Article body


“We live in an age drenched in data.”[1]

“On the Internet, we constantly live in a twilight between fact and fiction.”[2]

Over a decade ago, Justice Blair of the Court of Appeal for Ontario first posed and then answered a significant inquiry, in which he said:

Is there something about defamation on the Internet—“cyber libel”, as it is sometimes called—that distinguishes it, for purposes of damages, from defamation in another medium? My response to that question is “Yes.”

Communication via the Internet is instantaneous, seamless, interactive, blunt, borderless and far-reaching. It is also impersonal, and the anonymous nature of such communications may itself create a greater risk that the defamatory remarks are believed.[3]

This relatively recent predicament, extensively dealt with by Quebec courts of late, both summarizes and foretells this paper’s proposal to reorient Ontario cyber defamation analysis towards a civilian approach, whose hallmark flexibility and adaptability lends itself particularly well to the digital age. Indeed, as further discussed herein, harnessing the ordinary rules of negligence, and—in principle—foregoing defences, the civilian construction is chiefly interested in the contextual reasonableness of the impugned expression (rather than in its truth or falsity strictly speaking), in contradistinction to its somewhat categorical common law counterpart.[4] It is therefore recommended that defamation law evolve towards a “negligence standard” in common law parlance. This evolution in common law would require the plaintiff to make a showing of the contextual unreasonableness of impugned speech, an analysis which subsumes truthfulness and obviates the need for defences.

Such is the case in Quebec where, per article 1457 of the Civil Code of Québec’s (CCQ) three-pronged approach, statements deemed unreasonably expressed under the circumstances give rise to liability if they are found to have caused reputational harm.[5] What matters in this analysis is not the truthfulness of the statements but rather the unreasonable character of the expression under the circumstances and the injury that the statement causes. Furthermore, exemplary damages are available (and often awarded in the Internet context) by virtue of the Quebec Charter of Human Rights and Freedoms[6] through article 49(2) when the defamatory comments are found to have been made “intentionally” (a term generously interpreted to often include recklessly).[7] As courts struggle to strike a balance between protecting privacy and reputation, made fragile in the cyber era, with the value of freedom of expression protected in both the Canadian Charter and the Quebec Charter, the Quebec approach to defamation offers a legal framework that can better balance the Internet’s proclivity to exacerbate the reputational harms of statements made online.[8]

Compounding the importance of revisiting the matter of defamation online, a cyber-publication in one jurisdiction may be read and subsequently reposted anywhere in the world, thereby potentially causing reputational harm transcending traditional or national parameters. Canadian courts have grappled with the challenge of enforcing local rights infringed by conduct flowing from outside of Canada, wary of losing the ability to enforce local norms and policy by rectifying harm felt in Canada but originating elsewhere. The courts’ preoccupation with the potential for judicial helplessness in Internet cases is evidenced by the notably liberalized jurisdiction test in v. Goldhar[9] building on Breeden v. Black[10]. The Court endeavors to inject flexibility into the existing test and to adapt the steps set out to the complexity of an increasingly multi-jurisdictional landscape.[11] It is therefore essential to at least summarily touch upon the jurisdiction question, alongside that of intermediaries, if we are to have a true contextual understanding of cyber defamation as recommended herein.

Cognizant of the jurisdictional challenges faced in the information age, the objective is to carve a path forward for Ontario law. This paper offers contextual analysis as a basis for advancing the analytical framework in Ontario, modeled on the civilian approach in Quebec. Quebec courts may also be said to harness the principles of defamation law to hold third parties, most notably “data controllers” (search engines and the like) accountable, thereby pre-empting the weakening of local defamation law[12] in the borderless, “anonymous” digital age. This paper shines a comparative light on the civilian view with an eye towards informing the development of Ontario defamation law.

I. Part I: Introduction

A. Evolving Visions of Reputational Privacy and Expression in the Digital Age

Prior to proceeding to a more in-depth discussion of Quebec defamation law and its distinct pertinence to advancing the Ontario framework, it is necessary to first touch on the very nature of the social and technological change upon us. We are witnessing a culture of instantaneous sharing, enabling anyone to communicate random thoughts, potentially with everyone, literally worldwide, without the ability to correct, retract, control or contextualize subsequent dissemination, and they may further—to a certain degree—do so anonymously. Cognizant of these medium shifts, the Supreme Court of Canada has laudably evolved its approach to digital privacy as a means for protecting reputation, commonly predicated on the dignity and control theory discussed below, particularly in civilian thought.[13] It therefore stands to reason that defamation or “cyber libel” should similarly be revisited in light of the above-cited evolution in Canadian case law, as it pertains to “digital privacy” or innovation law more generally.[14]

Namely, the technological developments brought about by the Internet changed the circumstances of expression and the context in which defamation can occur. Social media is characterized by a “general lack of error correction, gate keeping (and its absence)”[15] and is conducted within a broader medium (the Internet) that rarely forgets or contextualizes. This, needless to say, is most salient for the purposes of revisiting the correct balance between freedom of expression and reputation/privacy in the digital age.

