Sydney Law Review
ANDREW T KENYON
Parties in civil defamation disputes very often disagree about what the publication in question means. While the differences may appear minor — for example, does the publication convey that the plaintiff is guilty of some discreditable action or merely suspected of such guilt — the way in which these differences are handled in law and litigation practice has great importance. Dealing well with the issue of meaning is central to litigation practices that are fair to both parties, respectful of limited court resources and responsive to the public interest in efficient and effective defamation litigation. This article examines three ways in which parties may differ about defamatory meaning, explains the contradictory labels that have been used for such differences in Australia, and considers how the Australian law on this issue has been modified in recent years. It then sets out two main contentions relevant to the uniform defamation statutes which now overlay the Australian common law on defamation. First, contrary to brief statements by the NSW Court of Appeal, the Australian common law does allow defences for truth in relation to some types of meaning that differ from those pleaded by plaintiffs. These defences are now available nationwide under the uniform law. They usually arise in relation to the defence of truth, but some decisions have also applied them to publications of opinion. Second, the common law approach to this issue, which is seen particularly in recent decisions from Victoria, South Australia and Western Australia, should be reconsidered in light of evidence from litigation practice in Australia and England. The most equitable and effective approach to defences of truth and opinion remains focused on the substance of the publication, rather than being caught in the intricacies of pleaded imputations, intricacies which have troubled defamation law in Australia for too long.
The enactment of largely uniform defamation statutes in all Australian jurisdictions, which took effect in 2006, followed decades of attempts at national reform. As disputes under the new law begin to appear before the courts, numerous issues will demand analysis. Many will involve questions of how, if at all, courts will draw on the former law and precedents of what were the Australian common law defamation jurisdictions, the code law defamation states, and New South Wales with its statutory-based action. The matter is not simply resolved by the uniform Defamation Acts providing that the general law of defamation applies as if the former defamation legislation had never been enacted.
This article examines one such issue that deserves careful analysis. It involves technical aspects of how defamation law deals with the meaning that is conveyed by a publication and the defences for truth and opinion. It concerns ‘Polly Peck’, to use the term that has been common in Australia. This article suggests in Part 2, however, that it would be useful to distinguish different aspects of the issue and label them distinctly as ‘Lucas-Box different meanings’ and ‘Polly Peck common sting meanings’. Research into litigation practice suggests this issue about meaning is of great practical importance. Dealing with it well is likely to foster defamation litigation that is more equitable, efficient and expeditious. Failing to deal with it appropriately is likely to increase technical arguments about the meanings conveyed by publications, shift disputes away from the substance of publications, and prompt more elaborate and lengthy analysis in court. That would not assist plaintiffs, defendants, or the public interest in civil litigation. Such an approach would continue some of the least successful aspects of NSW defamation law and practice under the Defamation Act 1974 (NSW), most notably the dangers of ‘diverting attention from the publication to the lawyer’s pleading of the alleged imputations’. The uniform Defamation Acts offer an important opportunity to improve, Australia-wide, the way in which defamation law treats the meaning that a publication conveys and the defences of truth and opinion. The issue is significant because of the importance of these defences in practice.
Defamation law, in outline, appears simple. Plaintiffs need only prove three things: material was published; the material identified them; and it conveyed a defamatory meaning. Damages have long been the usual remedy in civil defamation, and are presumed once the plaintiff’s case has been established. In many instances, the elements of publication and identification are met easily. ‘Publication’ means the material was made available by the defendant and received by a third person. ‘Identification’ need be neither express nor intentional. Rather, this element requires that recipients, including those who know the plaintiff, would think the plaintiff is being referred to by the publication. ‘Defamatory meaning’ is the element most often at issue in the plaintiff’s case: just what meaning does a publication convey? The test is framed in terms of the ordinary recipient, who reads between the lines of publications, interprets them in light of general knowledge, and is not ‘avid for scandal’. ‘Reasonableness’ is an apt shorthand description of the test. One result of the test for meaning is that a publication cannot convey inconsistent or contradictory meanings at once to the ordinary recipient, even if actual recipients are likely to have interpreted the publication in varied ways.
Note what defamation plaintiffs need not establish. Publications need not be proved false, nor need it be shown that defendants were at any fault in publishing. Plaintiffs in defamation face lesser hurdles than in many other civil actions. A strong contrast exists with malicious or injurious falsehood in which plaintiffs must prove that a false publication was made with malice which caused pecuniary loss. The easier position facing defamation plaintiffs helps to explain why defences are central in practice. Defences are the most ‘fascinating’, ‘important’ and ‘contentious’ area in defamation law.
Some defences relate to the defamatory meaning that is conveyed by a publication — for example, that meaning might be defended by being proved true. Others relate to the occasion on which the material was published under various forms of absolute, qualified and fair report privileges. These two approaches continue under the uniform Defamation Acts, with statutory defences existing in addition to general law defences. In relation to truth, for example, defendants could raise the common law defence of justification as well as its statutory version, and a statutory defence of contextual truth. While judges determine damages under the uniform law, where juries are used they decide all factual aspects of plaintiffs’ and defendants’ cases. In relation to those matters, judges play a gate-keeping role; for example, ruling out imputations that are incapable of being conveyed by the publication in question.
This article concentrates on defences related to meaning. Elsewhere, I have reviewed at length Australian case law to mid-2005 and explained how decisions such as David Syme v Hore-Lacy in Victoria and Nationwide News v Moodie in Western Australia have narrowed the availability of Lucas-Box different meanings and Polly Peck common sting meanings. Since then, NSW judges have stated that the defences are not part of the common law of Australia, notably in John Fairfax Publications v Zunter. In light of that history, the article sets out two arguments.
First, Australian common law does allow justification and opinion defences for some meanings that differ from the plaintiff’s pleaded imputations. The comments by the NSW Court of Appeal in Zunter appear to be a shorthand description for how other courts have modified the defences rather than removed them from Australian common law. The comments were made in a case under the Defamation Act 1974 (NSW), in which the cause of action was the pleaded imputation itself, rather than the publication. In addition, there was a split-trial process in which the jury only determined issues of publication, identification and whether the plaintiff’s pleaded imputations were conveyed and were defamatory. The judge dealt with all questions of defence and remedy. Neither the basis of the cause of action, nor the split trial process, left any room in practice to argue about the common law style of defence meanings. It is worth emphasising that the imputation-based cause of action became associated with great complexity in defamation practice compared with the common law. As I noted in 2004, when arguing for a nationally uniform cause of action in defamation to be based on the publication:
[T]he imputation-based cause of action in NSW has encouraged interlocutory fights about the form of pleaded imputations, as well as encouraging imputations to be considered separately, rather than in the context of the publications in question.
Under the Victorian common law approach — which, like the uniform Defamation Acts, based the cause of action on the publication — such interlocutory fights happened far less frequently and disputes remained more focused on the publications. The issues addressed in Zunter deserve fuller analysis under the uniform Defamation Acts, which remove these peculiarities of the former NSW law and should promote better practice.
The second argument advanced in this article is that the modified common law approach of cases like Hore-Lacy should itself be reconsidered. The evidence from litigation practice in Australia and England is that defence pleading of meaning promotes fair and efficient resolution of defamation disputes. Analysis of litigation strongly suggests that if Australian lawyers and judges do not use Lucas-Box meanings and do not use them well, Australian defamation practice will be mired in unhelpful technical battles removed from the substance of the publications in question. To that end, it is notable that many judicial comments in Hore-Lacy and similar decisions — including appellate judgments — express some reservation about how tightly defamation litigation should be constrained by the plaintiff’s pleaded imputations and how issues of defence meaning should best be analysed. This article seeks to clarify the role of a publication’s meaning in defamation law and practice, and it sets out a series of questions that should be asked in assessing relevant plaintiff and defence arguments.
After this introduction, Part 2 distinguishes three types of disagreement that might exist between plaintiff and defence cases and illustrates them through a hypothetical situation. Part 3 then addresses the first contention of this article — defence meanings of a particular form are available under Australian common law — through analysing recent NSW case law and decisions from other Australian states, such as Hore-Lacy. The analysis also demonstrates the significance, in the prevailing common law test, of defence meanings coming within the concept of ‘not substantially different from and not more injurious than’ the plaintiff’s imputation.
Part 4 sets out the article’s second contention: Hore-Lacy and related decisions deserve reconsideration. Litigation practices under differing approaches to defence meanings underline that the current Australian law does not take the best path for promoting effective and equitable litigation. Comparative litigation practices show that defamation is different from some other civil litigation, and its ‘special character’ justifies a particular approach to pleading. With regard to defence meanings, the approach that has been taken in England appears by far the best. It should be adopted in Australia. As Hedigan J noted soon after the Chakravarti decision, ‘the practices of past times must yield to the necessities of prompt and frugal disposition of litigation.’ Perfecting Polly Peck will further that aim. The evidence presented here offers powerful reasons to reconsider the approach to defence meanings in Australia — reasons which have not yet been addressed within appellate judgments, and reasons which may have been underappreciated by defamation litigators who have their greatest experience in one or more Australian jurisdictions or in England.
Parts 3 and 4 concentrate on defences for truth, which is when Lucas-Box different meanings and Polly Peck common sting meanings most commonly arise. Some English cases and existing Australian common law decisions also apply Lucas-Box and Polly Peck to opinion. However, other arguments can be made about defences for opinion. This is considered in Part 5, both in relation to common law fair comment and the statutory defence for honest opinion under the uniform Defamation Acts.
Parties may disagree about the meaning conveyed by a publication in at least three ways. These can be analysed in terms of meanings that differ, meanings that convey a common sting, and contextual meanings. Here, these are referred to using the terms ‘Lucas-Box’, ‘Polly Peck’ and ‘contextual truth’ respectively.
In the first type of disagreement, a defendant might argue that the publication does not convey the imputation pleaded by the plaintiff, but seek to justify a different meaning that is capable of arising from the publication. A classic example is where parties disagree about the level or degree of seriousness involved: for instance, does the publication convey that the plaintiff has committed some discreditable act or is reasonably suspected of committing that act? For the purposes of defamation law, the ordinary recipient is taken to understand such publications as meaning one or the other, but not both guilt and suspicion. They are alternative meanings. If the plaintiff pleads a meaning at the level of guilt, can the defendant prove the truth of the meaning at the level of suspicion? This difference can be described as a Lucas-Box defence, after the 1980s English Court of Appeal decision in Lucas-Box v News Group Newspapers. However, in England Lucas-Box meanings are not limited to a difference of degree such as that between guilt and suspicion. They encompass any meaning that is capable of arising from the publication and does not concern a separate and distinct allegation of which the plaintiff does not complain. In England, the ‘scope of the plea of justification does not depend upon the meanings of the words ... pleaded by the claimant but upon the meanings which the words are reasonably capable of bearing’.
In Australia, this type of defence meaning has been identified as the first Polly Peck proposition, but also as the second Polly Peck proposition, or merely as a Polly Peck defence. Using the label Lucas-Box follows common English terminology and clearly differentiates this type of disagreement between the parties about the publication’s meaning from the second type, examined below. One could substitute ‘first Polly Peck proposition’ wherever the term Lucas-Box is used in this article, which would follow recent NSW terminology. However, given the varied and contradictory labels that have been used in Australia to refer to this type of defence meaning, it appears better to separate clearly two types of defence meanings — the different meaning outlined here and the common sting meaning examined next — by using the label Lucas-Box for the first situation.
