Sydney Law Review
Before the High Court
Liability for the Publication of Third Party Comments: Fairfax Media Publications Pty Ltd v Voller
The liability of media outlets for third party comments posted on public Facebook pages raises difficult issues of basic principle for the tort of defamation. The appeal before the High Court of Australia in Fairfax Media Publications Pty Ltd v Voller involves a reconsideration of basic principles of publication. The media outlets argue that intention to publish is required to establish the element of publication as part of the cause of action in defamation. This column argues that the better view is that liability for publication is strict. This does not mean, however, that the media outlets are liable for the third party comments posted on their public Facebook pages. This column argues that it is necessary to analyse distinctly what the conduct amounting to communication of the defamatory matter is and the basis upon which the defendant is responsible for that conduct. It argues that the media outlets may not be liable for the third party comments, at least in the absence of actual notice, by virtue of publication by omission, the proper juridical basis of which is the defendant’s continuation of a third party’s wrong.
Please cite as:
David Rolph, ‘Liability for the Publication of Third Party Comments: Fairfax Media Publications Pty Ltd v Voller’  SydLawRw 9; (2021) 43(2) Sydney Law Review 225.
This work is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International Licence (CC BY-ND 4.0).
As an open access journal, unmodified content is free to use with proper attribution.
Please email firstname.lastname@example.org for permission and/or queries.
© 2021 Sydney Law Review and author. ISSN: 1444–9528
Almost two decades ago, the High Court of Australia handed down its judgment in Dow Jones & Co Inc v Gutnick. Gutnick was the first decision of a final appellate court on the issue of jurisdiction over internet defamation cases. The High Court reaffirmed the basic principles of publication in defamation law and confirmed their application to internet technologies. Their Honours refused the invitation to fashion special rules for internet defamation, thereby reaffirming the medium-neutral nature of the principles of publication in defamation law. In the intervening period, internet technologies have proliferated and, increasingly, publications by means of them have been sued upon in defamation cases. The diversity of internet technologies raises difficult issues of publication, with which courts throughout the common law world are grappling. These issues tended not to arise when mass media publications — newspapers, radio, television — were the dominant form of widespread publications. It is important to be clear on what has changed. For mass media publications, the composition and first dissemination of, and profit from, the defamatory matter are ordinarily integrated steps, for which the mass media publisher is responsible. As a consequence, hard questions about publication do not arise routinely. By contrast, internet technologies disaggregate these integrated steps, which complicates issues of publication. The High Court in Trkulja v Google LLC suggested that the law relating to publication is ‘tolerably clear’, but that its application to novel technologies may present some difficulties. It may be, however, that because the issue of publication was uncontentious in so many defamation cases for so long, the relevant principles are not as well-understood as they need to be and that defamation cases involving internet intermediaries expose issues of principle relating to the concept of publication that have not needed to be decided before.
The issue of publication arises for consideration in the forthcoming appeal in Fairfax Media Publications Pty Ltd v Voller. The case raises the difficult question of whether a media outlet is liable for third party comments posted on its public Facebook page, even in the absence of, and in advance of, actual notice of the presence of those comments. The course of argument has shifted as the proceedings have progressed through the various levels of the judicial hierarchy. On appeal to the High Court, the media outlets seek to argue that they cannot be liable for publication as they did not intend to publish the particular defamatory matter. They submit that intention is required for the element of publication in the tort of defamation. This column argues that the element of publication is a matter of strict liability but that, even if intention to publish were required, that may not necessarily assist the media outlets in this case to avoid liability. This column suggests that internet technologies, and the difficult issues of publication that they continue to present, necessitate distinct consideration of the conduct alleged to constitute communication of the defamatory matter, on the one hand, and the basis upon which a particular defendant should be held responsible for the communication of that defamatory matter, on the other hand. Viewing the problem in this way, it may be that the media outlets are not publishers of third party comments in the absence of actual notice of those comments, at a minimum.
Dylan Voller came to national prominence in ‘Australia’s Shame’, an episode of the Australian Broadcasting Corporation’s current affairs show, Four Corners. The program concerned the mistreatment of children and young people in the Northern Territory’s adult and juvenile detention system. Voller was shown, wearing a spit hood, shackled to a restraining chair, in an adult Alice Springs correctional centre. The public response to the Four Corners broadcast led the then Prime Minister, Malcolm Turnbull, to establish a royal commission into the protection and detention of children in the Northern Territory.
Three media outlets posted material to their respective public Facebook pages. The material they posted was not defamatory. However, it was possible for Facebook users to leave comments beneath the posts. Voller alleged that some of those comments were defamatory of him. Without giving prior notice to the media organisations, he commenced defamation proceedings against them in the Supreme Court of New South Wales (‘NSW’). The issue of whether the media organisations were publishers of the third party comments was ordered to be determined as a separate question. At first instance, Rothman J found that they were. The media companies appealed.
In the NSW Court of Appeal, Basten JA dealt with the issue of publication briefly. His Honour applied Ribeiro PJ’s judgment in Oriental Press Group Ltd v Fevaworks Solutions Ltd to conclude that the media outlets were publishers. Basten JA stated that: ‘They facilitated the posting of comments on articles published in their newspapers and had sufficient control over the platform to be able to delete postings when they became aware that they were defamatory.’
