[2001] 2 F.C. 636
A-276-00
2001 FCA 13
Smithkline Beecham Corporation (Appellant)
v.
Pierre Fabre Médicament (Respondent)
Indexed as: Pierre Fabre Médicament v. Smithkline Beecham Corp. (C.A.)
Court of Appeal, Desjardins, Décary and Noël JJ.A. —Ottawa, January 17 and February 13, 2001.
Trade-marks — Registration — Respondent applying for registration of trade-mark “Ixel” in connection with antidepressants — Appellant objecting on ground of risk of confusion with own trade-mark “Paxil” — Applicable test at issue in determining whether likelihood of confusion between two trade-marks within meaning of Trade-marks Act — Registrar of Trade Marks, Trial Judge wrong in relying on opinion of average bilingual consumer as independent test — Trade-mark cannot be registered where risk of confusion in either of two official languages — Only applicable test risk of confusion on part of average Francophone consumer, average Anglophone consumer or, in special instances, average bilingual consumer — New hearing required in interests of justice.
STATUTES AND REGULATIONS JUDICIALLY CONSIDERED
Trade-marks Act, R.S.C., 1985, c. T-13, ss. 6, 12(1)(b), 16 (as am. by S.C. 1994, c. 47, s. 195), 19 (as am. by S.C. 1993, c. 15, s. 60).
CASES JUDICIALLY CONSIDERED
APPLIED:
Produits Freddy Inc. v. Ferrero SpA (1986), 20 C.P.R. (3d) 61; 15 F.T.R. 16 (F.C.T.D.); affd (1988), 22 C.P.R. (3d) 346 (F.C.A.).
CONSIDERED:
Vins La Salle Inc. (Les) v. Vignobles Chantecler Ltée (Les) (1985), 6 C.P.R. (3d) 533 (T.M. Opp. Bd.); Boy Scouts of Canada v. Alfred Sternjakob GmbH & Co. KG et al. (1984), 2 C.P.R. (3d) 407 (F.C.T.D.); Scott Paper Co. v. Beghin-Say S.A. (1985), 5 C.P.R. (3d) 225 (F.C.T.D.).
APPEAL from a Trial Division decision ((2000), 8 C.P.R. (4th) 389; 182 F.T.R. 265) relying on the opinion of the average bilingual consumer in determining whether there is a likelihood of confusion between two trade-marks within the meaning of the Trade-marks Act. Appeal allowed.
APPEARANCES:
Shu-tai Cheng and Robert A. MacDonald for appellant.
Jean-François Buffoni for respondent.
SOLICITORS OF RECORD:
Gowling Lafleur Henderson LLP, Ottawa, for appellant.
Fasken Martineau DuMoulin, Montréal, for respondent.
The following is the English version of the reasons for judgment rendered by
[1] Décary J.A.: To what degree should the opinion of the average bilingual consumer be considered in determining whether there is a likelihood of confusion between two trade-marks within the meaning of the Trade-marks Act, R.S.C., 1985, c. T-13 (the Act)? That is the issue the Court must decide in this appeal from a decision of Pinard J., which is published at (2000), 8 C.P.R. (4th) 389 (F.C.T.D.).
[2] On December 9, 1993, the respondent Pierre Fabre Médicament applied for registration, on the basis of the planned use in Canada, of the trade-mark “Ixel” in connection with antidepressants. On November 22, 1994, the appellant Smithkline Beecham Corporation filed a statement of objection in which it cited the risk of confusion with its duly registered trade-mark “Paxil”.
[3] On September 11, 1998, the Registrar of Trade-marks (the Registrar) concluded that the respondent had not discharged its onus of demonstrating that there was no reasonable likelihood of confusion and he consequently refused the registration of the “Ixel” trade-mark. He said he was of the opinion, on the basis of observations made by the Registrar in Vins La Salle Inc. (Les) v. Vignobles Chantecler Ltée (Les) (1985), 6 C.P.R. (3d) 533 (T.M. Opp. Bd.), that in view of Canada’s bilingual character it was the average bilingual Anglophone or Francophone consumer who should serve as the standard for determining whether or not there was a reasonable likelihood of confusion. In his reasons, he lamented the lack of evidence in relation to the perception of the average bilingual consumer.
[4] To overcome this deficiency, each of the parties filed, on appeal before Pinard J., additional evidence essentially consisting of two affidavits subscribed by persons declaring themselves to be experts on bilingualism.
