I    INTRODUCTION
II   THE ORIGINS OF THE ORDER
III  THIRD PARTY INTERMEDIARIES: THE CLASS OF RESPONDENTS
        The Mere Witness Rule
        The "Only Practicable Source" Requirement
        Summary
IV   THE THRESHOLD FOR AN ORDER
V    SCOPE OF THE INFORMATION
VI   BALANCING OF INTERESTS
  




				

				
				 
				


      Privacy
        Confidentiality
        Breach of Confidence
        Defamation
VII  CONCLUSION

ABSTRACT

The Norwich order is an equitable remedy that enables an applicant to discover the identity of a covert tortfeasor from an intermediary in order to assist the applicant in launching an action This article examines the various rules that define the Norwich remedy and how well they facilitate access to justice while protecting certain social interests The article argues that the threshold for granting disclosure should be kept low to ensure that legal recourse be as widely available as possible; to achieve that objective, the class of respondents should not be constricted by an arcane stricture known as the mere witness rule, nor by requiring too literally that the respondent be the only source of information On the other hand, the article argues that the scope of information available in a Norwich order should be kept narrow to prevent it from becoming a general right of discovery or from imposing an undue burden on the respondent. With respect to countervailing interests, the article analyzes several cases that raised confidentiality or privacy concerns. The article suggests that the privacy arguments in BMG Canada Inc. v. John Doe mischaracterized the basis on which a Norwich order is granted, and that privacy cannot be a shield for wrongdoing. Moreover, the article notes that under the Wigmore framework, personal interests will seldom outweigh society's interest in the correct disposal of litigation," hence only significant social concerns may tip the balance towards non-disclosure. The article argues that such social interests are present in the need to protect journalist sources and character references, and that courts should pay greater heed to the chilling effect that may occur if such informants were unmasked with Norwich orders.

RESUME

L'ordre de Norwich est un remade equitable qui permet a un demandeur de decouvrir l'identite secret d'un auteur de delits civils par un intermediaire afin d'aider le demandeur a lancer une action civil. Cet article examine les diverses regles qui definissent le remede de Norwich et a quel point elles facilitent l'acces a la justice tout en protegeant certains interets sociaux. L'article discute que le seuil pour accorder la revelation devrait etre garde discret afin d'assurer que ce recours legal soit aussi largement disponible que possible; pour atteindre cet objectif, la classe des repondants ne devrait pas etre resserree par une restriction mysterieuse connue sous le nom de seule regle de temoin, ni en exigeant trop litteralement que le repondant soit la seule source d'information. D 'une part, l 'article argumente le fait que la portee des informations disponibles dans un ordre de Norwich devrait etre maintenue etroitement pour l'empecher de devenir un droit general de decouverte ou d'imposer un fardeau anormal au repondant. En ce qui concerne des interets de compensation, l'article analyse plusieurs cas qui ont souleve des inquietudes de confidentialite ou d'intimite. L 'article suggere que les arguments d'intimite de BMG Canada Inc. c. John Doe ont mal caracterise la base sur laquelle un ordre de Norwich est accorde, et que l'intimite ne peut pas etre un bouclier pour l'injustice. D'ailleurs, l'article note que sous le cadre de Wigmore, les interets personnels seront rarement superieurs a l'interet de la societe pour la disposition correcte du litige; par consequent seulement les soucis sociaux significatifs peuvent incliner l'equilibre vers la non-revelation. L 'article argumente le fait que de tels interets sociaux sont presents dans la necessite de proteger les sources des journalistes et des references de caractere, et que les cours devraient avoir une plus grande attention a un effet refroidissant qui peut se produire si de tels informateurs etaient demasques avec les ordres de Norwich.

I INTRODUCTION

It has often been said that litigation is the most civilized method of settling disputes. For this characterization to be true, there must first be a lawsuit--and to commence a lawsuit the plaintiff must be able to identify the defendant. Unfortunately, just as victims of a crime often do not know their wrongdoers, some tort victims are ignorant of their tortfeasors. For example, a person who has been defamed on the interact will be stymied from pursuing a proceeding in the case of an anonymous posting. Similarly, a corporation may find its confidential documents leaked to the press by an unknown individual. Parties such as these may find that despite the use of private detectives or other resources the name of the perpetrator of the harm still eludes them. Normally, it would be the end of the road for these tort victims; however, where an intermediary can identify the wrongdoer, the prospective plaintiff may resort to a legal device called a Norwich order to compel the intermediary to disclose the identity of the alleged tortfeasor.

The Norwich order was created in the 1974 case Norwich Pharmacal Co. v. Commissioners of Customs & Excise, (1) where the House of Lords revived and modified an old equitable remedy known as the pre-action bill of discovery. This decision allowed a pharmaceutical company to find out from the customs authorities the identity of the party that had been secretly importing the company's patented drug into the country. Although the Norwich decision seemed to offer prospective plaintiffs a powerful tool, use of the remedy in the next few decades was scant. Except for a handful of sensational whistle blowing cases in England, the Norwich order has been mostly deployed in intellectual property actions with factual situations that are very similar to the inaugural case. There are two potential explanations for the infrequent use of this remedy. First, it is rare to have situations where there is both a covert tortfeasor and a knowledgeable intermediary. Second, litigants in Canada can seek relief through a "John Doe" order through the provincial court rules (2) and move to discover the wrongdoer's identity.

Despite its infrequent use, the scope of information disclosed by a Norwich order has expanded to include information other than the wrongdoer's name. Furthermore, the internet may increase the relevance of this remedy as internet anonymity has increased the prevalance of the types of situtions requiring Norwich orders: torts committed by covert defendants whose identities are known by third-party intermediates. Although there were a few pre-internet defamation cases that utilized the Norwich remedy, online message boards have created a surge in applications for judicial assistance in tracking down the sources of libelous messages. (3) Similarly, the recent case of BMG Canada Inc. v. John Doe (4) is at the forefront of another on-line phenomenon--large-scale anonymous file downloading and sharing. BMG also re-established the relevance of the Norwich order in another way: the case was adjudicated using Norwich principles despite its origins as a "John Doe" action, thus raising the possibility that the Norwich standard will now apply to all pre-action discoveries.

But even before this newfound prominence, the Norwich order warranted closer examination. Since the identity of the defendant is essential to start a lawsuit, a Norwich order effectively acts as a gatekeeper to the courthouse, and thus its boundaries crucially affect access to justice. The limits delineated in the original Norwich decision have not always been clear or logical, and due to the discretionary nature of the remedy, subsequent decisions have made its limits even hazier. The objectives of this article are to examine the various rules that define the Norwich order and how well they facilitate access to justice, and to analyze whether the balancing process accurately weighs the various interests involved, both public and private.

Part II of this article outlines briefly the history of the Norwich order and the nature of an equitable remedy. Part III looks at the class of respondents that may be summoned before the court, specifically addressing two rules that constrict this class of persons. The discussion seeks the elusive purpose of these two rules and concludes that relaxation or abandonment of the rules is desirable. Parts IV and V examine the level of proof required to obtain disclosure via a Norwich order and the scope of information that may be compelled, while at the same time suggesting more appropriate parameters for these elements. Part VI explores the final step in the adjudication of a Norwich order: the balancing of interests. This discussion analyzes the extent to which privacy and confidentiality may outweigh a plaintiff's right to sue, with a special focus on how the Wigmore test (5) may apply to the different individuals who typically request Norwich orders.

II THE ORIGINS OF THE ORDER

The Norwich Pharmacal Company was in an unusual quandary in 1974. From reports published by the customs authorities, it had learned of unlicensed imports of its patented drug furazolidone into the United Kingdom over the previous 10 years. The company tried to discover the identity of the importers, but was unsuccessful. While it was aware that the Commission of Customs and Excise had importation records containing this information, the Commission had denied its request for disclosure on the grounds of confidentiality. Moreover, the company could not sue the Commission for the release of the information it needed as there was no cause of action against the Commission. As such, the company turned to the courts for equitable relief.

The case ultimately progressed to the House of Lords where the House of Lords faced a rare legal dilemma: discovery can only be permitted after an action has been commenced, but it was impossible to launch suit in this case without discovery. There was clear evidence that the appellant's right had been infringed, yet enabling the pursuit of redress might necessitate intruding upon the rights of an innocent third party.

In deciding the case, Lord Reid noted, "discovery as a remedy in equity has a very long history." (6) Although usually ordered during the litigation process, discovery was "extended at an early date to assist a person who contemplated litigation against the person from whom discovery was sought, if for various reasons it was just and necessary that he should have discovery at that stage." (7) The stumbling block in this case was that the traditional bill of discovery could only be ordered against the intended defendant, lest it violate the "mere witness" rule, a stricture against seeking information from those who would eventually testify as a witness. (8)

The court's solution to this obstacle was to enunciate the following exception to the mere witness rule:

   [A] person [who] gets mixed up in the tortious acts of others so as
   to facilitate their wrong-doing .may incur no personal liability
   but he comes under a duty to assist the person who has been wronged
   by giving him full information and disclosing the identity of the
   wrongdoers. (9)

Since the Commission had permitted the import of the unlicensed drug, its facilitation imposed a moral obligation on it to assist in uncovering the tortfeasor's identity. The court further held that if a plaintiff bonafide believes that the alleged culprits committed tortious conduct, and the respondents are "the only practicable source of information as to whom they should sue," (10) an order for disclosure would be available "unless there is some consideration of public policy which prevents that." (11) In Norwich, the information was neither secret nor confidential as it had passed through many hands, but even if a degree of confidentiality had been attached to the information, "the interests of justice in this case far outweigh any interest there may be in non-disclosure." (12)

A pre-action bill of discovery against non-parties was thus revived. In Canada, the first request for a Norwich order was considered in Re Johnston and Frank Johnston's Restaurants Ltd., (13) and rejected on the grounds that it violated the mere witness rule. It would take more than a decade, and a factual situation almost identical to Norwich, for a Canadian appellate court to finally embrace the remedy. (14) However, there has been scant use of the Norwich order since then, perhaps because potential applicants have been deterred by the defeat in the initial cases. (15)

While it has remained largely unchanged in Canada, the Norwich order experienced significant evolution in England. In Bankers Trust v. Shapira (16) Lord Denning decided that the remedy permits the discovery of any information, not just the name of the suspected wrongdoer, and in British Steel v. Granada Television (17) he held that the applicant need not have a definite intention to sue for the court to grant a Norwich order. These expansions to the legal boundaries of the Norwich order challenge the original purpose of this equitable remedy. To understand the limits of a Norwich order, it may be beneficial to briefly consider what an equitable remedy is.

