Mattel, Inc. v. 3894207 Canada Inc.
SUPREME COURT OF CANADA
The judgment of McLachlin C.J. and Bastarache, Binnie, Deschamps, Fish, Abella and Charron JJ. was delivered by
1 Binnie J. – The BARBIE doll is said by the appellant toy manufacturer to be an iconic figure of pop culture. And so, within limits, it is. The sale of various BARBIE products annually exceeds $1.4 billion worldwide, representing 35 percent of the appellant’s sales. The appellant advises that Canadian girls aged 3 to 11 years are given an average of two BARBIE dolls per year. The appellant therefore opposes the respondent’s application to register trade-marks in connection with its small chain of Montreal suburban “Barbie’s” restaurants on the basis that use of the name (albeit in relation to different wares and services) would likely create confusion in the marketplace. On a casual acquaintance with both marks, it is contended, there is a likelihood that consumers would think that the doll people had something to do with a restaurant called “Barbie’s”. Or, as the appellant framed its point in a consumer survey by asking the following question “Do you believe that the company that makes Barbie dolls might have anything to do with the restaurant identified by this sign or logo?” (Emphasis added.)
2 Merchandising has come a long way from the days when “marks” were carved on silver goblets or earthenware jugs to identify the wares produced by a certain silversmith or potter. Their traditional role was to create a link in the prospective buyer’s mind between the product and the producer. The power of attraction of trade-marks and other “famous brand names” is now recognized as among the most valuable of business assets. However, whatever their commercial evolution, the legal purpose of trade-marks continues (in terms of s. 2 of the Trade-marks Act, R.S.C. 1985, c. T-13) to be their use by the owner “to distinguish wares or services manufactured, sold, leased, hired or performed by him from those manufactured, sold, leased, hired or performed by others”. It is a guarantee of origin and inferentially, an assurance to the consumer that the quality will be what he or she has come to associate with a particular trade-mark (as in the case of the mythical “Maytag” repairman). It is, in that sense, consumer protection legislation.
3 The appellant advises that the name BARBIE and that of her “soul mate”, Ken, were borrowed by their original designer from the names of her own children. The name, as such, is not inherently distinctive of the appellant’s wares. Indeed, Barbie is a common contraction of Barbara. It is also a surname. Over the last four decades or so, however, massive marketing of the doll and accessories has created a strong secondary meaning which, in appropriate circumstances, associates BARBIE in the public mind with the appellant’s doll products.
4 The appellant’s argument is that the trade-mark BARBIE now transcends the products which originally it served to distinguish. Moreover, the appellant says, its fame goes beyond wares and services aimed at girls in the 3- to 11-year-old age group (its primary market) to the diverse products set out in various of its registrations (“[c]ologne, hand lotion and body lotion”), food products (“[s]pices, breads, cakes, cereal, coffee, crackers, flour, herbs, pies, ice cream, pizza”), as well as bicycles, backpacks, books and construction pads. Of course, nothing prevents the appellant from using its BARBIE trade-mark to boost (if it can) sales of everything from bicycles to cologne, or for that matter lawn mowers and funeral services, but the question is whether the appellant can call in aid trade-mark law to prevent other people from using a name as common as Barbie in relation to services (such as restaurants) remote to that extent from the products that gave rise to BARBIE’s fame.
5 Unlike other forms of intellectual property, the gravaman of trade-mark entitlement is actual use. By contrast, a Canadian inventor is entitled to his or her patent even if no commercial use of it is made. A playwright retains copyright even if the play remains unperformed. But in trade-marks the watchword is “use it or lose it”. In the absence of use, a registered mark can be expunged (s. 45(3)). There was no credible evidence that BARBIE has been used in Canada either by the appellant or by one of its licensees in connection with the services for which the respondent made trade-mark application, namely “restaurant services, take-out services, catering and banquet services”.
6 In opposition proceedings, trade-mark law will afford protection that transcends the traditional product lines unless the applicant shows the likelihood that registration of its mark will not create confusion in the marketplace within the meaning of s. 6 of the Trade-marks Act. Confusion is a defined term, and s. 6(2) requires the Trade-marks Opposition Board (and ultimately the court) to address the likelihood that in areas where both trade-marks are used, prospective purchasers will infer (incorrectly) that the wares and services – though not being of the same general class – are nevertheless supplied by the same person. Such a mistaken inference can only be drawn here, of course, if a link or association is likely to arise in the consumer’s mind between the source of the well-known BARBIE products and the source of the respondent’s less well-known restaurants. If there is no likelihood of a link, there can be no likelihood of a mistaken inference, and thus no confusion within the meaning of the Act.
7 The substance of the appellant’s argument (set out at para. 48 of its factum) is that “Mattel’s BARBIE trade-marks are famous in Canada and worldwide, instantly evoking the image of the brand in the minds of consumers. Having acquired such fame, marks such a…. BARBIE may not now be used in Canada on most consumer wares and services without the average consumer being led to infer the existence of a trade connection with the owners of these famous brands” (emphasis added). Some trade-marks may have that effect, but the Board found BARBIE’s fame to be tied to dolls and doll accessories. At this stage, its fame is not enough to bootstrap a broad zone of exclusivity covering “most consumer wares and services”. The Board was not required to speculate about what might happen to the BARBIE trade-mark in the future. It was required to deal with the respondent’s application on the facts established in the evidence.
8 The appellant relies on Professor J. T. McCarthy’s dictum, discussed below, that “a relatively strong mark can leap vast product line differences at a single bound” (McCarthy on Trademarks and Unfair Competition (4th ed. (loose-leaf)), at § 11:74). However, the evidence before the Board was that, for BARBIE, the present case was a leap too far, and that on the evidence prospective consumers will likely not infer that whoever owns the BARBIE doll trade-mark is associated in some way with the restaurants identified by the applied-for mark. This is a conclusion that was open to the Board.
9 As is permitted under s. 56(5) of the Act, the appellant sought to introduce fresh evidence before the applications judge about the likelihood of confusion but the proffered evidence was found to be unresponsive to the statutory test and on that account it was rightly rejected.
10 On this appeal, the Board’s decision should be upheld unless it is shown to be unreasonable. On the admissible evidence, the Board was not shown to be clearly wrong in its decision that the respondent restauranteur has established that it is not likely that prospective consumers will draw the mistaken inference. I would therefore affirm the reasonableness of the decision of the Board to accept registration of the respondent’s trade-mark and dismiss the appeal.
11 The respondent opened its first Barbie’s restaurant in Montreal in 1992. Over the years it has added two more (one of which subsequently closed). These are medium-priced “bar-and-grill” type operations, with meals including ribs, smoked meat, souvlaki, steak, chicken, seafood and pizza, much of which is barbecued (or, as the menu puts it, done on the “barbie-Q”), along with alcoholic beverages. There are also breakfast and lunch menus. The bulk of the business is sit-down clientele. The decor is predominantly geared to adults and the bar is an important feature of each location. The restaurants do offer a kids’ menu and a “kids eat free” promotion one night of the week but the Board found its target clientele to be adults. Compared to that of the appellant, it is a modest operation. Total restaurant sales for the six years 1992-1997 inclusive totalled about $11 million. Advertising expenses, including television, radio and newspaper advertising, amounted to about $616,000 for the same time period. In September 1993, an application (now assigned to the respondent) was made to register the trade-mark BARBIE’S & Design in association with “restaurant services, take-out services, catering and banquet services”. The applicant’s mark is prominently displayed on the exterior of the restaurants and on menus, napkins, matchbooks, receipts, order forms and business cards. The mark, as now applied for, looks like this:
12 There are clearly significant points of resemblance between the trade-marks of the appellant and the trade-marks applied for by the respondent. The appellant’s registered trade-mark looks like this:
13 The appellant’s trade-marks enjoy an extensive worldwide reputation for dolls and accessories primarily targeted at the market of 3- to 11-year-old girls. There are some adult collectors of BARBIE doll products. By 2001, sales, promotion, and advertising of BARBIE products across Canada generated annual sales revenues of approximately $75 million, annual licensing revenues of approximately $5 million, and annual advertising expenses of approximately $5 million. The appellant describes itself in these proceedings as being “in the business of building brand equity”. Licensing of the BARBIE trade-mark is commonplace and the wares or goods to which the trade-mark is attached are growing. The appellant sees licencing of the BARBIE trade-marks as an expanding and lucrative commercial opportunity. To date, BARBIE has not been used in Canada by the appellant or any of its licensees for restaurant services, take-out services, catering and banquet services. I generally will refer to the appellant’s trade-marks collectively as the BARBIE mark.