In distinction to the institutional press, which by most accounts boasts built-in safeguards (editorial oversight, and a civil if not purportedly neutral tone) bloggers and tweeters can share information free of intermediaries to a certain extent and in a manner that serves to radically compound the difficulties related to traditional defamation. Difficulties like punitive shaming, for example, may ultimately force us to recalibrate the balance reached between free speech, known as “very life blood of our freedom and free institutions”[16] and reputational considerations.

The ability of sheer vilification and distortions of information to instantaneously reach and mislead even the most educated is amplified by the lack of editorial oversight online.[17] Accordingly, a notable British Columbia case on defamatory statements on social media recognized the pernicious nature of the medium, opining that “the nature of Facebook as a social media platform and its structure mean that anyone posting remarks to a page must appreciate that some degree of dissemination at least, and possibly widespread dissemination, may follow.”[18] In light of the above, insisting on a multi-tiered approach as the common law does today may be unduly onerous and may inadvertently undermine the general objectives of defamation law in the internet age: shielding reputation.

B. A Notable Shift in Canadian Law: Privacy as Means for Protecting Reputation

Privacy, particularly as a means for safeguarding personal dignity, ultimately seeks to give people greater control over their reputations.[19] This control over the tidbits of decontextualized information that ultimately shape (or distort) reputation is of the essence in the age of Internet which, as leading privacy theorist Daniel Solove explains, “makes gossip a permanent reputational stain, one that never fades.”[20]

For defamation law, with its focus on reputation, this signifies a shift from a binary understanding of truthful and untruthful speech and presents an opportunity for reform. On the one hand, decontextualized “truths,” widely available through the Internet, can be just as destructive to one’s reputation as outright lies. Therefore, an emphasis on the “truth” or falsity of impugned communications is counterproductive, sanctioning a distorted view of an individual resulting from the decontextualization of tidbits of personal information. On the other hand, focusing on truth as a defence, rather than a mere element in the contextual analysis of reasonableness, places a disproportionate burden on defendant and unduly frustrates freedom of expression. Demonstrating that a post is “objectively true” or verifiable as a defence places a high bar on defendant. Assessing truth as part of reasonableness, as the civilian tradition endeavours to do, infringes less on freedom of expression by enquiring into the unreasonableness of the expression, rather than falsity. It further accounts for the Internet’s tendency to decontextualize, and therefore distort, truthful comments.

Privacy, as first recognized in the seminal piece by Samuel L. Warren and Justice Louis D. Brandeis, was born of the need to safeguard reputation in view of what were then new technologies, which speaks to the former’s importance let alone relevance to the latter.[21] What is more, the civilian/continental construction of privacy is itself predicated on human dignity,[22] a fundamental rationale underlying the protection of reputation.[23]

But the change in circumstances has caused—and continues to cause—considerable quandaries in terms of reconciling modern day, cross-jurisdictional problems with an outdated legal structure in place. For, as one commentator stated,

[t]he Internet is, at its core, a medium of instantaneous, long-distance communication. It makes communicating with a thousand, or a million, people no more difficult than communicating with a single person. For the first time, it brings mass communication to the masses: anyone with a computer and an Internet connection can utilize its potential. It facilitates communication in any combination of writing, sounds, and pictures. It knows no geographical boundaries: any Internet user can communicate globally, with a potentially limitless audience.

While other media of communication may have some of these qualities, their confluence in the Internet is unique. The Internet represents a communications revolution.[24]

Thankfully, AB v. Bragg, a unanimous landmark Supreme Court of Canada decision rendered in the fall of 2012, appears to herald a remarkable transformation, marking an important progression in both judicial attitudes and understanding of privacy harm. Moreover, Canada’s highest court appears to be taking serious notice of the indisputable impact that the Internet has had and continues to have on various areas of law, including defamation (or commonly called “cyber defamation”),[25] recognizing anonymity as a component of privacy.[26]

In effect, the Supreme Court of Canada appears primed for the first time to revisit prevailing concepts and approaches and, perhaps, reverse its own long-standing case law.[27] In devising a framework that will accommodate the vicissitudes of the digital age, and cross-border reputational harm in particular, the Court seems to echo the increasingly prevalent understanding that the common law should develop in a manner that is consistent with context—namely, the vicissitudes of the digital age.[28]

The following posits that, mindful of Canadian Charter values, Canadian courts should look to the civilian experience towards evolving defamation beyond common law defences, as one Saskatchewan court has arguably already done in Whatcott v. Canadian Broadcasting Corporation.[29]

Defamation law, positioned to protect reputation, is carved as an exception to freedom of expression and depends chiefly on perception.[30] With that in mind, Whatcott may be a source of comfort for Ontario courts, since the Saskatchewan court adopted a civilian methodology while attending to concerns about freedom of expression. It chose to focus contextually on whether the impugned communication created a false impression of the plaintiff’s views, rather than insisting on whether the words were actually false, recognizing that the latter approach would not properly satisfy the underlying rationale of defamation law, that of protecting reputation.