In their joint judgment, Meagher JA and Simpson AJA noted that defamation is a tort of strict liability. Their Honours observed that the Defamation Act 2005 (NSW) does not define ‘publication’, so that the common law principles apply. Meagher JA and Simpson AJA identified Isaacs J’s judgment in Webb v Bloch as ‘[t]he classic statement in this country of what constitutes publication’. Their Honours identified the particular act of publication for which the media outlets were responsible as the subscription to the public Facebook pages. According to Meagher JA and Simpson AJA this act made the media outlets instrumental and participants in the comments published on those pages. Thus, the NSW Court of Appeal unanimously and readily found that the media outlets were publishers of the third party comments on their public Facebook pages.
Two other related issues dealt with by the NSW Court of Appeal are worth noting. The first was innocent dissemination. Both judgments agreed that the only issue on appeal was publication and the trial judge had erred by holding that the media outlets were primary publishers. Yet there was a difference of opinion as to whether innocent dissemination at common law was properly regarded as a defence or as a plea of ‘no publication’. Basten JA acknowledged that there were conflicting obiter dicta about whether innocent dissemination was a defence or a denial of publication. His Honour found that the prevailing common law position was that of Brennan CJ, Dawson and Toohey JJ in Thompson v Australian Capital Television Pty Ltd, which appeared to accept the view that innocent dissemination should be viewed as a defence. This meant that the common law approach was consistent with the statutory defence of innocent dissemination under the national, uniform defamation laws.
The second related issue was the classification of primary and secondary publishers. Basten JA noted that, in light of the High Court’s obiter dicta in Trkulja v Google LLC, it was inappropriate to classify the media outlets as either primary or secondary publishers in advance of the media outlets pleading a defence of innocent dissemination. By contrast, Meagher JA and Simpson AJA concluded that the issue remained unresolved at common law. With respect, the position reached by Basten JA is to be preferred. It is consistent with the High Court’s approach in Trkulja v Google LLC, which did not strictly distinguish between innocent dissemination based on whether it arose at common law or under statute. There is also, with respect, an internal inconsistency in the reasoning of Meagher JA and Simpson AJA on this issue. Their Honours accepted that the only issue the trial judge should have determined was whether the media outlets were publishers of the third party Facebook comments and that the trial judge erred in holding that the media outlets were primary publishers. Yet if innocent dissemination remains a denial of publication, at least at common law, it was relevant to the resolution of the issue of publication, which was the only subject of the separate question. Equally, the proper classification of the media outlets as either primary or secondary publishers was germane to the issue of publication because, even at common law, only a subordinate distributor could rely upon a plea of innocent dissemination. Indeed, Meagher JA and Simpson AJA appeared implicitly to accept that the media outlets were primary publishers because they were ‘not relying on the “never published” principle as an answer to the separate question, no doubt because they accepted that they are not in the same position as, nor any position analogous to “book sellers, news vendors, messengers, or letter carriers”’. The proper classification of innocent dissemination is not an idle issue of taxonomy, but will determine whether innocent dissemination is relevant to the principles of publication as an element of the plaintiff’s cause of action.
On appeal to the High Court, the media outlets seek to rely upon an argument not raised at first instance or on intermediate appeal: that proof of intention is required to establish publication.
The tort of defamation is conventionally described as one of strict liability. Indeed, this was how Meagher JA and Simpson AJA described it in the NSW Court of Appeal in Voller (NSWCA). To say that defamation is a tort of strict liability ordinarily means that liability does not depend upon proof of fault; it does not depend upon proof of intention or negligence. Whether this overall characterisation of the tort of defamation is viable is debatable, given that many defences to defamation turn upon malice or reasonableness. Nevertheless, the elements of the plaintiff’s cause of action — defamatory meaning, identification and publication — have been understood to be matters of strict liability. This is demonstrated by Dixon J’s judgment in Lee v Wilson. In that case, the issue was whether intention was required to establish the element of identification. It turned upon whether the House of Lords’ decision in E Hulton & Co v Jones should be followed in Australia. In Hulton, Lord Loreburn LC found that, because it was not necessary for a plaintiff to prove intention to defame, it was also not necessary, by consistency of principle, to prove intention to identify the plaintiff. To resolve the issue under Australian law, Dixon J in Lee v Wilson returned to basic principle. Like Lord Loreburn LC in Hulton, his Honour reasoned that, if intention is not required to establish defamatory meaning or publication, then, by consistency of principle, intention is not required to establish identification. It would be anomalous for one element of the cause of action to require proof of intention when others do not.
In relation to the element of publication, Dixon J was explicit that it is a matter of strict liability. His Honour stated:
The cause of action consists in publication of the defamatory matter of and concerning the plaintiff. It might be thought, therefore, that, in any event, this warranted or required some investigation of the actual intention of the publisher. But his liability depends upon mere communication of the defamatory matter to a third person. The communication may be quite unintentional, and the publisher may be unaware of the defamatory matter.
As a matter of principle, it is understandable that defamatory meaning and publication would be dealt with on the same basis, as matters of strict liability, because they are interrelated elements of the cause of action. Defamatory meaning is concerned with whether something disparaging about the plaintiff’s reputation has been communicated and publication is concerned with whether it has been communicated.
The High Court of Australia in Gutnick endorsed Dixon J’s statement of principle from Lee v Wilson, affirming that defamation is a tort of strict liability. The element of the cause of action in defamation considered in Gutnick was publication.