Equitable remedies sprang out of frustration with the rigidity of common law rules prevailing before the 1700s. Litigants sought relief as a last resort from the monarch, and as the monarch grew weary of hearing these petitions the task was delegated to chancellors. A Court of Chancery was created and the chancellors were said to dispense equity. At first, adjudication proceeded on basic notions of fairness and conscience, with remedies being highly individualized to suit the circumstances. In opposition to the unyielding remedies of the common law, the Courts of Equity strove to keep the means of relief flexible and protean. Predictably the volume of cases rendered such an ad hoc approach impracticable and chancellors began systematizing equitable principles. (18) Today, equitable remedies are no longer highly idiosyncratic in form, but they are still said to be "of great width and elasticity." (19) Indeed, one academic authority has been highly critical of the entrenchment of rules, admonishing that specificity and rules of general application are inconsistent with equitable principles. (20) But the contrary view, that predictability and certainty in the law are also necessary to provide justice, has considerable merit given the frequent utilization of equitable remedies. (21) As the ensuing discussion shows there must be both doctrinal structure and flexibility in its application for the Norwich order to achieve its ultimate goal of facilitating access to justice, and for the court to achieve equity in any given case.

III THIRD PARTY INTERMEDIARIES: THE CLASS OF RESPONDENTS

A classic Norwich order is concerned with unmasking a wrongdoer's identity in situations where a tort has apparently been committed, but where the tortfeasor's identity is unknown. Because of the certainty of the what and the when for granting such disclosure, early cases focused on the who--that is, the question of who should be made to reveal the tortfeasor. One would think that this issue is not markedly different from determining who may be subpoenaed as a witness, yet the courts have resolved this issue in a complicated and counterintuitive manner. This unexpected direction is due to the existence of an obscure legal stricture known as "the mere witness rule," and to a less arcane dictum that the respondent be the "only practicable source of information." The following discussion shows how mechanical application of these rules has resulted in unnecessary confusion and harsh results, and how a more principled approach would yield more equitable outcomes.

The Mere Witness Rule

The use of the mere witness rule was first reported in the 1749 case Plummer v. May. (22) The rule prohibited discovery "from a person who will in due course be compellable to give that information either by oral testimony as a witness or on a subpoena duces tecum." (23) Even in its early inception, the rule contained exceptions for certain parties such as corporate officers, attorneys implicated in the transaction at issue, auctioneers, agents, and arbitrators. (24) A further exception was given for those with a "direct interest" in the pending action, which appears to be an early iteration of the "mixed up in the transaction" proviso. (25)

Nevertheless, the mere witness rule confounded the House of Lords. Much of the Norwich judgment was spent scrutinizing the few English cases that had provided some leeway on the rule. Lord Reid said that the authorities "were not very satisfactory, not always easy to reconcile and in the end inconclusive." (26) Commenting on one particular case, Lord Morris of Borth-y-Gest stated that they had "studied and re-studied" (27) the case to determine if its exception was reliable. Lord Kilbrandon characterized the topic as being one "of a technical nature involving an expert knowledge of English legal history in the nature of things denied to me." (28)

Lord Reid provided the most detailed guidelines on this issue. He acknowledged that the rule was well-settled, but held that to prevent frustration of the plaintiff's right to an action it should have "no application to a case like the present case." (29) He reiterated that "to apply the mere witness rule to a case like this would be to divorce it entirely from its proper sphere." (30) These words indicate clearly that the mere witness rule should be suspended in such cases, in other words, mere witnesses can be compelled to make disclosure in Norwich-like situations.

Had Lord Reid stopped there, subsequent confusion over the rule may not have ensued. Unfortunately, he proceeded to constrict the class of respondents:

   But that does not mean, as the appellants contend, that discovery
   will be ordered against anyone who can give information as to the
   identity of a wrongdoer. There is absolutely no authority for that.
   A person injured in a road accident might know that a bystander had
   taken the number of the car which ran him down and have no other
   means of tracing the driver. Or a person might know that a
   particular person is in possession of a libellous letter which he
   has good reason to believe defames him but the author of which he
   cannot discover. I am satisfied that it would not be proper in
   either case to order discovery in order that the person who has
   suffered damage might be able to find and sue the wrongdoer.
   Neither authority, principle [n]or public policy would justify
   that. (31)

This perplexing passage raises several questions. It appears that Lord Reid effectively substituted the mere witness rule with a "mere bystander" rule: whereas the original rule barred discovery of mere witnesses, Lord Reid's rule prevents the discovery of bystanders, such as those who stood by and saw a hit-and-run accident. This new formulation is confusing: is a bystander not a witness, and if not, how is a bystander distinguished from a witness? None of the other judgments in Norwich assist in understanding the distinctions. In fact, the other judges muddle the picture further by providing contradictory views. Viscount Dilhorne intoned that "someone involved in the transaction is not a mere witness" (32) while Lord Kilbrandon stated "the defendant is not a mere witness, or any kind of witness, because ... until the defendant has disclosed what he knows, there can be no litigation in which he could give evidence." (33)

The key to unraveling Lord Reid's reasoning is to realize that he had subdivided the universe of witnesses into two groups: bystanders and facilitators. A list of the parties involved may thus be presented as follows: (1) the victim; (2) the perpetrator; and (3) the witnesses as either (a) bystanders or (b) facilitators. This categorization is consistent with Lord Reid's dictums. While witnesses as a class are not barred from being discovered, only the subset comprising of facilitators--those who are "mixed up in the tortious act"--should be held accountable for their complicity by being compelled to give disclosure. But this leads to the next question: why should bystanders be immune from discovery? Why is it equitable to permit the lone bystander who had jotted down the license plate of a hit-and-run driver to remain silent since the damage to the plaintiff is no less injurious?

Again, none of the judgments in Norwich explicate this "protection" of the uninvolved bystander. No historical explanation exists for this additional prohibition against bystanders, as the traditional reasons for the mere witness rule were: (1) to prevent a litigant from concocting evidence in advance based on the witness' testimony; (2) to prevent hearsay evidence; (3) to prevent tactical delay of the proceedings; and (4) to prevent innocent persons from incurring trouble and expense. (34) However, as counsel for the appellant in Norwich argued, the first three reasons did not apply as the applicant was seeking simply the name of the tortfeasor. As for expenses, the applicant would give an undertaking to reimburse the respondent, as with all interlocutory requests. Lord Kilbrandon himself concurred by observing that expense was not a legitimate argument since a subpoenaed witness could be compelled to testify notwithstanding any inconvenience. (35)

Lord Reid's pronouncement that "neither authority, principle [n]or public policy" would justify an imposition on a bystander is particularly puzzling as the statement is completely unexplained. It would take several decades for the House of Lords to shed a dim ray of light on the matter. In Ashworth Security Hospital v. MGN Limited, (36) Lord Woolf suggested that the mere witness rule is based on "the general principle that at common law there is no legal duty to provide the police with information or otherwise to assist them with their inquiries." (37) In support of this proposition he cited Rice v. Connolly, a case in which the defendant was acquitted of "willful obstruction" in his refusal to identify himself when stopped by a police constable. (38) Although Rice did not specify a reason for the "no legal duty to assist rule," the underlying animus is easy to apprehend. The right to silence is well understood in common law to be motivated by a concern for self-incrimination. R. v. Hebert, (39) a leading Canadian case on the right to silence, confirms this rationale. (40)

The "no legal duty to assist" principle and its underlying "right not to self-incriminate" rationale do not adequately justify the "mere bystander" rule for several reasons. First, both of these concepts are applicable in criminal proceedings and not civil cases. Second, even if for argument's sake these rules were to apply, it is simply illogical to distinguish bystanders from facilitators on this basis. The risk of self-incrimination is nil, or fairly low at best, since a mere bystander is by definition a person who did not participate in the wrongdoing. The absurd effect of the mere bystander rule is to award special protection to the group that least requires it. Nor is the rule consistent with the practice of granting the right against self-incrimination to all suspects in criminal prosecutions, regardless of the strength of the case against each one. Moreover, "involvement" as a criterion suggests a certain degree of premature adjudication of a respondent's complicity. Perhaps this was Lord Kilbrandon's concern when he stated that "the respondent is not a mere witness, or any kind of witness, because the whole basis of the application is that, until the respondent has disclosed what he knows, there can be no litigation in which he could give evidence." (41) Perhaps Lord Woolf intended to invoke the common law rule against a duty to rescue, (42) instead of the "no legal duty to assist" rule. Although the former rule explains more sensibly why bystanders need not assist the tort victim, it is not clear that the "no duty to rescue" rule distinguishes between bystanders and facilitators as the principle is commonly understood to be applicable to bystanders generally. Finally, the distinction between "bystander" and "facilitator" cannot always be easily drawn. After all, a witness who remains silent about the identity of a wrongdoer is ultimately a facilitator of the wrongdoing.

Therefore, there is still no wholly satisfactory explanation for Lord Reid's "exception to the exception" that singles out bystanders as a group immune to discovery. If no legitimate justification can be found for the rule, then "judicial parsimony" (43) should apply to abolish it. After all, the categorical exclusion of any group of persons from a Norwich order has serious implications for individuals' ability to access justice.

The "Only Practicable Source" Requirement

The mere bystander rule is not the only constraint imposed on the class of respondents in a Norwich order. Courts have also invoked Lord Kilbrandon's criterion that the respondent be the "only practicable source of information." (44) In Re Johnston, for instance, MacDonald J. found that some of the respondents were not the only practicable sources of information, and that the plaintiff could have obtained divulgence from the one known tortfeasor by bringing suit against him. (45) Because this was the first case that considered the Norwich remedy in Canada, almost every subsequent case in this country has repeated the "only practicable source" requirement, (46) despite the fact that it was not a prerequisite stipulated by every judge in Norwich.

The proposition that the respondent be the party of last resort is not an unreasonable one. Requiring plaintiffs to exhaust other means before they impose their demand for information on a third party is a prudent measure, since the third party may have a host of interests that render disclosure an unpleasant or difficult exercise. The third party may have given a promise of confidentiality, for example, or be in a relationship with the alleged defendant that would make disclosure awkward. Despite the fact that this criterion is an eminently sensible one, an overly strict application can wreak devastating results, as demonstrated by Kenney v. Loewen. (47)

In that case, the plaintiff, Kenney, became aware that he was possibly defamed when he failed to secure a contract with the respondent Loewen Group. Through an intermediary, he learned that a Mr. Watts had passed on statements from an unknown party to the Loewen Group, statements alleging that Kenney was a spy for the respondent's competitor, Service Corporation International, amongst other nefarious deeds. Kenney thus sought disclosure from the Loewen Group for the identity of the person making the allegations. The British Columbia Court of Appeal rejected the request on the grounds that the respondent was not the only practicable source of information; the plaintiff could have sued Watts for defamation, and upon Watts' denial of being the source of the allegations, the true identity of the wrongdoer would be revealed.