II. Relevant Statutory Provisions
14 See Appendix.
III. History of the Proceedings
A. Canadian Intellectual Property Office, Trade-marks Opposition Board (Member Herzig for the Registrar of Trade-marks) (2002), 23 C.P.R. (4th) 395
15 The Board found that both marks possessed a relatively low degree of inherent distinctiveness as they would be perceived as a nickname for or truncation of the name Barbara. Mattel’s mark was very well known in Canada when used in association with dolls and doll accessories. The trade-mark sought to be registered by the applicant restauranteur had established some reputation in the vicinity of Montreal but the length of time of use of the marks favoured Mattel. The nature of the opponent’s wares and the applicant’s services were “quite different”. Mattel’s target market is children and, to some extent, adult collectors, whereas the applicant is in the restaurant business and its primary target market is adults. The marks were essentially the same aurally and in the ideas they suggested, and the overall visual impressions of the marks were essentially the same after the fairly non-distinctive design feature of the applicant’s mark was discounted.
16 Though the opponent (now the appellant) in this case had attempted to demonstrate a connection between food and food-related products sold under its own BARBIE trade-marks and the applicant’s restaurant services, the Board did not accept the submission that there was any real connection between the two. Although the opponent was not required to show any instances of actual confusion, and it had not done so, the absence of such evidence was only one circumstance among many to be considered.
17 In the circumstances, the Board found that the respondent’s mark was not likely to be confusing with any of Mattel’s BARBIE marks at the relevant times. The Board rejected the opposition and allowed the registration.
B. Federal Court (Rouleau J.) (2004), 248 F.T.R. 228, 2004 FC 361
18 The applications judge observed that “[i]t cannot be automatically presumed that there will be confusion just because [Mattel’s BARBIE] mark is famous” (para. 40). The test was that of reasonable likelihood of confusion and the fame of a mark “could not act as a marketing trump card such that the other factors are thereby obliterated” (para. 40). All of the relevant factors listed in s. 6(5) of the Trade-marks Act had to be evaluated, and “[o]ne of the key factors in this case is the striking difference between the wares” (para. 17). He stated that “confusion is less likely when the wares are significantly different, even when the mark is well known” and that “when the wares are significantly different, this factor must be given considerable weight” (paras. 17-18). In this case, he stated “the nature of the wares, as well as the nature of the business of both parties, could not be more different” (para. 21) and “there is nothing about these restaurants that is suggestive of toys, dolls or childhood” (para. 20).
19 The applications judge rejected Mattel’s application to introduce fresh evidence in the form of a survey proffered to show the likelihood of confusion between the marks because in his view the survey “ha[d] some blatant and determinative shortcomings that undermine[d] its relevance considerably” (para. 27). The result was that the survey could not be used to establish the existence of a real likelihood of confusion. It only served to show that Mattel’s BARBIE trade-marks were indeed famous. There was no evidence of any concrete case of confusion despite the co-existence of the marks in the Montreal area for 10 years. In the result, the applications judge dismissed the appeal.
C. Federal Court of Appeal (Létourneau, Noël and Pelletier JJ.A.) (2005), 329 N.R. 259, 2005 FCA 13
20 The appellate court concluded that no error had been committed by the applications judge in rejecting the survey evidence because of the way the survey firm had framed its questions. At best, the survey results might be probative of a possibility of confusion, which falls short of the threshold of a reasonable likelihood of confusion. In the result, the conclusion of the applications judge was agreed with. The appeal was dismissed.
21 Trade-marks are something of an anomaly in intellectual property law. Unlike the patent owner or the copyright owner, the owner of a trade-mark is not required to provide the public with some novel benefit in exchange for the monopoly. Here, the trade-mark is not even an invented word like “Kodak” or “Kleenex”. The appellant has merely appropriated a common child’s diminutive for Barbara. By contrast, a patentee must invent something new and useful. To obtain copyright, a person must add some expressive work to the human repertoire. In each case, the public through Parliament has decided it is worth encouraging such inventions and fostering new expression in exchange for a statutory monopoly (i.e. preventing anyone else from practising the invention or exploiting the copyrighted expression without permission). The trade-mark owner, by contrast, may simply have used a common name as its “mark” to differentiate its wares from those of its competitors. Its claim to monopoly rests not on conferring a benefit on the public in the sense of patents or copyrights but on serving an important public interest in assuring consumers that they are buying from the source from whom they think they are buying and receiving the quality which they associate with that particular trade-mark. Trade-marks thus operate as a kind of shortcut to get consumers to where they want to go, and in that way perform a key function in a market economy. Trade-mark law rests on principles of fair dealing. It is sometimes said to hold the balance between free competition and fair competition.
22 Fairness, of course, requires consideration of the interest of the public and other merchants and the benefits of open competition as well as the interest of the trade-mark owner in protecting its investment in the mark. Care must be taken not to create a zone of exclusivity and protection that overshoots the purpose of trade-mark law. As Professor David Vaver observes:
On the one hand, well-known mark owners say that people should not reap where they have not sown, that bad faith should be punished, that people who sidle up to their well-known marks are guilty of dishonest commercial practice. These vituperations lead nowhere. One might as well say that the well-known mark owner is reaping where it has not sown when it stops a trader in a geographic ormarket field remote from the owner’s fields from using the same or a similar mark uncompetitively.
(D. Vaver, “Unconventional and Well-known Trade Marks”,  Sing. J.L.S. 1, at p. 16)
23 The purpose of trade-marks is to create and symbolize linkages. As mentioned, s. 2 of the Trade-marks Act defines “trade-mark” to mean
(a) a mark that is used by a person for the purpose of distinguishing or so as to distinguish wares or services manufactured, sold, leased, hired or performed by him from those manufactured, sold, leased, hired or performed by other…. .
To the same effect is art. 15 of the World Trade Organization’s Agreement on Trade-Related Aspects of Intellectual Property Rights, 1869 U.N.T.S. 299, which defines “trade-mark” in part as
[a]ny sign, or any combination of signs, capable of distinguishing the goods or services of one undertaking from those of other undertaking…. .
24 As the Court put it in Kirkbi AG v. Ritvik Holdings Inc.,  3 S.C.R. 302, 2005 SCC 65, a trade-mark is “a symbol of a connection between a source of a product and the product itself” (per LeBel J., at para. 39). If, as the Board found, it is not likely that even casual consumers will make a connection between the source of BARBIE dolls and the respondent’s restaurants, then the appellant’s marks have received the protection to which the law entitles them.
25 The onus remained throughout on the respondent to establish the absence of likelihood, but the Board was only required to deal with potential sources of confusion that, in the Board’s view, have about them an air of reality.
26 The appellant’s aggressive defence of trade-mark protection is, of course, understandable. It not only seeks to exploit the BARBIE “brand equity” it has worked to establish but, like all trade-mark owners, is required by law to protect its trade-marks from piracy or risk having such marks lose their distinctiveness, and, potentially their legal protection:Aladdin Industries, Inc. v. Canadian Thermos Products Ltd.,  S.C.R. 845; Breck’s Sporting Goods Co. v. Magder,  1 S.C.R. 527.
27 At common law, the appellant’s recourse would have been to commence an action for “passing off” with its roots in the law of deceit. There is a good deal of that flavour about the appellant’s complaint. In its factum, it says the respondent offered “no credible explanation” for choosing the Barbie name for its restaurants which “arouses suspicion” and should be seen as “an attempt to trade on the goodwill and reputation of the famous trade-mark”. In an action for passing off, it would have been necessary for the appellant to show that the respondent restauranteur intentionally or negligently misled consumers into believing its restaurant services originated with the appellant and that the appellant thereby suffered damage (Consumers Distributing Co. v. Seiko Time Canada Ltd.,  1 S.C.R. 583, at p. 601; Kirkbi, at para. 68). Quite apart from the issue of damages, the disparity between dolls and restaurant services would have posed, in the context of a passing-off action, an uphill battle and the appellant has not even tried to climb it.