Although not a cyberdefamation case, the matter attests to the dangers of the decontextualization of facts, even when these facts, narrowly speaking, are not necessarily false. In his claim, Whatcott argued that he was defamed by a video clip aired by the CBC:

the plaintiff here complains about the defendant’s depiction of words that he authored in a flyer which he also printed and distributed. The essence of his argument is that the manner and context in which his words were presented seriously distorted and misrepresented his views, thereby giving the words a defamatory meaning.[31]

He attacked the prominence of certain clips—snippets of his flyers which, although accurate (he did not contest that the flyers pictured were indeed his), were depicted without referencing the disclaimer that followed on the second page of the flyer. This omission decontextualized the emphasized words, creating a false impression of the plaintiff and his message. As the court observed:

The plaintiff’s complaint focuses on that portion of the broadcast, near the beginning, where the defendant’s camera depicted the first page of the Alberta flyer. In his statement of claim, the plaintiff contends that by depicting the Alberta flyer in the manner it did, without referencing the disclaimer on the second page, the defendant conveyed the impression that the plaintiff advocated the killing of homosexual people, which was the opposite of the meaning he actually conveyed in the flyer, when read as a whole.[32]

Underscoring the danger of decontextualization a fortiori in the digital age, Justice Elson opined:

In my view, by focusing the camera’s attention on the phrase “kill the homosexual”, as it appeared in the Alberta flyer, and doing so early on in the broadcast, the defendant conveyed the impression that the plaintiff’s activism was considerably more extreme than it actually was. Indeed, it conveyed the impression that the plaintiff’s views extended to inciting violence against homosexual people. Despite the fact that the focus on these words was no more than five seconds long, I am satisfied that it was long enough to have injured the plaintiff’s reputation in the estimation of reasonable viewers. While the rest of the broadcast did nothing to support or reinforce this impression, it also did nothing to reduce it or to diminish the injury.[33] 

Consequently, and unknowingly applying a civilian framework of analysis, the court concluded that “selective editing” of otherwise accurate (or truthful) information may be deemed defamatory if decontextualized to create a false (harmful) and defamatory impression.[34]

C. Laying the Foundation

1. Understanding Reputational Privacy Conceptually

Digital privacy is not merely about seclusion. On the contrary, exposure seems to be the default, be it via social networking, YouTube videos, or Twitter.[35] It is not that we do not wish to be known or seen, but rather than we expect to be seen as we portray ourselves when we set out to bare our identities online. Where the idea is to share personal information—in the cyber world as in the “real” world—the intention is, not surprisingly, to expose what one considers an accurate rendering of oneself (whether it is precise or not). One might say that those sharing personal information online may wish to preserve control over their ability to represent themselves to the world.[36]

In most cases, people do not fear revealing even very personal information. Rather, they fear the often irreparable distortion and deformation of this information and its arguably indelible impact on reputation in the digital age.[37] Indeed, as Paul Schwartz observes, “[t]he weight of the consensus about the centrality of privacy-control is staggering.”[38] As Fairfield buttresses, privacy as control “has emerged as a dominant theory of informational privacy, in part because it promises individuals (rightly or wrongly) the ability to both disclose and control dissemination of information online” and “[m]odern privacy approaches have developed and intensified the emphasis on individual notice, choice, and control over information flows.”[39]

In other words, instead of isolation, people covet and what sociologist Erving Goffman labelled “impression management.”[40] According to Goffman, most people deploy significant efforts to control or manage their identity (or the perception thereof) through what he called “presentation of self.” Offline that is achieved by way of personal style, dress, body language and “the revealing and withholding of personal information to convey to the world who they are, or who they want to be taken to be.”[41]

This is not so in cyberspace. Accuracy, especially that relating to identity, is significantly contextual in a fragmented, inherently decontextualized networked environment. In cyberspace, depending on algorithm results, even otherwise exact information can easily convey a most misleading impression. Worse still, search results may yield maliciously stage-managed data that is otherwise “accurate.” Similarly, time-tested truths may be presented alongside blatant falsehoods to the point of being indistinguishable from one another.[42]

Given the nature of the digital environment, the end-result might well be to bring individuals into disrepute — not for a finite period or in a manner that might be corrected with reasonable effort. Worse still, an unassailable version of one’s identity, entirely incompatible with one’s one truth (or perhaps even “the truth”), might emerge and become entrenched as public record, upon which future thought is built. This entrenchment in the public record reflects what Goffman calls “virtual and actual social identity.”[43] An individual might thus, in Goffman’s words (discussing stigma more generally), be “reduced in our minds from a whole and usual person to a tainted, discounted one.”[44] What is more, identity, which Michel Foucault presented as a flexible construct,[45] is no longer (or is certainly less) malleable when perceived by others, as we become trapped in our deeds or even in the self-presentation of years past.[46] The capacity to reinvent oneself is therefore presumably either lost or severely compromised. Identity and its potential evolution is frozen in time, decontextualized, or in Goffman’s parlance “spoiled.”[47]

D. A Purposive Construction of Defamation in the Digital Age Informed by Comparative Approaches

As noted, defamation law endeavours to protect reputation and to reach a fragile equilibrium between two constitutionally protected values: freedom of expression on the one hand and privacy (and dignity) on the other, broadly speaking. As the Supreme Court of Canada clarified in Prud’homme: “freedom of speech is not absolute. It is limited by, inter alia, the requirements imposed by other people’s right to the protection of their reputation.”[48] Consequently, the rules of civil liability that apply to cases of alleged defamation act, in the Court’s words, as a “safety valve” to stop those who would take advantage of, and abuse, freedom of expression in order to defile another’s reputation, itself a fundamental attribute of personality rights per the civilian perspective.