Demanding such a circuitous approach for disclosure is troubling for several reasons. First, when it is clear that a plaintiff does not believe someone to be the actual tortfeasor, forcing him or her to sue this third party merely for the purposes of discovery seems unethical. Ironically, this point was raised by Saunders J. herself early in the judgment wherein she described such a tactic as "a sleight of hand" that is "highly improper." (48) Second, the plaintiff in this case had several pragmatic reasons for not pursuing this route: (1) Mr. Kenney was advised that the suit would be dismissed for its hearsay foundation; (2) the short limitation period in New York (where Mr. Watts resided) had already expired; and (3) Mr. Kenney was concerned that a lawsuit would further damage the reputation of his fledgling company. Justice Saunders rejected the hearsay argument, opining that "the hearsay issue is not a barrier to an action." (49) She found that the expiry of the limitation period was a consequence of Mr. Kenney's inaction due to a desire to suppress publicity of the allegations stating, "[w]hile this may be a good business reason not to pursue the action against Mr. Watts, it is not a basis on which to grant a bill of discovery." (50) Although Saunders J. cannot be faulted for following the dictates of the Norwich decision, her assessment may be overly dismissive of the practical realities a person confronts when contemplating litigation. The reasons advanced by Mr. Kenney were neither frivolous nor groundless, and the fact that he received counsel against litigation should have carried more weight as it demonstrates that his decision was informed.

In fact, the practical difficulties of obtaining discovery were brought to light in a postscript to the judgment. The court noted that Mr. Kenney had tried to subpoena Mr. Watts in relation to the Norwich order. Not only was he rebuffed by the New York District Court, but the court had admonished Mr. Kenney for not advising that it was an action for discovery and that the Loewen Group had committed no actionable wrong. Justice Saunders commented that there are obviously differences in practice between New York and British Columbia courts; in the latter court's eyes Mr. Kenney had not breached any standard. This is cold comfort, of course, to the fruitless plaintiff.

Despite a similar factual scenario, the English decision P. v. T. Ltd. (51) stands in stark contrast to Kenney. In this case, the plaintiff Mr. P. learned from his employers that very serious allegations of misconduct had been leveled against him by a third party. He was asked to defend himself at a disciplinary hearing despite management's steadfast refusal to specify the charges. Unable to meet the case against him, Mr. P was summarily dismissed. Although he was awarded damages by an industrial tribunal for unfair dismissal, the plaintiff was unable to find employment within the industry due to the well-known fact that he was terminated for impropriety. He consequently sought disclosure from his employers, and the court granted the order for essentially one reason namely, "justice demands that Mr. P should be placed in a position to clear his name." (52)

The factual differences between P. v. T. Ltd. and Kenney are few. While P. v. T. Ltd. may not have followed Norwich closely, its authority has never been questioned and it should have attracted greater precedential attention in Kenney. Arguably, the presence of other intermediaries in Kenney is the very wedge between the two cases, but the disquieting implication of this contention is that it may discourage candour in the plaintiff's pleadings: had Mr. Kenney not acknowledged his awareness of Mr. Watts' role in the affair, the British Columbia court may have concluded that the Loewen Group was indeed the only recourse for information.

Ergo, how should a court apply the "only practicable source" requirement? First, to reiterate, this criterion was issued by only one judge in the Norwich decision, and has not received a great deal of attention in English cases. Therefore, it should not be viewed as a fundamental cornerstone of a Norwich order. Second, it would seem that nothing more than a common sense approach is necessary to interpret the criterion. If an applicant can show that it has pursued available channels of inquiry in good faith then that should be sufficient. The kind of recourse demanded in Kenney--specifically, to sue a third party to acquire the name of the ultimate defendant--is hardly "practicable." Finally, it has been suggested that in the context of equitable remedies:

   [A] plaintiff is not ordinarily denied relief merely in view of
   prejudice or hardship to the public or third persons. Relief is
   refused by the court only in exceptional cases where considerations
   of this kind are so disproportionately great, as against
   prospective prejudice and hardship to the plaintiff, as to render
   specific performance unjust in all the circumstances. (53)

This comment pinpoints what is troublesome about the "only practicable source" dictum: it seems to contemplate some type of hardship for the respondent, but the rule is categorical in its reach.

Summary

Examining how a class of respondents is defined may seem at times to be an academic exercise, but as Kenney shows, even trivial procedural rules have mighty consequences. The mere witness rule may seem to be an historical curiosity; nevertheless, it was revived by the House of Lords some twenty years after Norwich in the Ashworth case. These artifacts from the Norwich decision have lingered on without much analysis, and it is time to challenge their purpose and place in the law.

Since the identity of the wrongdoer is a vital piece of the litigation puzzle, a Norwich order is effectively a gatekeeper to the courthouse. Because of its implication for access to justice, the class of respondents for a Norwich remedy should be as unfettered as possible. The mere witness rule and its derivative the mere bystander rule should be abolished; as well, the "only practicable source of information" criterion should not be rigidly enforced.

It should not be forgotten that the boundaries for the class of respondents are not the only filters for unwarranted Norwich applications. The court will also assess whether a bona fide claim of wrongdoing has been made, whether the scope of information sought is too wide, and whether the parties' interests are balanced. Concerns about disadvantaging the respondent are more suitable for the later stages of the adjudication; at the preliminary stage, the paramount concern should be that access to justice not be impeded for dubious reasons.

IV THE THRESHOLD FOR AN ORDER

If the class of respondents is to be as unrestricted as possible, it would seem fair to require the applicant to justify the intrusion on the respondent with solid evidence that a tort has occurred. The requirements of proof for a Norwich order, in fact, do not correspond to this proposition. All that is required is that the applicant swear to a bona fide belief in the existence of a tort. The laxity of this threshold has been prodigiously tested. For instance, while the plaintiff in P. v. T. Ltd had a good faith belief that a wrong had been perpetrated against him, he did not know if it actually occurred. (54) Similarly, while the plaintiff in BMG had some evidence to support his belief of wrongdoing, it was not clear that the impugned conduct constituted an infringement of copyright since the legality of digital file sharing had not at the time been decided in Canada. (55) These situations have disconcerting implications. If wrongdoing cannot even be described or substantiated, there would appear to be no basis for a legal action and thus no basis for disclosure.

A low threshold may also render the Norwich order vulnerable to abuse. As the court observed in Irwin Toy v. John Doe, (56) "the anonymity of the internet could be shattered for the price of the issuance of a spurious Statement of Claim." (57) One commentator has called for the imposition of the higher prima facie threshold for disclosure orders to prevent unscrupulous parties from unmasking online critics who may be speaking the truth or making fair comment, both of which are complete defences to a defamation suit. (58) The prima facie standard requires an applicant to demonstrate a case that is strong enough to win at trial. If the case presented is weak or conflicts with the affidavit evidence submitted by the defendant, the request will be denied.

On the other hand, the argument against a prima facie threshold is equally compelling. Until 1975, the prima facie case standard was the default standard for an interlocutory injunction. (59) This high standard of proof resulted in transforming the injunction hearing into essentially a preliminary trial on affidavit evidence. Specifically, once a respondent is restrained from exercising certain time-sensitive rights (e.g. picketing, publishing timely information, working for a competitor), the right is often permanently lost and the case never comes to trial. This led to Lord Diplock's opposition to the prima facie threshold in American Cyanamid Co. v. Ethicon Ltd., where he stated:

   It is no part of the court's function at this stage of the
   litigation to try to resolve conflicts of evidence on affidavit as
   to facts on which the claim of either party may ultimately depend
   nor to decide difficult questions of law which call for detailed
   argument and mature considerations. (60)

American Cyanamid effectively banished the prima facie standard as the default threshold for interlocutory measures. The exhortation against pre-judging was reiterated in both Glaxo Wellcome and BMG. The Glaxo Wellcome court noted that special difficulties exist for patent cases since defences such as lack of novelty, obviousness, or specificity are far too complicated to determine at the interlocutory stage, and thus it is nearly impossible to assess whether the plaintiff will win at trial. (61) What is true for patent cases also holds for other torts: an action for defamation is subject to a truth defence, and breach of confidence to a claim of qualified privilege. Hence in BMG, Sexton J.A. held that the prima facie standard was not appropriate for pre-action discovery since adjudicating any case before trial would be premature and unjust.

It is especially important not to erect too high a threshold for the Norwich order because of its "gatekeeper to the courthouse" role. This is a crucial distinction between the Norwich remedy and other interlocutory measures. It is still possible to air a dispute in court after an injunctive application has been defeated; it is not so for a Norwich application.

The fact that there are equally forceful arguments on both sides of the threshold debate is characteristic of the dilemma over interlocutory injunctions in general. Indeed, the question of threshold is even more vexing for other types of remedies as they can be more intrusive and onerous, often depriving a respondent of a right in total. Since it would be premature to adjudicate interlocutory orders on the merits, academics have found the other two American Cyanamid factors to be more important: (62) whether irreparable harm will result if the injunction is denied, and the balance of convenience in carrying out the order. These two factors have never been considered for the Norwich order because it is not a true injunction--it does not enjoin or compel either the respondent or the prospective defendant from a course of action, other than disclosure which has little relation to the underlying dispute. For example, if a drug importer has been illegally selling a patented pharmaceutical, a Norwich order does not demand the immediate cessation of sales. In contrast, an Anton Piller remedy orders the confiscation of the infringing product, a physical intrusion which decisively and immediately stops the importer from selling his or her goods. The difference in impact is precisely the reason why the scales tip in favour of a lower threshold for a Norwich order. Leubsdorf, who has posited an alternative model to the American Cyanamid test, argues that the test for an injunction should balance the potential risk of judicial error that the injunction was unwarranted against the irreparable loss of legal rights. (63) In other words, is it better to be over-inclusive or under-inclusive in the number of remedies to be granted? It is submitted that in the Norwich context, it is clearly preferable to be over-inclusive. The irreparable loss of the applicant's right to sue outweighs any inconvenience to the respondent. Moreover, the revelation of the defendant's identity does not preclude him or her from continuing his or her impugned activities.

The prima facie standard may be an illusory bulwark against spurious claims anyways. Suppose a corporation wishes to determine the identity of a critic who has been posting negative but true statements about the company on the internet. If the contention is that under a bona fide standard the corporation will succeed, it follows that the company is willing to be untruthful, as it is falsely stating that it has a good faith belief that it has been defamed. If that is so, then a prima facie threshold will hardly deter the corporation from crafting together a specious case that will pass muster. Even if the corporation is truthful, it is not difficult to assemble affidavit evidence that portrays the company's activities in a positive light and insinuates that the statements are incendiary and false. Who will present the defence of truth against the corporation? The respondent, the internet service provider, is hardly in no position to do so. Accordingly, the corporation will succeed regardless of the threshold test.