[5] Pinard J., describing the approach he intended to adopt, stated [at paragraph 8]:
Thus, since I agree with the Registrar himself that “[translation] given Canada’s bilingual character, it is the opinion of the average bilingual Anglophone or Francophone consumer that must be considered in assessing the issue of confusion,” and furthermore that “[translation] equal importance must be given to English and French in the evaluation of this issue,” I am of the opinion that with this crucial additional evidence, had it been presented, he would not have been able, in applying the test of the average bilingual consumer, to note as he did the flaw in the experts’ evidence in relation to the phonetic aspect of the trade-marks or to state that he “[translation] entertained some doubt about the confusion”.
And it was after making these remarks that he set about reviewing the new evidence and allowed the respondent’s appeal.
[6] I believe there was a mistake on the part of the Registrar, which was replicated by the Judge, concerning the applicable test. This mistake is such that it requires that the exercise be fully re-engaged, de novo, at first instance. I will explain.
[7] In Boy Scouts of Canada v. Alfred Sternjakob GmbH & Co. KG et al. (1984), 2 C.P.R. (3d) 407 (F.C.T.D.), Joyal J. expressed the following opinion, at page 413:
It could be argued that the criteria in the Trade Marks Act and the evidential findings thereunder as to confusion or deception should be measured not only with reference to English-speaking experience but to French-speaking experience as well. This would result in an inquiry as to the connotation or otherwise of certain words in a bilingual context, with each language having equal presence.
It is a fact that the policy of the Trade Marks Office and the practice of counsel and of agents before it are to check into and analyze the descriptive, misdescriptive, misleading, distinguishing and confusing consequences which flow from a French or English adaptation of any particular word or the use of it as a registered trade mark.
[8] Strayer J., then a judge in the Trial Division, held in Scott Paper Co. v. Beghin-Say S.A. (1985), 5 C.P.R. (3d) 225 (F.C.T.D.), at page 231, that:
I have no doubt that the Registrar of Trade Marks and the court should be alert to the possibility of confusion between trade marks in either or both of Canada’s official languages. This is not only required by the constitutional and legal status of both languages at the federal level, but is also a reflection of the fact that there are several million bilingual Canadians who may associate words in one official language with their equivalent in the other. In the present case, however, I can see no possibility of such confusion between these two trade marks in the minds of unilingual anglophones, and for those who know French or both languages I can see no real probability of such confusion.
[9] Relying on these two decisions, the Registrar, in Vins La Salle Inc. (Les) v. Vignobles Chantecler Ltée (Les) (1985), 6 C.P.R. (3d) 533, concluded at pages 535 and 536 that it was necessary, in deciding the issue of confusion, to choose between either of the following methods:
… i) assess the question of confusion in the context of unilingual francophones, unilingual anglophones and bilingual persons and then if two trade marks are confusing to the average member of any of these groups conclude that the trade marks are confusing, or ii) assess the question of confusion in the context of bilingual persons only.
He chose the second method, that of the average bilingual consumer.
[10] With respect, I think the Registrar, in Les Vins La Salle Inc., misread what Joyal and Strayer JJ. were saying. They were simply observing that in the Canadian linguistic context the perception of the Francophone consumers warranted just as much attention as that of the Anglophone consumers, and once there was a reasonable probability of confusion among either of them, the mark could not be registered. Both judges added, in an effort to adequately cover all the bases in light of the special features of the cases before them, that neither the average French-speaking consumer nor the average English-speaking consumer might be confused but the average bilingual consumer might be, in which case this risk of confusion sufficed by itself to rule out the registration. Nowhere was there any suggestion of a test that would consider only the perception of the average bilingual consumer as opposed to the perceptions of the average French-speaking and English-speaking consumers.
[11] Joyal J., a few months later, in Produits Freddy Inc. v. Ferrero SpA (1986), 20 C.P.R. (3d) 61 (F.C.T.D.), had occasion himself to issue a warning after the Registrar’s decision in Les Vins La Salle Inc. was cited to him. He stated, at page 65, that:
It seems to me that it is not sufficient to simply apply the bilingual version test in determining the issue of confusion between one trade mark and another
and at page 68, that:
… I feel that caution must be exercised when applying the bilingual equivalence test to coined words.