Recent years have seen a convergence between legal systems, thereby rendering comparative analysis more attractive than ever. Thus, as Justice Dov Levin, formerly of the Supreme Court of Israel, observes:

systems faithful to the Common Law tradition, based on the adversarial system, appear less “adversarial” then they were, whereas the supposedly inquisitorial Civil Law systems have lost some of their “inquisitorial” character. Judges too travel more and more, thereby allowing them to interact with their counterparts internationally and familiarize themselves with judicial training methods used abroad- models, which often differ from country to country. More and more, judges communicate via judicial exchange programs. Moreover, international conferences are becoming increasingly prevalent, particularly regionally, with the objective of advancing and improving judicial training techniques and achieving a certain degree of harmony between the various approaches.[49]

Cross-pollination of systems, or—at the very least—comparative analysis, is natural here in Canada, home to a meeting of two great legal traditions in the western world.[50] Whereas Quebec civil law was traditionally heavily influenced by its common law counterpart, the reverse is increasingly—however incrementally and implicitly—true in Canada, as evidenced by a number of Supreme Court cases citing the former towards evolving the latter.[51]

Moreover, it is helpful to have Canadian Charter rights conceived uniformly nationwide, all while maintaining the particularities of each tradition, as Justice Dechamps opined in Crookes v Newton: “[i]n order to give guidance, I would prefer to outline a rule that is consistent with the common law and the civil law of defamation and that will also accommodate future developments in Internet law.”[52]

Coherence and flexibility are of the essence and the civilian approach has always prided itself on malleable, adaptable principles rather than pigeonholed categories.[53] It is therefore all the more fitting for Canadian jurisdictions such as Ontario to seek out inspiration from sister provinces for defamation renewal. Particularly instructive is the civil law view of reasonable truthfulness as “the Internet has a distinctive cause instantaneous, and irreparable, damage to the business reputation of an individual or corporation by reason of its interactive and globally all-pervasive nature and the characteristics of Internet communications”.[54]

II. The Civil Law View

A. A Word on the Civilian Understanding of Reputational Privacy

Broadly speaking, civilian privacy generally and reputational privacy as an aspect thereof are matters of affirmative rights, and consist of two parts. First, privacy can be conceived of as the right to engage in individual self-definition and self-invention, rather than a right to be secluded or free from surveillance. Second, adopting civilian parlance, which correlates rights with duties, privacy is also the responsibility not to unnecessarily compromise one’s own information in the naïve hope that the information will not be misused.

Furthermore, in the civilian tradition, privacy is considered to be a “personality right,” a concept alien to the common law.[55] Therefore, in civil law jurisdictions, privacy attaches to persons rather than property, irrespective of property or special constraints.[56] In other words, “[p]ersonality rights focus on the être—the being—in contrast with the avoir—the having” and are significantly divorced from territory”.[57] Privacy, as a personality right, is predicated on dignity and control of one’s identity.[58] For example, article 2 of the Basic Law for the Federal Republic of Germany provides that: “everyone shall have the right to the free development of his personality insofar as he does not violate the rights of others or offend against the constitutional order or against morality.”[59] In the privacy context, the concept of dignity in Germany is encompassed within “the right to the free unfolding of personality.”[60] In Quebec, personality rights are enshrined in articles 3 to 5 and 35 to 41 of the CCQ, which explicitly recognizes the right to reputational privacy and enumerates examples of violations.[61] Personality rights are also reflected in articles 4 and 5 of the Quebec Charter.

While very important differences exist between the approaches discussed above, conceiving of the right to privacy as a personality right, free of spatial or property constraints, generally allows the civilian legal method to grasp privacy as a zone of intimacy delineated by the basic needs of personhood, rather than by space or ownership. In effect, “personality allows one to define oneself in relation to society”[62] and can, therefore, be a very important “impression management” tool in the Internet age due to the difficulties set out herein. As Resta observes more generally:

In Continental Europe, by contrast [to the common law world], the evolution [of privacy] has been different. Instead of breaking up the traditional category of personality rights, courts have resorted to techniques of dynamic interpretation to adapt old provisions on name, image and privacy rights to changing social and economic landscapes. They have favoured, in other words, a functional evolution (Funktionswandlung) of the category of personality rights, rather than a radical paradigm shift, like the one implied in the recognition of a full-scale intellectual property right in one’s own identity. It should be underlined that this development has been feasible only because the continental law of personality has—from the very beginning—maintained a deeper and more ambiguous connection with the universe of property rights than a Warren’s style right to privacy... At stake in these cases was the value of autonomy, which lies at the core of the continental system of personality protection.[63]

Returning to above-referenced duties, as Popovici opines: “personality rights, as subjective rights, comprise both an active and corresponding passive side. The active side is the ‘power’ of the right’s holder over the object of the right; the passive side is the ‘duty’ of others to respect this very same object.”[64] This becomes particularly relevant in Part III, discussing intermediary liability.