Judges in other jurisdictions when dealing with difficult issues of whether particular internet intermediaries are publishers for the purposes of defamation law have stated that liability for publication is strict. In Godfrey v Demon Internet Ltd, dealing with whether an internet service provider was a publisher of third party content posted and stored on a news server operated and controlled by it, Morland J stated: ‘At common law liability for the publication of defamatory material was strict.’ In Oriental Press Group, dealing with whether the operators of an internet discussion forum were liable for comments posted in it, Ribeiro PJ stated that: ‘Until mitigated by the common law defence of innocent dissemination which evolved in the late nineteenth century, liability for publishing a libel was strict and could lead to harsh results.’ It should also be noted that the Law Commission of Ontario, following its wide-ranging four-year review of defamation law, recommended that the concept of a ‘publisher’ should be defined in statute to require an intentional act of communicating a specific expression. The purpose of this recommendation is to clarify and narrow the scope of liability for publication at common law. The fact that a legislative amendment would be required to achieve this result suggests that the common law position does not require proof of an intention to communicate the particular defamatory matter.
The argument that intention is required for publication relies heavily upon the obiter dicta of Isaacs J in Webb v Bloch. In his collection of statements of principle from case law and treatises, his Honour stated that:
The meaning of ‘publication’ is well described in Folkard on Slander and Libel, 5th ed. (1891), at p. 439, in these words: ‘The term published is the proper and technical term to be used in the case of libel, without reference to the precise degree in which the defendant has been instrument to such publication, if he has intentionally lent his assistance to its existence for the purpose of being published, his instrumentality is evidence to show a publication by him.’
It should be noted that the sole reference to intention here occurs with reference to the notion of lending assistance to a publication. The reference to assistance more readily connotes the extension of legal responsibility, a form of accessorial liability, rather than the imposition of liability for the original communication of defamatory matter. Thus, the extent to which intention is required for all forms of liability for publication is questionable. The collation of principle undertaken by Isaacs J in Webb v Bloch needs to be considered ‘secundum subjectam materiam’ (according to the subject matter). The issue of publication in that case was whether the members of a committee who authorised a solicitor to prepare a circular were publishers of that circular, in circumstances where the solicitor was actuated by malice. Given that the statements of principle in Isaacs J’s judgment are directed to that issue, it is open to doubt whether they should be taken as a comprehensive and definitive distillation of the principles of publication in defamation law. The significance of Isaacs J’s judgment in Webb v Bloch on the issue of publication can be overstated if its factual context is forgotten.
The case law dealing with innocent dissemination supports the conclusion that liability for publication is strict. If liability for publication were fault-based, then the doctrine of innocent dissemination would not have needed to have been developed to overcome its harshness. In Emmens v Pottle, Lord Esher MR stated that by handing the newspaper to other people, the newsvendor was prima facie liable. In Vizetelly v Mudie’s Select Library Ltd, AL Smith LJ was more explicit, finding that: ‘The defendants having lent and sold copies of the book containing that libel, prima facie they published it. What defence, then, have they? None, unless they can bring themselves within the doctrine of Emmens v Pottle.’
Romer LJ adopted a similar position in his judgment in Vizetelly, stating that:
The law of libel is in some respects a very hard one. ... For many years it has been well settled law that a man who publishes a libel is liable to an action, although he is really innocent in the matter, and guilty of no negligence. That rule has been so long established as to be incapable of being altered or modified, and the Courts, in endeavouring to mitigate the hardship resulting from it in many cases, have only been able to do so by holding that, under the circumstances of cases before them, there had been no publication of the libel by the defendant.
What his Lordship makes clear is that, at common law, liability for publication is strict; the courts were unable to depart from the established common law position that liability for publication is strict; to overcome the unfairness of that position in particular cases, the court devised innocent dissemination to deem conduct that would, on established principle, amount to publication not to be publication.
It is true that Lord Esher MR in Emmens v Pottle rejected classifying innocent dissemination as a form of privilege, instead classifying it as a plea of ‘no publication’. The extent to which that should be taken as an expression of high principle is doubtful. Lord Esher MR in Emmens v Pottle was overt that he was not concerned with issues of taxonomy and principle. In fashioning the doctrine of innocent dissemination, his Lordship was candid that he was doing so as a matter of policy and pragmatism, to deal with what he perceived as the unfairness of the outcome dictated by a strict application of principle. By the time of Vizetelly, AL Smith LJ could refer to innocent dissemination as a defence, without the qualification of quotation marks. In Vizetelly, Romer LJ was explicit that innocent dissemination was not grounded in principle, opining that ‘the decisions on the subject have not been altogether logical or satisfactory on principle’ and, whilst accepting that Emmens v Pottle ‘worked substantial justice’, did not think that the way in which those cases were decided was ‘altogether logical or satisfactory on principle’ and thought that the judgment did not ‘very clearly indicate on which principle Courts ought to act in dealing with similar cases in the future’. There are similar criticisms of the unprincipled nature of innocent dissemination in the High Court of Australia’s judgments in Thompson. Brennan CJ, Dawson and Toohey JJ referred to the ‘somewhat muddied origins’ of innocent dissemination and, in relation to Vizetelly, observed that ‘the judgments ... hardly offer a satisfactory statement of principle’. Similarly, Gummow J describes innocent dissemination as ‘difficult to reconcile with other principles of the tort and rest[ing] upon “expediency”.’