Perhaps the debate over threshold stems from the discomforting amorphousness of the terms "bona fide" and "prima facie." The bona fide standard seems particularly impoverished: does it really entail nothing more than an honest belief that a tort occurred? In Norwich, Lord Kilbrandon made a fleeting remark that the appellants must "bona fide believe" that wrongdoing occurred, but subsequent cases have used the phrase "bona fide claim" instead. The latter suggests that some evidence must be presented to make out a case, and this was the position in Re Johnston; (64) as well, in BMG the court defined a bona fide claim as necessitating an intention to sue and that "there is no improper purpose" (65) for the application. It would appear then that the bona fide standard in practice requires some evidence of wrongdoing and thus is an adequate safeguard against spurious claims. The only anomaly is P v. T. Ltd., where despite being unable to recount the allegedly defamatory statements, the hapless applicant was nevertheless able to obtain divulgence.

Mr. P.'s predicament, it turns out, is not unique. The jobseeker in Straka v. Humber River Regional Hospital et al. (66) also did not know what comments were made in his employment reference letters, but, like Mr. P., he knew that they were derogatory and that they impaired his job prospects. Justice Morden observed that if a narrow approach were to govern the issue of threshold, the applicant could be said to fall short of the bona fide standard as he does not know whether the impugned statements were true. The judge rejected the contention that Dr. Straka was "fishing" for a case as the consequences of the letters made it obvious that damage was inflicted, and that he was legitimately seeking to clear his name through legal proceedings. Justice Morden decided the issue of threshold thusly:

   On these facts, I do not think that the appellant should be
   "non-suited" because his claim is not a bona fide one, i.e. that
   his claim should fail because the threshold requirement of a bona
   fide claim has not been shown. As I have said, we are concerned
   with an equitable remedy the granting of which involves the
   exercise of a discretion. The general object is to do justice.
   Accordingly, I do not think that a rigid view should be taken of
   the elements of the claim. With this approach in mind, I think that
   it is reasonable to accept that sufficient bona tides has been
   shown to justify consideration of the case as a whole. The nature
   and apparent strength of the appellant's case is a factor to be
   weighed together with the other relevant factors in arriving at the
   final determination of the claim. (67)

This resolution is consonant with Spry's view that "to attempt to lay down an inflexible standard for the proof of particular matters involves a misunderstanding of equitable principle" and "the court is hence required to exercise a discretion that involves the weighing of a large number of matters, none of which if regarded by itself and out of context, is necessarily or intrinsically decisive." (68) However, the problem posed by a liberal approach to a bona fide standard is the difficulty of preventing a person from demanding one neighbour to name another because he suspects the latter has uttered unkind words about him. Arguably Straka and P. v. T. Ltd. can be distinguished by the causal link between the unknown statements and the damage they suffered: in both cases the men were directly told that their employment fiascos were the result of negative comments. However, in neither instance does the causal link pertain to the key element of a case for defamation: the truthfulness of the statements. Hence these two cases are no different than a situation where idle chatter was suspected to have caused some detriment. The only difference relates to the quantum of the injury.

This discussion suggests there are two routes that a court could take. One solution is to take a de minimus approach, wherein trivial slights would be filtered out as being "frivolous." The other approach is to hold fast to the requirement that an applicant demonstrate the elements of his or her case, which necessitates pinpointing the "tortiousness" of the impugned conduct. This means an applicant must be able to recount the statements leveled against her and show why they are defamatory; hence the disclosure order should have been denied in P. v. T. Ltd.

This latter approach is certainly a brutally callous stance, but it is a part of life's vicissitudes that one cannot control negative opinions of oneself. The legal system cannot remedy every suspicion of malicious gossip and undeserved criticism, and unfortunately this means the prospective plaintiff must gather sufficient evidence to mount a case. Mr. P.'s employer indisputably acted egregiously, but recourse against this behaviour should have been sought through the channels that address specifically this type of wrong: employment law. It is unclear why the judgment in Mr. P.'s favour at the industrial tribunal did not exonerate him in the eyes of other employers, but there was always the possibility of greater vindication through a civil action against his former employees. As for Dr. Straka, there were other reasons for denying disclosure in his case, which are discussed in Part VI.

In addition to the discussion above, there remains the issue of novel claims, and whether pre-trial or interlocutory orders should be permitted when the allegations may not even constitute a tort. This is the concern in internet cases such as BMG where the legality of file downloading was unclear at the time.

To label the purported wrongdoing in BMG an unrecognized tort, however, is a mischaracterization. Although the law on internet file sharing is unresolved, the issue largely concerns the interpretation of the Copyright Act. (69) The debate revolves around whether large-scale file swapping constitutes "private use" or "distributing, whether or not for the purpose of trade." (70) As a legal matter to be decided at trial, it is no different than any other copyright case that requires determination as to whether the conduct can be excused as fair dealing or some other acceptable defence.

But what if the claim is truly an unrecognized wrong in law? If novel causes of actions have been permitted to proceed, albeit infrequently and very cautiously, then to reject a pre-trial motion on the basis that it is not grounded in an established tort would be to adjudicate the claim prematurely. The Norwich request is simply a novel claim in its embryonic form, and as long as it appears justiciable the application should be decided just as any other application would be. This argument is consistent with the position adopted by R. Grant Hammond, who warns that "the court should be careful not to equate 'no present right' with frivolity" (71) and that the court should examine if there are objectively arguable reasons on at least plausible grounds. It has often been said that the categories of tort are not closed; therefore, to thwart novel claims at the pre-action stage would hamper the evolution of law.

V SCOPE OF THE INFORMATION

In Norwich, Lord Cross of Chelsea distinguished the case from one that "open[s] the door to 'fishing requests' by would-be plaintiffs who want to collect evidence." (72) At its inception, the primary constraint of the Norwich order was the scope of the information sought: only the purported wrongdoer's identity was to be divulged. That quickly changed. Not only can an applicant request any information that would assist him or her in starting an action, but it is now possible to ask for financial information such as the movement of funds between bank accounts. The question, once again, is whether the scope of a Norwich order should be reined in.

By declaring that a facilitator is obliged to assist the victim by "giving him full information and disclosing the identity of the wrongdoers," (73) Lord Reid unwittingly opened the door to an expansion of the Norwich remedy. Plaintiffs have not been hesitant to seize on the words "full information" to procure all sorts of documents. For example, in Guinness UDV North America Inc. v. Flexitrace Ltd. (74) the court ordered a trademark infringer to divulge the names of its supplier and customers along with any documents evidencing its dealings in the infringing goods. There was no indication in the case that the plaintiff could not have acquired the information during the normal discovery process, that it tried to independently seek this information, or that the information was crucial to launch an action.

The applicant in Societe Romanaise de la Chaussure SA v. British Shoe Corporation Limited (75) was even bolder in its request. In that case, the applicant, a shoe manufacturer in the course of suing a major retailer for importing a shoe that infringed one of its designs, demanded that the respondent, a British retailer, disclose a wide range of document so that the applicant could ascertain whether the respondent's Italian supplier had further inventory of the shoe in question. The plaintiff was explicit about its motives: it did not wish to waste money on an Italian lawyer and engage in futile litigation in Italy if there was no likelihood that the offending shoe was still being sold by the manufacturer in that country. Justice Millett permitted the order, adding, "it is not sufficient to disclose the identity of the wrongdoer, the obligation extends to giving full information. That must include all information necessary to enable the plaintiff to decide whether it is worth suing the wrongdoer or not." (76)

The most striking expansion of the Norwich jurisdiction is the development, in the English courts, of an offshoot branch known as the Bankers' Trust order. This device is now used extensively to trace funds in cases of fraud, but in the eponymous decision Bankers Trust v. Shapira, (77) the applicants were able to obtain disclosure of the balance in the bank accounts, the movement of the funds in the last six months, and the identities of recipients of any transfers. The wide scope of information granted in Bankers Trust alarmed the court in Arab Monetary Fund v. Hashim. (78) In that case, Hoffmann J. (as he then was) enunciated some additional restrictions (79) on Bankers Trust orders (which have not been universally accepted (80)) and made these comments:

   The reference to "full information" has sometimes led to an
   assumption that any person who has become "mixed up" in a tortious
   act can be required not merely to disclose the identity of the
   wrongdoer but to give general discovery and answer questions on all
   matters relevant to the cause of action. In my view this is wrong.
   The principle upon which Lord Reid distinguished the "mere witness"
   rule was that unless the plaintiff discovered the identity of the
   wrongdoer, he could not commence proceedings. The reasoning of the
   other members of the House is the same. The Norwich Pharmacal case
   is no authority for imposing upon "mixed up" and third parties a
   general obligation to give discovery or information when the
   identity of the defendant is already known. (81)

This passage not only casts Societe Romanaise in doubt--it evinces a trepidation that the Norwich order will become a general right of discovery. Has the remedy in fact become a general right of discovery, given the outcomes in Bankers Trust, Guinness, and Societe Romanaise? The Bankers Trust order certainly appears to be an ill-fitting member of the Norwich family since the identities of the culprits are usually known in these situations. A better fit would have been to relegate the device under the law of tracing, or to append it to the Mareva remedy (82) since it is frequently used to support that type of injunction. Therefore, since the Bankers Trust order stands virtually in a separate category from the Norwich jurisdiction, it should be regarded as a different device altogether and inapposite to the discussion. The more pertinent changes come from the range of information granted in cases like Guinness and Societe Romanaise.

Obviously, scope is closely linked to purpose. "Fishing expeditions" occur because they stray from the Norwich mandate to assist an applicant in launching a lawsuit by providing the name of the would-be defendant. What is disconcerting about Guinness and Societe Romanaise is not that the scope of disclosure went beyond identity, but that the applicants in both situations were perfectly capable of suing the purported wrongdoer. Instead they wanted the court's assistance in starting secondary lawsuits and fact-gathering, which ultimately amounts to preempting the discovery process.

Arguably, the objective of an equitable remedy is to aid a litigant in pursuing justice in any manner reasonably necessary, and this should encompass any ancillary purposes as long as they pertain to redressing an inequity. However, the central idea underpinning the Norwich order is to nudge the applicant just over the bar set for launching a lawsuit. This approach is consonant with the principle that an equitable remedy should be a remedy of last resort. As Jeffrey Berryman states, "equity is only available after proof that the common law is deficient or inadequate." (83) The law recognizes that a plaintiff may not have all the pieces of the litigation puzzle; the discovery process is intended to address that deficiency. There is no reason to duplicate this activity so that it is available both pre-action and post-action.

While British Steel removed the itnention to sue requirement, this act by the House of Lords appears to have been a hasty pronouncement in reaction to the facts of the case, in which the respondent hinted that the applicant may dismiss an employee instead of suing him for breach of confidence. The intention of the Law Lords appears to have been that the remedy should be indifferent as to when a plaintiff may launch suit--now, later, or never--but that its purpose remains to assist an aggrieved person to obtain redress where a cause of action exists. In any event, despite British Steel, a court is unlikely to grant a request for disclosure if the purpose is totally aberrant, or if it is not related to litigation in some way.