B. From the General to the Specific

With particular regard to defamation, the Quebec “hybrid” approach, anchored in French law but heavily influenced by its common law counterparts, may be summarized thusly: defamation, or diffamation, in Quebec, as in the civilian tradition more broadly, is subsumed within the general realm of civil liability under the “catch all” negligence provision, article 1457 CCQ. Diffamation is therefore not a separate civil delict with distinct rules and defences as in the common law tradition, where headings are compartmentalized.

Article 1457 lays out a tripartite requirement of fault, injury and causation that must be proven by the plaintiff on a balance of probabilities. The standard for fault is that of the “reasonable person” assessed contextually. Article 1457 imposes a duty of reasonable conduct onto every person, thereby doing away with any need for classic common law “duty analysis” (i.e., determining whether a duty was owed by the defendant to the specific plaintiff). Instead, the breach of this general overarching duty to act reasonably, assessed by an objective standard, applies to all. The breach (either through commission or omission) thereof constitutes a fault, which if causally linked to injury (pecuniary and/or non-pecuniary) gives rise to damages. As a general rule, article 1457 CCQ provides that the tortfeasor is held to repair all “bodily, material and moral” injury attributable to him. Finally, as the proximity requirement is alien to the civil law and a broad duty of care is owed to all, rather than a specific “neighbour,” recognizing a broad category of potential claimants appears to pose few difficulties.

Within this general mechanism of civil responsibility, defamation “constitutes one of the most frequent and most serious breaches of the obligation to respect the reputation of others.”[65] Because reputation is also protected by the Quebec Charter, infringing thereupon may entitle the causally injured plaintiff to both ordinary and punitive damages under section 49 of the Quebec Charter (in combination with article 1621 CCQ).[66] What is more, under Quebec civil law, unlike its common law counterparts, these damages may be awarded in a broader range of circumstances: the question is not whether the defamatory statements are true or false but rather whether expressing them under the circumstances would unreasonably bring the plaintiff into disrepute according to an objective standard.[67] Accordingly, for example, it may be unreasonable to report an extramarital affair of a colleague on a social network, even if it did in fact occur, given the disastrous potential associated with such an announcement, unless circumstances dictate otherwise.

It bears repeating, truth is not a determining factor.[68] Perhaps surprisingly to common law readers, a lie, reasonably told in the circumstances, does not meet the defamation threshold while truth, unreasonably disclosed publicly, thereby causing harm to reputation, does meet the threshold. Thus, for instance, in Piquemal c. Cassivi-Lefebvre the Quebec Court of Appeal reiterates:

The impugned facts can be false or true. If they are false, the author is liable; if, on the other hand, they are true, the author may similarly be held liable if he did not have a duty or a serious, legitimate interest in reporting/disclosing these facts. I add that, if the author has a duty or a serious and legitimate interest in reporting the facts, whether he felt any degree of satisfaction in so doing is irrelevant [for purposes of determining fault].[69]

So too in Laforest c. Collins,[70] where the Superior Court of Quebec confirmed that the mere fact that impugned statements are believed to be true does not mitigate liability for comments expressed unreasonably and thereby causing reputational harm.[71]

C. Context

What is determinative, therefore, is neither truth nor public interest as independent determining elements, but rather context. Context is important to strike the proper balance between freedom of expression and the protection of reputation (under the overarching Charter values of human dignity and privacy as above noted[72]) in the Internet information context. We are, as Solove sets out, “drenched in data” and the “digital scarlet letter”[73] can forever follow the subject of defamatory expression.

In his edifying critique, Robert Danay faults common law defamation for “repeatedly equating” the Internet with “traditional broadcast media” notwithstanding significant differences and accordingly failing to strike the correct balance between the above-cited values.[74] While a full discussion of this critique exceeds the scope of this endeavour, Danay further impugns the outdated approach to defences observing that

the medium of an allegedly defamatory communication has always been—and indeed continues to be one of the most important determinants of how a plaintiff in a common law defamation action will fare in meeting the requisite elements of the tort, fending off any defences that the defendant might raise, and collecting a significant damage award at the close of proceedings...

Canadian courts have seemed to view their role in cyber-libel actions as that of a final bulwark against the defamatory excesses of those members of the general public who might abuse the tremendous new power entrusted to them by the Internet. This has led the courts to hold that when defamatory words are transmitted using the Internet, the availability of any qualified privilege that would otherwise have immunized the defendant from liability under traditional defamation principles will be vitiated, and will substantially increase any resulting award of damages...