28 Under the Trade-marks Act, however, the appellant’s protected commercial space is not so limited. It relies in particular on the 1953 amendments and the antecedent Report of the Trade Mark Law Revision Committee (January 20, 1953) chaired by the redoubtable Dr. Harold G. Fox, Q.C. (“Fox Report”). The appellant contends that “[t]he scope of protection of famous marks in Canada was an important justification for the enactment by Parliament and proclamation” of the amendment and notes the following passage in the Fox Report:
Some trade marks are so well known that the use of the same or similar trade marks on any wares of any kind would cause the general purchasing public to believe that the original user and owner of the trade mark was in some way responsible for the wares to which the use of the mark has been extended. [Emphasis added; p. 23.]
Parliament recognized the truth of that statement in 1953 and the subsequent experience of more than 50 years has borne out its wisdom. The problem is to apply the broad principle to particular situations in a way that is fair to all concerned.
29 In my view, with respect, the appellant’s case is based on an overgeneralization. The fact that Parliament has recognized that some trade-marks are so well known that use in connection with any wares or services would generate confusion is not to say that BARBIE has that transcendence. As the Fox Report also stated:
In a proper case this [new] ambit of protection can be widened to include the whole of the course of trade or restricted to a field limited by the use which has been made of a trade mark or trade name and the reputation acquired by it. The particular ambit of protection will in the future so far as applies to registration, be a matter for determination, having regard to all the circumstances, by the Registrar in the first instance, and by the Exchequer Court on appeal. [p. 27]
30 No doubt some famous brands possess protean power (it was submitted, for example, the distinctive red and white “Virgin” trade-mark has now been used in connection with such a diversity of wares and services that it knows virtually no bounds), but other famous marks are clearly product specific. “Apple” is said to be a well-known trade-mark associated in separate markets simultaneously with computers, a record label and automobile glass. The Board’s conclusion that BARBIE’s fame is limited to dolls and dolls’ accessories does not at all mean that BARBIE’s aura cannot transcend those products, but whether it is likely to do so or not in the context of opposition proceedings in relation to restaurant, catering and banquet services is a question of fact that depends on “all the surrounding circumstances” (s. 6(5)). Neither the “Virgin” nor “Apple” situations are before us and I make no pronouncement on either except to note them as illustrations that surfaced in the course of argument.
A. The Legal Framework
31 The respondent is not entitled to registration of its trade-mark unless it can demonstrate that use of both trade-marks in the same geographic area will not create the likelihood of confusion, i.e. mistaken inferencesin the marketplace. If, on a balance of probabilities, the Board is left in doubt, the application must be rejected.
B. Standard of Review
32 The Board found the respondent had demonstrated that if granted, its trade-mark would be unlikely to create confusion with that of the appellant. While this is essentially a question of mixed fact and law, the appellant says the Board’s consideration was fundamentally flawed by the erroneous interpretation given to s. 6 of the Act by the Federal Court of Appeal in Pink Panther Beauty Corp. v. United Artists Corp.,  3 F.C. 534, and Toyota Jidosha Kabushiki Kaisha v. Lexus Foods Inc.,  2 F.C. 15 (“Lexus”).
33 In choosing the proper standard of review from the available options (correctness, reasonableness, or patent unreasonableness) the Court has regard to the elements of the test set out most recently in Dr. Q v. College of Physicians and Surgeons of British Columbia,  1 S.C.R. 226, 2003 SCC 19, and Law Society of New Brunswick v. Ryan,  1 S.C.R. 247, 2003 SCC 20. These elements have not greatly altered since U.E.S., Local 298 v. Bibeault,  2 S.C.R. 1048, where Beetz J., speaking for the Court, said at p. 1088:
…the Court examines not only the wording of the enactment conferring jurisdiction on the administrative tribunal, but the purpose of the statute creating the tribunal, the reason for its existence, the area of expertise of its members and the nature of the problem before the tribunal.
34 I turn then to the elements of the Dr. Q test.
(1) Presence or Absence of a Privative Clause or Statutory Right of Appeal
35 The Act provides for a full right of appeal to a Federal Court judge who is authorized to receive and consider fresh evidence (ss. 56(1) and 56(5)). There is no privative clause. Where fresh evidence is admitted, it may, depending on its nature, put quite a different light on the record that was before the Board, and thus require the applications judge to proceed more by way of a fresh hearing on an extended record than a simple appeal (Philip Morris Inc. v. Imperial Tobacco Ltd. (No. 1) (1987), 17 C.P.R. (3d) 289 (F.C.A.)). Section 56 suggests a legislative intent that there be a full reconsideration not only of legal points but also of issues of fact and mixed fact and law, including the likelihood of confusion. See generally Molson Breweries v. John Labatt Ltd.,  3 F.C. 145 (C.A.), at paras. 46-51; Novopharm Ltd. v. Bayer Inc. (2000), 9 C.P.R. (4th) 304 (F.C.A.), at para. 4, and Garbo Creations Inc. v. Harriet Brown & Co. (1999), 3 C.P.R. (4th) 224 (F.C.T.D.).
(2) The Board’s Expertise
36 The determination of the likelihood of confusion requires an expertise that is possessed by the Board (which performs such assessments day in and day out) in greater measure than is typical of judges. This calls for some judicial deference to the Board’s determination, as this Court stressed in Benson & Hedges (Canada) Ltd. v. St. Regis Tobacco Corp.,  S.C.R. 192, at p. 200:
In my view the Registrar’s decision on the question of whether or not a trade mark is confusing should be given great weight and the conclusion of an official whose daily task involves the reaching of conclusions on this and kindred matters under the Act should not be set aside lightly but, as was said by Mr. Justice Thorson, then President of the Exchequer Court, in Freed and Freed Limited v. The Registrar of Trade Marks et al [ 2 D.L.R. 7, at p. 13]:
…reliance on the Registrar’s decision that two marks are confusingly similar must not go to the extent of relieving the judge hearing an appeal from the Registrar’s decision of the responsibility of determining the issue with due regard to the circumstances of the case.
37 What this means in practice is that the decision of the registrar or Board “should not be set aside lightly considering the expertise of those who regularly make such determinations”:McDonald’s Corp. v. Silcorp Ltd. (1989), 24 C.P.R. (3d) 207 (F.C.T.D.), at p. 210, aff’d (1992), 41 C.P.R. (3d) 67 (F.C.A.). Reception of new evidence, of course, might (depending on its content) undermine the factual substratum of the Board’s decision and thus rob the decision of the value of the Board’s expertise. However, the power of the applications judge to receive and consider fresh evidence does not, in and of itself, eliminate the Board’s expertise as a relevant consideration:Lamb v. Canadian Reserve Oil & Gas Ltd.,  1 S.C.R. 517, at pp. 527-28.
(3) The Purpose of the Trade-marks Act and in Particular the Trade-mark Registration Scheme
38 In Dr. Q, the Chief Justice pointed out that “[a] statutory purpose that requires a tribunal to select from a range of remedial choices or administrative responses, is concerned with the protection of the public, engages policy issues, or involves the balancing of multiple sets of interests or considerations will demand greater deference from a reviewing court” (para. 31). An inquiry into the likelihood of the “mistaken inference” does not call for the exercise of discretion. Nor is the Board in this respect making public policy decisions or allocating scarce resources. Essentially, the Board is deciding a lis between the parties in a procedure that looks like an informal version of an everyday court case.
(4) Nature of the Question in Dispute
39 While the appellant frames its argument as a challenge to the correctness of the interpretation given to s. 6 by the Federal Court of Appeal in Pink Panther and Lexus, I think that in reality, for reasons which I will develop, its challenge is directed to the relative weight to be given to the s. 6(5) enumerated and unenumerated factors. The legal issue is not neatly extricable from its factual context, but calls for an interpretation within the expertise of the Board. In answer to the Bibeault question (“Did the legislator intend the question to be within the jurisdiction conferred on the tribunal?” (p. 1087)), I think the answer is yes, within reasonable limits. See also Pezim v. British Columbia (Superintendent of Brokers),  2 S.C.R. 557, at p. 595; Canadian Broadcasting Corp. v. Canada (Labour Relations Board),  1 S.C.R. 157, at para. 43; Canada (Director of Investigation and Research) v. Southam Inc.,  1 S.C.R. 748, at para. 43; Moreau-Bérubé v. New Brunswick (Judicial Council),  1 S.C.R. 249, 2002 SCC 11, at para. 61.