[12] This Court, upholding Joyal J.’s decision in Produits Freddy Inc. v. Ferrero SpA (1988), 22 C.P.R. (3d) 346 (F.C.A.), did not make the average bilingual consumer test an independent test. Quite the contrary, Marceau J.A. noted that the risk of confusion was a finding “of concrete fact to be verified in real life, and not one to be inferred from the constitutionally established bilingual nature of the country” (at page 350). Lacombe J.A., for his part, commented as follows, at page 354:
It is only where a mark is borrowed from everyday speech or derived from a common noun that a question arises as to the possible effect of its transposition into the other language: see Boy Scouts of Canada v. Alfred Sternjakob GmbH & Co. KG et al. (1984), 2 C.P.R. (3d) 407, 4 C.I.P.R. 103 (F.C.T.D.); 101482 Canada Inc. v. Registrar of Trade Marks (1985), 7 C.P.R. (3d) 289, 6 C.I.P.R. 222 (F.C.T.D.). It is a matter which must be considered as one of the tests to be applied in deciding whether such a mark creates confusion with another mark in the mind of the public. This requirement is necessary not only to safeguard the official status of both languages but also the integrity and statutory protection of trade marks in respect of those who speak or understand English and French. However, this factor should not be taken too far, obliterating the other tests mentioned in s. 6(5) of the Act or ignoring the rules laid down by the courts.
[13] The approach adopted by the Court is easily comprehensible. French and English are of equal value in Canada. The Trade-marks Act applies throughout Canada. Section 6 states that confusion may result from the use of a trade-mark in only one area of Canada. Paragraph 12(1)(b) provides that a trade-mark is not registrable if it is “either clearly descriptive or deceptively misdescriptive in the English or French language of the character or quality of the wares or services”. A trade-mark may be used anywhere in Canada (see, for example, section 16 [as am. by S.C. 1994, c. 47, s. 195]) and its registration, according to section 19 [as am. by S.C. 1993, c. 15, s. 60], gives the owner “the exclusive right to the use throughout Canada of the trade-mark”, subject to the cited exceptions.
[14] It follows that once there is a risk of confusion in either of the country’s two official languages, a trade-mark cannot be registered. The particular problem with which Joyal and Strayer JJ. were confronted was the possibility that a trade-mark that does not create any confusion in a Francophone or in an Anglophone might create confusion in a bilingual person through the use of usual, distinct words in French and in English but, to someone who knew what it meant in both languages, referring to the same reality. For example, in Produits Freddy Inc., the word “noixelle” might mean nothing to an English-speaking person, and the word “nutella” might be meaningless to a French-speaking person, but it was not excluded that the use of either of these words would confuse a bilingual person who knew the meaning in both languages. It was solely to guard against this possibility that the test was extended to the average bilingual consumer.
[15] In short, the Trial Judge and, before him, the Registrar, erred in transforming the third prong of the only applicable test (is there a risk of confusion in the average Francophone consumer, the average Anglophone consumer or, in some special instances, the average bilingual consumer?) into an independent test. What is more, the case at bar did not even involve a special case in which the perception of an average bilingual consumer would be relevant, in the sense that Lacombe J.A. understood it in Produits Freddy Inc.
[16] Counsel for the respondent submits that, while the Judge may have erred in law when he said he was applying the sole test of the average bilingual consumer, nevertheless on the facts he applied the right test since he implicitly held that there was no risk of confusion in the average Anglophone consumer. (It is common ground in this case that there is no risk of confusion in French.) I do not think that the Judge applied to the facts any test other than the one he said he was applying in law. On the one hand, in his reasons he refers only to the test of the average bilingual consumer. On the other hand, the only evidence on which he relies, which was not before the Registrar, is the evidence emanating from the two experts who were asked, in their own words, to compensate for the deficiencies that the Registrar alleged might confuse the average bilingual consumer. In any event, the Judge’s error of principle was such that it could only cast suspicion on his approach and his conclusion, especially since the additional evidence that had been presented to him had to a large degree distorted the issue.
[17] The appellant’s counsel, for her part, submits that this Court is as favourably situated as the Trial Judge to review the evidence and arrive at a conclusion without the need to send the matter back to be retried. I do not agree. If the evidence presented to the Trial Judge might have been distorted by the Registrar’s error, this Court would be ill-advised to rely on it on appeal. In the interests of justice a new hearing is required at first instance on the basis of such evidence as the parties consider appropriate to file in addition to the evidence previously filed with the Registrar, and in terms of the test explained in these reasons.
[18] In the circumstances, I am not ruling in any way on the merits of the application for registration.
[19] The appeal should be allowed, the decision of the Trial Judge should be set aside and the matter should be sent back for a trial de novo before another judge of the Trial Division on the basis of such evidence as the parties consider appropriate to file in addition to the evidence previously filed with the Registrar. I would allow the appellant its costs both at trial and on appeal.
Desjardins J.A.: I agree.
Noël J.A.: I agree.