Even if innocent dissemination is deemed to be a plea of ‘no publication’, it is, in substance, a policy-based exculpation of the defendant for responsibility for the publication — which is to say, it is a defence. It does not assist in determining the principles of publication because it is pragmatic and directed towards avoiding liability for publication, which is prima facie imposed.
Cases of accidental or unintentional publication are also unhelpful in establishing that proof of intention to publish is required as a matter of general principle in all cases of defamation. These cases, such as that involving the curious butler or the eavesdropping employee, are highly particular and do not supply a sound principled basis for establishing a general principle within the concept of publication.
Recent judicial attempts to integrate intention into the element of publication in a cause of action for defamation have not been wholly successful. This can be seen in the judgments of the Supreme Court of South Australia in Duffy v Google Inc, at first instance and on appeal. These treated only publication as requiring proof of intention, with defamatory meaning and identification treated as matters of strict liability. With respect, these statements of general principle are difficult to follow and seem contrary to principle and authority. In particular, given that they turn upon a putative global distinction between primary and secondary publishers, these statements are inconsistent with the obiter dicta of the High Court in Trkulja v Google LLC on this issue.
Even if the element of publication requires proof of intention, a range of questions follow, which do not find a ready answer in the case law on defamation. The concept of intention in tort law, while superficially simple, is by no means free from difficulty. Does it mean merely a voluntary act on the part of the defendant? Is the requisite intention subjective? May the requisite intention be objectively inferred? Is it only the act itself that must be intended or can liability extend to consequences that are substantially certain to result, even if they are not desired? Will recklessness suffice to establish a form of imputed intention? If the element of publication is fault-based, not strict liability, can publication be established upon proof of negligence? Who bears the onus of proof or disproof of fault?
If the element of publication required proof of intention, one might have expected these issues to have been explored in defamation cases more than they have been. In most Australian defamation cases, the issue of intention is rarely discussed in relation to publication, still less recklessness or negligence. This may be contrasted with trespass to the person, where issues of the availability of negligent trespass under Australian law and the onus of proof or disproof of fault have generated a not insubstantial case law. The absence of any detailed or consistent discussion of intention in relation to publication in Australian defamation case law is consistent with the view that publication, like the other elements of the cause of action, is a matter of strict liability, turning upon the fact of the communication of the relevant matter, not whether the communication has been proved to be intentional or otherwise fault-based.
Even if the element of publication requires proof of intention, this may not necessarily assist the media outlets’ case. If proof of intention to do the act is required, there is a question as to what, in fact, constitutes the relevant act. Whether the defendants intended to publish cannot be determined unless the act of publication that they were alleged to have intended is identified. In Voller, there are different ways of characterising the conduct constituting publication. For instance, if the NSW Court of Appeal is correct, and the act of publication is the act of subscribing to the public Facebook pages, then it may be, on one analysis, that the requisite intention is established because the media outlets intended to subscribe to the public Facebook pages, which carried with it the substantially certain consequence that third parties would post comments on the pages. If this characterisation of the relevant act of publication is correct, it has consequences for the liability of other potential publishers, beyond the defendants in Voller. The reasoning of the NSW Court of Appeal in Voller (NSWCA) could not be confined, as a matter of principle, only to commercial publishers. The principles of publication apply generally — the common law has avoided having specific rules for different types of publishers — and do not depend upon the motive or purpose of the publication. Thus, any intentional subscription to a social media page that permitted third parties to comment would expose the person or entity so subscribing to liability for the publication of third party comments. If correct, the reasoning of the NSW Court of Appeal in Voller (NSWCA) would also suggest that Beach J’s analysis in Trkulja v Google LLC (No 5) of the liability of search engine operators for the publication of search results generated by third party users is to be preferred over McCallum J’s analysis of the same issue in Bleyer v Google Inc. Beach J reasoned that, because the search engine operator set up and programmed the search engine to operate in the way in which it did, the search engine operator is liable for the publication of search engine results from the point at which they are generated by third party users because the search engine is operating as intended.
On appeal, the media outlets characterise the act of publication as the ‘deliberate, purposeful’ communication of the particular matter complained of. They argue they could not have had such an intention, given that, at the time the comments were published, they did not know of their existence. Several points may be made about this. The difficulties of defining intention in tort law have already been noted. On this approach to the relevant act of publication, intention is defined in a way that does not seem to find a sound basis in the case law. Similarly, the requirement that there should be an intention to communicate the particular matter complained of is not emphasised in the case law. It finds some support in cases ordinarily characterised as accidental or unintentional publication, but such cases are atypical. Fundamentally, though, the media outlets, through their approach to intention, seek to address the absence of actual notice of the defamatory matter in the attempt to fix them with liability for publication. Whether it is apposite to deal with the absence of actual notice through the issue of intention to publish is questionable.
What constitutes the conduct amounting to communication of the relevant matter is contestable. Until that issue is settled, it will be difficult to decide whether the media outlets intended that conduct. There is another way of analysing the facts in Voller to reach the conclusion that the media outlets are not publishers of third party comments at least in the absence of actual notice, which does not rely upon introducing intention into the element of publication. Rather than the relevant conduct amounting to publication being the positive act of subscribing to the public Facebook pages, the relevant conduct could be characterised as the failure to deal with third party comments. Viewed in this way, the conduct amounting to publication is not a positive act, but an omission.