The purpose of a Norwich order, therefore, should be limited to its original mandate: to enable the commencement of a particular action once the applicant has demonstrated a bona fide case. If this mandate is followed, the range of material to be divulged will not be great. It is well-established that the threshold for starting an action is set low to ensure that plaintiffs get their day in court; only the material facts demonstrating the elements of the cause of action need to be pleaded. The typical Norwich applicant will be seeking that missing piece of the litigation puzzle to nudge him or her over the threshold for starting a claim. Aside from the identity of the perpetrator, the material fact sought by the applicant is unlikely to be anything more than an eyewitness account of the events that transpired, as a statement of claim cannot contain pleadings of evidence. (84) It should not be unduly burdensome for the respondent to provide information in this scenario, and hence there is no reason to distinguish between the name of the defendant and other information that may assist in starting litigation. It is only due to the vagaries of fate that an applicant is ignorant of the name of the wrongdoer rather than, say, the exact date that the tort occurred.

Therefore, under this analysis, the scope of disclosure need not be restricted to identity, provided that the information is absolutely crucial to launching suit. Nevertheless, there are several reasons why courts should keep the scope of disclosure narrow. First, as a matter of fairness, it would be inequitable to impose too onerous a burden on the respondent, who is after all not the wrongdoer but an innocent third party. Recall that one of the purposes of the mere witness rule is "to prevent innocent persons from incurring trouble and expense," and that the rule was suspended only for extraordinary Norwich-type situations. Notwithstanding Lord Reid's offhand remark about "full information," the fact that disclosure was confined only to identity in Norwich was influential in allowing the remedy.

Second, a relaxed scope of disclosure will heighten the risk of exposing the respondent's confidential information. Although the respondent may plead privilege or other reasons, there is always the possibility that the information may not qualify for suppression because it does not satisfy the rigorous requirements for privilege. The names of suppliers or customers, for instance, are usually considered confidential; however, this information was acquired in Guinness. Naturally, information should be divulged if it is indeed a material fact needed to commence litigation, but there is always debate over what is material. The point is that a liberal rather than restrictive approach may lead to a creeping expansion in scope.

Finally, while a court has the inherent jurisdiction to decline particular requests for disclosure, to be continually adjudicating questions of privilege is to essentially engage in the discovery process pre-emptively, which will bring to fruition Hoffmann J.'s fear that the Norwich order will gradually become a general right of discovery.

It should be noted that the other option for disclosure, a John Doe order, is in fact a general right of discovery, though technically it is a post-action right. However, the Federal Court of Appeal in BMG saw fit to superimpose Norwich principles on the John Doe method, (85) a move that may herald the end of what was seen to be an easier route to a non-party's information. The court expressed a particular concern about an overly broad scope of disclosure, although the apprehension in this case pertains to incidental disclosure. In any event, the court's concern about scope was subsumed under the general question of whether disclosure ought to be permitted, a question which was resolved through a balancing of interests. Once again, the opportunity to rein in the scope of disclosure was not realized, though the court could hardly be faulted as it was not a serious issue in BMG.

One may argue that the current approach is entirely satisfactory as the balancing stage should allay any concerns about scope. That is certainly true on an individual case basis, but general rules of application are useful in that they lend predictability to the process. As the next section shows, back loading issues into the balancing stage has the disadvantage of yielding divergent and unexpected results. Furthermore, prescribing an overarching approach to the problem helps crystallize the interests at play. For this particular issue, it can be seen that compellable information need not be restricted to one thing, but that, conversely, a lenient approach will be both undesirably burdensome and intrusive for a non-party. It is submitted that purpose should dictate the scope of disclosure, and that this in turn suggests that a restrictive approach is warranted.

VI BALANCING OF INTERESTS

The recurring theme of this article is to advocate liberal parameters for the issuance of a Norwich remedy while limiting the scope of disclosure. Thus the final stage of a Norwich application may be where adjudication truly begins. Due to the overwhelming certainty of tortious conduct in the original Norwich decision, it was perhaps not the best factual scenario to create a template for the balancing process. Nevertheless, the House of Lords considered the question of prejudice to the importer and held that it did not lie in the mouths of tortfeasors to complain. Indeed, the Law Lords found that the identity of the importer was neither secret nor confidential as it had passed through many hands and was contained in bills of lading. They rejected the contention that disclosure would discourage candour, as not only was this unlikely, but "any honest trader who was disturbed at the thought [of disclosure] in circumstances such as exist here would be a most unreasonable man." (86) In the absence of any remaining public policy that would militate against disclosure, it was an easy step from them to order disclosure.

Subsequent cases presented much more confounding situations. Historically, an equitable remedy may be denied on several grounds, including unclean hands, laches, abandonment or acquiescence, and hardship. Curiously, none of these traditional defences has ever been used. There have been primarily two objections to a Norwich order: privacy and confidentiality. The examination of the final stage is thus divided between these twin concepts and how each weighs against the applicant's right to pursue legal redress.

Privacy

Privacy is not to be confused with confidentiality, though the two may overlap. Privacy is an endogenous concept that relates to the intrinsic sensitivity of certain matters that compel concealment. (87) Confidentiality, on the other hand, is an exogenous concept that arises from an agreement between two parties, either explicitly or implicitly, to keep a piece of information secret. (88) Thus, medical records are automatically considered private without any agreement having been made, while a shopping list of no significance may be rendered confidential by virtue of an agreement.

Until BMG, Norwich disputes were exclusively concerned with confidentiality because the respondent, as an intermediary, could not assert privacy rights since the requested information did not intrude upon her personal realm. The respondent could only claim that he or she had promised the unknown party to keep this person's identity or information secret. The "intermediary obstacle" was overcome in BMG when an intervener was allowed to speak on behalf of the purported wrongdoers' rights to privacy. BMG is also an unprecedented decision in another respect: it is the first Norwich application to encounter the Personal Information Protection and Electronic Documents" Act (PIPEDA). (89)

The issue of privacy dominated the BMG decision, as shown by the beginning of the judgment when Sexton J.A. stated, "this case illustrates the tension existing between privacy rights of those who use the Internet and those whose rights may be infringed or abused by anonymous Internet users." (90) To adjudicate the contest between these two rights, Sexton J.A. started by articulating some of the concerns held by privacy advocates. Internet records, he observed, enable an investigator to learn among other things a person's occupation, wealth, shopping history, and reading habits. By piecing these morsels of information together, a person's views and beliefs can be subject to "untenable scrutiny." (91)

On the other side of the equation is the long recognized need to protect intellectual property. Copyright law provides an incentive for innovators to create. As Sexton J.A. noted, "if [innovators] are robbed of the fruits of their efforts, their incentive to express their ideas in tangible form is diminished." (92) Indeed, both Glaxo and Norwich were cited as "compelling" (93) support for this societal concern. Sexton J.A. concluded that "although privacy concerns must also be considered, it seems to me that they must yield to public concerns for the protection of intellectual property rights in situations where infringement threatens to erode those rights." (94) However, due to an evidentiary defect in the case, disclosure was denied.

The BMG court's conclusion has miscast a Norwich application as primarily a battle between privacy and intellectual property, leading music publishers to declare the decision as a vindication of copyright protection. (95) Sexton J.A.'s judgment creates the impression that privacy was ceded in this instance only because of the strong policy reasons to uphold intellectual property rights when in fact privacy must yield in every case to an injured party's quest for discovery, irrespective of the kind of right the claimant holds. A court need not explain, for example, why defamation deserves to be a tort or the policy implications of a negligence action, before granting an order for disclosure: otherwise, it would lead to the indefensible implication that one type of tort is more worthy of a Norwich remedy than another. The plaintiff's right, be it protection of intellectual property or a right against trespass, is ipso facto assumed to be worthy of legal protection where there is a valid cause of action. The type of tort at issue is simply irrelevant.

More importantly, privacy cannot be a shield for wrongdoing. One cannot commit a crime or tort and assert the right to privacy as a means of staying hidden. It is important to remember that a Norwich order simply seeks the name of the wrongdoer; if a person's name has a freestanding expectation of privacy, then no one will ever be held accountable for his or her actions as no prosecution could be issued in the person's name without consent. Certainly there are instances where a person's name could be withheld for privacy reasons, but this must be with respect to certain activities that may be personal in nature or potentially embarrassing--for example, a patient's name may be withheld to conceal the fact that he or she is undergoing psychotherapy. There simply cannot be an expectation of privacy when a person is robbing a bank, selling counterfeit goods, or participating in any other type of criminal activity.

The main thrust of the privacy argument in BMG concerns the ability to profile an internet user through his or her computer trails. Just as it would be absurd to argue in Norwich that a disclosure order should be denied as it might reveal other documents in the importer's possession, the issue of user profiling (96) is equally irrelevant to the analysis in BMG. Justice Sexton was right to caution the respondents against releasing more information than necessary about its customers, but perhaps it should have been stressed that since a Norwich order only seeks the identity of the file sharer, the contents of that person's computer are not germane. Even if the contents were to be revealed to a computer technician for the purposes of ascertaining identity, that information is protected by the implied or deemed undertaking rule. (97) However, had the contents been included in the subject of the order--in other words, included in the scope of the order--then privacy would truly be an issue. As it stands, BMG should not be regarded as a case about interact privacy.

Finally, while Sexton J.A. stated, "privacy concerns may be met if the court orders that the user only be identified by initials, or makes a confidentiality order," (98) this trepidation is unwarranted. The practice of using initials is employed to protect young offenders, sexual assault victims, and those involved in family law disputes. Except for these sensitive areas, the name of a respondent has always been public information in accordance with the principle of open courts. There is, after all, no freestanding right to be immune from lawsuits.

Confidentiality

In Norwich applications, confidentiality is a precept more commonly invoked by respondents than privacy. To assert a claim of confidentiality, the information in question must either fall into a category of class privilege or be subjected to a case-by-case analysis to qualify for common law privilege. The categories of class privilege are few--the more well-known ones being solicitor-client privilege, litigation privilege, settlement privilege, and Crown privilege (99)--and no Norwich application to date has ever had this type of claim. Class privilege, in any event, will result in the omission of the balancing process since a court must automatically deny disclosure once the claim of privilege is made out. Common law privilege, on the other hand, requires a Canadian court to apply the Wigmore test, which stipulates that the following conditions must be met: (1) the communications must originate in a confidence that they will not be disclosed; (2) this element of confidentiality must be essential to the full and satisfactory maintenance of the relation between the parties; (3) the relation must be one which in the opinion of the community ought to be sedulously fostered; and (4) the injury that would inure to the relation by the disclosure of the communications must be greater than the benefit thereby gained for the correct disposal of litigation. (100)

The first two conditions of the Wigmore test are of little concern, as they merely verify the existence of a highly valued confidence. The third condition is slightly more contentious, but with respect to the relationships that commonly arise in Norwich applications, these tend to be normal commercial relationships (e.g. employer-reference provider, customs-importer) that can hardly be said to be unsavory or undesirable. The fourth prong of the Wigmore framework is clearly the decisive factor and it is where the real balancing of interests is carried out. The bar for passing this condition has been set extremely high since A.M. v. Ryan. (101) In A.M., the Supreme Court of Canada denied total suppression of the plaintiff's psychiatric notes, on the basis that the material in the records may be pertinent to the correct disposal of litigation. This was a case in which privacy and confidentiality overlapped: the psychiatric notes were created in confidence, and contained the extremely sensitive subject of the plaintiff's sexual history. The fact that both these interests could not outweigh the defendant's right to full answer and defence is an indication of how difficult it is to pass the fourth prong of the Wigmore test.