In the second type of disagreement, a defendant might argue that the multiple imputations conveyed by a publication are not distinct, but have a common sting, and that proving the truth of one aspect of the common sting will justify the entire publication. For example, a publication alleging separate acts of adultery may convey a common sting of adulterous promiscuity. In Khashoggi v IPC Magazines, a proposed publication arguably conveyed this common sting. If one allegation of adultery could not be proved true but the common sting could be justified, the publication could be defended even if the plaintiff only complained about the one allegation that could not be proved true. This difference between the parties — where the defence argues common sting in response to the plaintiff’s particular imputation — can be described as a Polly Peck common sting defence, after the 1980s English Court of Appeal decision in Polly Peck (Holdings) v Trelford.
In Australia, this type of defence meaning has been referred to as the second Polly Peck proposition, but also as the first Polly Peck proposition, or merely as the Polly Peck defence. This variation in labels risks confusing Lucas-Box and Polly Peck defences in an unhelpful manner. That is significant because Lucas-Box different meanings, in particular, are very common in England and have an important role in promoting good litigation practice.
It is notable that the above two types of defence meaning could only arise for argument if plaintiffs are not held precisely to their pleaded imputations. That question — of how closely plaintiffs are held to their imputations — has been a key issue in cases that have modified the Australian approach to these defences. This issue, however, is left to one side here; it can be addressed more easily after litigation practice is examined below in Part 4.
In the third type of disagreement, a defendant might argue that a publication conveys one or more imputations in addition to and at the same time as the imputation complained of by the plaintiff and, by reason of the truth of those contextual imputations, the plaintiff’s reputation is not further harmed by the plaintiff’s imputation. For instance, a publication may state that the plaintiff is a blackmailer and overstayed a visa in Australia. If the plaintiff sues over the visa-related statement, the defendant might seek to raise the contextual imputation, about blackmail, the truth of which could mean the plaintiff’s reputation is not harmed by the visa-related statement. This can be described as a contextual truth defence. It existed in NSW for several decades under the 1974 Act, and now exists throughout Australia under the uniform Defamation Acts.
The following situation may help clarify the three types of disagreement about meaning and illustrate how a publication can raise separate and distinct allegations. A media publication reports that the tax office is investigating a business person over suspected non-filing of tax returns. The publication also suggests the business person had some involvement in the mysterious death of another named individual. Disagreements could well exist about just what meaning was conveyed in relation to each allegation: had the business person actually not filed a tax return, was this merely suspected, or just worthy of investigation? Was the business person guilty of illegal conduct in relation to the death and, if so, of what sort? Was the death murder; was it accidental? Was the business person the primary perpetrator or an accessory, and so forth?
First, if the business person sued the publisher only about alleged involvement in the death, a Lucas-Box different meaning might be raised by the defence. For example, the plaintiff might plead a direct culpability for murder as the meaning of the publication, and the defendant might argue the truth of a lesser involvement in the death. Such a difference of degree is commonplace in Lucas-Box defences in England. Judgments there have emphasised that publications may convey three different levels of defamatory meaning: guilt, reasonable grounds to suspect guilt, or grounds to investigate. Parties frequently differ about the level of meaning conveyed by a publication. However, English Lucas-Box different meanings are not limited to that sort of difference of degree. The key questions are always whether each party’s meaning is capable of being the meaning of the allegation, whether the defence meaning concerns a separate and distinct allegation, and whether there are particulars of fact that could establish the truth of the Lucas-Box meaning if supported by evidence. The judge determines such matters. Then capable meanings, which do not concern separate and distinct allegations and which might be proved true by evidence, go forward to the finder of fact. (The position in Australia is examined below in Parts 3 and 4.)
Second, if the business person sued over both allegations, again Lucas-Box different meanings might be raised in relation to each one. However, it would seem extremely unlikely that the allegations about death and taxes conveyed one common sting. In almost all instances, a Polly Peck common sting defence would not be arguable for this publication. This conclusion would be reached by the judge asking whether the plaintiff’s meanings could be subsumed in one common sting meaning, which is capable of being the meaning of the allegation, and whether appropriate particulars of fact have been pleaded. If so, the Polly Peck argument would be left to the fact finder.
Third, if the business person sued only about the tax aspects of the publication, a contextual truth defence might be raised. If there was evidence to establish the plaintiff’s involvement in the other person’s death — and the involvement was of a serious kind — that could well amount to a contextual truth defence. An imputation about the death would be carried by the publication in addition to the plaintiff’s complaint about taxes, and the facts, matters and circumstances establishing the truth of the imputation about death would mean that the issues about tax did not further harm the plaintiff’s reputation.
In relation to Lucas-Box different meanings and Polly Peck common sting meanings, a defence cannot raise a separate and distinct allegation in response to the plaintiff’s imputations. If the plaintiff only set out imputations related to the tax return issue, for example, there could be no defence — except for contextual truth — based on the death because they would be separate and distinct allegations. It is true that the question of whether something is separate and distinct can be related to the level of abstraction adopted in analysing a publication. It is a question of degree, but one that remains focused on the publication in context, and is a task inherent in determining meaning. Evidence from practice shows it is useful to bear in mind a distinction between conveyed imputations, and the allegations, charges or statements that a publication contains. Pleaded imputations distil the meaning of an allegation, charge or statement, and the plaintiff determines which allegations in a publication are complained about. But the plaintiff cannot — under traditional common law — confine the meanings that can be argued to arise from the allegation or charge:
[A]lthough the defendant is not entitled to select for justification a separate and distinct charge which the plaintiff has not pleaded, the defendant can, in respect of a distinct charge which the plaintiff has pleaded, justify that charge in a different meaning.
This is clearly the position in England — and evidence from practice shows that it works very well for the parties, the courts and the public interest in effective litigation. As is explored below, it should also be the position in Australia.
In addition, when considering whether Polly Peck common sting meanings — rather than Lucas-Box different meanings — arise, the English case of Haslam v Times Newspapers may offer useful guidance. In Haslam, Gray J suggests considering the sectors of the plaintiff’s reputation on which the imputations bear, whether there is a connection between the two sides’ imputations, whether they have common features, and whether the common sting meaning is probative of the plaintiff’s complaint. The questions will remain ones for judgment, but they are far preferable for the operation of defamation law to questions focused on the specifics of pleaded imputations.
Two points warrant further attention in light of recent NSW case law, which was decided under the former Defamation Act 1974 (NSW). First, Australian common law decisions have not ruled out the availability of Lucas-Box different meanings or Polly Peck common sting meanings, contrary to a brief statement by the NSW Court of Appeal in 2006. Second, if decisions like Hore-Lacy are followed and defendants are able to raise different meanings or common sting meanings only where they are not substantially different from and not more injurious than the plaintiff’s imputations, courts need to consider carefully what comes within the concept of ‘not substantially different’. As is explained below, the concept of ‘not substantially different’ is broader under the Hore-Lacy test than under the former Defamation Act 1974 (NSW).
In Zunter, the NSW Court of Appeal stated that the common law of Australia does not recognise what in this article are called Lucas-Box different meanings and Polly Peck common sting meanings. Here, after outlining the case, the statement is compared with the leading Australian common law decision that was considered by the NSW Court of Appeal.
Zunter concerned a newspaper report about a bushfire and the actions of John Zunter who managed a caravan park on the south coast of NSW. When the bushfire threatened the caravan park, Zunter lit a series of smaller fires to consume the foliage and other fuel that would have fed the approaching bushfire. This backburn helped protect the caravan park from the bushfire. However, the Sydney Morning Herald published an article entitled ‘Illegal backburn that went wrong ruined our strategy, fire fighters say’. The article suggested Zunter lost control of the backburn, which fire fighters were unable to bring back under control, and he would be questioned by police about the incident. The jury found two defamatory imputations were conveyed by the article: Zunter ‘lost control of his own backburn’ and ‘wrecked the main strategy of the Shoalhaven Fire Control Officer’.
Fairfax pleaded two contextual imputations: Zunter ‘carried out an illegal backburn’ and ‘carried out an illegal backburn in circumstances of extreme fire danger’. At trial, the plaintiff conceded that the first contextual imputation was conveyed and was substantially true and Simpson J accepted that the second was also conveyed and substantially true. However, in accordance with John Fairfax Publications v Blake, she rejected the contextual truth defence after balancing the seriousness of each of the plaintiff’s imputations against the seriousness ‘of the facts, matters and circumstances that establish the truth of the contextual imputations’. While those facts, matters and circumstances were serious, Simpson J held that Zunter’s reputation was further harmed by the imputations which he had pleaded and the jury had found to be conveyed. The Court of Appeal agreed. Thus, the contextual truth defence was not available for this publication.
Fairfax also pleaded, in what it called a Polly Peck defence, that the plaintiff ‘carried out an illegal backburn in circumstances of extreme fire danger’. This appears to have been offered as a common sting meaning. Simpson J dismissed this defence without detailed consideration. She stated that decisions such as Hore-Lacy ‘make it very difficult for Fairfax to succeed in this defence. It could do so only if I declined ... to follow the decisions of appellate courts in other states of Australia.’ It is worth noting that, in this passage, Simpson J does not state that defences based on Lucas-Box different meanings or Polly Peck common sting meanings are not recognised in Australian common law, nor does she detail precisely why the defence would fail in this instance.
The issue also came before the Court of Appeal, which went further as to the effect of decisions like Hore-Lacy on Polly Peck common sting meanings:
[T]here is no need to consider this defence in detail because at the present time the common law of Australia ... does not recognise it. It was rejected in Chakravarti ... by Brennan CJ and McHugh J, in dicta which did not receive the express endorsement of the other members of the Court. However those dicta have been followed by intermediate appellate courts in Victoria ... Queensland ... Western Australia ... and South Australia ... This Court should follow this line of authority and it would be inappropriate for us to re-examine the question. The Judge rightly rejected this defence.
It is correct that the Queensland Court of Appeal held the defence was not recognised under the state’s former defamation code. However, that code had significant differences from the common law and based the cause of action on the plaintiff’s pleaded imputations. This meant that it more closely resembled the former NSW position than it did the common law. The former Queensland approach is not relevant to the common law and the uniform Defamation Acts. Decisions in other Australian states — such as Victoria, South Australia and Western Australia — do recognise a form of Lucas-Box different meaning and Polly Peck common sting meaning. The most influential decision on the issue in those states has been Hore-Lacy which modifies but does not remove these defences. Under the Hore-Lacy test, a defendant can defend a meaning that is not substantially different from, and not more injurious than, the plaintiff’s imputation.