(5) Conclusion on the Standard of Review
40 Given, in particular, the expertise of the Board, and the “weighing up” nature of the mandate imposed by s. 6 of the Act, I am of the view that despite the grant of a full right of appeal the appropriate standard of review is reasonableness. The Board’s discretion does not command the high deference due, for example, to the exercise by a Minister of a discretion, where the standard typically is patent unreasonableness (e.g. C.U.P.E. v. Ontario (Minister of Labour),  1 S.C.R. 539, 2003 SCC 29, at para. 157), nor should the Board be held to a standard of correctness, as it would be on the determination of an extricable question of law of general importance (Chieu v. Canada (Minister of Citizenship and Immigration),  1 S.C.R. 84, 2002 SCC 3, at para. 26). The intermediate standard (reasonableness) means, as Iacobucci J. pointed out in Ryan, at para. 46, that “[a] court will often be forced to accept that a decision is reasonable even if it is unlikely that the court would have reasoned or decided as the tribunal did.”The question is whether the Board’s decision is supported by reasons that can withstand “a somewhat probing” examination and is not “clearly wrong”:Southam, at paras. 56 and 60.
41 The foregoing analysis of the proper standard of review is consistent with the jurisprudence of the Federal Court of Appeal:see in particular Molson v. Labatt, per Rothstein J.A., at para. 51; Novopharm, per Strayer J.A., at para. 4; Polo Ralph Lauren Corp. v. United States Polo Assn. (2000), 9 C.P.R. (4th) 51, per Malone J.A., at para. 13, and Isaac J.A., at para. 10; Christian Dior, S.A. v. Dion Neckwear Ltd.,  3 F.C. 405, 2002 FCA 29, per Décary J.A., at para. 8, and Purafil, Inc. v. Purafil Canada Ltd. (2004), 31 C.P.R. (4th) 345, 2004 FC 522, per MacKay D.J., at para. 5.
C. Admission of the Fresh Survey Evidence
42 Before the applications judge, the appellant sought to adduce fresh evidence in the form of a survey which purported to show that:
– For 57 percent of the participants, BARBIE dolls came to mind when they saw the Barbie’s restaurant logo.
– 36 percent of the participants believed that the company that manufactured BARBIE dolls might have something to do with the logo of Barbie’s restaurant.
– 99.3 percent of the participants were familiar with the BARBIE dolls. (Rouleau J., at para. 10)
43 Until comparatively recently, evidence of public opinion polls was routinely held to be inadmissible because it purports to answer the factual component of the very issue before the Board or court (i.e. the likelihood of confusion), and in its nature it consists of an aggregate of the hearsay opinions of the people surveyed who are not made available for cross-examination, see, e.g., Building Products Ltd. v. BP Canada Ltd. (1961), 36 C.P.R. 121 (Ex. Ct.); Paulin Chambers Co. v. Rowntree Co. (1966), 51 C.P.R. 153 (Ex. Ct.). The more recent practice is to admit evidence of a survey of public opinion, presented through a qualified expert, provided its findings are relevant to the issues and the survey was properly designed and conducted in an impartial manner.
44 The principal attack on the survey evidence in this case rests on relevance. The issue in these opposition proceedings is the likelihood of confusion. The survey question (“Do you believe that the company that makes Barbie dolls might have anything to do with the restaurant identified by this sign or logo?” (emphasis added)) addresses the wholly different issue of possibilities. If the survey is not responsive to the point at issue, it is irrelevant and should (as the Federal Court of Appeal held) be excluded on that ground alone.
45 As to the usefulness of the results, assuming they are elicited by a relevant question, courts have more recently been receptive to such evidence, provided the survey is both reliable (in the sense that if the survey were repeated it would likely produce the same results) and valid (in the sense that the right questions have been put to the right pool of respondents in the right way, in the right circumstances to provide the information sought). See Canadian Schenley Distilleries Ltd. v. Canada’s Manitoba Distillery Ltd. (1975), 25 C.P.R. (2d) 1 (F.C.T.D.), at p. 9; Joseph E. Seagram & Sons Ltd. v. Seagram Real Estate Ltd. (1990), 33 C.P.R. (3d) 454 (F.C.T.D.); Walt Disney Productions v. Fantasyland Hotel Inc. (1994), 20 Alta. L.R. (3d) 146 (Q.B.). Thus, in Cartier Inc. v. Cartier Optical Ltd. (1988), 20 C.P.R. (3d) 68 (F.C.T.D.), the court accepted as helpful a survey found to be properly designed and impartially administered and whose findings were directly relevant to the likelihood of confusion. This was also the case in Sun Life Assurance Co. of Canada v. Sunlife Juice Ltd. (1988), 22 C.P.R. (3d) 244 (Ont. H.C.J.).
46 On the other hand, surveys have been excluded
(i) where the individuals surveyed did not constitute the relevant population, see Joseph E. Seagram, at p. 472; New Balance Athletic Shoes, Inc. v. Matthews (1992), 45 C.P.R. (3d) 140 (T.M.O.B.); National Hockey League v. Pepsi-Cola Canada Ltd. (1992), 70 B.C.L.R. (2d) 27 (S.C.), at para. 44; McDonald’s Corp. v. Coffee Hut Stores Ltd. (1994), 55 C.P.R. (3d) 463 (F.C.T.D.), at p. 475, or only included a subset of the relevant population:Unitel Communications Inc. v. Bell Canada (1995), 92 F.T.R. 161, at para. 117;
(ii) where the trade-mark used in the survey was not precisely the trade-mark applied for:Canada Post Corp. v. Mail Boxes Etc. USA, Inc. (1996), 77 C.P.R. (3d) 93 (T.M.O.B.);
(iii) where the wrong question is asked, e.g. a survey done for Coca-Cola was excluded on the basis that it was not directed to the issue of confusion but rather to measure public recognition of the word Classic as part of soft drink names:Coca-Cola Ltd. v. Southland Corp. (2001), 20 C.P.R. (4th) 537 (T.M.O.B.). See also: Molson Companies Ltd. v. S.P.A. Birra Peroni Industriale (1992), 45 C.P.R. (3d) 28 (T.M.O.B.); Molson Breweries v. Swan Brewery Co.,  T.M.O.B. No. 253 (QL); Toys “R” Us (Canada) Ltd. v. Manjel Inc. (2003), 229 F.T.R. 71, 2003 FCT 283;
(iv) where the survey had not been carried out in an impartial and independent manner:Unitel Communications, per Gibson J., at para. 120.
47 In the present case, the appellant retained a reputable and qualified firm to conduct the survey. However, the evidence showed that the persons questioned by the pollsters were merely shown without any context the respondent’s Barbie’s design logo, subsequently had it removed from their sight, and were then asked a series of questions, some of them suggestive. The key question, which the Federal Court of Appeal found to be irrelevant, was “Do you believe that the company that makes Barbie dolls might have anything to do with the restaurant identified by this sign or logo?” (emphasis added). The Montreal respondents were posed the question in French: “Croyez-vous que la compagnie qui fabrique les poupées Barbie pourrait avoir quelque chose à faire avec le restaurant identifié par cette enseigne ou ce logo?” (Emphasis added.)
48 Other alleged shortcomings included:
(i) the lack of information provided to those who responded to the survey;
(ii) the survey excluded anyone who was even aware of the respondent’s Barbie’s restaurants. It is pointed out that such people were part of the relevant universe of individuals in the market for the respondent’s services and are therefore potentially some of the people who may or may not “likely” be misled by the applied-for trade-mark. The survey need not be limited to such people of course, but nor should they be excluded. A similar objection was sustained in both McDonald’s Corp. v. Coffee Hut Stores Ltd. and in the Fantasyland Hotel case which referred with approval to Safeway Stores, Inc. v. Safeway Insurance Co., 657 F. Supp. 1307 (M.D. La. 1985), in this respect;
(iii) the survey methodology included questions suggestive of the answers the appellant apparently wanted to hear.
