In today’s global economy we can be exposed to new products and service providers on a daily basis who may originate from anywhere in the world. Many of these businesses have originated in countries where the written language is based on something other than the roman alphabet. When businesses expand their brand into Canada from a non-roman alphabet language country, they face unique challenges. These include issues with translations and transliterations and the potential confusion with other foreign language marks.

A. Non-Latin Alphabet Marks in the Trade-marks Act

Non-Latin alphabet marks in Canada have a long and inconsistent history. They are a separate group of marks within foreign language marks. Foreign language marks are largely dealt with within section 29 of the Trade-marks Act, which states:

  1. a translation into English or French of any words in any other language contained in the trade-mark;
  2. where the trade-mark contains matter expressed in character other than Latin characters or in numerals other than Arabic or Roman numerals, a transliteration of the matter in Latin Characters or Arabic numerals; and
  3. a specimen of the trade-mark as used

Section 29(b) deals specifically with non-Latin alphabet marks, requiring both a translation and a transliteration of the mark. The transliteration is often where trade-mark owners encounter problems because the correctness of the transliteration is not monitored by the Trade-marks Office – and incorrect translations and transliterations exist on the register.

B. Registration of Foreign Language Marks: General

The prohibition against registration of clearly descriptive and deceptively misdescriptive trade-marks under section 12(1)(b) of the Act applies to foreign language marks. The caveat is that the trade-marks or transliteration of the marks must be clearly descriptive or deceptively misdescriptive in English or French. If the trade-mark is clearly descriptive or deceptively misdescriptive in a language other than English or French it remains registrable- so long as it is not the actual name of the goods or services with which it is associated. Specifically, this means a word that is descriptive to someone who speaks the foreign language, but is not for an English and French speaker, remains registrable; for example in B. Manischewitz Co. v. Gula, [1946] Ex.C.R. 570, 6 C.P.R. 4, the trade-mark Tam Tam was found to be registrable even though in Yiddish, it is equivalent to the term “taste” or “tasty” in English.

The courts have also considered marks consisting of foreign words that have become integrated into either English or French to such an extent that they are recognized and understood by French or English speakers. In such situations the marks are unregistrable.[I] These include generic terms such as “wurst”, German for sausage; “gelato”, Italian for ice cream; “anorak”, Inuktitut for parka.

C. The Matter of Confusion

On the matter of confusion between foreign language marks, the courts have been split on the issue; and have reached different decision based on the facts of the case and evidence presented or lack thereof.

Issues and Historical Treatment of Non-Latin Alphabet Marks:

One of the earliest cases was the case of B. Jadow and Sons, Inc. v. Grupo Cyanomex, 1992 FCJ 957, 45 CPR (3d) 161, which dealt with the trade-mark “Kola Loka” which is an exact translation of “Krazy Glue” in Spanish. The Court found that there was no likelihood of confusion since the test for confusion is a matter of first impression to the average consumer having vague or impact recollection. The two marks were distinctive from each other, having no similarities in either sound or appearance. More importantly, the court found there was no likelihood of confusion because there was only a minimum proportion of Canadians speaking or understanding Spanish and therefore capable of making the translation from Spanish to English.

This issue of whether the average consumer can understand and translate the foreign language has been a central issue in the determining whether there is a likelihood of confusion when considering foreign language trade-marks.

In the 1990s, confusion was initially difficult to establish for non-Latin alphabet trade-marks. The confusion analysis was performed from the perspective of the average Canadian consumer who was assumed to only understand English and French; therefore when confronted with a non-Latin alphabet mark, the court and Trade Mark Opposition Board could only conclude that the average consumer, understanding only English and French, would view the mark as a collection of meaningless scribbles.

The problem with establishing confusion is made more difficult because transliterations could be incorrect or provide poor guidance in pronunciation. With non-Latin alphabet marks, transliterations and translations are not monitored by the Trade-marks Office and trade-mark applicants are allowed to place on the record whatever they wish. This obviously becomes an issue with languages such as Chinese. The Chinese language has two main dialects but only one written language (although there is both a traditional and simplified form). The two main dialects of Chinese are Mandarin and Cantonese, which are vastly different as spoken languages but are largely identical when written. This means that transliterations of the same characters in Mandarin and Cantonese can be very different but no less correct. It also leads to problems where the transliteration from Chinese characters does not provide a clear pronunciation in the native language. In such cases the transliteration will be pronounced in different ways. Take for example the trade-mark and trade-name of the Chinese company Huawei. Mandarin speakers will pronounce it as “HOO-A –WAY” but Cantonese speakers will pronounce it “WAH-WAY” and English speakers will tend to make up their own pronunciations such as “HOW-WAY”. This clearly leads to difficulties when assessing such marks for confusion.

The fact that vague translations and incorrect or partially correct transliterations were allowed to be registered for non-Latin alphabet trade-marks meant that the register was unreliable at best and obviously wrong at the worst. Coupled with the fact that the confusion test was applied from the perspective of the average Canadian consumer who could only understand English and French it meant that removing confusing non-Latin alphabet marks was extremely difficult. The test when applied under these conditions left marks that were nearly identical to the average consumer who could understand the foreign language in which the marks were rendered on the register.

The Current Approach:

The Courts have addressed the confusion analysis as it pertains to non-Latin marks in more details in recent years; however the case law is still murky.

The Courts have at least reached a consensus on the perspective from which the analysis for confusion of non-Latin alphabet marks is to be conducted and have accepted that the analysis can deviate from the normal course of the average consumer who is assumed to only understand English and French.

Instead, in the case of non-Latin alphabet marks, the analysis is now slightly more complex – the analysis can be conducted using the perspective of an average consumer who understands the foreign language in which the mark is rendered if the facts of the case can establish that there is an appreciable/significant number of consumers who can understand the language in which the mark is rendered.[ii] This essentially adds an additional step to the analysis, namely making a factual decision about whether or not an appreciable number of the consumers understand the language in which the mark is rendered.

Where the Court finds there is no evidence that an appreciable/significant number of the consumers can understand the language the decision has been that the average consumers would view the non-Latin alphabet mark as no more than random symbols with no particular meaning and no similarity when sounded. This determination is determined case by case and relying on case where evidence that a portion of the Canadian population can speak and understand a particular language was adduced and accepted does not appear to be sufficient.[iii]

In the case where there is a significant number of consumers who understand the language in which the mark is rendered the Court then needs to assess how the relevant consumer would understand the mark. This is extremely costly to trade-mark owners as expert testimony and reports are usually required to inform the Court how a relevant consumer would understand the trade-mark in question. How well this evidence is presented to the Court and how the Court interprets the expert testimony and reports will often determine the outcome of the case.

The case of Saint Honore Cake Shop Limited v Cheung’s Bakery Products Ltd., 2015 FCA 12, demonstrates that cases involving non-Latin alphabet marks require costly evidence to demonstrate two things; one that an appreciable number of consumers understand the foreign language and two that the marks would be construed in a certain way by the relevant consumer segment.

This case in particular highlights the difficulties associated with non-Latin alphabet marks in languages where it is impossible to “sound out” a word – specifically Chinese. Convincing the Courts that a person who can speak and understand a language as spoken but not be able to read or understand the written language appears to be an uphill battle.