On the test set out in Hore-Lacy, a judge would ask four questions of a pleaded defence meaning. (All except the third question also arise under the English common law approach to defence meanings.) First, is the defence meaning capable of arising from the publication? If the meaning is incapable of arising because, for example, only an unreasonable recipient who was ‘avid for scandal’ would think it was the publication’s meaning, then the defence meaning cannot be argued. Second, does the defence meaning arise from a separate and distinct allegation in the publication, about which the plaintiff does not complain? If it does, the defence meaning cannot be argued. For instance, an allegation about not paying taxes cannot be answered under these defences by proving the truth of a separate and distinct allegation concerning murder. Third, is the defence meaning not substantially different from and not more injurious than the plaintiff’s imputation? This is the question added by Hore-Lacy to the assessment of defence meanings, and just what ‘not substantially different’ encompasses is considered further below. Fourth, could proposed evidence establish the truth of the defence meaning or, more accurately, are proper particulars of fact provided capable of supporting the defence? If there are no particulars capable of going to the truth of the Lucas-Box different meaning or Polly Peck common sting meaning, the meaning should be removed from the case. Alternatively, if a great deal of complex or unwieldy evidence appears to be available to prove the truth of the imputation, managing that complexity should be faced directly. A potentially unruly body of evidence is not a reason to deny the availability of a particular imputation, but the scope of evidence that may properly be admitted in any case is an important matter for case management.
The comment by the Court of Appeal in Zunter may have been intended to serve simply as a statement that the defence was not arguable on the basis of the publication in question. In any event, it is difficult to see how a Polly Peck common sting or a Lucas-Box different meaning could have been raised in the split-trial process under s 7A of the Defamation Act 1974 (NSW) which was used in Zunter. In effect, the judge at the trial of defences would have to overrule the earlier jury determination of what defamatory imputations actually were conveyed by the publication. It could have been better for the Court of Appeal here simply to note that under the Defamation Act 1974 (NSW) the defence was bound to the jury finding on meaning in relation to all states and territories. Stating that Australian law does not recognise Lucas-Box different meanings or Polly Peck common sting meanings might be true for the law of former code states — such as Queensland — but it is only a shorthand approximation of the law that existed in the common law states before the uniform Defamation Acts, and it is an inaccurate statement of the law in all Australian jurisdictions under the uniform statutes.
Under the former NSW law, plaintiffs were held to their pleaded imputations and could only seek verdicts on imputations ‘not substantially different’ from those pleaded. This followed from the cause of action being the pleaded imputation itself. In Woodham v John Fairfax Publications, this ‘not substantially different’ requirement of the former NSW law was interpreted as being equivalent to the ‘not substantially different and not more injurious’ meaning discussed in Hore-Lacy and other cases. Given the obvious similarities in the wording of the tests this is hardly surprising, but it deserves further consideration because a narrower range of imputations were included within the former NSW law than have been allowed under the Hore-Lacy test.
Two quotations from the judgment of Nicholas J make clear the interpretation in Woodham:
Necessarily, the plaintiff’s imputation comprehends imputations which do not differ in substance, or are less injurious, or which are but shades, nuances, and gradations of meaning of substantially similar imputations.
I am not persuaded ... that a different approach is to be taken in deciding the question of ‘substantially different from’ for the purpose of a Polly Peck defence to the approach taken in New South Wales in respect of the plaintiff’s imputations. The verbiage used in, eg, Chakravarti ... including “nuance”, “less serious”, “variant”, “shades”, referable to a Polly Peck imputation is identical to that used in, eg, Morosi ... referable to the plaintiff’s imputations.
However, there is an important difference between how the former NSW law and the Australian common law after Hore-Lacy treat the concept of ‘not substantially different’. Imputations of guilt and suspicion were treated as substantially different under the 1974 Act but, on the better view, they are not substantially different for the purposes of defence pleas under Hore-Lacy. This arises from the example given by Ormiston JA in Hore-Lacy as well as from the facts of Chakravarti itself, both of which are examined below. The approach could be contrasted with some later examples in which meanings at the level of guilt and suspicion have been treated as substantially different under the common law. Elsewhere, I have analysed why such examples should not be followed: if lesser imputations are to be ruled out in a given case, it should be on the basis of the meaning being incapable of arising from the publication in question, concerning a separate and distinct allegation, or not being supported by any pleaded particulars.
Under the Defamation Act 1974 (NSW), each pleaded imputation was seen to include all meanings that did not differ in substance. This included imputations of slightly lesser seriousness. For example, the plaintiff could not have an imputation of promiscuity removed from the jury by arguing the publication only ‘bore imputations of sexual immorality of a slightly lesser degree’. The imputation encompassed both promiscuity and ‘sexual immorality of a slightly lesser degree’. However, the allowable degree of difference did not extend to matters of guilt and suspicion. Suspicion was treated as a substantially lesser imputation than guilt, meaning the plaintiff had to plead both imputations if both levels of seriousness were to be open to the jury.
The point was explained by Hunt J in Hepburn v TCN Channel Nine:
The jury may, of course, be invited to find an imputation which has not been pleaded if that imputation does not differ in substance from an imputation which has been pleaded ... But .... [t]he whole course of authority in relation to [the NSW court rules for defamation] has made it clear that different gradations of seriousness must be pleaded in separate imputations.
By the phrase ‘different gradations of seriousness’, Hunt J meant a difference such as that between guilt and suspicion. For example, he commented in reference to contextual imputations:
A defendant must obviously be entitled to plead contextual imputations ... which are but different gradations of seriousness of the same basic assertion (for example — the plaintiff is guilty of a crime, the plaintiff has been convicted of that crime, and the plaintiff has so conducted himself as to cause the police to suspect that he was guilty of that crime, etc). The defendant’s entitlement to do so rests upon the same basis as the plaintiff’s entitlement to do so.
Thus, under the former NSW law, guilt and suspicion were treated as substantially different. But this is not the position at common law in Australia, even after Hore-Lacy, and it is not the position under the uniform Defamation Acts. In Hore-Lacy, Ormiston JA said:
[M]any articles in the press ... are devised on the “no smoke without fire” premise, so that many allegations take a form which might be construed ... as alleging highly improper activity though on detailed analysis ... the allegation would appear less serious. It is this sort of case which might go to the jury with the plaintiff pleading imputations of high impropriety and the defendant asserting ... less serious peccadillos which it wished to justify. The “smoke” could therefore be justified but it would remain for the jury ... to decide whether the imputation was still one of “fire”.
The words ‘smoke’ and ‘fire’ echo Lord Devlin’s comments in Lewis v Daily Telegraph, which concern the difference between suspicion and guilt:
It is not ... correct to say as a matter of law that a statement of suspicion imputes guilt. It can be said as a matter of practice that it very often does so ... . [Publishers who want] to talk at large about smoke may have to pick [their] words very carefully if [they want] to exclude the suggestion that there is also a fire.
As Ormiston JA explained, guilt and suspicion should not be treated as substantially different meanings for the Hore-Lacy form of Lucas-Box meanings; a defendant could choose to prove the truth of ‘smoke’, but it would be for the jury to decide if the meaning conveyed was actually one of ‘fire’.
It is also notable that a similar point about guilt or suspicion was made in Chakravarti itself. Between them, the parties raised questions of guilt and suspicion of financial misconduct. This meant the defendant could not be prejudiced by the plaintiff relying on a non-pleaded meaning of suspicion, when the plaintiff had pleaded at the level of guilt. The meanings of guilt and suspicion, although they have differences, were not treated as substantially different meanings for the purposes of common law defamation.
This type of Lucas-Box different meaning is also clearly available in England. It may be ruled out in a particular case if it is incapable of arising from the publication in question, but not because of how close or distant it is from the plaintiff’s pleaded imputation. The plaintiff’s imputation is relevant, but only for determining whether the defence meaning concerns a separate and distinct allegation from that about which the plaintiff complains. The key referent in deciding whether the Lucas-Box plea is open is the publication, not the precise words of the plaintiff’s pleaded imputation. Under the uniform Defamation Acts, the same focus on the publication is open and warranted.
The approach to defence meanings that has developed through cases like Hore-Lacy, Nationwide News v Moodie and Advertiser-News Weekend Publishing v Manock might seem a plausible way to deal with the question of what meaning a publication conveys. It could be seen as ‘conducive to the fair conduct of a trial’. And it avoids concerns that defendants justifying different meanings could lead to ‘the introduction of evidence that will increase the length of the trial, may tend to cloud the issues, and may work to the unfair prejudice of the plaintiff’. Thus, one might expect the Hore-Lacy approach to work well: the plaintiff needs to make its case clear to the other side and to the court. Holding plaintiffs to pleaded imputations in general terms — that is, to imputations that are not substantially different from, and not more injurious than, the pleaded imputations — might be thought to further the aim of dealing effectively with the question of meaning. Its analytical appeal is clear. Surely it is appropriate to hold plaintiffs, in this general sense, to their cases?
That is correct, but only to a limited degree. It certainly needs to be clear whether some distinct allegations are not being complained about. For example, to return to the example from Part 2, is the plaintiff complaining about allegations concerning both death and taxes, or only one of them? In addition, the plaintiff will be constrained in general terms by their pleaded imputations in how they present the case at trial where that is necessary to avoid prejudice to the other side. But at common law, the finder of fact is not constrained by the plaintiff’s imputation; the fact finder determines the meaning conveyed by a publication. As the evidence outlined below makes abundantly clear, that flexibility for the finder of fact is very important in supporting better defamation practice.
The situation may appear strange and be thought likely to produce unwarranted complexity. However, evidence from comparative litigation practice demonstrates the reverse. It strongly supports such freedom for the finder of fact and emphasises that pleaded imputations should not be given any great power to define the question of meaning in defamation litigation. The meaning of publications is the central concern of defamation law. Because of the way the legal tests for defamatory meaning seek to mirror the interpretation of the ordinary audience members, defamation law is different from much civil litigation. Questions about meaning should remain focused on the publication, rather than the pleaded imputations. Australian judgments since the late 1990s have moved a long way from this approach, without explanation for why the significant difference from the English law and practice on this point is warranted. Evidence from practice shows it is not.
Elsewhere, I have detailed the findings of doctrinal analysis and empirical research about defamation law and practice in Australia and England. The research concerned litigation practice before the introduction of Australia’s uniform Defamation Acts, and largely predated the changes from decisions like Hore-Lacy. Interviewing more than 100 experienced defamation litigators and judges in London, Sydney and Melbourne over several years, and examining more than 130 court files from NSW and Victorian Supreme and intermediate courts, produced a detailed picture of litigation practices in defamation. Some of the central points from that comparison are outlined here. This summary is just a snapshot of that research, which engaged closely with the practices and understandings of experienced defamation lawyers with experience in one or another of the jurisdictions discussed.
In English practice each side tended to plead relatively few meanings, their specificity was rarely challenged pre-trial, Lucas-Box pleading was normal practice and was seen as a necessary, useful part of litigation. The difference between the parties about meaning was often substantial — for example, quite different evidence would be relevant under the defence plea. That difference was made clear to the other side well before trial through the pleaded defence. And the availability of the defence plea was not dependant on the precise wording of the plaintiff’s imputation. Interviewees showed absolutely no support for the concern raised in some Australian judgments that Lucas-Box different meanings must cause difficulties in practice. Instead, interviewees described Lucas-Box as ‘absolutely vital’ and obvious in its benefits. Polly Peck common sting meanings were also supported by English practitioners, but they arose far less frequently in practice. This is not surprising, as they require both a publication that conveys a common sting and a plaintiff that limits its claim to only one or some elements of that common sting.