49 The appellant has not made any application to adduce the survey testimony or results as fresh evidence in this Court, and in my view the exclusion of that evidence by the applications judge does not constitute reversible error in light of the broad power given to him under s. 56(5), which provides:
(5) On an appeal under subsection (1), evidence in addition to that adduced before the Registrar may be adduced and the Federal Court may exercise any discretion vested in the Registrar.
The use of the word “might” (or “pourrait”) in the survey question was directed to a mere possibility, rather than a probability, of confusion. The respondents who answered “yes” to this question may merely have believed that it was within the realm of possibility that the appellant had something to do with the restaurant, rather than actually inferring from the trade-marks that it was likely that the two trade-marks represented wares or services emanating from the same source. It is relevant if even a limited percentage of the population surveyed is confused, but evidence that a lot of people might or could possibly make the mistaken inference is not. The other grounds of objection go to the weight rather than to admissibility of the survey evidence and, subject to assessing that weight (or the lack of it), the survey evidence could have been admitted over the respondent’s other objections. Lack of relevance, however, was fatal.
50 At most, the proferred survey indicated that the BARBIE trade-mark is well known. Even without the survey however, the courts below were content to proceed on the basis that BARBIE is a “well known” if not “famous” mark, and I do so as well.
D. The Statutory Test of “Confusion”
51 Trade-mark confusion is a term defined in s. 6(2) and arises if it is likely in all the surrounding circumstances (s. 6(5)) that the prospective purchaser will be led to the mistaken inference
that the wares or services associated with those trade-marks are manufactured, sold, leased, hired or performed by the same person, whether or not the wares or services are of the same general class. [Emphasis added.]
This is not to say the nature of the wares or services is irrelevant. Section 6(5)(c) specifically identifies “the nature of the wares, services or business” as a relevant consideration. The point of the underlined words in s. 6(2) is simply to lay it down in clear terms that the general class of wares and services, while relevant, is not controlling.
52 It is also relevant that the appellant’s mark is symbolic of wares while the respondent’s applied-for mark identifies services. It is possible to have confusion between the source of wares and services (Building Products) but that distinction, as well, is but one of the surrounding circumstances that may be taken into consideration.
53 The appellant argued that the courts below erred in looking at the respondent’s actual operations rather than at the terms set out in its application for the proposed trade-mark. It is quite true that the proper focus is the terms of the application, because what is at issue is what the registration would authorize the respondent to do, not what the respondent happens to be doing at the moment. Still, the appellant itself led a great deal of evidence (as is the practice) about the actual operation of the respondent’s restaurants, including many photographs, numerous sample menus and clippings of various advertisements. In these circumstances, it is not surprising that the Board and the applications judge felt it appropriate to comment on the respondent’s operation, based largely on the evidence the appellant itself had adduced. That said, I do not think the Board or the courts below were under misapprehension about the nature of the dispute. The terms of the respondent’s application (“restaurant services, take-out services, catering and banquet services”) were referred to by both the Board and the applications judge, and reading their respective reasons as a whole, I do not think they misapprehended the question before them.
54 Within the “all the surrounding circumstances” test, s. 6(5) of the Act lists five factors to be considered when making a determination as to whether or not a trade-mark is confusing. These are: “(a) the inherent distinctiveness of the trade-marks or trade-names and the extent to which they have become known; (b) the length of time the trade-marks or trade-names have been in use; (c) the nature of the wares, services or business; (d) the nature of the trade; and (e) the degree of resemblance between the trade-marks or trade-names in appearance or sound or in the ideas suggested by them”. The list of circumstances is not exhaustive and different circumstances will be given different weight in a context-specific assessment. See Gainers Inc. v. Marchildon (1996), 66 C.P.R. (3d) 308 (F.C.T.D.).In opposition proceedings, as stated, the onus is on the applicant (here the respondent) to show on a balance of probabilities that there is no likelihood of confusion.
55 Evidence of actual confusion would be a relevant “surrounding circumstance” but is not necessary (Christian Dior, at para. 19) even where trade-marks are shown to have operated in the same market area for ten years:Mr. Submarine Ltd. v. Amandista Investments Ltd. (1987), 19 C.P.R. (3d) 3 (F.C.A.). Nevertheless, as discussed below, an adverse inference may be drawn from the lack of such evidence in circumstances where it would readily be available if the allegation of likely confusion was justified.
(1) The Casual Consumer Somewhat in a Hurry
56 What, then, is the perspective from which the likelihood of a “mistaken inference” is to be measured?It is not that of the careful and diligent purchaser. Nor, on the other hand, is it the “moron in a hurry” so beloved by elements of the passing-off bar:Morning Star Co-Operative Society Ltd. v. Express Newspapers Ltd.,  F.S.R. 113 (Ch. D.), at p. 117. It is rather a mythical consumer who stands somewhere in between, dubbed in a 1927 Ontario decision of Meredith C.J. as the “ordinary hurried purchasers”:Klotz v. Corson (1927), 33 O.W.N. 12 (Sup. Ct.), at p. 13. See also Barsalou v. Darling (1882), 9 S.C.R. 677, at p. 693. In Delisle Foods Ltd. v. Anna Beth Holdings Ltd. (1992), 45 C.P.R. (3d) 535 (T.M.O.B.), the Registrar stated at pp. 538-39:
When assessing the issue of confusion, the trade marks at issue must be considered from the point of view of the average hurried consumer having an imperfectrecollection of the opponent’s mark who might encounter the trade mark of the applicant in association with the applicant’s wares in the market-place.
And see American Cyanamid Co. v. Record Chemical Co.,  F.C. 1271 (T.D.), at p. 1276, aff’d (1973), 14 C.P.R. (2d) 127 (F.C.A.). As Cattanach J. explained in Canadian Schenley Distilleries, at p. 5:
That does not mean a rash, careless or unobservant purchaser on the one hand, nor on the other does it mean a person of higher education, one possessed of expert qualifications. It is the probability of the average person endowed with average intelligence acting with ordinary caution being deceived that is the criterion and to measure that probability of confusionthe Registrar of Trade Marks or the Judge must assess the normal attitudes and reactions of such persons.
57 Having repeated that, I fully agree with Linden J.A. in Pink Panther that in assessing the likelihood of confusion in the marketplace “we owe the average consumer a certain amount of credit” (para. 54). A similar idea was expressed in Michelin & Cie v. Astro Tire & Rubber Co. of Canada Ltd. (1982), 69 C.P.R. (2d) 260 (F.C.T.D.), at p. 263:
…one must not proceed on the assumption that the prospective customers or members of the public generally are completely devoid of intelligence or of normal powers of recollection or are totally unaware or uninformed as to what goes on around them.
58 A consumer does not of course approach every purchasing decision with the same attention, or lack of it. When buying a car or a refrigerator, more care will naturally be taken than when buying a doll or a mid-priced meal:General Motors Corp. v. Bellows,  S.C.R. 678. In the case of buying ordinary run-of-the-mill consumer wares and services, this mythical consumer, though of average intelligence, is generally running behind schedule and has more money to spend than time to pay a lot of attention to details. In appropriate markets, such a person is assumed to be functionally bilingual: Four Seasons Hotels Ltd. v. Four Seasons Television Network Inc. (1992), 43 C.P.R. (3d) 139 (T.M.O.B.). To those mythical consumers, the existence of trade-marks or brands make shopping decisions faster and easier. The law recognizes that at the time the new trade-mark catches their eye, they will have only a general and not very precise recollection of the earlier trade-mark, famous though it may be or, as stated in Coca-Cola Co. of Canada Ltd. v. Pepsi-Cola Co. of Canada Ltd.,  2 D.L.R. 657 (P.C.), “as it would be remembered by persons possessed of an average memory with its usual imperfections” (p. 661). The standard is not that of people “who never notice anything” but of persons who take no more than “ordinary care to observe that which is staring them in the face”: Coombe v. Mendit Ld. (1913), 30 R.P.C. 709 (Ch. D.), at p. 717. However, if ordinary casual consumers somewhat in a hurry are likely to be deceived about the origin of the wares or services, then the statutory test is met.