Two features of Voller have tended to be overlooked or downplayed. First, the issue in Voller — as it was, in part, in Murray v Wishart — is whether the defendants were liable for third party comments. So to state the issue should raise a concern that, as a general principle, a defendant is not ordinarily liable for the wrongs of a third party. The common law will hold a defendant liable for his or her own wrongs but will ordinarily require an exceptional reason to hold a defendant liable for the wrongs of a third party. This is made clear by Dixon J in Smith v Leurs, wherein his Honour stated:
[A]part from vicarious responsibility, one man may be responsible to another for the harm done to the latter by a third person; he may be responsible on the ground that the act of the third person could not have taken place but for his own fault or breach of duty. There is more than one description of duty the breach of which may produce this consequence. For instance, it may be a duty of care in reference to things involving special danger. It may even be a duty of care with reference to the control of actions or conduct of the third person. It is, however, exceptional to find in the law a duty to control another’s actions to prevent harm to strangers. The general rule is that one man is under no duty of controlling another man to prevent his doing damage to a third. There are, however, special relations which are the source of a duty of this nature.
Second, and related to this, is the differential treatment of acts and omissions at common law. The common law is comfortable imposing liability on a defendant for his or her own positive acts, but is averse to imposing liability on a defendant for mere omissions. Where a defendant is sought to be held liable for an omission, something more than the defendant’s failure to act is required.
The proper characterisation of the defendant’s conduct therefore becomes important because different legal consequences will follow, depending upon how the relevant conduct is identified. The suggestion that the media outlets’ positive acts of subscribing to public Facebook pages constituted the relevant act of publication and the consequences that flowed from that have already been considered. It may be, however, that the conduct for which the media outlets may be held responsible is the failure to deal with the third party comments. Rather than being positive acts on the part of the media outlets, the conduct amounting to communication for which they should be held responsible is an omission. Given that the common law is averse to imposing liability for mere omissions, the question then becomes what are the conditions under which the media outlets may be held liable for their failure to act.
In this regard, the English Court of Appeal’s decision in Byrne v Deane assumes central importance. In Byrne v Deane, both counsel and the Court were astute to the novelty of the issue presented and the need to resolve this by recourse to basic principles of tort law. Both were, in particular, attentive to the distinction between acts and omissions. For instance, at the outset of his judgment, Greene LJ rejected the broad proposition that a person could never be liable as a publisher of defamatory matter due to an omission. His Lordship then proceeded to examine the circumstances in which a person could be held liable for a failure to act with respect to a defamatory matter.
Properly understood, Byrne v Deane is concerned with establishing publication by omission. Because a mere failure to act with respect to a third party’s communication of defamatory matter cannot make the person so failing to act responsible as a publisher of that matter, the issue becomes under what conditions the person can be rendered liable. To make the Deanes potentially liable for the unknown poster’s defamatory doggerel required actual notice of its presence. The Deanes’ actual notice of the presence of the defamatory matter on the golf clubhouse walls was transformative of their legal liability. Prior to actual notice of the presence of the defamatory matter, the Deanes could not be held liable for the wrong committed by a third party. It was only after they had actual notice of its presence that it was possible for them to be held liable as publishers of the defamatory matter. Actual notice alone was not sufficient, though. Then, the Deanes also needed to have the power and capacity to remove the defamatory matter, but fail to do so within a reasonable time. The failure to act in that context would render the Deanes liable for the publication of the defamatory matter by a third party. This was because they continued the third party’s tort and thereby made it their own.
The continuation of a third party’s tort should be regarded as the principled basis for Byrne v Deane. This was the basis of tortious responsibility by which the Deanes became responsible for a third party’s tort. Byrne v Deane is sometimes treated as establishing a special category for landowners or property owners. The line of authority it originates is sometimes styled the ‘trespass cases’. This obscures what should be regarded as the proper juridical basis of Byrne v Deane. Only Greene LJ referred to the third party’s act here as a trespass. The reference to trespass here should not be taken in isolation. His Lordship, on more than one occasion in his judgment, referred to the Deanes as continuing the third party’s tort. This provides a principled basis for holding a defendant liable for a third party’s tort, consistent with basic principles of tort law. In doing so, it follows the position in relation to private nuisance. In private nuisance, the occupier of land can become liable for a nuisance created by a trespasser if the defendant: becomes aware of the nuisance; has the power and capacity to deal with the nuisance; and, within a reasonable period of time, fails to do so. In such circumstances, the defendant, as the occupier of the land, becomes liable for the third party’s tort because the defendant continues that wrong.
This analysis can apply equally to internet intermediaries. In his judgment in Oriental Press Group, Ribeiro PJ cogently analysed the defence of innocent dissemination and Byrne v Deane. His Honour ultimately distinguished Byrne v Deane from the case of an operator of an internet discussion forum on the basis that the Deanes were not in the business of soliciting communications, whereas the operator of the internet discussion forum was. With respect, this is not a satisfactory basis upon which to distinguish Byrne v Deane. The flaw in the reasoning is that it treats the particular defendant’s purpose in participating in the communication as relevant to whether the defendant is responsible for the publication. If the defendant’s purpose were relevant to deciding whether the defendant were a publisher, many defamation cases would have been decided differently. Defendants with a non-commercial purpose may have been able to avoid liability. Responsibility for publication turns upon the fact of the defendant’s participation in the communication of the defamatory matter, not the defendant’s purpose. As such, an analysis based on Byrne v Deane should be equally available in cases involving internet intermediaries.