Practice under the Defamation Act 1974 (NSW) was very different. In NSW, plaintiffs usually pleaded multiple meanings at varied degrees of seriousness, such as the levels of guilt or suspicion. Because plaintiffs were bound closely to their imputations, defendants had no ability to respond with Lucas-Box different meanings or Polly Peck common sting meanings. This accentuated the ability of plaintiff lawyers to craft imputations that would not be struck out for being incapable of arising from a publication, but were particularly difficult to defend. It was also notable from the research that defences related to truth and opinion appeared to have become tightly focused on plaintiffs’ imputations. This may be understandable given the status of the imputation as the cause of action under s 9 of the Defamation Act 1974 (NSW). But the defences appeared to be much less accessible than at common law. All this meant that defendants considered challenging plaintiffs’ pleaded imputations more often and pre-trial challenges were frequently made about the form and specificity of pleaded imputations. Attacking plaintiffs’ pleaded imputations in this way could be seen to operate as a major style of defence. While aspects of this position were accentuated by the split-trial process under s 7A of the 1974 Act, it was also clear that the attention paid to pleaded imputations — and the way in which plaintiffs were held to imputations — gave rise to much of it. The imputation-centred approach of the former NSW law did not appear to promote equitable, efficient or expeditious litigation focused on the substance of the dispute. The common law was more effective.
Under the common law in Victoria, plaintiffs commonly pleaded multiple imputations whose form was rarely challenged pre-trial, parties were not held particularly close to pleaded meanings, and Lucas-Box different meanings were frequently pleaded. Justification and opinion defences were far more commonly pleaded than the corresponding NSW defences, and it appeared that defendants in NSW may have been unreasonably limited in raising them because of the imputation-centred approach of the former NSW law. All these factors resulted in meaning largely being left until trial in Victoria, for the finder of fact to determine from the publication in context.
Figures from NSW and Victorian court files supported the interview findings. Changes to pleaded meanings occurred far more commonly in NSW than Victoria — more than five times as often. Defences related to opinion and truth, were much less frequent in NSW, being raised nearly twice as often in Victoria. And about one third fewer cases settled or were disposed of as quickly in NSW as in Victoria. It is also noteworthy that trials did not appear to run for any longer under the Victorian approach than in NSW. The practicality of the trial hearing — when it is eventually reached — has always been a key argument for the imputation-centred approach, but the fieldwork did not find support for that analysis. In addition, the interviews make clear that reaching trial promptly is a central aim for plaintiffs and the evidence suggests it was better supported by the common law.
What is suggested by this research for how Australian common law should deal with the issue of meaning in defamation? Here, one point appears particularly important: the Hore-Lacy ruling moves Australian common law an appreciable distance towards the former NSW imputation-centred approach, a shift that was implicit in the comments of Brennan CJ and McHugh J in Chakravarti itself. If Lucas-Box different meanings and Polly Peck common sting meanings are not available for defences as they are in England, weaknesses of the former NSW law and practice may recur Australia wide. The evidence from litigators and files in NSW and important comparative jurisdictions strongly suggests that the change to the common law in Hore-Lacy is unwarranted. It should be reconsidered. At the least, Australian common law should interpret ‘not substantially different’ in the manner discussed in Part 3 — something which is open to courts in all Australian jurisdictions under the uniform Defamation Acts and decisions like Hore-Lacy. But it would be better to reconsider those cases.
The better approach recognises that each party in defamation has an interest in shaping the area of dispute to their advantage. As Gillard AJA commented in Herald & Weekly Times v Popovic:
[A] plaintiff’s legal practitioner may through ignorance, a misunderstanding of the language, carelessness, or seeking to confine the plaintiff’s case within narrow limits or for some other reason, plead ... imputations which are inadequate or in some way do not properly or fully convey the true defamatory meaning of the words complained of. Common sense and justice demands that the defendant be permitted to plead the true imputation conveyed by the words complained of and in that meaning prove that they are true and correct.
Gillard AJA continued that ‘fairness and justice’ require defendants to alert plaintiffs to the imputations sought to be defended:
Under the modern practice the plaintiff pleads the imputations. It is necessary to do justice that the defendant plead the imputations, if they are different, which [the defendant] intends to justify. Of course the imputations must be confined to the words actually complained of by the plaintiff subject to context.
These statements closely track the best practice as regards defamatory meaning. Each party should set out the meaning being argued, each meaning must be capable of arising from the publication, and the meanings must not concern separate and distinct allegations, such as the example about death and taxes in Part 2 above. Defence meanings are not tested by their closeness to plaintiffs’ imputations but by the publication in context. And beyond ignorance, misunderstanding of language or carelessness, it is worth emphasising that both parties have tactical interests in ‘pushing’ meaning in particular directions: plaintiffs as much as defendants. The evidence from litigation practice in England and Australia shows that plaintiffs’ imputations will be shaped quite intentionally by their lawyers. And this will happen even more if plaintiffs are held closely to their imputations and defendants limited in how they can respond.
It is true that defendants may try to convert ‘a modest and narrow claim ... into a wide-ranging expansive and extensive inquiry’. However, equally, plaintiffs may aim to limit defendants’ ability to argue the substance of the publications: they may seek ‘to use a blue pencil upon words published’. The law should ensure that each party has an equivalent opportunity to present its case within the context of the publication in question. That is best done by allowing Lucas-Box different meanings and Polly Peck common sting meanings in the manner of the English law and practice. There is nothing in the uniform Defamation Acts preventing this approach.
Instead of the rule in Hore-Lacy and similar cases, the questions for evaluating any defence meaning should be three. First, is the defence meaning capable of arising from the publication? Second, does it concern a separate and distinct allegation from those complained of by the plaintiff? Third, are there proper particulars of justification — or by way of shorthand, is there evidence which could establish the defence meaning’s truth? For the defence meaning to be arguable, the judge must answer these questions: yes, no and yes. The extra question that results from Hore-Lacy — about defence meanings being not substantially different from, and not more injurious than, the plaintiff’s imputation — is not merely superfluous, but appears harmful to good defamation practice. While Hore-Lacy is an understandable attempt to respond to an important issue, it is an issue which has already been tackled by the second question above. That is, a defendant cannot respond to a plaintiff’s case by raising a separate and distinct allegation. Hore-Lacy is not needed: defence meanings will be dealt with best by remaining focused on the publication in context, not the plaintiff’s pleaded imputations in their particular and specific wording.
Some defences based on Lucas-Box different meanings or Polly Peck common sting meanings may need to be controlled in terms of their effects on trial length. However, that control should not be based on the plaintiff’s pleaded imputations. The control should consider the meanings that are capable of arising from the publication — and not arising from separate and distinct allegations — on which each party deserves a fair opportunity to present its case. And the control should be exercised on the basis of good case management practices, which was something considered by O’Connor LJ in Polly Peck itself.
It was noted above, in Part 2, that the whole question of defence meanings relates to the degree to which plaintiffs are held to pleaded imputations. One option could be for plaintiffs to be held precisely to their imputations, which would leave defendants with no room to argue about different meanings. As has been noted in some judgments, holding plaintiffs to their imputations would appear more consistent with traditions of civil pleading. It was also the position of plaintiffs under the Defamation Act 1974 (NSW) when the cause of action was the pleaded imputation rather than the publication. It was long a common argument that requiring NSW plaintiffs to plead specific and unambiguous imputations, to which they were held, fostered better litigation practice than the common law. With evidence from practice that view is seen to be incorrect. The evidence strongly suggests that parties should not be bound closely to the level of any pleaded imputations. At trial, each party may be constrained in presenting its case — if that is needed to avoid injustice to the other side — but the tribunal of fact should not be restricted in that way. It should focus on the publication and the allegations that it conveys which have been complained about. That will foster equitable litigation focused on the substance of defamatory publications.
The above analysis has concentrated on truth defences, but it is important to recognise that some English and Australian cases have applied Lucas-Box and Polly Peck to fair comment. The treatment of common law fair comment is due to be addressed by the Australian High Court in Channel Seven Adelaide v Manock. Before the background is discussed below, it is worth noting two outcomes that appear possible. One option is that the High Court, in agreement with the appellant publisher, determines that the common law defence of fair comment responds to the words published rather than the plaintiff’s imputation. Under that approach, the words published would be analysed and the defence would succeed if the words are comment rather than fact, the elements of public interest, truthful basis, and fairness are made out, and the plaintiff does not establish that the publisher was actuated by malice. In effect, the defendant would argue ‘that the way he or she chose to express the opinion’ is defensible, rather than the imputations as such. This would give common law fair comment a wide scope, in line with the long-recognised role for fair comment in protecting freedom of speech. Under this approach, Lucas-Box and Polly Peck would remain matters for truth defences in Australia and the analysis set out above in Parts 3 and 4 would not be significant for fair comment. The four questions under Hore-Lacy, or the three under a revised approach, would arise in relation to justification. Alternatively, however, if the High Court determines that the comment defence addresses — in some sense — the plaintiff’s imputation, there are important questions about just what would be involved.
Given this article’s focus on Lucas-Box, Polly Peck and contextual truth, the initial question raised in Manock — about whether common law fair comment addresses the words published or the plaintiff’s imputation — is left aside here. (Other issues in the appeal are also passed over here, such as when and how a factual basis for the comment needs to be indicated to the recipients of a publication.) Instead, this Part considers what approach Australian law should take if the fair comment defence addresses the meaning of a publication. After outlining the issues in Manock and considering that question, the statutory defence for honest opinion under the uniform Defamation Acts is also examined. The argument is advanced that if common law fair comment is not directed to the words published, it must be directed to the meaning that is actually found to be conveyed. Equivalent issues to those set out above in Parts 3 and 4 then arise. In addition, under the uniform Defamation Acts, a wider defence may be arguable, a defence of the sort advocated by the publisher in Manock. Whatever the scope of the statutory opinion defence, some of the difficulties that faced opinion under the former Defamation Act 1974 (NSW) need not arise under the uniform law. This is an important development, although uncertainties also clearly exist for the new statutory defence.
Manock v Channel Seven Adelaide concerned a brief television promotion for an upcoming current affairs broadcast. The publication, which predated the uniform Defamation Acts, contained these words: ‘The new Keogh facts. The evidence they kept to themselves. The data, dates and documents that don’t add up. The evidence changed from one court to the next.’ While those words were spoken, a picture of the plaintiff was displayed. The plaintiff, a forensic pathologist, pleaded that the publication meant he ‘had deliberately concealed evidence’ from the trial and retrial in which Keogh faced a charge of murder.
In the Full Court of the South Australian Supreme Court, a strike out application succeeded against defence particulars for a common law fair comment defence. The argument was framed in terms of whether the fair comment defence had to address the imputation pleaded by the plaintiff. In May 2007, the High Court granted leave to hear an appeal on this issue, with the appellant publisher foreshadowing an argument that fair comment ‘responds to the words used by the defendant’ — that is, the publication itself — rather than fair comment needing to ‘address the plaintiff’s imputation’. The publisher argued the words quoted above were fair comment and set out particulars of public interest related to ‘a suggested wrongful conviction for murder arising from questionable forensic investigations by and evidence from Dr Manock’. The particulars did not concern the deliberate concealment of evidence. In addition, lengthy particulars of fact were set out — they cover five pages in the authorised South Australian report — by way of factual basis for the defence. However, as noted in the Full Court:
None of the alleged facts particularised appear to be directed to or to address the imputation pleaded by Dr Manock — the deliberate concealment of evidence from murder trials.