By treating the vast and diverse world of Internet communications as an undifferentiated and uniformly menacing whole, the courts improperly favour plaintiffs in most cyber-libel cases to the detriment of vibrant online free expression, and devalue individual dignity and reputation in cases involving other, less-feared media.[75]

It naturally follows that context is key. Accordingly, in Prud’homme, Justices L’Heureux-Dubé and Lebel remark:

[t]his is not to say that whether the statements were true or of public interest is irrelevant. Truth and public interest are factors that are taken into account when engaging in the global contextual evaluation of fault under the CCQ. They are but pertinent examples in the puzzle but do not necessarily play a determining role in all circumstances.[76]

Unlike the common law, Quebec civil law does not, strictly speaking, recognize any defences to defamation, other than the absence of fault or causation. This is a relatively recent state of affairs following many years of confusion due to the historically prevalent common law influence on Quebec civil law, as common law defences were routinely—albeit erroneously—raised (and often times accepted) by Quebec courts. Since Prudhomme and Néron, courts have clarified the distinct approach, and the practical result of this distinction is as follows: the Quebec approach may at times consider defences of common law ilk as “part of the puzzle”, though not in a determinant manner, and lends itself particularly well to digital age. This is because, as the above noted examples illustrate, when truthful comments are decontextualized, misread, and widely disseminated via the Internet, they can be just as harmful as lies.[77]

1. An Affirmative Duty to Take Reasonable Precautions

It may further be said that in the Internet age, a reasonable person, cognizant of the “volatile” nature of social media (including but not limited to the potential of instantaneous worldwide dissimilation with difficult recall), must take reasonable precautions to avoid forever tarnishing another’s reputation lest he be held liable for defamation (negligent breach the general duty enshrined in 1457).[78] In consequence, civilian reasoning foresees that a reasonable person would not in most circumstances disseminate even true information that is expected to cause reputational harm in the delicate Internet context. Needless to say, he who purposefully sets out to tarnish another’s reputation will a fortiori be judged unreasonable and indeed held liable should this conduct be found to have directly engendered reputational harm.[79] For purposes of determining fault, the reputational impact—unlike the conduct itself—is assessed from plaintiff’s perspective, once fault (of omission or commission) and causal link have been established.

D. Damages

While a thorough discussion of damages is beyond the scope of this paper, a few words in summary are helpful. The Quebec approach to damages in the defamation context is relatively similar to the Common Law’s, save the willingness to compensate purely emotional harm and any third party who satisfies the tripartite inquiry, irrespective of remoteness. In terms of assessing damages, Quebec courts proffer a number of recurring criteria, namely:

  1. the intrinsic severity of the defamatory act/ statement

  2. its particular scope in relation to the victim (subjective element in terms of impact);

  3. the extent/ significance of its publication;

  4. the sort of individuals who gained knowledge thereof and the impact that these defamatory statements may have had on their impression of the plaintiff;

  5. The degree of lowering in esteem as compared to prior status;

  6. The reasonably foreseeable eventual and projected duration of the damages and disrepute caused;

  7. Duty to mitigate: the victim’s contribution to their own damages;

  8. The external circumstances that may have independently probably caused damages, independent of the alleged fault, thereby contributing independently at least in part to the damages.[80]

III. Matters of Falsity and Context/Doing Away with “Unreasonable Truth” in the Digital Age? A Recommendation

Mark Twain is reputed (perhaps incorrectly) to have observed that “a lie can make its way halfway around the world before the truth has a chance to put its boots on.”[81] This statement rings true in the digital age when:

[c]ommunication via the Internet is instantaneous, seamless, inter-active, blunt, borderless and far-reaching. It is also impersonal, and the anonymous nature of such communications may itself create a greater risk that the defamatory remarks are believed...[82]

And yet, common law defamation analysis “is [still] premised on falsity...It therefore bears only a passing relationship to an individual’s interest in preserving his private life, or more specifically, in keeping certain true aspects of his life within the private realm.”[83] As Kary explained years ago, incredulously to civilian ears:

[t]ruth is an absolute defence to a defamation claim in the common law. If the defendant can prove that the statement was true, the lawsuit fails. Justifiable error or innocent mistakes, however, are not excuses. No matter how reasonable it may have been to believe in the truth of the statement, one will be liable if it turns out to have been false.[84]

That is problematic, since “English defamation can justifiably be said no longer to protect reputation in the wider sense of fama, i.e. whether justified or not, but rather deserved reputation, sometimes also called ‘reputation founded in character.’” In turn, this evolution transformed the common law of defamation into something quite different from Roman iniuria. The focus shifted from the protection against insulting or outrageous words to something akin to a right not to be lied about.[85]

Perhaps the time has come to give Cyber libel “special status” as in the civil law, whereby truth is not a determinant element but rather a “piece” in the proverbial puzzle (as above stated), assessed contextually.[86] In a word, “responsible communication” as a standard of conduct rather than as a mere defence. Accordingly, the burden firmly affixes to the plaintiff to show negligence (omission or commission) on a balance of probabilities, rather than compelling the defendant to assert reasonableness in her defence, possibly an undue burden on Canadian Charter rights.

A. A Word on Anonymity and Defamation

In R v. Spencer, the Supreme Court of Canada, under the pen of Justice Cromwell, explicitly recognized anonymity as a necessary component of modern Charter privacy for the first time.[87] Although handed down in the criminal law context, it stands to reason that this landmark recognition builds on AB v. Bragg, where the Court came alive to the importance of protecting anonymity in certain circumstances.[88]

Perhaps reflecting the aforementioned need to restore context to the digital realm, Justice Cromwell underscored that the analysis pertaining to informational privacy must address the totality of the circumstances. Consequently, while “anonymity may, depending on the totality of the circumstances, be the foundation of a privacy interest that engages constitutional protection,”[89] given its relevance to personal growth and dignity, it might also do the exact opposite.

