Eather that her conduct in relation to the costs application gave her a motive to lie.

27. [59]

    I reject the appellant’s reliance on Ms Eather’s conduct during the costs application as a ground for impugning her credit. There is no reason to doubt the primary judge’s assessment of Ms Eather’s credit.

28. [60]

    The appellant also sought to disturb the primary judge’s unfavourable findings about Ms Eppijnga’s credit. Mr Muriniti conceded “there was an issue” about his client’s evidence about whether she believed, on 25 January 2019, that Dr Kalil was Muslim. In answer to this question Ms Eppinga was initially adamant that she had no knowledge, at the time, of Dr Kalil’s religion. She persisted in that assertion even when confronted by the Facebook posts explained above at [27], and only after an extended exchange conceded with great reluctance that she did know Dr Kalil’s faith at the relevant time:

29. [61]

    The primary judge was correct to conclude that those answers reflected very poorly on Ms Eppinga’s credit.

30. [62]

    Mr Muriniti’s next complaint concerning Ms Eppinga’s credit was that Mr Smark SC, who appeared for the respondents below and in this Court, did not put to Ms Eppinga in cross-examination that any aspect of her evidence was untruthful. This complaint is without merit. The purpose of the rule in Browne v Dunn (1893) 6 R 67 is procedural fairness. Complying with the rule ensures the witness may respond to the basis on which they are to be contradicted or discredited. Here, it was squarely put to the appellant that she had picked up a bottle of medication, was seen holding it by Dr Kalil, and was then asked to leave the Vet Hospital. These questions put the appellant on notice that the respondents’ version of events was contradictory to hers. In answering the questions asked, Mr Eppinga had ample opportunity to put her version of events forward. It was not necessary or helpful in those circumstances formally to “put” to Ms Eppinga that her evidence was “untruthful”. Ms Eppinga was confronted with the basis on which she was to be contradicted and was allowed a proper opportunity to make her response.

31. [63]

    Mr Smark SC did not submit that a positive finding should be made that Ms Eppinga was lying in a particular part of her evidence. The submission was, rather that Ms Eather’s evidence should be preferred to Ms Eppinga’s in cases of conflict. No error has been shown in the approach of the primary judge in preferring the evidence of Ms Eather to that of Ms Eppinga based on credit.

32. [64]

    There is no reason to disturb the primary judge’s factual findings based on credit. The primary judge was entitled to conclude that Ms Eppinga was not a witness of credit and that Ms Eather’s evidence should be preferred to Ms Eppinga’s in cases of conflict.

33. [65]

    Ex C10 was Dr Kalil’s statement to the police. It was tendered without restriction by Mr Muriniti and was available for all purposes. In Ex C10, Dr Kalil deposed to seeing Ms Eppinga holding medication, and explicitly stated his belief that she was attempting to steal the medicine in question. Having tendered Ex C10, without restrictions, Mr Muriniti faced a formidable hurdle in establishing that Dr Kalil had no honest belief Ms Eppinga was trying to steal drugs from his practice.

34. [66]

    On appeal, the appellant’s principal submission was that there was an inconsistency between the account given by Dr Kalil in Ex C10 and the contents of the matter complained of. This inconsistency was said to undermine the probative value of Ex C10 or permit an inference that Dr Kalil had fabricated some or all of his version of events.

35. [67]

    In his police statement (Ex C10), Dr Kalil stated (at [5]):

36. [68]

    The matter complained of said:

37. [69]

    The differences between Ex C10 and the matter complained of do not establish that Dr Kalil knew or believed that what he published was false. Crucially, both accounts allege an attempted theft (“going to steal” and “caught red handed” respectively). The purpose of Ex C10 was for Dr Kalil to give a short account to police of the events of 25 and 26 January 2019, and in particular, the threatening calls and negative online publicity that began on 26 January 2019. Dr Kalil’s impression that Ms Eppinga was trying to steal medicine was one relevant topic among several.