(2) Did Pink Panther Skew the Test?
59 The argument for the appellant is that in the case of a famous mark like BARBIE the words in s. 6(2) “whether or not the wares or services are of the same general class” take on added significance. Pink Panther, it is said, laid undue weight on the similarity or dissimilarity of the wares and services. Modern commerce is driven to a significant extent by famous brands. The Virgin conglomerate, appellant’s counsel argues, has diversified from records to airlines to insurance to superstores. The marketing strategy predicts that very different services will benefit from cross-selling. The message is that the people who can fly you cheaply across the Atlantic can probably get you a good deal on your house insurance. The same, he suggests, is true of BARBIE.
60 The appellant is correct that, following introduction of the 1953 amendment, “famous brands” received a significantly broader ambit of protection. In some cases, the courts emphasized that a significant dissimilarity in wares or services was no longer fatal. Thus, in Carson v. Reynolds,  2 F.C. 685 (T.D.), it was held that the use of the mark “Here’s Johnny” for portable trailers, outhouses and lavatory facilities would likely suggest to a “significant number of people in Canada” (p. 690), a connection with Johnny Carson and the Tonight Show. In John Walker & Sons Ltd. v. Steinman (1965), 44 C.P.R. 58, the Registrar of the Trade Marks refused an application for the mark “Johnny Walker” for sporting equipment over a prior registration of the mark “Johnnie Walker” for whisky. In James Burrough Ltd. v. Reckitt & Colman (Canada) Ltd. (1967), 53 C.P.R. 276 (Reg. T.M.), the owner of the mark “Beefeaters” for gin defeated an application to trade-mark “Beefeater” for sauce mixes, Yorkshire pudding mixes, spices and condiments. In Leaf Confections Ltd. v. Maple Leaf Gardens Ltd.(1986), 12 C.P.R. (3d) 511 (F.C.T.D.), aff’d (1988), 19 C.P.R. (3d) 331 (F.C.A.), the Federal Court, Trial Division rejected an application to register a mark consisting of the word “leaf” in a maple leaf for bubble gum because of the likelihood of confusion with the Toronto Maple Leafs, the judge holding that “I am convinced that a young person seeing bubble gum packaged in a wrapper which bore as its principal identifying mark the appellant’s LEAF & Design would be misled into thinking that the product was associated with the Toronto Maple Leafs hockey club and would confuse the appellant’s mark and design with those of the respondent” (p. 521). In Danjaq, S.A. v. Zervas (1997), 75 C.P.R. (3d) 295 (F.C.T.D.), the court refused to allow a pizza shop to register the trade-marks “007”, “007 Pizza & Subs and Design” and “007 Submarine and Design” because of the likelihood it would lead to confusion with the respondent’s trade-mark related to the famous and irresistible James Bond.
61 According to the appellant, the continuity of this line of cases was interrupted by Pink Panther which the appellant says wrongly limited the scope of protection available to a famous mark to cases where there is a “connection” or “similarity” between the wares or services associated with the famous mark and the wares or services of “the newcomer”. The appellant criticizes particularly the following passages of the Pink Panther decision (factum, at para. 66):
The wide scope of protection afforded by the fame of the appellant’s mark only becomes relevant when applying it to a connection between the applicant’s and the opponent’s trade and services. No matter how famous a mark is, it cannot be used to create a connection that does not exist.
…in each case a connection or similarity in the products or services was found. Where no such connection is established, it is very difficult to justify the extension of property rights into areas of commerce that do not remotely affect the trade-mark holder. Only in exceptional circumstances, if ever, should this be the case.
What the Trial Judge did not give sufficient weight to is that, not only were the wares in each case completely disparate, but there is no connection whatsoever between them. As I stated earlier, where no such connection exists a finding of confusion will be rare. [Emphasis added by appellant; paras. 44, 46 and 51.]
62 In Pink Panther, a majority of the Federal Court of Appeal allowed the registration of “Pink Panther” as a trade-mark for a line of hair care and beauty product supplies over the opposition of an existing trade-mark holder which had produced a very successful series of films of the same name starring Peter Sellers as Inspector Clouseau. McDonald J.A., dissenting, complained that “it is precisely because of the fame and goodwill associated with the name ‘The Pink Panther’ that the appellant has chosen that name for its business” (para. 58). However, his objection was rejected by the majority, which observed at para. 50:
…the issue to be decided is not how famous the mark is, but whether there is a likelihood of confusion in the mind of the average consumer between United Artists’ mark and the one proposed by the appellant with respect to the goods and services specified. That question must be answered in the negative. There is no likelihood of confusion as to the source of the products. The key factor here is the gaping divergence in the nature of the wares and in the nature of the trade. It is not a fissure but a chasm.
63 After referring to a number of cases where the famous trade-mark prevailed, Linden J.A. stated that
…in each case a connection or similarity in the products or services was found. Where no such connection is established, it is very difficult to justify the extension of property rights into areas of commerce that do not remotely affect the trade-mark holder. Only in exceptional circumstances, if ever, should this be the case. [para. 46]
I agree with the appellant that the “exceptional circumstances, if ever” test puts the bar too high and may be seen as an attempt to impose rigidity where none exists. If the result of the use of the new mark would be to introduce confusion into the marketplace, it should not be accepted for registration “whether or not the wares or services are of the same general class”(s. 6(2)). The relevant point about famous marks is that fame is capable of carrying the mark across product lines where lesser marks would be circumscribed to their traditional wares or services. The correct test is that which Linden J.A. earlier stated at para. 33:
The totality of the circumstances will dictate how each consideration should be treated.
64 The appellant argues that the rigidity of the Pink Panther approach was repeated in Lexus.In that case, the applicant sought to register the mark “Lexus” in association with canned foods and canned fruit juice. Toyota argued this would likely cause confusion with its mark “Lexus” in association with automobiles. Linden J.A. stated, at paras. 7 and 9:
The use of the phrase “whether or not the wares or services are of the same general class” as used in subsections 6(2), 6(3) and 6(4) does not mean that the nature of the wares is irrelevant in determining confusion; they suggest only that confusion may be generated with goods that are not in the “same general class”, but still have some resemblance or linkage to the wares in question.
Famousness alone does not protect a trade-mark absolutely. It is merely a factor that must be weighed in connection with all the rest of the factors. If the fame of a name could prevent any other use of it, the fundamental concept of a trade-mark being granted in relation to certain wares would be rendered meaningless. [Underlining added.]
65 I believe Linden J.A. misspoke to the extent he suggested that, for confusion to occur, there must be “some resemblance or linkage to the wares in question”, i.e. to the wares for which registration of a trade-mark is sought. Resemblance is clearly not a requirement under s. 6. On the contrary, the point of the legislative addition of the words “whether or not the wares or services are of the same general class” conveyed Parliament’s intent that not only need there be no “resemblance” to the specific wares or services, but the wares or services marketed by the opponent under its mark and the wares or services marketed by the applicant under its applied-for mark need not even be of the same general class.
66 The International Trademark Association, which intervened in the Veuve Clicquot Ponsardin v. Boutiques Cliquot Ltée,  1 S.C.R. 824, 2006 SCC 23 appeal, heard concurrently, said with respect to the Pink Panther and Lexus decisions that the Court of Appeal “made the same error in both cases of requiring a connection between the wares and services when the real issue in a s. 6 analysis is consumer confusion” (factum, at para. 30).
67 I agree with the appellant and the intervener that the issue in s. 6 is consumer confusion (or, in other words, the likelihood of a mistaken inference), but I do not agree that in their dispositive passages (as opposed to stray dicta) Pink Panther or Lexus departed from the totality of the circumstances test. What Linden J.A. actually held in Pink Panther was that
[t]he fact that the opponent’s mark was world-renowned could not be a factor so important as to make the differences in wares and services irrelevant.
…I do not see how the fame of the mark acts as a marketing trump card such that the other factors are thereby obliterated. [Emphasis added; paras. 45 and 52.]