One may even posit that if the detriment suffered from disclosure is limited to the respondent and a few other parties--in other words, there is only a private interest in non-disclosure and not a public one--there is little chance that the privilege will be granted, given the high threshold set by A.M. In the context of Norwich applications, this theory may explain why there have been no vigorous assertions of confidentiality in intellectual property infringement cases. The dearth of confidentiality claims has been somewhat surprising given that trade secrets and non-disclosure agreements are common in the practice of intellectual property. However, it is almost certain that commercial competitiveness and other business reasons will be insufficient to defeat the interest in justice being done, given that these motives are far less sensitive or dire than a person's sexual history. In fact, it is difficult to imagine any scenario where the personal interests of the respondent will be more exigent than the ones in A.M.

Logically, confidentiality claims with societal implications should fare better at meeting the Wigmore test since they have a greater countervailing weight against the "correct disposal of litigation" interest. However, the case law has not always borne this out as the interests of the parties are difficult to frame properly.

Breach of Confidence

So far, it has been argued that as long as tortious behaviour can be established, details of the wrongdoing, including the type of tort, are largely irrelevant to the determination. Such issues are to be decided at trial. There is an exception to this position: disclosure should not be granted where the tort is a breach of confidence, specifically a breach of confidence that aims to expose governmental or corporate malfeasance. In other words, disclosure should not be granted in "whistle blowing" cases.

The rationale for this exception lies in the well-recognized societal interest in uncovering misconduct and in suppressing the chilling effect that occurs when a whistle blower is unmasked. As Lord Denning remarked in British Steel, if newspapers were to reveal their informants their sources would dry up and thus wrongdoing would not be disclosed, charlatans would not be exposed, unfairness would go unremedied, and misdeeds in the corridors of power (either in companies or in government departments) would never be known. (102) He observed that a policy of protecting journalists' sources has "for over 100 years ... been a settled rule" (103) in England and that although the informant may be guilty of wrongdoing, "this is not a reason why his name should be disclosed." (104) He further stated that similar endorsements for whistle blower protection can be found throughout Anglo-American jurisprudence. (105)

Hence it is surprising to find that in the actual instances where Norwich was invoked to unmask an informant, the cases have consistently resulted in disclosure. While the decisions vary in their reasons for ordering disclosure, it would appear that the courts may not have understood or appreciated a concept crucial to "journalistic privilege": that the court must prima facie give protection to the wrongdoer and only grant relief to the applicant in exceptional circumstances. Instead of this "reverse balancing process," the Norwich cases appear to be adjudicated on the traditional premise that the wrongdoer must be held accountable unless the wrong is justified. This orthodox approach led the court to the wrong conclusion in several cases.

To wit, if one were to accept the importance of journalistic privilege, one must accept the occasional necessity of breaching confidence to expose iniquity. This fundamental principle was not truly recognized in the first two Norwich applications before the House of Lords. In the first case, British Steel, the court ordered revelation of the person responsible for leaking confidential papers of the British Steel Corporation to Granada Television. This leak was during a bitter strike by steel miners in the early 1980s, and the documents contained a settlement offer that may have been suppressed by government interference. The news appeared to be the very type of explosive information that creates the need for whistle blowing, yet the court was unconcerned by the public interest in the story. Instead, there was outrage at the fact that Granada Television tampered with the documents to conceal the informant's identity, and outrage at the possibility of leaving the informant "undetected and still apparently a trusted employee [of British Steel], enjoying his pay and earning his pension." (106) Some of the judges even showed distaste for the television program on which the papers were discussed. For example, Lord Denning called the conduct of the presenter "deplorable," (107) and Sir Robert Megarry V.C. went so far as to dismiss it as bereft of public interest. (108) Lord Justice Templeman did not even appear to fully grasp the concept of whistle blowing:

   If information is truly confidential it does not cease to be
   confidential merely because it relates to matters of public
   interest. In the present case the B.S.C. documents and the contents
   of those documents which were quoted by Granada were truly
   confidential albeit that they related to matters of public concern
   and Granada was not entitled to conceal the source and break the
   confidence. (109)

The X Ltd. v. Morgan-Grampian Ltd. (110) decision showed a similar judicial repulsion for a "snitch" going unpunished. In this case, an employee of a major corporation leaked refinancing plans to the press to show that a supposedly stable company was in financial difficulty. According to the court, the informant was a "ticking ... time bomb" (111) and "if an employer of a large staff is suffering grave damage from the activities of an unidentified disloyal servant, it is undoubtedly in the interests of justice that he should be able to identify him in order to terminate his contract of employment." (112) What may have been the ultimate deciding factor in the case, though, was the financial viability of the plaintiff. If the refinancing were not granted, a significant number of jobs would have been lost, and thus this case was "not a situation in which the court should be or be seen to be impotent in the absence of compelling reasons." (113)

In both these cases the court not only exhibited excessive umbrage at the informant's lawbreaking, it would seem that it also misapprehended its role in this kind of adjudication. It is simply not the judiciary's role to act as a corporate manager, and so it is not its task to root out a disloyal servant. For the same reason, the threat of insolvency and job cuts in X Ltd. should not have figured into the decision. The court's undue concern for the plaintiff's plight results from its failure to use a "reverse balancing process" that foremost protects the informant's identity.

Indeed, in the wake of the British Steel decision, the British legislature sought to strengthen protection for informants by passing the Contempt of Court Act 1981, in which s. 10 essentially mandates a "reverse balancing process":

   No court may require a person to disclose, nor is any person guilty
   of contempt of court for refusing to disclose, the source of
   information contained in a publication for which he is responsible,
   unless it be established to the satisfaction of the court that
   disclosure is necessary in the interests of justice or national
   security or for the prevention of disorder or crime. (114)

The Act thus allows disclosure only in exceptional circumstances, but the phrase "necessary in the interests of justice" has proven to be its undoing. Lord Bridge in X Ltd. construed the phrase to mean, "that persons should be enabled to exercise important legal rights and to protect themselves from serious legal wrongs," (115) which in his view implies a balancing process. (116) This approach effectively returns the adjudication to the traditional weighing of interests, which led to Lord Bridge's conclusion that the detection and termination of a disloyal servant to be the weightier concern in the case. Lord Donaldson interpreted "necessary in the interests of justice" more narrowly to mean "the administration of justice generally," (117) but this too ultimately requires a balancing exercise, as the proper weighing of interests is necessary to ensure the public confidence in the administration of justice Lord Donaldson held that journalistic privilege in this case could not outweigh the threat of the "ticking time bomb" unless the leaked documents exposed malfeasance of some sort. In this case, "no iniquity was shown," (118) "no shareholders have been kept in the dark" (119) and hence the information did not tip the scales in favour of non-disclosure.

Lord Donaldson's review of the documents' public interest quotient raises a question: should the newsworthiness of the information be a factor in the determination? At first blush, newsworthiness appears to be a legitimate concern as it seems to be the very basis for shielding the informant, but this is a misconception of journalistic privilege. It is not so much the information that is at stake as the need to combat the chilling effect. As Laws L.J. at the Court of Appeal explained in Ashworth:

   [T]he chilling effect ... is in no way lessened, and certainly not
   abrogated, simply because ... the information actually published is
   of no legitimate, objective interest. Nor is it to the least degree
   lessened or abrogated by the fact (where it is so) that the source
   is a disloyal and greedy individual, prepared for money to betray
   his employer's confidences. The public interest in the
   non-disclosure of press sources is constant, whatever the merits of
   the particular publication, and the particular source. (120)

This dictum is perhaps better understood by comparison to solicitor-client privilege, where the content of the communications between a lawyer and client is irrelevant to upholding the privilege. Similarly, to encourage informants to come forward they must be assured of anonymity a priori without being afraid that their information may not stand up to scrutiny.

Moreover, as Lord Donaldson's own assessment demonstrates, the judiciary is simply not an infallible arbiter of what is relevant, important, or interesting. A refinancing plan may appear to be a mundane private affair, but it may have enormous ramifications in the business world. Just as the court should not act as a corporate manager, it should not usurp the role of journalist or editor. This contention was affirmed in Financial Times Ltd. v. Interbrew SA, (121) but it left the court faced with a bit of a dilemma. Interbrew was a case in which the press had received doctored copies of a brewery's plan to take over another company where some parts of the plan were false. The newspapers did not print the inaccurate portions but did report the brewery's intended takeover bid. The court was clearly concerned about the "lethal cocktail of fact and falsehood," (122) but as it had established that it should not assess the value of the information, it settled on another criterion to hold the informant responsible: purpose. Because the informant in Interbrew had obviously altered the documents to damage the brewery's reputation, Sedley J. found this malicious purpose tipped the scales towards disclosure.

The issue of purpose is highly dubious in terms of relevance. Purpose or motive may be relevant in deciding an informant's liability, but like newsworthiness, it has neither any bearing on the chilling effect nor constitutes a pressing need for compelling disclosure. Justice Sedley reasoned that purpose is relevant in that there is a public interest in ensuring only truthful and accurate information gets published, (123) but this seems to contradict the view that the quality of the information is irrelevant to the determination. One may also question how a court can accurately and reliably divine the purpose of a silent and unknown person.

The real issue posed by Interbrew was one raised by Lord Denning in British Steel: if a publisher knowingly relies on an untrustworthy source, has it forfeited its entitlement to journalistic privilege? According to Lord Denning, the answer is yes, (124) but a strong case can be made that a reporter's source should still be accorded protection. First, disclosing the identity of a source punishes directly the informant and only indirectly the publisher. It may be effective in deterring mendacious informants, but since it also discourages truthful informants from coming forward, the overall effect is undesirable. Disclosure is not required to discipline publishers as they already have the incentive to be careful due to the threat of defamation lawsuits. Second, defamatory statements should be dealt with in a defamation action; a Norwich order should not be used to pre-judge such a claim, nor should it be used to mete out punishment to a reckless publisher. Third, unlike the situation in other Norwich claims, the plaintiff still has recourse in this situation if he or she is injured, as he or she can sue the publisher for defamation. If publication will result in detriment that cannot be cured by the time of trial, the plaintiff can obtain an injunction against publication. An injunction, for example, was obtained in X Ltd.