38. [70]

    The purpose of the matter complained of, by contrast, was to defend the Vet Hospital against the attacks made on it by the appellant. It was to be expected in those circumstances that the matter complained of would dwell in more detail on Ms Eppinga’s perceived motive for her attacks, and the events which in Dr Kalil’s view demonstrated that motive.

39. [71]

    Further, Dr Kalil’s account to police was taken by police officers, the day after the events in question. Its language was understandably careful and reflected fresh recollections of the incidents. Ex C10 was written by Dr Kalil or his associate and posted online as a reply to a heated attack on the Vet Hospital, written some eight or nine days later. It was to be expected that the matter complained of would contain more colourful language. None of those features demonstrate that Dr Kalil knew it was untrue that Ms Eppinga was attempting to steal drugs. None of these features suggests that Dr Kalil altered his story or fabricated any part of it.

40. [72]

    Rather, Ex C10 (which the appellant tendered without restriction) is powerful evidence against the appellant’s submission that Dr Kalil knew that any part of the matter complained of was false. Further evidence elicited by the appellant in cross-examination of Ms Eather strengthened that conclusion:

41. [73]

    This was first-hand hearsay evidence of Dr Kalil’s belief that the appellant had been intending to steal medication. It was adduced by the appellant. This evidence supports the account Dr Kalil gave in Ex C10. The primary judge found as much at [73]:

42. [74]

    Mr Muriniti submitted that paragraph [73] contained an error of fact. That submission is inconsistent with the transcript of Ms Eather’s cross-examination, extracted above at [72]: Ms Eather’s evidence, in terms, was that Dr Kalil had “most likely” told her he believed Ms Eppinga was trying to steal the drugs.

43. [75]

    Nonetheless, Mr Muriniti submitted that Ex C10 and Ms Eather’s cross-examination demonstrated, at best, that Dr Kalil believed on 25 and 26 January 2019 that Ms Eppinga was attempting to steal drugs. That evidence, he said, could not prove his belief a few days later, on 3 February 2019, the day he published the matter complained of. I reject the appellant’s submission. The onus of proving malice rested on the appellant. The appellant led no evidence demonstrating that Dr Kalil’s beliefs had changed between 26 January and 3 February 2019. The suggested textual differences between Ex C10 and the matter complained of showed no such thing. It was not for the respondents to show that Dr Kalil held an honest belief. Ex C10 and Ms Eather’s evidence on this subject were both adduced by the appellant and not subject to any restriction. That evidence was fatal to the appellant’s prospects of demonstrating malice.

44. [76]

    I have already rejected the appellant’s principal arguments on the question of malice. Nonetheless, some remaining ancillary challenges should be addressed.

45. [77]

    The appellant maintained her position, already advanced at trial, that the primary judge should have drawn a Jones v Dunkel inference from Dr Kalil’s failure to give evidence. In particular, the appellant argued that the “inconsistency” between Ex C10 and the matter complained of called for an explanation that only Dr Kalil could give. Since Dr Kalil did not do so, the appellant said, the Court should have inferred that the respondents’ case would have been harmed by whatever explanatory evidence Dr Kalil might have given.

46. [78]

    This submission should be rejected. As the High Court explained in ASIC v Hellicar (2012) 247 CLR 345; [2012] HCA 17, adverse inferences of two kinds can flow from the unexplained failure of a party to call a witness whom that party would be expected to call. One is that the trier of fact may infer that the evidence of the absent witness would not assist the case of that party. The other is that the trier of fact may draw an inference unfavourable to that party with greater confidence. Importantly, the High Court emphasised that Jones v Dunkel does not enable the trier of fact to infer that the evidence of the absent witness would have been positively adverse to that party.

47. [79]

    A recent consideration of the circumstances in which a Jones v Dunkel inference might be drawn and the consequences of doing so is found in Ling v Pang [2023] NSWCA 112 at [20]–[34].