68 In Lexus, Linden J.A. stated:
In conclusion, taking into account all the factors listed in subsection 6(5), I am of the view that the Trial Judge erred and that the Registrar was correct when he decided that the appellant met the onus of proof resting on it and allowed the registration of the trade-mark “Lexus” in connection with certain canned foods. [Emphasis added; para. 12.]
69 In my view, Linden J.A. was disputing what he took to be the argument of the filmmaker and car maker that fame “trumps” all other factors. (It is perhaps of interest that at an earlier date, Toyota had successfully argued in the United States that its Lexus mark could co-exist with the Lexis legal database without confusion: see Mead Data Central, Inc. v. Toyota Motor Sales, U.S.A., Inc., 875 F.2d 1026 (2d Cir. 1989).) Pink Panther was thus understood by the applications judge in the present case:
It cannot be automatically presumed that there will be confusion just because the applicant’s mark is famous. Under the circumstances, if we keep in mind that the test to satisfy was the likelihood of confusion (and not the possibility of confusion), the fame of the mark could not act as a marketing trump card such that the other factors are thereby obliterated. [para. 40]
70 It may be, as the appellant argues, that “the nature of wares or services should have less weight because the famous mark more likely will lead to the inference that the source of the two is the same” (factum, at para. 70) and “…fame itself can and does create a connection in the mind of the ordinary consumer who first sees a famous mark in a new context” (factum, at para. 73 (emphasis in original)), but in Pink Panther, Linden J.A. clearly said that “[t]he key factor here is the gaping divergence in the nature of the wares and in the nature of the trade” (para. 50 (emphasis added)). The jurisprudence is clear that different factors may be given different weight in different situations. For example, Professor Mostert points out that
[c]ollectors of engraved, high quality shotguns may, for instance, be few and far between in any given country but the PURDEY mark is undoubtedly exceptionally well-known among the circle of such shotgun cognoscenti.
(F. W. Mostert, Famous and Well-Known Marks: An International Analysis (1997), at p. 26)
Even so, I doubt that even the cognoscenti would think the world famous shotgun specialist is likely associated with the well-known Vancouver purveyor of chocolates, such is the divergence in the type of wares and channels of trade.
71 To the extent Linden J.A. held that the difference in wares or services will always be a dominant consideration, I disagree with him, but given the role and function of trade-marks, it will generally be an important consideration. The appellant contends that some of Linden J.A.’s obiter statements can be read virtually to require a “resemblance” between the respective wares and services. In that respect, the obiter should not be followed.
72 On a more general level, however, it seems to me that it is the appellant, not the Federal Court of Appeal, which seeks to sidestep the “all the surrounding circumstances” test in the case of a famous trade-mark and place fame in the ascendent. I agree with theappellant that a difference in wares or services does not deliver the knockout blow, but nor does the fame of the trade-mark. Each situation must be judged in its full factual context.
73 I acknowledge that the result of Pink Panther has been somewhat controversial on the facts. Professors Gervais and Judge, for example, comment that
[g]iven current marketing practices, such as the increasing use of character merchandising on a broader range of goods outside the typical entertainment field, consumers may come to expect (or at least not be surprised) that a single company will be selling disparate products (like shampoo and movies) and the likelihood of confusion when similar marks are used on different products may increase.
(D. Gervais and E. F. Judge, Intellectual Property: The Law in Canada (2005), at p. 212)
Be that as it may, the view is correct that “all of the surrounding circumstances” must be taken into consideration but that, in some cases, some circumstances (such as the difference in wares) will carry greater weight than others.
E. Applying the Section 6(5) Circumstances to the Facts of This Case
74 It follows from the foregoing discussion that, while the fame of the BARBIE brand is a “surrounding circumstance” of importance, the scope of its protection requires consideration of all the circumstances, including the enumerated s. 6(5) factors. The Board found that while the BARBIE trade-mark is indeed famous, its fame in Canada is “in association with dolls and doll accessories”, not the other items listed in the appellant’s registrations, including coffee, pizza and other food products. Nor, in particular, is its fame likely to generate any confusion in relation to the mark applied for by the respondent as a symbol of “restaurant services, take-out services, catering and banquet services”. It is important to keep in mind, as the appellant says, that the issue is not the scope of the respondent’s existing business but the scope of protection it seeks by its trade-mark application.
(1) The Inherent Distinctiveness of the Trade-Marks or Trade-Names and the Extent to Which They Have Become Known
75 “Distinctiveness is of the very essence and is the cardinal requirement of a trade-mark”: Western Clock Co. v. Oris Watch Co.,  Ex. C.R. 64, per Audette J., at p. 67. The word “Barbie” is an everyday expression not originated by the appellant, and on that account would normally receive less protection “than in the case of an invented or unique or non-descriptive word” (like Kleenex), per Rand J. in General Motors, at p. 691, to which one might add:“No person is entitled to fence in the common of the English or French languages and words of a general nature cannot be appropriated over a wide area”:K. Gill and R. S. Jolliffe, Fox on Canadian Law of Trade-marks and Unfair Competition (4th ed. (loose-leaf)), at p. 8-56. I accept, as discussed earlier, that BARBIE has now acquired a strong secondary meaning associated with the appellant’s doll products, and on that account has achieved considerable distinctiveness. While the mark as registered is just the plain word BARBIE, its use in advertising and packaging is accompanied with distinctive designs and graphics.
76 The mark applied for by the respondent has become somewhat known within the area where both parties’ marks are used. Its applied-for trade-mark is not only the word “Barbie’s” but the totality of the effect, including the script in which it is written, and the surrounding design:Henry K. Wampole & Co. v. Hervay Chemical Co. of Canada,  S.C.R. 336, and British Drug Houses, Ltd. v. Battle Pharmaceuticals,  Ex. C.R. 239. In assessing the issue of confusion, attention should be paid to the differences as well as the similarities in the marks said to be confusing. However, as none of the s. 6(5) circumstances “trump” the others, this is but one consideration to be weighed in the balance of “all the surrounding circumstances”.
(2) The Length of Time the Trade-Marks and Trade-Names Have Been in Use
77 Dr. Fox noted that “[l]ength of user is only important in considering the question of fact, whether the trade mark has really and truly become distinctive”: H. G. Fox, The Canadian Law of Trade Marks and Unfair Competition (3rd ed. 1972), at p. 133. The appellant’s marks have been widely publicized since the early 1960s, the respondent’s only since 1992. The appellant’s mark has deeper roots, and has been much more highly publicized over a much larger geographic area.
(3) The Nature of the Wares, Services or Business
78 The traditional rule was that “[i]f [a manufacturer] does not carry on a trade in iron, but carries on a trade in linen, and stamps a lion on his linen, another person may stamp a lion on iron”:Ainsworth v. Walmsley (1866), L.R. 1 Eq. 518, at pp. 524-25. While I agree with Professor McCarthy, quoted earlier, that “a relatively strong mark can leap vast product line differences at a single bound” (§ 11:74), nevertheless it is implicit in Professor McCarthy’s statement that the “product line” will generally represent a significant obstacle for even a famous mark to leap over. The doll business and the restaurant business appeal to the different tastes of a largely different clientele. As was found by the Board:
The nature of the opponent’s wares and the applicant’s services are quite different. In this regard, the opponent has established that its mark is very well known, if not famous in Canada, in association with dolls and doll accessories. The opponent’s target market are children and to some extent adult collectors. By contrast, the applicant is in the restaurant business and its target market are [sic] adults. [p. 400]
79 There is no evidence that adult consumers would consider a doll manufacturer to be a source of good food, still less that the BARBIE trade-mark would be understood to guarantee, as the 1953 Fox Report (at p. 26) put it, “character and quality”. The appellant suggests that the BARBIE doll has become part of pop culture, and there is some truth to that, but the meaning is not necessarily a positive recommendation for all wares and services:
Barbie doll /…/ noun Colloquial a female who is superficially attractive in a conventional way, especially with blue eyes and blonde hair, but who lacks personality.
(The Macquarie Dictionary (rev. 3rd ed. 2002), at p. 147)
In that regard the association of the BARBIE doll with food might be taken as a warning of blandness.