If neither newsworthiness nor purpose is relevant to the adjudication, then on what basis is a court to decide if disclosure is "necessary in the interests of justice"? According to Ashworth, the new mantra today is "necessity and proportionality." These two concepts mean nothing

more than that "the disclosure ordered was necessary and not disproportionate as a response" (125) to the reason for disclosure, and furthermore, the distinction between the two is slight: "the requirements of necessity and proportionality are here separate concepts which substantially cover the same area." (126) Practically speaking, the two terms add little to the adjudication except to indicate that disclosure must be a remedy of last resort, or near last resort. The real delimiter comes from the stipulation in Goodwin v. United Kingdom, (127) the decision from which these terms were extracted, that the reason for granting disclosure must be "a pressing social need." (128)

This is a commendable requirement, as it states explicitly that the applicant's individual interests should not outweigh the societal interest in protecting informants. This test was illustrated in Ashworth, a case in which a mental asylum sought the name of the employee who had leaked the health records of an infamous murderer to the press. The court ordered disclosure not because the threat of a disloyal servant was sufficient to do so, but because expert testimony had indicated that psychiatry, more than any other branch of medicine, depends on a trusting relationship between therapists and patients and that the situation was exceptional as the care of mental patients was fraught with difficulty and danger. (129)

The balancing exercises in these English cases have been a study in evolution, starting from a traditional stance that demanded justification for the wrongdoing, and moving to the view that only exceptional reasons should pierce the anonymity of informants. One obstacle in this progression has been the failure to recognize that the applicant's private interests should not be placed on par with the public interest in protecting journalists' sources, a failure that persisted even with the codification of a "reverse balancing process." Certainly it should be acknowledged that a breach of confidence of this species is particularly challenging to adjudicate--"a difficult problem of weighing one public interest against another," (130) as Lord Bridge commented--but the recognition that it should be the battleground of competing public interests is key to understanding why this tort must be distinguished from others that rely on the Norwich jurisdiction.

Defamation

Defamation claims that trigger the use of the Norwich jurisdiction tend to fall into two categories. The first group shares the problem of internet defamation. These are, on the whole, rather straightforward cases, where a defamatory statement is posted online, and the subject of the statement demands the name of the writer from the internet service provider. Irwin Toy exemplifies such an action and the judgment shows how simple the balancing of interests can be. Since the plaintiff can produce the impugned statement and make a bona fide, if not prima facie, case for defamation, the court will easily grant the request since there is no prejudice to the internet service provider. Therefore, although the internet may be a novel medium for the broadcast of defamatory statements, this type of defamation is no different than an anonymous article in a newspaper, and thus presents no novel issues.

The second group of cases presents a much more complex problem. These are the character reference cases such as P. v. T. Ltd., Kenney, and Straka. Unfortunately, two of these decisions are not helpful in understanding the interests that are to be weighed in defamation cases. Specifically, Kenney never reached the balancing stage, and the court in P. v. T. Ltd. did not engage in any balancing exercise beyond finding the applicant's position "intolerable." (131) Only Straka underwent a full Wigmore analysis.

The facts of Straka fall into the classic pattern of an employee who failed to get a position because of deprecating character references. The applicant was an anesthesiologist who was denied hospital privileges due to reference letters that were critical of his competence. He did not deny that confidentiality attached to the letters, nor the fact that the relationship between requesters and responders of hospital reference letters was one that, in the opinion of the community, should be sedulously fostered. Because of these concessions by the applicant, the court passed over the first and third criteria of the Wigmore test and proceeded to examine only the second and fourth criteria. The second Wigmore step requires demonstration that confidentiality is essential to the full and satisfactory maintenance of the relation between the parties. Here, the court accepted on the evidence of the hospital's Chief of Staff that confidentiality is essential to candour in reference letters. (132)

For the final criterion, the court methodically split the analysis into the injury that would arise from disclosure, and the benefit of the correct disposal of the litigation. Justice Morden found that disclosure would have a serious impact on "the integrity and thoroughness of the methods hospital boards use" in appointments, and moreover, "the lives and health of members of the public are directly affected." (133) The prejudice to the applicant, on the other hand, was alleviated by the fact that he had recourse to the Public Hospitals Act, (134) which allowed review of hospital appointments, whereby the respondents would have to reveal the details of the reference letter, and possibly the identity of its writer. Ironically, had the applicant maintained his original position that he only sought to clear his name and not to obtain a position at the hospital, Morden J.A. would have ruled out this alternative. (135) In any event, the availability of this option tipped the scales in favour of upholding privilege.

Straka is a bit unsatisfactory as a precedent since most jobseekers do not have the luxury of a board review. Moreover, the hospital setting of the case, with its life-and-death implications, may not translate well to other employment situations. Nevertheless, Straka drew out the salient interests in character reference cases, namely the need for candour in the references and how this need for candour is systemic across hospital boards. This language, of course, reflects the classic rationale for privilege--namely, that parties can rely on their communications being kept confidential.

In fact, the need for confidentiality is highly important in the case of honest character references. Character references have a social significance that should not be underestimated. They are vital not only to jobseekers and employers, but also to everyone in the work force because they are a means to screen out violent, dangerous or incompetent employees. (136) Unfortunately, reference providers are prone to the same chilling effect as whistleblowers. Over the past twenty years, fears of litigation from disgruntled jobseekers has resulted in employers either declining to give references or providing opinion-less "name, rank and serial number" references. (137) This unhappy outcome has in turn provoked the enactment of statutes that furnish partial immunity from lawsuits to reference providers. (138) This legislative reaction underscores the tremendous public interest in protecting reference providers.

Ergo, for the same reason that privilege is extended to journalists' sources--to combat the chilling effect that threatens the availability of future information--there should be a judicial disinclination towards uncloaking reference providers. In fact, a blanket or near-blanket privilege may well be more workable in the context of character references than journalists' sources, as the range of adverse consequences is narrower. There is no concern here of a disloyal servant running amuck in the plaintiff's organization, and there are no national security or criminal investigations at stake. Instead, a refusal to disclose only produces a disappointed jobseeker or employee, but such a person takes on the risk of eliciting negative opinions whenever he or she applies for a job that requires references. This assumption of risk, it is submitted, should lessen the weight attached to the applicant's interests and tilt the balancing of interests towards the respondent.

Whatever form the protection takes, be it an absolute privilege or otherwise, the important point to recognize is that the dispute is not a purely private matter. The arrangement for a reference is essentially a social contract. The reference provider gains no direct benefit from giving the reference, except as an incentive for the employee to perform well to earn a positive review, and the expectation of reciprocal courtesy from other employers. Although it may seem at first glance to be a private affair between the jobseeker and the reference provider, the provision of character references is, in fact, part of a larger social agreement that facilitates job mobility and ensures effective hiring practices.

VII CONCLUSION

As Cardozo J. wrote, "the law of discovery has been invested at times with unnecessary mystery. There are few fields where considerations of practical convenience should play a larger role." (139) While Cardozo J. was addressing a pre-trial, but post-action, application for equitable relief, his comment is no less apposite for the pre-action equitable bill of discovery. Certainly the Norwich jurisdiction has been plagued with "unnecessary mystery," as manifest in the mere witness rule. It is argued that this arcane stricture should be banished as it restricts access to justice for no satisfactory reason. It is also argued that a strict adherence to the "only practicable source of information" dictum is inimical to the practical convenience advised by Cardozo J., and similarly undesirable. There is little justification for constricting the class of respondents on procedural grounds, since neither privacy nor inconvenience can outweigh the plaintiff's right to justice. For the same reason, the threshold for a Norwich application cannot be set too high lest it screens out potential actions prematurely.

Naturally, dispensation of Norwich orders cannot be an unfettered exercise either. As an equitable remedy, it should not evolve so far as to become a general right of discovery, as Lord Hoffmann advised. The scope of a Norwich order should be confined to its intended purpose of uncovering wrongdoers, and not as a means of pre-proceeding discovery. This necessarily means that a plaintiff's case should be firmly established, though it need not be as high as a prima facie standard. The intrusiveness of an interlocutory order for discovery can only be justified if it is clear that there has been tortious injury inflicted.

Although the particular prejudice to a respondent should always be considered, there is some force to the argument that certain promises of confidentiality warrant broad protection. Journalistic sources and employee references both serve socially important functions, and should be guarded against a chilling effect. Whether a blanket privilege should be extended to these breaches of confidences is debatable, but there should be no controversy in urging greater recognition of the policy issues at stake.

The covert wrongdoer is a problem that is addressed in the Canadian legal system largely through criminal law. In tort law there was no mechanism to uncover tortfeasors until Norwich. The extension of the equitable bill of discovery against third parties created a powerful remedy that was soon applied to a variety of situations beyond patent infringement. As this discussion shows, the simplicity of a disclosure order belies some thorny issues about the conflicting rights of the plaintiff, the conduit, and even the wrongdoer. While these questions have not always been addressed satisfactorily by the courts, this article provides a new beginning to homogenize the use of Norwich orders.

* The author thanks Peter Ruby, David Crerar, and the editors of the University, of Toronto Faculty of Law Review for their assistance with this article

(1) [1974] A.C. 133 (H.L.) [Norwich].

(2) The rules of civil procedure that permit pre-action discovery are as follows: Ontario, Rules of Civil Procedure, R.R.O. 1990, Reg. 194, r. 31.10; Alberta, Alberta Rules of Court, Alta. Reg. 390/1968, r. 193; British Columbia and Yukon, Supreme Court Rules', B.C. Reg. 221/90, rr. 26(11) and 28; Manitoba, Court of Queen's, Bench Rules', Man. Reg. 553/88, r. 31.10; New Brunswick, Rules of Court, N.B. Reg. 82-73, r. 32.12; Newfoundland and Labrador, Rules of the Supreme Court, S.N.L. 1986, c. 42, Sch. D, r. 30.01; Northwest Territories and Nunavut, Rules of the Supreme Court, N.W.T. Reg. 010-96, rr. 219, 225, 229; Nova Scotia, Judicature Act, R.S.N.S. 1989, c. 240, and Civil Procedure Rules, r. 18.02(c); Prince Edward Island, Supreme Court Act, R.S.P.E.I. 1988, c.S-10, and Rules of Civil Procedure, r. 18.02(1); Saskatchewan, Queen's Bench Act, R.S.S. 1978, c. Q-l, and Queen's Bench Rules, r. 222A; Quebec, Art. 331 C.C.P.

(3) See Philip Services Corp. v. John Doe (24 June 1998), Hamilton Court File No. 4592/98; Ontario First Nations Partnership v. John Doe (3 June 2002), File No. 02-CV-229617-CM3 (Ont.S.C.J.); and Canadian Blood Services v. John Doe (17 June 2002), No. 02-CV-20980 (Ont.S.C.J.). These are just a sample of the many "John Doe" applications.