48. [80]

    A Jones v Dunkel inference is to be drawn against a party only when that party bears the legal or at least evidential burden of proving an issue: Payne v Parker [1976] 1 NSWLR 191 at 200G-201A per Glass JA. Here, the legal burden of proving malice lay squarely on the appellant. No evidential burden arose. It must be recalled that there was a presumption in favour of the respondents, namely that Dr Kalil acted with honesty of purpose: Hubba Bubba at [59]-[61] per Payne JA (with whom Basten and White JJA agreed). Evidence was needed to rebut that presumption. As I have explained, however, the evidence (particularly evidence adduced by the appellant) weighed strongly against a finding that Dr Kalil acted for an improper purpose. Nor did the asserted “inconsistency” between Ex C10 and the matter complained of call for explanation, since the differences between the two were neither substantial nor unexpected. There was no error by the primary judge in failing to draw a Jones v Dunkel inference.

49. [81]

    In this case it is unremarkable that Dr Kalil chose not to give evidence. In Fraser v Holmes [2009] NSWCA 36; 253 ALR 538, a letter was published on an occasion of qualified privilege, but the trial judge had found malice: at [4]. The defendant did not give evidence, although as in the present case, there was evidence from other sources of his state of mind: [43]-[44]. The Court of Appeal concluded that on the evidence, malice had not been established. There was no Jones v Dunkel issue. The same conclusion should be reached here.

50. [82]

    The appellant advanced a ground of appeal attacking the primary judge’s application of the principle in Briginshaw v Briginshaw (1938) 60 CLR 336. The appellant simply reiterated her argument that the inconsistencies between Ex C10 and the matter complained of revealed an improper purpose, and that Ms Eather’s version of events should not be believed. For the reasons set out above, those submissions should be rejected.

51. [83]

    The appellant submitted that the imputations conveyed by the matter complained of “could have no basis in truth” and were therefore ipso facto malicious. That submission misunderstood the nature of qualified privilege. It is commonplace for a qualified privilege defence to succeed where no truth defence is raised, or where it fails: see Fraser v Holmes; Cush v Dillon (2011) 243 CLR 298; [2011] HCA 30; Murray v Raynor [2019] NSWCA 274; Hubba Bubba. That is because the defences are concerned with different issues. Truth is concerned with proof of facts and events, while qualified privilege is concerned with the relationship between an interest and duty (the occasion of privilege) and subjective belief (malice). So long as the matter complained of was published on an occasion of privilege and without malice, it will still enjoy the protection of qualified privilege.

52. [84]

    Finally, the appellant submitted that the respondents’ belief in truth of the matter complained of was not “reasonable”. For example, she argued that Dr Kalil “implausibly jumped to the conclusion” that Ms Eppinga was about to steal drugs.

53. [85]

    The appellant submitted that “[t]he question of reasonableness is interwoven with the question of malice to the extent that both serve to limit the application of privilege”. This is not a correct statement of the law. “Reasonableness” of conduct is not, of itself, relevant to common law qualified privilege: Roberts v Bass at [5] and [13] per Gleeson CJ; Cush v Dillon at [10] per French CJ, Crennan and Kiefel JJ. Nor does an unreasonable belief amount, on its own, to malice.

54. [86]

    Mr Muriniti tried to distinguish reply to attack cases like Fraser v Holmes on the basis they involved arms-length public political disputes, which left no scope for the defendants to clarify the motivations of those attacking them. In the present case, he submitted, the respondents could have asked Ms Eppinga why she was acting the way she did, rather than inferring she was trying to steal medicine. This submission should be rejected. To establish common law qualified privilege, there is no necessary requirement for a defendant to make reasonable inquiries before publishing. In this way, the defence is different from that of qualified privilege under s 30 of the Defamation Act 2005 (NSW).

55. [87]

    A failure to make inquiries is normally only relevant in a case like the present if that failure can support a finding that the publisher was wilfully blind about a relevant matter. Wilful blindness is a species of actual knowledge. Here the appellant’s case on malice was that Dr Kalil knew the contents of the matter complained of were false. Wilful blindness was not pleaded. The respondents drew the primary judge’s attention to this in closing submissions. Alleged wilful blindness was not an issue in the case. In any event, assuming favourably to the appellant that the issue was litigated, the primary judge found at [41]-[42] that the respondent was not wilfully blind. That conclusion was correct. The appellant failed to prove that Dr Kalil knew or was wilfully blind to the fact that the contents of the matter complained of were false.