In relation to publication, defamation law has tended to conflate the issues of the conduct constituting communication of the defamatory matter, on the one hand, and the basis of responsibility for that communication, on the other hand. Yet, where responsibility is sought to be imposed on a defendant for a communication originating from another person, the basis of responsibility is a distinct and important issue. Properly understood, the collection of principle by Isaacs J in Webb v Bloch is directed not to the issue of what constitutes publication, but the bases upon which a person who does not originate the communication may be held responsible for it. The references to assistance, conducing, concurring, assent, approval, authorisation, encouragement and inducement and accessorial liability are directed to the bases of extending responsibility to a person for a communication, not a prescription of what constitutes publication.
This conflation of the issues of what constitutes communication of the relevant matter and the basis upon which a given defendant could be held responsible for it may not have been particularly problematic when mass media publications predominated. Where a defendant is being sued for his or her own positive act, there is no real issue as to the conduct amounting to communication of the relevant matter or the basis of the defendant’s responsibility for it. However, the disaggregation of the steps involved in the dissemination of defamatory matter brought about internet technologies has meant that these issues arise more regularly. To deal with the difficult issues of publication, now more regularly posed by internet technologies, which flow as a consequence of this disaggregation, it may be prudent also to disaggregate the issues of what constitutes publication and who is responsible for the publication, so as to give distinct, proper and principled consideration to each of these issues.
In Voller, a fact that has not been given sufficient emphasis is that the proceedings were commenced without giving notice to the media outlets that there were allegedly defamatory third party comments posted on their public Facebook pages. Such an approach will not be possible once the reforms to the national, uniform defamation laws commence, as defamation proceedings will not be able to be instigated unless a concerns notice has been given to the prospective defendant. If the conduct constituting communication of the defamatory matter is not the positive act of subscribing to the public Facebook page, but the failure to deal with third party comments, then, in the absence of actual notice, the media outlets could not begin to be held liable for the wrongs of a third party. Analysing the facts as a publication by omission suggests that, if the media outlets had actual notice of the third party comments and, having the power and capacity to deal with those third party comments, they failed to do so within a reasonable time, then they could be held liable as publishers of those comments on the basis that they continued the torts of the third parties, thereby making them their own.
Internet technologies continue to proliferate and, in doing so, will continue to present difficult issues for defamation law, in particular, but not limited to, the element of publication. The most effective way of dealing with these complex questions is by recourse to basic principle. Internet technologies tend to expose how little consideration needed to be given to the concept of publication when mass media predominated. In such cases, proof of publication was usually uncontentious. Internet technologies compel reconsideration and clarification of basic principles of defamation law. The best explanation for the absence of detailed and sustained consideration of intention in the context of publication in defamation cases generally is that proof of intention is not required to establish the element of publication. Publication turns upon the fact of the communication of the matter to a person other than the plaintiff, not upon proof of intention to communicate the matter, and, as such, is a matter of strict liability. However, liability for publication can depend not only upon the fact of communication, but also upon notions of responsibility for that communication. By disaggregating the steps involved in the dissemination of defamatory matter, internet technologies expand the number of persons and entities who can be held responsible for publication of that matter. To think clearly about the issue of publication, distinct consideration should be given first to identifying the precise conduct alleged to constitute communication of the defamatory matter and then to articulating the precise basis upon which the particular defendant should be held liable for that communication.
* Professor, The University of Sydney Law School, The University of Sydney, Camperdown, NSW, Australia. Email: email@example.com.
 Dow Jones & Co Inc v Gutnick (2002) 210 CLR 575 (‘Gutnick’).
 Ibid 605 – (Gleeson CJ, McHugh, Gummow and Hayne JJ), 629–35 – (Kirby J) 649–50  (Callinan J).
 Kylie Pappalardo and Nicolas Suzor, ‘The Liability of Australian Online Intermediaries’ (2018) 40(4) Sydney Law Review 470, 472–3.
 Trkulja v Google LLC  HCA 25; (2018) 263 CLR 149, 163  (per curiam) (‘Trkulja v Google (HCA)’).
 Ibid 163–4  (per curiam).
 Fairfax Media Publications Pty Ltd v Voller; Nationwide News Pty Ltd v Voller; Australian News Channel Pty Ltd v Voller (High Court of Australia, Case Nos S236/2020, S237/2020, S238/2020) (‘Voller’).
 Royal Commission and Board of Inquiry into the Protection and Detention of Children in the Northern Territory (Final Report, November 2017) vols 1–4 <https://www.royalcommission.gov.au/
 Voller v Nationwide News Pty Ltd  NSWSC 766.
 Fairfax Media Publications Pty Ltd v Voller  NSWCA 102; (2020) 380 ALR 700 (‘Voller (NSWCA)’).
 Ibid 712 –.
 Oriental Press Group Ltd v Fevaworks Solutions Ltd  HKCFA 47; (2013) 16 HKCFAR 366 (‘Oriental Press Group’).
 Voller (NSWCA) (n 9) 712 .
 Ibid 720 .
 Ibid 720 –.
 Webb v Bloch  HCA 50; (1928) 41 CLR 331, 363–4.
 Voller (NSWCA) (n 9) 722 .
 Ibid 723 –.
 Ibid 710–11 –.
 Ibid 710–11  quoting Thompson v Australian Capital Television Pty Ltd  HCA 38; (1996) 186 CLR 574, 586 (‘Thompson’).