Thus the ‘core’ issue on appeal was described as being whether the fair comment defence and particulars had to address the plaintiff’s imputation or the words published. As noted above, the appellant’s arguments for a broader common law defence may succeed. However, another view of fair comment would position it between the plaintiff’s imputation and the publication — that is, between the options foreshadowed by the parties in the Manock appeal. On that view, fair comment defends the defamatory comment that is conveyed by the publication. The defence does relate to meaning, but not solely to the plaintiff’s imputation. As the analysis in earlier Parts of this article should make clear, if fair comment does defend meaning, it defends the defamatory comment that is conveyed by the publication. It is not confined to the very imputation of which the plaintiff complains. The disutility of that approach is evident from experience in NSW under the Defamation Act 1974. However, the defendant is not free to defend any and all meanings that are conveyed by a publication as fair comment. Two limits are noteworthy here. Any meaning defended by fair comment must be comment rather than fact, and the defence cannot protect a separate and distinct allegation about which the plaintiff does not complain.
In simple terms, the plaintiff in a situation like Manock argues that the publication means ‘deliberate concealment’ of evidence, while the defence is directed towards ‘questionable’ forensic investigation, failing to meet professional standards, inconsistencies in providing expert evidence, and inadequate professional practice. The analysis in earlier Parts suggests the following three questions would be the best for a judge to use in such a situation, if it is determined that fair comment does defend meaning. Under the Hore-Lacy test, an extra question would be added. Of course, the defence could only succeed in relation to matters of comment, not fact. And there may be questions such as whether the facts have been appropriately stated in the publication, whether they were indicated in it or whether they were notorious.
First, is the defence meaning capable of arising from the publication? Can the short current affairs promotion in Manock convey to the ordinary recipient the meaning that the plaintiff’s forensic practice and expert evidence were inadequate? It seems possible that, in the context of the publication, these meanings are alternative to the plaintiff’s ‘deliberate concealment’ meaning. That is, the ordinary recipient would think the advertisement meant the plaintiff was guilty of deliberate concealment or guilty of questionable techniques and evidence, but the ordinary recipient would not understand these two things as arising at once from the publication. In the terms set out in earlier Parts, the defence would be raising a Lucas-Box meaning in its fair comment defence.
Second, does the defence meaning concern a separate and distinct allegation from that complained of by the plaintiff? That does not appear very likely — it is unlike the example in Part 2 of a publication that raises issues about death and taxes. However, this publication may be one in which the context of sound and images is decisive in determining whether separate and distinct allegations are being raised. If they do relate to separate and distinct allegations, one cannot be used to defend the other.
Third, are there adequate particulars capable of supporting the defence meaning? In relation to Manock, it appears likely that many of the particulars would be capable of doing that. However, before reaching this issue the judge would face another question under the Hore-Lacy approach: is the defence meaning not substantially different from, and not more injurious than, the plaintiff’s imputation?
The scope of ‘not substantially different’ has been examined above in Part 3. While differences of guilt and suspicion have generally been seen to come within the Hore-Lacy test, deliberate concealment and inadequate professional performance might be thought to be substantially different. For example, they would make different evidence relevant in justification. But attempts to confine the scope of ‘not substantially different’ by considering the evidence that would be relevant to each meaning are problematic. In Hore-Lacy, Charles JA said what was substantially different ‘would presumably be tested by asking whether the defendant would have been entitled to plead a different issue, adduce different evidence or conduct the case on a different basis ... or possibly whether the justification would be substantially different’. However, even proving guilt and suspicion — the simplest difference in meaning, which arises from Hore-Lacy itself — can involve quite different evidence. For example, if a publication about a person’s arrest, in context, conveys the suspicion that the person is guilty of a crime, the defendant must usually prove three things to justify the publication. First, the arrest and charge occurred. Second, the police had a suspicion of guilt, with reasonable cause. Third, the plaintiff’s conduct giving rise to the suspicion occurred. If the publication conveys a meaning at the level of guilt, much more must be proved. Unless the parties are confined to the very nuance of the plaintiff’s pleaded imputation, different evidence is likely to be relevant. Limiting parties to the plaintiff’s pleaded imputation would implement an imputation-centred approach like that of the former NSW law, which was not at all the aim of the uniform law. If a test like Hore-Lacy is to continue in Australian law, judgments need to explain more about what differences are excessive and why. The necessary distinction to be drawn in defamation may not be a substantial difference in imputations but whether the parties are arguing about separate and distinct allegations.
This again illustrates how the extra question that results from Hore-Lacy — about defence meanings needing to be not substantially different from and not more injurious than the plaintiff’s imputations — is not the best question. Focusing on the three questions set out above and in Part 4 will promote an approach to defences that keeps disputes focused on the substance of publications, in a way which is fairest to both parties and promotes efficient litigation practice.
If the meaning advanced by each party in a case like Manock is capable of being the meaning that is conveyed by the publication — which would also mean they do not concern separate and distinct allegations — then doctrinal analysis and evidence from litigation practice strongly supports each party being allowed to argue its case on its meaning. That will promote effective litigation, with fewer interlocutory skirmishes and consequently less delay and expense before a plaintiff is capable of receiving a remedy. It will also promote litigation directed to the substance of the dispute, not the technicality of pleaded imputations. However, if one of the meanings advanced by each party is incapable of being conveyed by the publication — or if the defence meaning concerns a separate and distinct allegation in the publication about which the plaintiff does not complain — then the meaning should be struck out.
It is also worth briefly examining the statutory honest opinion defence under the uniform Defamation Acts. The uniform law, which is a modification of the former NSW Act, provides this defence for honest opinion:
(1) It is a defence to the publication of defamatory matter if the defendant proves that:
(a) the matter was an expression of opinion of the defendant rather than a statement of fact, and
(b) the opinion related to a matter of public interest, and
(c) the opinion is based on proper material.
For material to be ‘proper’, it must be substantially true or published on an occasion of absolute, qualified or fair report privilege. Where only some of the material is proper — that is, substantially true or privileged — the defence can succeed if the opinion might reasonably be based on that proper material. The defence is defeated if the plaintiff proves the defendant did not honestly hold the opinion when it was published. Similar provisions protect the publication of opinion of the defendant’s employees, agents or third party commentators (such as writers of letters to the editor).
Even if the common law fair comment defence is directed towards the meaning conveyed by a publication, this statutory defence may have a wider ambit of the sort argued for by the publisher in Manock. That is, it might be directed to the publication itself. The wording of the honest opinion defence relates to published matter, not to the plaintiff’s pleaded imputations, unlike case law interpretation of the earlier NSW statutory defence. Section 32(1) of the former Defamation Act 1974 (NSW), for example, provided ‘a defence as to comment’ where ‘the comment is the comment of the defendant’. This was interpreted as limiting the defence to the very imputation pleaded by the plaintiff. If plaintiffs’ lawyers crafted factual imputations that were not incapable of arising from a publication — even where the publication could more easily be thought to convey matters of comment — the defendant had no possibility of using the comment defence. Under the former NSW law, it was ‘extraordinarily difficult to succeed on a defence of comment’. The evidence from practice underlined how very difficult the NSW defence was in operation. In interviews, experienced defamation lawyers, without being prompted, raised concerns about the comment defence under the 1974 Act. Comment was seen as being where the former NSW approach, with its imputation-based cause of action, ‘works the harshest’. The defence was seen as a ‘complete mess’ with defendants being ‘frightened’ of using it. Plaintiff lawyers pleaded imputations that prevented comment being raised in matters that were thought clearly to involve comment.
When the cause of action in defamation is the published matter, as it now is Australia-wide, there is no reason to tie the defence to the plaintiff’s pleaded imputations as such. The statutory wording under the uniform Defamation Acts should allow those problems from the former NSW law to be avoided. In relation to both the current common law and statutory defences for comment and opinion, it is worth repeating that to the degree that decisions like Anderson v Nationwide News apply the Hore-Lacy test to defences, they risk importing the same problems that were experienced in NSW. The danger to avoid is ‘the form of the imputation’ defeating ‘the substance of the defence’.
The above analysis might call to mind this observation of Nicholas J in the Woodham case:
One may well wonder whether Lord Justice O’Connor could ever have anticipated that for two decades gallons of ink would be spent in judicial and academic analysis of his judgment in Polly Peck to no certain conclusion.
However, the doctrinal analysis outlined in this article, informed as it is by extensive research into defamation litigation both in Australia and England, suggests some conclusions are now clear. The interpretation and application of ‘Polly Peck’ — better labelled Lucas-Box different meanings and Polly Peck common sting meanings — should be revised in Australia. Lucas-Box meanings, in particular, should be a common and useful part of litigation practice. ‘Far from raising a false issue’, such defence pleading ‘directs attention to the very essence of the dispute.’ While it may appear counter-intuitive to some, the evidence strongly suggests that it will support cheaper and quicker litigation, focused more on the substance of the publication in question. Such speed, economy and focus will assist plaintiffs as much as defendants. And it will assist the public interest in effective defamation litigation.
The uniform Defamation Acts should not sound any ‘death knell’ for Lucas-Box and Polly Peck. Indeed, if any bell tolls for defence meanings it may lead to a reprisal of comments made some years ago by another experienced defamation judge about the death of defamation. Justice David Levine, speaking extra-judicially, lamented the ‘excruciating and sterile technicalities’ with regard to defamatory meaning that were seen in the NSW practice of the late 1990s:
The tort of defamation is intended to provide a remedy for a person whose reputation has indefensibly been injured by the publication of something disparaging of that person’s good name. It boils down to determining what the publication means. Or it should. The amount of the Court’s time, let alone litigants’ resources, expended profligately in the determination of what words, sentences and phrases mean is positively scandalous.
Case law developments have improved the NSW situation somewhat since then. But the challenge remains real for defamation practice to achieve ‘speedy and efficient and fair resolution of disputes’. Dealing well with issues of meaning is central to that challenge.
* Director, CMCLCentre for Media and Communications Law and Associate Professor, Faculty of Law, University of Melbourne; Editor, Media & Arts Law Review. Thanks to the anonymous referees for helpful suggestions, the Australian lawyers who assisted in obtaining research materials and those who commented on an earlier draft. Arguments laid out in this article were also presented to the London barristers’ chambers, 5 Raymond Buildings, in September 2007. Thanks to the lawyers at 5RB for their hospitality and useful comments.
 Referred to as ‘uniform Defamation Acts’ below. See Civil Law (Wrongs) Act 2002 (ACT) as amended by Civil Law (Wrongs) Amendment Act 2006 (ACT); Defamation Act 2005 (NSW); Defamation Act 2006 (NT); Defamation Act 2005 (Qld); Defamation Act 2005 (SA); Defamation Act 2005 (Tas); Defamation Act 2005 (Vic); and Defamation Act 2005 (WA). The state Acts all commenced on 1 January 2006, the ACT Act on 23 February 2006, and the NT Act on 26 April 2006. Not only do the commencement dates differ, the numbering of some sections varies. For a brief overview, see David Rolph, ‘Uniform at Last? An overview of uniform, national defamation laws’ (2006) 76 Precedent 35; and for a general background to the history of earlier reform efforts, see Andrew T Kenyon, Defamation: Comparative Law and Practice (2006) at 362–364.