56. [88]

    The appellant failed to show that the primary judge’s reasoning on malice was affected by error. The appellant failed to prove Dr Kalil was actuated by malice in publishing the matter complained of. In circumstances where the appellant conceded that the matter was otherwise protected by qualified privilege, it follows that the primary judge was correct that the defence was made out and that Ms Eppinga’s cross-claim should be dismissed.

57. [89]

    My conclusions about malice dispose of the appeal. Even if the appellant had succeeded on issues 3 and 4, the respondents’ successful plea of qualified privilege would remain a complete answer to the appellant’s cross-claim. I will, nonetheless, briefly consider those issues.

Issue 2: Failure to give adequate reasons

1. [90]

   The appellant’s written submissions complain that the primary judge gave inadequate reasons, reflecting ground 23 of appeal. This argument was not developed orally. The essence of the submission was to reiterate matters already addressed concerning malice. The substance of the complaint was that the primary judge failed to explain why:

   1. (1)

      the alleged “inconsistency” between Ex C10 and the matter complained of did not allow an inference that Dr Kalil had fabricated his story about Ms Eppinga;

   2. (2)

      Ms Eather’s evidence should be preferred to Ms Eppinga’s.

2. [91]

   I have explained why these aspects of the primary judge’s reasoning were correct. There was no failure by the primary judge to give adequate reasons for his decision. This ground of appeal should be rejected.

Issue 3: Imputations

1. [92]

   The appellant submitted that the primary judge was wrong to find that four out of the 14 pleaded imputations were not conveyed. Both parties agreed that nothing turned on whether these imputations were carried in addition to the 10 that were found to be carried, given that I have found that the matter complained of was published on an occasion of qualified privilege.

2. [93]

   Had the appellant established malice, the imputation challenge may have been relevant to the assessment of damages. Since the appellant failed on those grounds, the imputation challenge is not strictly necessary to decide. Nonetheless, I will address it briefly: Kuru v State of New South Wales (2008) 236 CLR 1; [2008] HCA 26 at [12] and Boensch v Pascoe (2019) 268 CLR 593; [2019] HCA 49 at [8].

3. [94]

   The appellant challenged the primary judge’s finding that imputations (b), (d), (h) and (j) were not conveyed:

   1. (1)

      Imputation (b): the appellant “is a thief”. The appellant submitted that, from the matter’s allegation of theft, the reasonable reader would conclude the appellant “was a thief on this particular occasion”, and that if a person is caught stealing, they are “justifiably called a thief”.

   2. (2)

      Imputation (d): the appellant “was a racist”. The “extremely emotive and highly suggestive” words “range of vile and racist profanity” would, the appellant argued, caused the reasonable reader to view Ms Eppinga as a racist.

   3. (3)

      Imputation (h): the appellant “was an abuser of drugs”. The appellant submitted that the matter had to be considered as a whole, particularly the allegation that only after 20 minutes the appellant was caught “red-handed” going through drawers and attempting to steal drugs from a veterinary practice. “A picture emerged”, the appellant said, of a person “desperate to get her hands on drugs” to feed a drug habit.

   4. (4)

      Imputation (j): the appellant “was the subject of a police prosecution”. The appellant focused on the words “attempting to steal drugs from a veterinary practice is a serious criminal offence and so we were obligated to call the police. If anyone is curious about the incident, I urge them to call Dee Why police and ask”. Since Ms Eppinga was referred to the police on suspicion of a serious offence, the appellant said, the ordinary reasonable reader would infer that a police prosecution would inevitably follow.

4. [95]

   There are well-known principles for assessing whether imputations are conveyed. The central construct is the “ordinary reasonable reader”, in whose eyes the meaning of published matter is judged: Reader's Digest Services Pty Ltd v Lamb (1982) 150 CLR 500 at 506. This reader has certain traits: they are of fair, average intelligence, are neither perverse, nor suspicious of mind or avid for scandal: Lewis v Daily Telegraph Ltd [1964] AC 234 at 260. They do not live in an ivory tower and they can read between the lines in light of their general experience of worldly affairs. Applying these principles, none of the appellant’s challenges to the imputations succeeds:

   1. (1)

      Imputation (b): The allegation was of a single act of attempted theft in very particular circumstances. The primary judge was correct to conclude that the ordinary reasonable person would not read the matter complained of as carrying an allegation that the appellant was “a thief”.