(4) (2005), 252 D.L.R. (4th) 342 [BMG].

(5) As explained in detail later, the Wigmore test is used to determine whether certain information of a party or witness may remain confidential.

(6) Norwich, supra note 1 at 173.

(7) Ibid.

(8) The mere witness rule was first cited in Plummer v. May (1749) 1 Ves. Sen. 426.

(9) Ibid. at 175.

(10) Norwich, supra note 1 at 205.

(11) Ibid. at 175.

(12) Ibid. at 190.

(13) Nfld. & P.E.I.R. 341 [Johnston].

(14) Glaxo Wellcome PLC v. Minister of National Revenue, [1998] 4 F.C. 439 [Glaxo].

(15) See footnote 46 for the cases that comprise most of the reported requests for Norwich orders in Canada, in addition to Re Comeau (1986), 77 N.S.R. (Zd) 57; Straka v. v. Humber River Regional Hospital et al (2000), 51 O.R. (3d) 1 [Straka]; and Irwin Toy, supra note 56.

(16) [1980] 3 All E.R. 353 (C.A.) [Bankers].

(17) [1981] A.C. 1096 (H.L.) [British Steel].

(18) Jeffrey Berryman, The Law of Equitable Remedies (Toronto: Irwin Law, 2000) at 3.

(19) I.C.F. Spry, The Principles of Equitable Remedies (Agincourt: Carswell, 1997) at 1.

(20) Ibid. at 4-25.

(21) B.M. McLachlin, "The Place of Equity and Equitable Doctrines in the Contemporary Common Law World: A Canadian Perspective" in D. Waters, ed., Equity, Fiduciaries and Trust (Scarborough: Carswell, 1993) at 37.

(22) (1749), 1 Ves. Sr. 426 (Ch.).

(23) Norwich, supra note 1 at 174. A duces tecum is a writ directing a person to appear in court and to bring a document described in the writ.

(24) Ibid. at 157.

(25) Ibid.

(26) Ibid. at 175.

(27) Ibid. at 180.

(28) Ibid. at 201.

(29) Ibid. at 174.

(30) Ibid.

(31) Ibid.

(32) Ibid. at 188.

(33) Ibid. at 203.

(34) Ibid. at 166.

(35) Ibid. at 203.

(36) [2002] 4 All E.R. 193 [Ashworth].

(37) Ibid. at para. 58.

(38) [1966] 2 Q.B. 414.

(39) (1990), 57 C.C.C. (3d) 1.

(40) Ibid. at para. 18.

(41) Norwich, supra note 1 at 203.

(42) At common law there is no general duty to rescue someone from peril. See Horsley v. MacLaren, [1972] S.C.R. 44.

(43) "Judicial parsimony" refers to the concept that a purposeless law or rule should be abolished. See Pressed Steel Car Co. v. Union Pacific Co., 240 F. 135 (S.D.N.Y. 1917).

(44) Norwich, supra note 1 at 205.

(45) Johnston, supra note 13 at paras. 34-35.

(46) See Leahy v. A.B. (1992), 113 N.S.R. (2d)417 at para. 11; Kenney, supra note 47 at para. 33; National Bank of Canada v. Mann, [1999] O.J. No. 3064 (Ont. Gen. Div.) (QL) at para. 25; Alberta (Treasury, Branches) v. Leahy (2000), 270 A.R. 1 at para. 106;; Canadian Derivatives Clearing Corp. v. EFA Software Services Ltd., 2001 ABQB 425, [2001] A.J. No. 653 at para. 37; Glaxo, supra note 14 at para. 24; BMG, supra note 4 at para. 15.

(47) (1999), 64 B.C.L.R. (3d) 346; [Kenney].

(48) Ibid. at para. 30.

(49) Ibid. at para. 38.

(50) Ibid. at para. 37.

(51) [1997] 4 All E.R. 200 [P.].

(52) Ibid. at para. 29.

(53) Spry, supra note 19 at 202.

(54) P., supra note 51 at 4.

(55) BMG, supra note 4 at paras. 7-12 and 46-54.

(56) (2000), 12 C.P.C. (5th) 103 (Ont. S.C.J.) [Irwin].

(57) Ibid. at para. 17.

(58) See Jonathon T. Feasby, "Who Was that Masked Man? Online Defamation, Freedom of Expression, and the Right to Speak Anonymously" (2002) 1:1 C.J.L.T. 2, online: Canadian Journal of Law and Technology <http://cjlt.dal.ca/vol1_no1/>.

(59) R. Grant Hammond, "Interlocutory Injunctions: Time for a New Model?" (1980) 30 U.T.L.J. 241 at 249-254.

(60) [1975] A.C. 396 (H.L.).at 407.

(61) Glaxo, supra note 14 at para. 30.

(62) See Berryman, supra note at 24; and Jeffrey B. Berryman et al, Remedies: Cases, Materials and Commentary, 4th ed. (Toronto: Emond Montgomery, 2001) at 845.

(63) See John Leubsdorf, "The Standard of Preliminary Injunctions" (1978) 91 Harv. L. Rev. 524.

(64) Johnston, supra note 13 at paras. 21-30.

(65) BMG, supra note 4 at para. 34.

(66) Straka, supra note 15.

(67) Ibid. at para. 53.

(68) Spry, supra note 19 at 15.

(69) R.S.C. 1985, c. C-42.

(70) BMG, supra note 4 at paras. 46-53.

(71) Hammond, supra note 59 at 278-281.

(72) Norwich, supra note 1 at 199.

(73) Ibid. at 175.

(74) 2001 WL 1476319 (WeC).

(75) [1991] FSR 1.

(76) Ibid. at 5.

(77) [1980] 3 All E.R. 353 (C.A.) [Bankers].

(78) (No. 5), [1992] 2 All ER 911 (Ch.) [Arab].

(79) The first requirement is that the plaintiff must demonstrate a real prospect that the information may lead to the location or preservation of assets to which he is making a proprietary claim. The second requirement is that the third party should be entitled to the same specificity in the documents be is asked to produce as he would be if served with a subpoena.

(80) See Marc" Rich & Co. Holdings GmbH v. Krasner and others, [1999] EWCA Civ 581.

(81) Arab, supra note 78 at 914.

(82) A Mareva remedy is an injunction to freeze the assets of a party where there is a risk that the assets may be dissipated. See Mareva Compania Naviera SA v. International Bulkcarriers SA, [1980] I All E.R. 213.

(83) Berryman, supra note at 2.

(84) See for example Rules of Civil Procedure. R.R.O. 1990, Reg. 194, r. 25.06 (1).

(85) BMG. supra note 4 at paras. 28-34.

(86) Norwich, supra note 1 at 199.

(87) In Canadian jurisprudence, privacy is pertains to whether the information is related to a person's dignity, integrity, and autonomy. See R. v. Plant, [1993] 3 S.C.R. 281.

(88) See Lac Minerals Ltd. v. International Corona Resources Ltd.,[1989] 2 S.C.R. 574; and Cadbury Schweppes Inc. v. FBI Foods Ltd., [1999] 1 S.C.R. 142.

(89) S.C. 2000, c. 5. As it turns out, this statute has no impact on Norwich adjudications as the statute permits disclosure under court orders.

(90) BMG, supra note 4 at para. 1.

(91) Ibid. at para. 4.

(92) Ibid. at para. 40.

(93) Ibid.

(94) Ibid. at para. 41.

(95) "We welcome the court's confirmation that Canada isn't a piracy haven," says Canadian Recording Industry Association president Graham Henderson. "This was the key issue on which we appealed, and we're delighted that the court agreed with us." The president of Sony BMG Music (Canada) Inc. also said, "'We are encouraged that the Court of Appeal recognized the value of music and provided a roadmap for how we can defend that value from those who would seek to steal it." See Canadian Recording Industry Association, News Release, "'Canadian Recording Industry Welcomes Music Piracy Decision" (19 May 2005), online Canadian Recording Industry Association <http://www.cria. ca/news/190505_n.php>.

(96) User profiling refers to the practice of developing a computer user's profile by interring interests and habits from the websites that the user visited, as well as any information about the user available from other sources such as commercial databases.

(97)The implied undertaking rule is a common law requirement that all parties and counsels undertake not to use testimony or documents gained in the discovery process for purposes other than the proceeding in which the information was produced. The rule is called the deemed undertaking rule in Ontario as it has been codified in Rules of Civil Procedure, R.R.O. 1990, Reg. 194. r. 30.1.

(98) Ibid. at para. 45.

(99) See generally: Robert W. Hubbard, Susan Magotiaux, & Suzanne M. Duncan, The Law of Privilege in Canada, looseleaf (Aurora: Canada Law Book, 2006).

(100) See John Henry Wigmore. Evidence in Trials at Common Law, vol. 8 (Boston: Little, Brown and Company, 1961) at 527. The Wigmore test was first adopted in Canada in Slavntych v. Baker, [1976] 1 S.C.R. 254. The test is not used in the United Kingdom.

(101) [1997] I S.C.R. 157 [A.M.].

(102) British Steel, supra note 17 at 1129.

(103) Ibid. at 1128.

(104) Ibid. at 1129.

(105) Ibid. at 1128-1129. The cases cited by Lord Denning are as follows: Hennessy v. Wright (No. 2)(1888), 24 Q.B.D. 445n; Hope v. Brash, [1897] 2 Q.B. 188; Plymouth Mutual Co-operative and Industrial Society Ltd v. Traders' Publishing Association Ltd., [1906] 1 K.B. 403; Lyle-Samuel v. Odhams Ltd., [1920] 1 K.B. 135 and Georgius v. Oxford University Press, [1949] 1 K.B. 729; Baker v. F. & F. Investment (1972) 470 F. 2d 778, 785; Democratic National Committee v. McCord (1973) 356 F. Supp. 1394, 1396; and Branzburg v. Hayes (1972) 408 U.S. 665, 7.

(106) Ibid. at 1138.

(107) Ibid. at 1130.

(108) Ibid. at 1119.

(109) Ibid. at 1132.

(110) [1991] 1 AC 1 (H.L.) [X Ltd.].

(111) Ibid. at 28.

(112) Ibid. at 43.

(113) Ibid. at 28.

(114) (U.K), 1981, c. 49, s. 10.

(115) Ibid. at 43.

(116) Ibid.

(117) Ibid. at 28.

(118) Ibid. at 29.

(119) Ibid. at 29.

(120) Ashworth Security Hospital v. MGN, [2000] EWCA Civ 334 at para. 101 [Ashworth].

(121) SA [2002] EWCA Civ 274 at para. 54 [Interbrew].

(122) Ibid. at para. 28.

(123) Ibid. at paras. 54-56.

(124) British Steel, supra note 17 at 1130-1131.

(125) Ashworth, supra note 36 at para. 61.

(126) Ibid.

(127) (1966) 22 E.H.R.R.