Cognizant of these complications and of their potential to undermine justice and render the domestic regime ineffectual, Justice Rochon offered in Rawdon that “[n]ot all forms of expression merit the same protection.”[90] Rather, he said that exceptional remedies aimed at unmasking anonymous defendants need to be carefully tailored or “chiseled” to the circumstances.[91]

As commentator Michaël Poutré observes “it is the combination of anonymous insults which seems most problematic.”[92] This is a classic scenario in the defamation context. In consequence, Poutré proposes that we only address the impact of anonymity after having established fault (the defamatory character of the expression). He cautions, anonymity per se cannot constitute or indicate fault. Accordingly, he puts forward: “it is best to broadly interpret the protection favoring anonymity and limit its scope thereafter...anonymity can subsequently be sanctioned when it is misused.”[93]

That approach, involving an independent assessment helping to distinguish defamatory speech as an exception to the general rule protecting freedom of expression, as discussed below, comports with addressing the serious issues related to the thorny implications for members of vulnerable populations in particular (for example historically targeted minorities, political objectors, LGBTQ community members, etc.).[94]

Whereas a more fulsome discussion of the significant issues attaching to and flowing from emerging anonymity rights far exceeds the scope of this modest endeavour, cyber defamation, along many related areas of “Internet law” cry out for a robust exchange on point.

B. Intermediaries: Affirmative Duties as Corollaries to Rights

Whereas the same robust exchange is imperative as it pertains to the role of intermediaries and jurisdictions, the scope of this present endeavour permits us only to say this: it is incumbent upon us to reassess third-party liability in the defamation context in light of worldwide access to Internet communications.

Suffice it merely to note that, broadly speaking, civilian privacy is a matter of affirmative rights, and consists of two parts. First, privacy, a personality right, can be conceived as the right to engage in individual self-definition and self-invention, rather than a right to be secluded or free from surveillance.[95] Second, adopting civilian parlance, which correlates rights with duties, privacy is also the responsibility not to unreasonably compromise one’s own information or that of another in the naïve hope that the information will not be misused.

For instance, in Corriveau c. Canoe inc.[96], the Superior Court of Quebec analogized an online news source to a broadcaster, holding it responsible for its failure to remove defamatory comments that appeared on an independently-run blog it hosted. The defamatory comments were posted on the blog by members of the public in response to the blog manager’s provocative invitation to discuss the plaintiff’s activities as a lawyer. Canoe had agreed with the blog manager to take sole responsibility for compliance with its own internal by-laws respecting defamatory content. The court interpreted this agreement as the website reserving the right to make editorial choices about its hosted content. The court’s decision to hold the website responsible recognized that some online intermediaries have an obligation to take reasonable steps to suppress the defamatory content they host, even if they are not directly responsible for its initial publication. That said, oversight, perhaps best embedded in the process, might be a necessary corollary to such obligations.

Even in the common law world, courts are more recently inclined to hold intermediaries liable based on what this paper labels “contextual control”. In other words, knowingly publishing defamatory material despite explicit and verified requests to investigate/suppress such materials. For instance, consider the case of Duffy v. Google[97] where, citing the Canadian case of Crookes v. Newton,[98] the Supreme Court of South Australia found that Google was a publisher of, and therefore potentially liable for, defamatory content published in search results, auto-complete suggestions and even third-party websites to which it provides linked search results.[99] In Duffy, the plaintiff Dr. Duffy had discovered that excerpts of professionally defamatory material had become especially and immediately visible via Google searches, which not only displayed this material whenever her name was searched but that the algorithm even suggested derogatory terms when her name was searched, using auto-complete (e.g., “Janice Duffy Psychic Stalker”).

She complained vociferously but the intermediary to her account fell short of taking the necessary steps. Rather than attempt to sue the multiple and geographically diverse sources of communication (presumably in light of the difficulties associated with both anonymity and jurisdictional enforcement), she proceeded to sue the search engine itself, which she deemed the publisher of defamatory material. As one commentator opined, reflecting the Court’s opinion:

if Google personnel were made aware of the existence of the defamatory website, “snippets” generated by Google’s own software programs and failed to remove them, their continuing existence thereafter was the direct result of human action or inaction rather than merely the result of machine operation. The same analysis applied in respect of the autocomplete suggestions.