 Those courts may include the Federal Court in defamation actions under its accrued jurisdiction, of actions that involve Australia’s implied constitutional protection for political communication or concern publications of interstate or international scope; see Steven Rares, ‘Defamation and Media Law Update 2006: Uniform National Laws and the Federal Court of Australia’ (2006) 28 Australian Bar Review 1.
 Before the uniform Defamation Acts, the civil defamation law of the ACT, NT, SA, Vic, and WA was based on the common law, with relatively small legislative amendments; NSW operated under the Defamation Act 1974 (NSW) which made far more substantial changes to the common law; and Qld and Tas operated under 19th century defamation codes. For an outline of the differing laws see, for example, Des Butler & Sharon Rodrick, Australian Media Law (2nd ed, 2004) at 26–27 (see also 3rd ed, 2007).
 See, for example, Defamation Act 2005 (NSW) s 6(3).
 After Polly Peck (Holdings) v Trelford  QB 1000.
 John Fairfax Publications v Rivkin  HCA 50; (2003) 201 ALR 77 at 96 (Kirby J).
 Injunctions are also available, although difficult to obtain: see, for example, Australian Broadcasting Corporation v O’Neill  HCA 46; (2006) 227 CLR 57. The uniform Defamation Acts abolish exemplary damages and limit ordinary compensatory damages to approximately $250,000, see, for example, Defamation Act 2005 (NSW) ss 35, 37. The Acts also abolish the common law distinction between libel and slander and remove the requirement for proof of special damage, which at common law applied to some types of slander, see, for example, Defamation Act 2005 (NSW) s 7. In addition, the uniform Defamation Acts aim to promote the resolution of disputes without litigation and provide an offer of amends process under which the publisher may offer to make amends before, or as an alternative to, litigation: see, for example, Defamation Act 2005 (NSW) ss 12–19.
 See, for example, Jones v Amalgamated Television Services (1991) 23 NSWLR 364 at 367 (Hunt J).
 See, for example, E Hulton & Co v Jones  UKLawRpAC 57;  AC 20.
 See, for example, Capital and Counties Bank v George Henty & Sons (1882) 7 App Cas 741 at 745 (Lord Selbourne LC).
 See, for example, Lewis v Daily Telegraph  AC 234 at 258, 260 (Lord Reid, with whom Lord Jenkins agreed); Mirror Newspapers v Harrison  HCA 50; (1982) 149 CLR 293 at 301 (Mason J); Chakravarti v Advertiser Newspapers (1998) 193 CLR 519 (hereafter Chakravarti) at 572-573 (Kirby J).
 Amalgamated Television Services v Marsden  NSWSC 4; (1998) 43 NSWLR 158 at 165 (Hunt CJ at CL, with whom Mason P and Handley JA agreed).
 For example Charleston v News Group Newspapers  UKHL 6;  2 AC 65 at 71–72 (Lord Bridge, with whom Lords Goff, Jauncey and Mustill agreed), 73–74 (Lord Nicholls).
 Such elements have become central to public plaintiffs in US defamation since New York Times v Sullivan,  USSC 40; 376 US 254 (1964); see, for example, Rodney A Smolla, Law of Defamation (2nd ed, 1997 updated); Kenyon, above n1 at 239–280.
 For brief overviews of malicious falsehood, see, for example, David Price & Korieh Duodu, Defamation: Law, Procedure and Practice (3rd ed, 2004) at 45–50; Michael Gillooly, The Law of Defamation in Australia and New Zealand (1998) at 4–7. The situation for plaintiffs in malicious falsehood is particularly relevant since the uniform Defamation Acts have removed the ability of most corporations to sue in defamation: see, for example, Defamation Act 2005 (NSW) s 9.
 Michael Gillooly, The Third Man: Reform of the Australian Defamation Defences (2004) at 6. Note that Australian terminology is used throughout this article; English claimants, for example, are called plaintiffs as they would be in Australia. See further, Kenyon, above n1 at 7–8.
 The possible treatment of comment and opinion defences — as being directed to the meaning that is conveyed by the publication or being directed to the publication itself — is considered below in Part 5.
 See, for example, Defamation Act 2005 (NSW) ss 27–30. The uniform Defamation Acts also provide defences of innocent dissemination and triviality: see, for example, Defamation Act 2005 (NSW) ss 32–33.
 See, for example, Defamation Act 2005 (NSW) s 24(1) which reads ‘A defence under this Division is additional to any other defence or exclusion of liability available to the defendant apart from this Act (including under the general law) and does not of itself vitiate, limit or abrogate any other defence or exclusion of liability’.
 See, for example, Defamation Act 2005 (NSW) s 25. The uniform Defamation Acts only require substantial truth to be shown. They do not require an element of public interest or public benefit, unlike the former NSW, Queensland and Tasmanian law. On the former law, see, for example, Gillooly, above n15 at 112–116.
 See, for example, Defamation Act 2005 (NSW) s 26.
 See Kenyon, above n1 at 281–325; David Syme v Hore-Lacy  VSCA 24;  1 VR 667 (hereafter Hore-Lacy); Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314.
  NSWCA 227 (hereafter Zunter) at – (Handley JA, with whom Spigelman CJ and McColl JA agreed).
 Under Defamation Act 1974 (NSW) s 7A; see, for example, Patrick George, Defamation Law in Australia (2006) at 227–230; for instances of s 7A in operation, see Kenyon, above n1 at 137–152; and for analyses of difficulties that emerged under it, see David Rolph, ‘Perverse Jury Verdicts in New South Wales Defamation Trials’ (2003) 11 Torts Law Journal 28 and David Rolph, ‘Simple Questions, Difficult Juries: Perversity in Australian Defamation Trials after John Fairfax Publications Pty Ltd v Rivkin’ (2004) 18(3) Commercial Law Quarterly 9.
 For example, as Simpson J noted in Jones v John Fairfax Publications  NSWSC 1133 at , there was no room for arguments about such defence meanings because ‘the jury [had] already determined that the meaning attributed to the publication by the plaintiff was conveyed. The defendant [was] bound by that determination’.
 Andrew T Kenyon, ‘Imputation or Publication: The Cause of Action in Defamation Law’  UNSWLawJl 5; (2004) 27 University of NSW Law Journal 100 at 102. [Emphasis in original.]
 Historical analysis can also be used to support the reconsideration of Hore-Lacy and similar cases: see Kenyon, above n1 at 287–293. However, the limits of reliance on 19th century defamation practice have been well-noted: Hore-Lacy, above n22 at 687 (Charles JA).
 See, for example, comments by Steytler J in Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314 at 327-328 who explicitly referred to the lack of any clear High Court majority on the issue in Chakravarti, above n11 and noted the situation remained dependant on further guidance. See also comments by McLure JA in that and other judgments, including Buckeridge v Walter  WASCA 19 at –.
 The concept is discussed in Hore-Lacy, above n22 at 673-674 (Ormiston JA), 686-687 (Charles JA), 690 (Callaway JA).
 Des Butler, ‘Is Polly Peck at Risk of Losing Its Sting?’ (2000) 5 Media & Arts Law Review 259 at 263.
 This also means that some criticism of the law on defence meanings, focussed on general principles of civil litigation and pleading, is much less persuasive than it otherwise might be; for example, the material presented on litigation practices in Part 4 of this article could be compared with Anthony J H Morris, ‘Polly Peck Defence: Its Future in Australia’ (2000) 74 Australian Law Journal 760. The effectiveness in England, indeed the indispensability, of the approach recommended in this article undercuts such analyses.
 Hore-Lacy v David Syme  VSC 96 at .
 See, for example, Jameel v Wall Street Journal Europe Sprl  EMLR 6 at  (Simon Brown LJ); Berezovsky v Forbes  EWCA Civ 1251;  EMLR 45 at  (Sedley LJ for English Court of Appeal); Burstein v Times Newspapers  EWCA Civ 338;  1 WLR 579 at 589 (May LJ, with whom Sir Christopher Slade and Aldous LJ agreed); Carlton Communications v News Group Newspapers  EWCA Civ 1644;  EMLR 16 at  (Latham LJ).
  1 WLR 147.
 Patrick Milmo & W V H Rogers (eds), Gatley on Libel and Slander (10th ed, 2004) at [11.14].
 See, for example, John Fairfax Publications v Jones  NSWCA 205 at  (Hodgson JA); and see Polly Peck (Holdings) v Trelford  QB 1000 at 1032 (O’Connor LJ, with whom Robert Goff and Nourse LLJ agreed).
 See, for example, Woodger v Federal Capital Press of Australia (1992) 107 ACTR 1 at 24 (Miles CJ), and in particular point (3) of the five points listed on 23–24.
 See, for example, Sands v Channel Seven Adelaide  SASC 182; (2005) 91 SASR 466 at 478 (White J) in which it was noted that the label Polly Peck is used ‘in Australia colloquially’ to refer to Lucas-Box.
 See, for example, John Fairfax Publications v Jones  NSWCA 205 at  (Hodgson JA).
  1 WLR 1412.
 Carlton Communications v News Group Newspapers  EWCA Civ 1644;  EMLR 16 provides another example in which allegations of faking a number of television current affairs programs arguably raised a common sting meaning so evidence of faking one program could be used in defence of the whole publication.
 See, for example, Polly Peck (Holdings) v Trelford  QB 1000 at 1032 (O’Connor LJ, with whom Robert Goff and Nourse LLJ agreed); Hart v Wrenn and Australian Broadcasting Corporation  NTSC 7; (1995) 5 NTLR 17 at 24-25 (Mildren J).
  QB 1000.
 See, for example, John Fairfax Publications v Jones  NSWCA 205 at  (Hodgson JA).
 See, for example, Woodger v Federal Capital Press of Australia (1992) 107 ACTR 1 at 23–24 (Miles CJ), and in particular point (2) of the five points listed on 23–24.
 See, for example, Chakravarti, above n11 at 530-532 (Brennan CJ & McHugh J), 546 (Gaudron & Gummow JJ), 578-580 (Kirby J); Kenyon, above n1 at 54–60.
 This example is given in Jackson v John Fairfax  1 NSWLR 36 at 39 (Hunt J) and discussed in Gillooly, above n16 at 48–49. Hunt J also notes the use of the example by John G Fleming, The Law of Torts (6th ed, 1983) at 525 in Hepburn v TCN Channel Nine  1 NSWLR 386 at 397.
 Defamation Act 1974 (NSW) s 16; see, for example, John Fairfax Publications v Blake  NSWCA 434; (2001) 53 NSWLR 541; TCN Channel 9 v Antoniadis (1998) 44 NSWLR 682 at 691–692 (Handley, Beazley & Stein JJA); Waterhouse v Hickie (1995) Aust Torts Reports 62,486 at 62,489-62,491 (Priestley JA, with whom Mahoney and Powell JJA agreed); Hepburn v TCN Channel Nine  1 NSWLR 386 at 405 (Hunt J); Jackson v John Fairfax & Sons  1 NSWLR 36.
 See, for example, Defamation Act 2005 (NSW) s 26.
 It is adapted from Kenyon, above n1 at 24.