   2. (2)

      Imputation (d): The statement the appellant “began using a range of vile and racist profanity” would not convey to the ordinary reasonable person that Ms Eppinga was a racist. The ordinary reader would understand that a person in a heated dispute like Ms Eppinga might use an intemperate slur without necessarily being “a racist”.

   3. (3)

      Imputation (h): The ordinary reasonable person would not conclude from the statement that the appellant was “attempting to steal drugs”, that in context, she “was an abuser of drugs”.

   4. (4)

      Imputation (j): The statement that the Vet Hospital had been “obligated to call the police” was followed by the remark “we urge them to call Dee Why police and ask”. The ordinary reasonable reader would understand that an investigation is distinct from a prosecution. The imputation that the appellant “was the subject of a police prosecution” was not conveyed.

5. [96]

   Accordingly, the primary judge was correct that none of the four additional imputations was carried.

Issue 4: Damages – extent of publication

1. [97]

   Again, this issue is strictly unnecessary to address. Nevertheless, for the same reasons as above, I will do so briefly. At trial, the appellant particularised two online readers who had accessed the matter complained of. This proved publication of the matter complained of to at least one person. The primary judge cited Sims v Joost [2016] WASCA 83 for the proposition that, with online media, there is no inference of publication as there is with traditional mass media. The primary judge then held, at [27], “that the extremely limited extent of publication was relevant to damages”.

2. [98]

   The appellant challenged this approach, especially the primary judge’s interpretation of Sims v Joost, in the following terms:

3. [99]

   This challenge is without merit. A plaintiff might rely on different forms of evidence to prove on the one hand publication as an element of the tort and, on the other hand, the extent of publication as a factor going to damages. However, for online publications, proof is still required, for the extent of publication as well as the element of publication. No inference is to be drawn that a matter was widely disseminated merely because a plaintiff clears the threshold of proving it was published to at least one other person. In this case, the only evidence the appellant led to prove the extent of publication was the two particularised readers. In closing submissions, there was no dispute between the parties that that was the extent of proven publication. In the absence of other evidence, the primary judge was entitled to find that the scope of publication was very limited.

4. [100]

   For that reason, the appellant was wrong to characterise the primary judge’s approach as a finding that harm to Ms Eppinga’s reputation was not presumed. At common law, harm to reputation ensues as soon as a defamatory matter is published, however few third parties it was published to: Bristow v Adams [2012] NSWCA 166 at [20]-[31] per Basten JA (see now, however, the serious harm threshold in s 10A of the Defamation Act 2005 (NSW)). But the extent of that harm is not presumed. A matter published to only two people will very often cause less reputational harm than the same matter published more widely.

5. [101]

   The primary judge found that Ms Eppinga’s reputation was harmed, but that the narrow scope of publication limited the damages she would have recovered, had the qualified privilege defence failed.

6. [102]

   I therefore reject this challenge to the primary judge’s reasoning. In line with that reasoning and the very limited scope of publication, it would have been appropriate to award damages of only a small amount, much less than $100,000, had the qualified privilege defence failed.

Conclusion and orders

1. [103]

   For the foregoing reasons, the primary judge was correct to conclude that the appellant did not establish malice. The respondents were therefore entitled to rely on a defence of common law qualified privilege, which was a complete answer to the appellant’s cross-claim. Accordingly, the primary judge was correct to dismiss the appellant’s cross-claim with costs.

2. [104]

   I propose the following orders:

   1. (1)

      Leave to appeal granted;

   2. (2)

      Appeal dismissed;

   3. (3)

      Appellant to pay the respondents’ costs of the application for leave and the appeal.

3. [105]

   KIRK JA: I agree with Payne JA.

4. [106]

   STERN JA: I agree with Payne JA.