80 The appellant points to the broad scope of some of its licences, and the potential growth of BARBIE into new fields of commerce, but as the Fox Report points out,
there might be many users of the same trade mark with many variations in the quality of the wares or services available under the trade mark with consequent public confusion and deception. [p. 39]
81 The appellant relies on s. 50(1) of the Act which states that provided the trade-mark owner retains direct or indirect control of the character or quality of the wares or services use by the licensee is deemed to have the same effect as use by the owner. Acknowledging this to be so, it is also true that profligate use by the owner alone can destroy the distinctiveness of the mark, and licences granted too widely and unwisely may aggravate its problem. Nevertheless, the recognition in our society that “famous brands” are widely licenced for wares and services not traditionally associated with the mark must be kept in mind as another “surrounding circumstance”.
82 The Board did not accept that the BARBIE mark is “famous” or distinctive of anything other than dolls and dolls’ accessories. The very specificity of BARBIE dolls in the mind of the consumer arguably constrains its capacity for growth. In Joseph E. Seagram, the appellant argued that the natural tendency towards corporate diversification would lead the consumer to presume that its alcoholic beverage business was connected with the respondent’s real estate business. MacKay J. dismissed this argument, stating at pp. 467-68:
I do not agree with this proposition. In my view, consideration of future events and possibilities of diversification is properly restricted to the potential expansion of existing operations. It should not include speculation as to diversification into entirely new ventures, involving new kinds of wares, services or businesse…. .
83 The point, I think, is that the law of trade-marks is based on use. In an earlier era it was not possible to register a “proposed” use. Here, expansion of the BARBIE mark is more than just speculation, but if the BARBIE mark is not famous for anything but dolls and doll accessories in the area where both marks are used and there is no evidence that BARBIE’s licensees, whoever they may be, are in the marketplace using the BARBIE mark for “restaurant services, take-out services, catering and banquet services”, it is difficult to see the basis on which the mistaken inference is likely to be drawn.
84 On this point, the Board found that the appellant’s
average annual advertising expenditure of $3.5 million, for the years 1990-1995, was dedicated solely to doll and doll accessory items. I do not accept [the appellant’s] submissions that there is any real connection between [its] wares and [the respondent’s] services.In my view, the connections postulated by the [appellant] are too tenuous to be anything other than theoretical. [Emphasis added; p. 402.]
85 This assessment is supported by the evidence. The appellant’s real complaint is that the Board’s negative view on this point was given too much weight in the Board’s “weighing up” of “all the surrounding circumstances”.
(4) The Nature of the Trade
86 The parties operate in different and distinct channels of trade within which theirrespective wares and services do not intermingle. In McDonald’s Corp. v. Coffee Hut Stores Ltd., at p. 474, McKeown J. pointed out that even though both parties sold coffee, a specialty coffee store occupies a different market niche than a fast food outlet (aff’d (1996), 68 C.P.R. (3d) 168 (F.C.A.)). The nature and kind of customer who would be likely to buy the respective wares and services has long been considered a relevant circumstance: General Motors, at p. 692; Gill and Jolliffe, at pp. 8-38 to 8-40.
87 In the present case, quite apart from the great difference between the appellant’s wares and the respondent’s services, they occupy different channels of trade and the increased potential for confusion that might arise through intermingling in a single channel of trade does not present a serious problem.
(5) The Degree of Resemblance Between the Trade-Marks or Trade-Names in Appearance or Sound or in the Ideas Suggested by Them
88 Both marks use the name “Barbie” but the respondent’s applied-for mark wraps the name in a design, whereas the appellant’s mark as registered does not. On the other hand, if the appellant’s mark as used in packaging and advertising is taken into account, there is a considerable resemblance.
(6) Other Surrounding Circumstances
89 No doubt, as an abstract proposition, the appellant’s mark is “famous” whereas the respondent’s applied-for mark is not. The question, however, is whether there will likely be (or has been) confusion in the marketplace where both may operate. In that respect, evidence of actual confusion, though not necessary, would have been helpful (ConAgra, Inc. v. McCain Foods Ltd. (2001), 14 C.P.R. (4th) 288, 2001 FCT 963; Panavision, Inc. v. Matsushita Electric Industrial Co. (1992), 40 C.P.R. (3d) 486 (F.C.T.D.), but it was not forthcoming. Décary J.A. commented in Christian Dior, at para. 19:
While the relevant issue is “likelihood of confusion” and not “actual confusion”, the lack of “actual confusion” is a factor which the courts have found of significance when determining the “likelihood of confusion”. An adverse inference may be drawn when concurrent use on the evidence is extensive, yet no evidence of confusion has been given by the opponent.
I agree. The lack of any evidence of actual confusion (i.e. that prospective consumers are drawing the mistaken inference) is another of the “surrounding circumstances” to be thrown into the hopper:Pepsi-Cola Co. of Canada, Ltd. v. Coca-Cola Co. of Canada, Ltd.,  S.C.R. 17, at p. 30; General Motors Corp. v. Bellows,  Ex. C.R. 568, at p. 577, aff’d  S.C.R. 678; Freed & Freed Ltd. v. Registrar of Trade Marks,  Ex. C.R. 431; Monsport Inc. v. Vêtements de Sport Bonnie (1978) Ltée (1988), 22 C.P.R. (3d) 356 (F.C.T.D.), at p. 360; Multiplicant Inc. v. Petit Bateau Valton S.A. (1994), 55 C.P.R. (3d) 372 (F.C.T.D.), at p. 379.
90 The appellant contended in effect that the respondent was a deliberate free rider who had no reasonable explanation for adopting its trade-mark. The obvious conclusion, argues the appellant, is that the respondent seeks to register a mark which is designed to pirate whatever goodwill it can from the mark of the appellant. It points to a subsequent trade-mark application by the respondent (now abandoned) that included in its design the head and shoulders of a female waitress with a “B” monogram on her blouse. It seems to me there is some justice to this complaint, but the relevant perspective of s. 6(2) of the Trade-marks Act is not that of the respondent but rather the perception of the relevant mythical consumer. Mens rea is of little relevance to the issue of confusion: Lexus. It has been established since Edelsten v. Edelsten (1863), 1 De G. J. & S. 185, 46 E.R. 72, at pp. 78-79, that a trade-mark is a proprietary right. If, as the appellant says, the respondent’s activities have trespassed on the marketing territory fenced off by its BARBIE trade-marks, it would be no defence for the respondent that it did not intend to trespass. Equally, however, if the respondent’s activities did not in fact trespass, evidence that it may have wished to do so does not constitute confusion: Fox, at p. 403. Historically, courts have been slow to conclude that a demonstrated piratical intent has failed to achieve its purpose, but in this case no such intent was found as a fact by the Board.
91 The Board held that, notwithstanding the fame of the appellant’s trade-mark, it was satisfied by the respondent that there was no likelihood of confusion in the marketplace having regard to all the surrounding circumstances. In the absence of fresh evidence to shed further light on the correctness of this conclusion, I would not be prepared to say that in reaching its conclusion the Board was unreasonable.
92 The appeal will therefore be dismissed with costs.
The following are the reasons delivered by
93 LeBel J. – I agree that Mattel’s appeal should be dismissed. The Trade-marks Opposition Board at the Canadian Intellectual Property Office had to weigh a number of factors in its assessment of the distinctiveness of the two marks. Its decision was entitled to deference, but had to be reasonable. On the facts of this case, as demonstrated by the analysis of the legal and factual issues in the reasons of Binnie J., I agree that the decision of the Board was indeed reasonable ((2002), 23 C.P.R. (4th) 395).
Appeal dismissed with costs.
Solicitors for the appellant: Gowling Lafleur Henderson, Ottawa.
Solicitors for the respondent: Desjardins Ducharme, Montréal.
*Major J. took no part in the judgment.
This page has been generated using Mincov Legal Markup Language (“MLML”). While every effort has been made to ensure accuracy and currency of this document, Mincov Law Corporation makes no warranty, representation or undertaking, expressed or implied, nor does it assume any liability, direct or indirect, or responsibility for the accuracy, currency, usefulness or continued availability of this document.The official version of this document is located at http://scc.lexum.org/en/2006/2006scc22/2006scc22.html.