 See, for example, Bennett v News Group Newspapers  EMLR 39 at  (Robert Walker LJ for English Court of Appeal); Chase v News Group Newspapers  EMLR 11 at  (Brooke LJ, with whom Rix and Keene LJJ agreed); Musa King v Telegraph Group  EMLR 23.
 Perhaps a common sting argument would be open if the publication in question suggested a tax official had died as a result of the tax inquiry.
 The other question asked in relation to Lucas-Box different meanings — namely, whether the defence meaning concerns a separate and distinct allegation — is necessarily answered in the negative when a common sting meaning is found capable of being conveyed by a publication.
 See, for example, John Fairfax Publications v Blake  NSWCA 434; (2001) 53 NSWLR 541 (Court of Appeal) for a discussion of the similar contextual truth defence under the Defamation Act 1974 (NSW). Such a contextual truth argument is not available in England; it arises only under Australian statutes.
 Butler, above n30.
 See, for example, Gumina v Williams (No 2) (1990) 3 WAR 351 at 364 (Seaman J) in relation to ‘charges’; Khashoggi v IPC Magazines  1 WLR 1412 at 1417 (Sir John Donaldson MR, with whom Slade LJ agreed) in relation to ‘allegations’; Polly Peck (Holdings) v Trelford  QB 1000 at 1032 (O’Connor LJ, with whom Robert Goff and Nourse LLJ agreed) in relation to ‘statements’.
 Laurence W Maher, ‘Defamatory Meaning and the Defence of Truth: Polly Peck in Australia’ (1993) 1 Torts Law Journal 215 at 217.
 (Unreported, Queen’s Bench Division, Gray J, 15 November 2001) (hereafter Haslam); see Kenyon, above n1 at 323–324.
 Zunter, above n23 at .
 There is also a question about whether the ‘not more injurious’ test is at all appropriate for situations of Polly Peck common sting meanings: in principle, common stings would seem likely to often be more serious or injurious; see, for example, Kenyon, above n1 at 323.
 For background on the Hore-Lacy test, see, for example, Hore-Lacy, above n22; Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314; Advertiser-News Weekend Publishing v Manock  SASC 82; (2005) 91 SASR 206; Kenyon, above n1 at 296–299.
 Zunter, above n23.
 For a discussion of the background facts, see Zunter, above n23 at – (Handley JA, with whom Spigelman CJ and McColl JA agreed); see also Zunter v John Fairfax Publications  NSWSC 759 at – (Simpson J).
 Zunter, above n23 at  (Handley JA, with whom Spigelman CJ and McColl JA agreed).
 Zunter, above n23 at  (Handley JA, with whom Spigelman CJ and McColl JA agreed)..
 Zunter v John Fairfax Publications  NSWSC 259 at  (Simpson J). They were pleaded under s 16 of the Defamation Act 1974 (NSW).
 Zunter v John Fairfax Publications  NSWSC 259 at  (Simpson J).
 John Fairfax Publications v Blake  NSWCA 434; (2001) 53 NSWLR 541.
 John Fairfax Publications v Blake  NSWCA 434; (2001) 53 NSWLR 541 at 543 (Spigelman CJ, with whom Rolfe AJA agreed).
 Zunter v John Fairfax Publications  NSWSC 259 at  (Simpson J).
 Zunter, above n23 at – (Handley JA, with whom Spigelman CJ and McColl JA agreed). The Court of Appeal noted that contextual truth, apart from the element concerning no further harm to reputation, was established at trial and not challenged on appeal. Thus, the Court of Appeal did not consider in any detail whether the contextual imputations were conveyed in addition to and at the same time as the plaintiff’s imputations.
 Zunter v John Fairfax Publications  NSWSC 259 at  (Simpson J).
 It appears to have been pleaded as ‘a single additional imputation’ conveyed by the publication: ibid. Thus it could not be a Lucas-Box different meaning (which, in any event, could not be argued for once the jury had found the meaning that was conveyed by the publication).
 Zunter v John Fairfax Publications  NSWSC 259 at –.
 Id at .
 Zunter, above n23 at  (Handley JA, with whom Spigelman CJ and McColl JA agreed). The passage cited these cases as following the obiter comments of Brennan CJ & McHugh J in Chakravarti, above n11 at 526–530: Hore-Lacy, above n22; Robinson v Laws  1 Qd R 81; Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314; and Advertiser-News Weekend Publishing v Manock  SASC 82; (2005) 91 SASR 206.
 Robinson v Laws  1 Qd R 81 at 92-95 (de Jersey CJ), 99-102 (Williams JA), 106-108 (Mackenzie J); Defamation Act 1889 (Qld) ss 4 (definition of ‘defamatory matter’), 8.
 It is worth noting that in each judgment in Robinson v Laws, the defence was rejected with express reference to the basis of the cause of action lying in the plaintiff’s pleaded imputations under the Defamation Act 1889 (Qld). On this aspect of the former Queensland law, see Bellino v Australian Broadcasting Corporation  HCA 47; (1996) 185 CLR 183 at 230 (Dawson, McHugh & Gummow JJ), 237 (Gaudron J), approving Pervan v North Queensland Newspaper Co  HCA 64; (1993) 178 CLR 309 at 333 (McHugh J). However, the basis of the cause of action has changed under the uniform Defamation Acts to raise the importance of the publication: see, for example, Defamation Act 2005 (Qld) ss 4 (definition of ‘matter’), 6, 8. The advent of the uniform defamation laws means the availability of Lucas-Box and Polly Peck defences should be revisited in Queensland.
 See, for example, George, above n24 at 102–103.
 See, for example, Hore-Lacy; Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314; Advertiser-News Weekend Publishing v Manock  SASC 82; (2005) 91 SASR 206.
 See Lewis v Daily Telegraph  AC 234 at 260 (Lord Reid, with whom Lord Jenkins agreed).
 See, for example, Milmo & Rogers, above n35 at 283.
 See above n24.
 It is worth noting that while leave to appeal to the High Court was refused in Zunter, the issues of Lucas-Box and Polly Peck pleading were not raised by the applicant publisher. The grounds of application centred on the interpretation of ‘reasonableness’ under the statutory privilege in Defamation Act 1974 (NSW) s 22 and the assessment of further harm to reputation under the contextual truth defence in s 16 of that Act. These defences resemble, in general terms, statutory privilege and contextual truth defences under the uniform Defamation Acts; see, for example, Defamation Act 2005 (NSW) ss 30, 26. Leave was refused by Gummow and Hayne JJ, with Gummow J noting the result depended on the particular ‘facts and circumstances’ of the case, which ‘would not provide a convenient vehicle for determining any wider question’ about the uniform law raised by the publisher:  HCATrans 64 (9 February 2007). For a detailed examination of the s 30 defence for reasonable publication, see Kim Gould, ‘The More Things Change, the More They Stay the Same ... or Do They?’ (2007) 12 Media & Arts Law Review 29.
 See, for example, Woodham v John Fairfax Publications  NSWSC 1204 (hereafter Woodham) at  (Nicholas J).
 See, for example, Hore-Lacy, above n22; Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314; Advertiser-News Weekend Publishing v Manock  SASC 82; (2005) 91 SASR 206.
 Woodham, above n85 at  (Nicholas J).
 Woodham, above n85 at  (Nicholas J) referring to Chakravarti, above n11; Morosi v Mirror Newspapers  2 NSWLR 749.
 Hore-Lacy, above n22.
 Chakravarti, above n11.
 See, in particular, Gutnick v Dow Jones (No 4) (2004) Aust Torts Reports 65.777 (Supreme Court of Victoria) and Elliott v West Australian Newspapers  WASC 149.
 For a fuller analysis of the case law to mid-2005, which also examines why decisions like Gutnick v Dow Jones (No 4) should not be followed, see Kenyon, above n1 at 281–325 and in particular 310–312.
 See, for example, Morosi v Mirror Newspapers  2 NSWLR 749 at 771.
 See, for example, Anderson v Mirror Newspapers (No 2) (1986) 5 NSWLR 735 at 738 (Hunt J).
 Morosi v Mirror Newspapers  2 NSWLR 749 at 771.
  1 NSWLR 386 at 398.
 Hepburn v TCN Channel Nine  1 NSWLR 386 at 397-398. For the history of the treatment of ‘precision’ in NSW pleading, see the references listed in Amalgamated Television Services v Marsden  NSWSC 4; (1998) 43 NSWLR 158 at 161–162 (Hunt CJ at CL, with whom Mason P and Handley JA agreed); see also Greek Herald v Nikolopoulos (2002) 54 NSWLR 165.
 Hore-Lacy, above n22 at 675-676.
 Lewis v Daily Telegraph  AC 234 at 285 (Lord Devlin).
 Chakravarti, above n11.
 See Chakravarti, above n11 at 530-532 (Brennan CJ & McHugh J), 546 (Gaudron & Gummow JJ), 578-580 (Kirby J); Kenyon, above n1 at 54–60.
 Hore-Lacy, above n22.
  WASCA 273; (2003) 28 WAR 314.
  SASC 82; (2005) 91 SASR 206.
 Advertiser-News Weekend Publishing v Manock  SASC 82; (2005) 91 SASR 206 at 220 (Doyle CJ).
 See, for example, National Mutual Life Association of Australasia v GTV Corporation  VicRp 66;  VR 747 at 768 (Ormiston J); Barclay v Cox  VicRp 87;  VR 664.
 See, for example, repeated comments from publications of the NSW Law Reform Commission about the precision that was then believed to be achieved by the imputation-centred approach of the Defamation Act 1974 (NSW): NSW Law Reform Commission, Defamation, Report No 75 (1995) at [4.33]–[4.35]; NSW Law Reform Commission, Defamation, Discussion Paper No 32 (1993) at [4.37]–[4.41] and see Kenyon, above n1 at 338–344 discussing these and other Commission publications.
 See, for example, United Kingdom (Faulks Committee), Report of the Committee on Defamation, Cmnd 5909 (1975) at .
 See, for example, Amalgamated Television Services v Marsden  NSWSC 4; (1998) 43 NSWLR 158 at 165 (Hunt CJ at CL, with whom Mason P and Handley JA agreed). This test sees the law sidestep interesting questions of communication (and wider) theory. Roy Baker’s work provides one sociolegal engagement with some of those communication issues: see, for example, Roy Baker, ‘Third Person Singular? Instructing the Defamation Jury’ (Paper presented at the Jury Research Conference, Sydney, 17 October 2003). For an accessible overview of wider approaches from communication studies, which also has a good awareness of issues in critical theory, see John Durham Peters, Speaking in the Air: A History of the Idea of Communication (1999).
 It was notable that English barristers, with very extensive experience in acting for each party in defamation, expressed incredulity at the law in Hore-Lacy, above n22, when attending a seminar on the matter in 2007, let alone at the restrictive approach to ‘not substantially different’ that has been taken in a few later Australian decisions (above n91). They saw the Australian approach as radically different from English practice. My own research, detailed in this article and in Kenyon, above n1, strongly supports that assessment and underscores the preferability of the English approach for promoting equitable, efficient and effective defamation litigation.
 Kenyon, above n1. The study also included extensive research of the US situation, which is left aside for the purposes of this article.
 For details of the research, see id at 102–108, 131–134, 166–168, 393–401. For overviews of the results, see id at 101–193, 293–295.