Additionally, in an even more significant finding, [the Court] held that as the Google website is programmed to automatically cause the user’s web browser to display third party webpages by clicking on the relevant hyperlinked search result, Google was also a secondary publisher of that [defamatory] webpage if it failed to remove the hyperlink after reasonable notice by the person alleged to be defamed by it.[100]

With regard to the onerous duty of surveillance, the court proposed what this author deems a “contextual control” test. In effect, duties are triggered: “upon being notified that a particular website is considered to be defamatory, a search engine provider which fails to remove the website from its search results within one month, may be held separately liable for defamation every time someone clicks on a link to that material from the search results.”[101]

Parenthetically, an approach such as this, cognizant of the hardships of pursuing individual defamers in the porous and “anonymous” digital age, may serve to keep at bay the spread of controversial solutions such as the much-criticized “Right to be Forgotten”, which for its part unnecessarily infringes upon freedom of expression.[102] This is similarly true of a liability-based approach popular in Europe post-GDPR, which may unintentionally provide a misbegotten incentive to needlessly suppress more speech than minimally necessary by platforms fearful of regulators and the gargantuan fines that they impose.[103]

Instead, greater equipoise, respectful of the Charter, might be reached via a Duffy analysis. Duffy permits seasoned courts of law, rather than unaccountable and untrained corporate actors, to adjudicate what material is truly defamatory and therefore must be deindexed. It does so by adopting and explicitly citing the Crookes obiter for when third parties go beyond mere hyperlinking. In other words, as the Supreme Court remarks in Crookes, there could be liability for intermediaries under different facts if the latter goes beyond simply hyperlinking and fails to heed reasonable warnings. Pierre Trudel’s description of the underlying rationale, although offered in a different context and prior to the Duffy ruling, is edifying:

[t]his reasoning corresponds to the principle advanced by case law in several jurisdictions whereby an owner is not, in principle, liable for the faults/ misdeeds committed by his tenants. That said, a hotel, for its part, that is knowingly at the centre of illicit activities will be liable for damages caused as would the owner of a website that endorses defamatory communications transmitted by its users.[104]

IV. Preliminary Conclusions: Integrating Lessons from the Civil Law Approach

In her instructive paper on law and technological evolution generally, Lyria Bennet Moses explores the merits of revisiting norms in light of technological change: “[e]xisting rules were not formulated with new technologies in mind. Thus, some rules in their current form inappropriately include or exclude new forms of conduct.”[105] This is particularly true of defamation, as traditionally conceived at common law. Fundamentally, and as previously noted, the common law’s pigeonholed, categorical approach is problematic in many ways, as the Supreme Court itself recognized in Grant v. Torstar Corp.[106] Although it ultimately affirmed the “familiar categories” and traditional framework,[107] the Court significantly moved towards what Goffman called a “broader paradigm”,[108] attempting then to soften notoriously rigid classifications towards greater fairness, practicality and with an eye towards striking equilibrium between the relevant Charter values. This equilibrium between freedom of expression and reputation/reputational privacy in the digital age is precisely our objective and underlay the thinking animating the foregoing.

Indeed, the Internet, as Danay cautions in his instructive critique of defamation law, cannot be equated with traditional broadcast media for many purposes including defamation. More specifically, whereas a binary, categorical understanding of “truth” may have had a determinative role to play hitherto, the preceding recommends a more nuanced, contextual understanding of defamatory expression, which better lends itself to striking a constitutional amenable balance between digital expression and reputation (including reputational privacy).

As a bisystemic country and cognizant of the above-mentioned convergence between legal systems upon us, Ontario would do well to look to its civilian counterparts (Quebec and beyond) for practical and conceptual inspiration, particularly in the notion of the contextual reasonableness of the impugned expression. This is a far more malleable approach than the “truth defence” and perhaps even defences more generally; it accounts for the promise and perils of unvetted, often instinctive gradations of digital expression and is better equipped to remedy the sort of “overdeterrence”[109] that some lament as undermining the otherwise empowering digital speech uniquely imparted by technological advances.[110]

Accordingly, under Quebec civil law, unlike its common law counterparts, the question—perhaps somewhat surprisingly—is not necessarily whether the purportedly defamatory statements are true or false. But rather whether expressing them under the circumstances was reasonable according to an objective standard of reasonableness (an ex post facto built-in fair comment defence, one might posit). Circumstances, of course, can and must account for the vicissitudes of the digital variety where pre-existing reputational difficulties can be compounded by punitive shaming in an era of “infinite memory.” Context therefore is key. Such a standard can assist in reintroducing much needed context oft divested by the digital medium or “recontextualizing” digital expression, thereby aligning it with a purposive understanding of the very Charter values that inform defamation law post-Grant.

At this juncture, it is important to note that this essay by no means purports to answer or even fully touch on many of the issues raised by the topic. These challenges should be addressed in future discussions. In particular, although not limited to, the potential relevance of a “hybrid” model, which marries civil and common law approaches, one where context and good faith supplant “truth” and other defences with an eye towards better balancing Charter values.[111]

Two interrelated recommendations were advanced: first, to adopt a “contextual reasonableness” approach, as a Canadian court has already done (however inadvertently) in Whatcott. Second, in line with rejecting a pigeonholed approach ill-suited to the digital age, to retreat from the truth defence, which places undue burden on freedom of expression by requiring defendant to show verifiable truth (rather than for plaintiff to show lies or unreasonable expression). In common law parlance, moving to a negligence standard in defamation law, particularly post-reasonable communication defence post-Grant.

The foregoing argument suggests that reviewing and revising defamation law is a promising way to balance freedom of expression with privacy and reputational concerns in a changing technological and societal landscape. The rigid view of the common law, practically unworkable in the digital age, can be made more adaptable by turning to the civil law’s broad and flexible time-tested principles.