 See Civil Law (Wrongs) Act 2002 (ACT) s 139C; Defamation Act 2005 (NSW) s 32; Defamation Act 2006 (NT) s 29; Defamation Act 2005 (Qld) s 32; Defamation Act 2005 (SA) s 30; Defamation Act 2005 (Tas) s 32; Defamation Act 2005 (Vic) s 32; Defamation Act 2005 (WA) s 32.
 Trkulja v Google (HCA) (n 4) 164–5 – (per curiam).
 Voller (NSWCA) (n 9) 712–14 –.
 Ibid 720–21 –.
 Vizetelly v Mudie’s Select Library Ltd  UKLawRpKQB 79;  2 QB 170, 180 (Romer LJ) (‘Vizetelly’).
 Voller (NSWCA) (n 9) 721 .
 For recent judicial statements to this effect, see, eg, John Fairfax Publications Pty Ltd v Obeid  NSWCA 60; (2005) 64 NSWLR 485, 503 , 506  (McColl JA; Sheller JA and McClellan A-JA agreeing); Wookey v Quigley  WASC 284,  (Hasluck J); Ten Group Pty Ltd v Cornes (2012) 114 SASR 46, 88 , 92  (Blue J); Opai v Culpan  NZHC 1036;  NZAR 1142, 1160  (Katz J); Durie v Gardiner  NZHC 377;  3 NZLR 72, 84  (Mallon J); Cummings v Fairfax Digital Australia & New Zealand Pty Ltd  NSWCA 325; (2018) 99 NSWLR 173, 197  (McColl JA); Bailey v Bottrill (No 2) (2019) 14 ACTLR 108, 113  (McWilliam AsJ). Contra Murray v Wishart  NZCA 461;  3 NZLR 722, 751  (‘the well-understood nature of the tort of defamation as an intentional tort’).
 Voller (NSWCA) (n 9) 720 .
 Benning v Wong (1969) 122 CLR 249, 298–9 (Windeyer J). See also Peter Cane, The Anatomy of Tort Law (Hart Publishing, 1997) 45; Carolyn Sappideen and Prue Vines (eds), Fleming’s The Law of Torts (Lawbook, 10th ed, 2011) [1.50]; WE Peel and J Goudkamp, Winfield and Jolowicz on Tort (Sweet & Maxwell, 19th ed, 2014) [3-011].
 Lee v Wilson  HCA 60; (1934) 51 CLR 276, 286–95.
 E Hulton & Co v Jones  UKLawRpAC 57;  AC 20 (‘Hulton’).
 Ibid 23–4.
 Lee v Wilson (n 29) 288 (Dixon J).
 David Rolph, ‘The Ordinary, Reasonable Search Engine User and the Defamatory Capacity of Search Engine Results in Trkulja v Google Inc’  SydLawRw 24; (2017) 39(4) Sydney Law Review 601, 610–11.
 Gutnick (n 1) 600  citing Lee v Wilson (n 29) 288 (Dixon J).
 Godfrey v Demon Internet Ltd  QB 201, 204–5.
 Ibid 207.
 Oriental Press Group (n 11) 377 .
 Law Commission of Ontario, Defamation Law in the Internet Age (Final Report, March 2020) 80 (Recommendation 35) <https://www.lco-cdo.org/wp-content/uploads/2020/03/Defamation-Final-Report-Eng-FINAL-1.pdf>.
 Webb v Bloch (n 15) 363–4 (emphasis in original).
 Commonwealth v Bank of New South Wales  HCA 47; (1949) 79 CLR 497, 637–8 (Privy Council).
 As to the facts of this case, see Webb v Bloch (n 15) 335–43 (Starke J).
 For criticisms of the breadth of liability for publication in defamation law, see Crookes v Newton  3 SCR 269, 283  (Abella J), 295–6  (Deschamps J).
 Emmens v Pottle (1885) 16 QBD 354, 356.
 Vizetelly (n 24) 175. For an earlier example to similar effect, see Day v Bream  EngR 916; (1837) 2 M & Rob 54; (1837) 174 ER 212.
 Vizetelly (n 24) 178–9.
 Emmens v Pottle (n 43) 357.
 Ibid 357–8.
 Vizetelly (n 24) 175.
 Ibid 179.
 Ibid 180.
 Ibid 179.
 Ibid 180.
 Thompson (n 19).
 Ibid 586.
 Ibid 618.
 Huth v Huth  3 KB 32.
 McNichol v Grandy  1 DLR 225.
 For a further discussion of the impact of cases of accidental or unintentional publication on general principles of publication at common law, see David Rolph, ‘The Concept of Publication in Defamation Law’ (2021) 27(1) Torts Law Journal (forthcoming).
 Duffy v Google Inc  SASC 170; (2015) 125 SASR 437, 472–3 (Blue J); Google Inc v Duffy (2017) 129 SASR 304, 332–8; – (Kourakis CJ), 456–63; – (Hinton J).
 Trkulja v Google (HCA) (n 4) 164–5 –.
 Kit Barker, Peter Cane, Mark Lunney and Francis Trindade, The Law of Torts in Australia (Oxford University Press, 5th ed, 2012) [2.3.2].
 Fede v Gray  NSWCA 316; (2018) 98 NSWLR 1149, 1179–80  (Basten JA).
 Sappideen and Vines (n 28) [2.60]; RP Balkin and JLR Davis, The Law of Torts, (LexisNexis Butterworths, 5th ed, 2013) [3.2]–[3.3].