 See, for example, Chakravarti, above n11 at 527-528 (Brennan CJ & McHugh J).
 An interviewee with more than 20 years experience as a barrister.
 An interviewee with a similar length of experience as a solicitor. For a fuller analysis, see material cited above n113.
 See above n24.
 See, for example, above n108.
 Michael Gillooly, ‘False Innuendoes and the Polly Peck Defence’ (1999) 7 Tort Law Review 189 at 192-193.
 Herald & Weekly Times v Popovic  VSCA 161; (2003) 9 VR 1 at 62 (Warren AJA agreed).
 See Kenyon, above n1 at 304–310 for a longer analysis.
 That is the point raised by ‘the words actually complained of by the plaintiff subject to context’ in the final sentence quoted from Gillard AJA, above n120.
 See, for example, Kenyon, above n1 at 155 in relation to plaintiff pleading in NSW and 72–74 in relation to mitigation and defence meanings in general.
 Woodger v Federal Capital Press of Australia (1992) 107 ACTR 1 at 21 (Miles CJ).
 Polly Peck (Holdings) v Trelford  QB 1000 at 1023 (O’Connor LJ, with whom Robert Goff and Nourse LLJ agreed). The use of such metaphors in relation to each party in defamation judgments could be analysed further in light of metaphors in other areas of law and their influence on doctrinal development; for an accessible analysis, see Patricia Loughlan, ‘Pirates, Parasites, Reapers, Sowers, Fruits, Foxes ... The Metaphors of Intellectual Property’  SydLawRw 12; (2006) 28 Sydney Law Review 211 at 213–216.
 See, for example, in relation to truth defences, Defamation Act 2005 (NSW) s 25 reads: ‘It is a defence to the publication of defamatory matter if the defendant proves that the defamatory imputations carried by the matter of which the plaintiff complains are substantially true.’ There could be argument under s 25 as to whether substantial truth must be proved for the imputations of which the plaintiff complains, or for the imputations conveyed by the matter (with the plaintiff’s complaint being linked to the matter rather than the imputation by the final words of s 25). Doctrinal analysis and evidence from practice under English and earlier Australian law strongly suggest that the best interpretation is the latter one: the truth of the meaning actually conveyed must be established, so long as that is not a separate and distinct allegation of which the plaintiff does not complain. In any event, common law defences are preserved by the uniform law — for example Defamation Act 2005 (NSW) s 24(1) — under which Lucas-Box and Polly Peck can be argued.
 Polly Peck (Holdings) v Trelford (1985)  QB 1000 at 1021 (O’Connor LJ, with whom Robert Goff and Nourse LLJ agreed). See also Kenyon, above n1 at 355–357.
 See above n46 and accompanying text.
 See, for example, Chakravarti, above n11 at 527-528 (Brennan CJ & McHugh J).
 Defamation Act 1974 (NSW) s9(2); see, for example, Anderson v Mirror Newspapers (No 2) (1986) 5 NSWLR 735 at 738 (Hunt J).
 See, for example, above n108.
 See, for example, Control Risks v New English Library  1 WLR 183; Anderson v Nationwide News  VSC 335; (2001) 3 VR 619. See also, for example, Matthew Collins, The Law of Defamation and the Internet (2nd ed, 2005) at [9.17].
 See  HCATrans 252 (25 May 2007);  HCATrans 414 (7 August 2007);  HCATrans 416 (8 August 2007); Manock v Channel Seven Adelaide  SASC 322; (2006) 95 SASR 462 (hereafter Manock).
 See, for example, the comments of Samuels JA in Petritsis v Hellenic Herald  2 NSWLR 174 at 193, which arose in relation to the former Defamation Act 1974 (NSW) but suggest how a broad view of fair comment could be sustained: ‘[A] defence of comment, accepting that the comment is defamatory, is not concerned with the precise nature of the defamatory meaning or imputation. It asserts that, whatever the defamatory character of the matter — or so much of it as is alleged to be defamatory — the words complained of are comment ... and are, therefore, not actionable. The defence does not challenge that the matter has a defamatory meaning ... or what those meanings are. It is directed to the character of the vehicle by which those meanings, whatever they are, are conveyed; that is by a statement of fact or by a statement of opinion. It must, therefore, penetrate beyond the alleged meanings to the raw material of the actual words employed.’
 The characterisation of material as comment or fact is often an important matter for the tribunal of fact; see, for example, O’Shaughnessy v Mirror Newspapers  HCA 52; (1970) 125 CLR 166. The statement in Clarke v Norton  ArgusLawRp 97;  VLR 494 at 499 (Cussen J) that ‘comment’ is a deduction, inference, conclusion, criticism, judgment, remark or observation is a useful yardstick. Price & Duodu, above n15 at [9–03] provide examples from English case law. The English Court of Appeal has recently analysed this and other elements of the defence in Associated Newspapers v Burstein  EWCA Civ 600 (21 May 2007).
 Fairness here should be understood as meaning the opinion is not one beyond that which an honest person would be capable of holding, given the underlying factual or privileged basis; see, for example, Albert Cheng v Tse Wai Chun Paul (2000) 10 BHRC 525 at 529 (Lord Nicholls).
 Steven Rares, ‘No Comment: The Lost Defence’ (2002) 76 Australian Law Journal 761 at 767.
 See, for example, the jury direction of Diplock J in Silkin v Beaverbrook Newspapers  1 WLR 743 at 745–746.
 See, for example, Defamation Act 2005 (NSW) s 31.
 Manock , above n134.
 Manock, above n134 at 463 (Gray & Layton JJ).
 Manock, above n134 at 464 (Gray & Layton JJ).
 Manock, above n134 at 470-471 (Gray & Layton JJ).
 Channel Seven Adelaide v Manock  HCATrans 252 (25 May 2007). The appeal was heard in August 2007; see above n134.
 Manock, above n134 at 465 (Gray & Layton JJ).
 Manock, above n134 at 465 (Gray & Layton JJ).
 Manock, above n134 at 470 (Gray & Layton JJ).
 Manock, above n134 at 472 (Gray & Layton JJ).
 See, for example, Rares, above n138; Andrew T Kenyon, ‘Defamation, Artistic Criticism and Fair Comment’  SydLawRw 9; (1996) 18 Sydney Law Review 193 at 209–216.
 Manock, above n134 at 481 (Gray & Layton JJ).
 See above n136.
 This is where the concept of ‘congruency’ could be usefully applied. The term has been prominent in NSW decisions about comment under the Defamation Act 1974 (NSW); for example, New South Wales Aboriginal Land Council v Perkins (1998) 45 NSWLR 340. It has also been used elsewhere; for example, Anderson v Nationwide News  VSC 335; (2001) 3 VR 619 at 635 (Ashley J). However, under the uniform Defamation Acts, ‘congruency’ should be understood as being equivalent to the concept that a separate and distinct allegation cannot be raised in fair comment or honest opinion defences. See further, Kenyon, above n1 at 299–301.
 See discussion of earlier decisions in Manock, above n134 at 470-472 (Gray & Layton JJ).
 Hore-Lacy, above n22 at 686 (Charles JA) referring to comments by Brennan CJ and McHugh J, as well as comments by Gummow and Gaudron JJ, in Chakravarti, above n11.
 See, for example, Kenyon, above n1 at 75–76.
 See, for example, Mirror Newspapers v Harrison  HCA 50; (1982) 149 CLR 293 at 302 (Mason J); Gumina v Williams (No 2) (1990) 3 WAR 351 at 371 (Seaman J).
 See, for example, Berezovsky v Forbes  EWCA Civ 1251;  EMLR 45. Note that legislation allows earlier convictions to be used as evidence of crimes; see, for example, Defamation Act 2005 (NSW) s 42.
 See, for example, Defamation Act 2005 (NSW) s 31.
 Defamation Act 1974 (NSW) ss 29–35.
 See, for example, Defamation Act 2005 (NSW) s 31(1). It appears that the use of the term ‘opinion’ in the uniform Defamation Acts rather than ‘comment’ in the former NSW Act should not be treated as a substantive difference, with the analysis under Clarke v Norton  ArgusLawRp 97;  VLR 494 at 499 (Cussen J) still being applied; see above n136. The distinction being sought remains that between factual and non-factual material.
 See, for example, Defamation Act 2005 (NSW) s 31(5). The statute does not explicitly require the facts to be stated in the publication, indicated in the publication, or notorious. This may provide a greater ability to prove a factual basis for opinions. It appears to differ from the position under the Defamation Act 1974 (NSW) s 30(1) which, for the purposes of determining what was proper material on which comment could be based, referred back to the common law requirement for facts to be stated, indicated or notorious: see, for example, Sims v Wran  1 NSWLR 317 at 322. For the common law fair comment defence at issue in Manock, the facts were arguably notorious because of extensive media coverage and public debate surrounding the death, investigation, murder charge, trial and retrial. For an outline of the common law position as to the requirement for facts to be truly stated, indicated or notorious see, for example, George, above n24 at 341–344.
 See, for example, Defamation Act 2005 (NSW) s 31(6).
 See, for example, Defamation Act 1974 (NSW) s 31(4).
 See, for example, Defamation Act 2005 (NSW) ss 31(2)–(4). In relation to third party commentators, the uniform Defamation Acts, like the 1974 NSW Act, aim to overcome complexities illustrated by Canadian and English decisions: compare Cherneskey v Armadale Publishers  1 SCR 1067 at 1079–1081 (Ritchie J, with whom Laskin CJC, Pigeon & Pratte JJ agreed); Telnikoff v Matusevich  UKHL 2;  2 AC 343 at 354–355 (Lord Keith, with whom Lords Brandon & Oliver agreed).
 Defamation Act 1974 (NSW) ss 29–35. Under the uniform Acts, ‘matter’ has a broad, inclusive definition, see, for example, Defamation Act 2005 (NSW) s 4.
 See, for example, New South Wales Aboriginal Land Council v Perkins (1998) 45 NSWLR 340. The NSW courts took a more workable approach to the issue of deciding whether a plaintiff’s pleaded imputation was fact or comment. That was determined in the context of the whole publication; see, for example, Goldsworthy v Radio 2UE Sydney  NSWSC 547 at .
 Rares, above n138 at 761.
 Senior counsel from the NSW bar.
 Two other senior counsel from the NSW bar.
  VSC 335; (2001) 3 VR 619.
 These aspects are considered at greater length in Kenyon, above n1 at 299–304. See also the discussion of Nationwide News v Moodie  WASCA 273; (2003) 28 WAR 314 in Kenyon, above n1 at 315–320.
 Rares, above n138 at 768.
 Woodham, above n85 at  (Nicholas J).
 In comparison, common sting meanings and contextual truth meanings arise in far fewer publications which become the subject of suit, which is not surprising when their particular qualities are kept in mind.
 Gillooly, above n119 at 193.
 Compare Woodham, above n85 at  (Nicholas J).
 Justice David Levine, ‘The Future of Defamation Law’ (Speech given on the occasion of the publication of the first issue of the UTS Law School Law Review, Sydney, 31 August 1999) at .
 Id at .
 See, for example, Greek Herald v Nikolopoulos (2002) 54 NSWLR 165 (Court of Appeal) and Kenyon, above n1 at 330–336.
 Levine, above n178 at .