 Sappideen and Vines (n 28) [2.60].
 See, eg, Weaver v Ward (1616) Hob 134; (1616) 80 ER 284; Stanley v Powell  UKLawRpKQB 153;  1 QB 86, 88–94 (Denman J); Blacker v Waters  NSWStRp 46; (1928) 28 SR(NSW) 406, 410 (Street CJ); McHale v Watson (1964) 111 CLR 384, 386–88 (Windeyer J); Venning v Chin (1974) 10 SASR 299, 310–16 (Bray CJ); Platt v Nutt (1988) 12 NSWLR 231, 235–40 (Kirby P), 243–6 (Clarke JA); Croucher v Cachia  NSWCA 132; (2016) 95 NSWLR 117, 122–5 – (Leeming JA); New South Wales v Ouhammi  NSWCA 225; (2019) 101 NSWLR 160, 165–8 – (Basten JA), 173–88 – (Brereton JA), 202–3 – (Simpson AJA).
 Trkulja v Google LLC (No 5)  VSC 533.
 Bleyer v Google Inc  NSWSC 897; (2014) 88 NSWLR 670, 685 .
 Trkulja v Google LLC (No 5) (n 67) .
 Fairfax Media Publications Pty Ltd, Nationwide News Pty Ltd and Australian News Channel Pty Ltd, ‘Appellants’ Submissions’, Submission in Voller (n 6), 17 February 2021, .
 Murray v Wishart (n 26) 740 – (O’Regan P and Ellen France J).
 Smith v Leurs  HCA 27; (1945) 70 CLR 256, 261–2. For further judicial statements to similar effect, see Weld-Blundell v Stephens  AC 956, 986 (Lord Sumner); McHale v Watson (n 66) 386–7 (Windeyer J); Jaensch v Coffey  HCA 52; (1984) 155 CLR 549, 578 (Deane J); P Perl (Exporters) Ltd v Camden London Borough Council  EWCA Civ 9;  QB 342, 359–60 (Robert Goff LJ).
 See, eg, Dorset Yacht Co Ltd v Home Office  UKHL 2;  AC 1004, 1060 (Lord Diplock); Smith v Littlewoods Organisation Ltd  UKHL 3;  AC 241, 271 (Lord Goff); Sutherland Shire Council v Heyman  HCA 41; (1985) 157 CLR 424, 477–9 (Brennan J); Michael v Chief Constable of South Wales  UKSC 2;  1 AC 1732, 1760 (Lord Toulson JSC).
 Byrne v Deane  1 KB 818. See generally Barbara McDonald, ‘Byrne v Deane (1937)’ in David Rolph (ed), Landmark Cases in Defamation Law (Hart Publishing, 2019) 59–80.
 Byrne v Deane (n 75) 837.
 Google Inc v Duffy (n 60) 341–3 – (Kourakis CJ), 461  (Hinton J). See also Stoltenberg v Bolton  NSWCA 45; (2020) 380 ALR 145, 189–90 – (Gleeson JA) (‘property cases’).
 Byrne v Deane (n 75) 837.
 See, eg, ibid 838.
 See, eg, Saxby v Manchester, Sheffield, and Lincolnshire Railway Co  UKLawRpCP 5; (1869) LR 4 CP 198; Barker v Herbert  UKLawRpKQB 119;  2 KB 633, 636–7 (Vaughan Williams LJ), 642–3 (Fletcher Moulton LJ), 645 (Farwell LJ); Job Edwards Ltd v Birmingham Navigations Proprietors  1 KB 341, 360 (Scrutton LJ); Noble v Harrison  2 KB 332, 338 (Rowlatt J); Sedleigh-Denfield v O’Callaghan  UKHL 2;  AC 880, 893–4 (Viscount Maugham), 897, 899 (Lord Atkin), 904–5 (Lord Wright), 912–13 (Lord Romer), 919 (Lord Porter); Wringe v Cohen  1 KB 229, 233; Cushing v Peter Walker & Son (Warrington & Burton) Ltd  2 All ER 693, 702 (Hallett J); Fennell v Robson Excavations Pty Ltd  2 NSWLR 486, 492 (Glass JA), 495 (Samuels JA); Montana Hotels Pty Ltd v Fasson Pty Ltd (1986) 69 ALR 258, 261–2 (Privy Council); City of Richmond v Scantelbury  VicRp 44;  2 VR 38, 40 (Kaye J); Owners of Strata Plan No 13218 v Woollahra Municipal Council  NSWCA 92; (2002) 121 LGERA 117, 136 (Young CJ in Eq). As to the definition of ‘continuation’ and its differentiation from other bases of tortious liability, such as adoption, see Sedleigh-Denfield v O’Callaghan  UKHL 2;  AC 880, 894 (Viscount Maugham).
 Oriental Press Group (n 11) 387.
 Webb v Bloch (n 15) 362–6 (Isaacs J).
 This emerges clearly from the chronology filed in the High Court appeal: Fairfax Media Publications Pty Ltd, Nationwide News Pty Ltd and Australian News Channel Pty Ltd, ‘Appellants’ Chronology’, Submission in Voller (n 6), 17 February 2021.
 This is the effect of the new s 12B of the Model Defamation Provisions: see Parliamentary Counsel’s Committee, Model Defamation Provisions (27 July 2020) <https://www.pcc.gov.au/uniform/2020/