Déjà brew: the registrability of non-English trade marks post-Cantarella
As the demographic of the average Australian consumer, and their understanding of non-English words, continue to grow, it is likely that there will likewise be a similar growth in both registrations and cases involving challenges to the distinctiveness of trade marks composed of non-English words.
The High Court of Australia’s decision in Cantarella Bros Pty Ltd v Modena Trading Pty Ltd (2014) 254 CLR 337 remains the leading Australian authority for determining the distinctiveness of non-English word trade marks. In the more than seven years since the decision, there remains debate and uncertainty regarding how to apply the majority’s two-step test, particularly in the context of marks comprising non-English word marks. In this article, we look at reported decisions involving non-English word marks that have followed Cantarella and consider what lies ahead.
Almost as soon as it was handed down over seven years ago, the High Court’s decision in Cantarella became the subject of intense debate among commentators, with some critical of what the decision left unclear. The two-step test, which formed the ratio decidendi of the majority decision, has been repeatedly applied in Australia both by the Trade Marks Office and the Federal Court of Australia. There have, however, been relatively few reported decisions since then that have grappled with the specific challenges that arise when applying the test to trade marks composed of non-English words. An analysis of those decisions reveals some practical difficulties in applying the test and also highlights the uncertainty that remains as to a number of matters that arise specifically in the context of such trade marks.
After providing an overview of the distinctiveness requirement in section 41 of the Trade Marks Act 1995 (Cth) and the Cantarella litigation, this article analyses subsequent decisions that have applied the test to non-English words. In particular, we examine some of the practical considerations that may arise in attempting to secure acceptance, to defend an opposition to registration, or to defend a registration in the context of an expungement action. In the final section we look at what remains unclear and consider whether the potential issues raised in early critiques of Cantarella have been addressed or remain unanswered.
Distinctiveness for the purposes of section 41
Section 41 provides that an application for the registration of a trade mark must be rejected if the “sign” (trade mark) is “not capable of distinguishing” the applicant’s goods or services for which registration is sought from the goods or services of other persons.
Marks that are inherently adapted to distinguish ipso facto meet the threshold. Those that have some inherent adaptation to distinguish may be brought up to the requisite threshold by evidence of acquired distinctiveness (through use or intended use) and/or other circumstances. Marks that lack any inherent adaption to distinguish must be shown to be distinctive in fact at the application date by reason of their use.
Inherent adaptation to distinguish
The concept of inherent adaption to distinguish was carried over from the Trade Marks Act 1955 (Cth) and previous British legislation on which it was modelled. What Gibbs J said in Burger King Corporation v Registrar of Trade Marks(1973) 128 CLR 417, in relation to paragraph 26(2)(a) of the Trade Marks Act 1955 (Cth) remains good law:
"Inherent adaptability is something which depends upon the nature of the trade mark itself – see Clark Equipment Co v Registrar of Trade Marks [(1964) 111 CLR 511,515] – and therefore is not something that can be acquired; the inherent nature of the trade mark itself cannot be changed by use or otherwise."
Generally, trade marks that are not inherently adapted to distinguish goods or services are mostly trade marks that consist wholly of a sign that is ordinarily used to indicate the kind, quality, quantity, intended purpose, value, geographical origin, or some other characteristic, of goods or services or the time of production of goods or of the rendering of services.
The High Court in Cantarella confirmed that there is no special test for non-English words. Cantarella reiterated that the governing principle for admitting trade marks to registration and allowing them to remain on the Register is a public interest consideration that has been constant in Anglo-Australian trade mark law for more than a century, as restated by Lord Diplock in 1976:
"[L]ong before the reference to inherent adaptability had been incorporated in the current statutes dealing with trade marks, it had been held upon grounds of public policy that a trader ought not to be allowed to obtain by registration under the Trade Marks Act a monopoly in what other traders may legitimately desire to use. The classic statement of this doctrine is to be found in the speech of Lord Parker in [Du Cros] ... The reference to ‘inherently adapted’ in s 9(3) of the Consolidation Act of 1938, which was first enacted in 1937, has always been treated as giving statutory expression to the doctrine as previously stated by Lord Parker." (Smith Kline & French Laboratories Ltd v Sterling-Winthrop Group Ltd [1975] 1 WLR 914; [1975] 2 All ER 578)
Cantarella Bros Pty Ltd v Modena Trading Pty Ltd
Federal Court – Cantarella Bros Pty Ltd v Modena Trading Pty Ltd (2013) 299 ALR 752
Cantarella Bros Pty Ltd (“Cantarella Bros”) is an Australian importer and supplier of European foods and drinks. Cantarella Bros was and remains the owner of Australian registered trade marks for the words ORO and CINQUE STELLE, each of which is registered in respect of a range of coffee products. “Oro” is the Italian word for “gold” and “Cinque Stelle” is Italian for “five stars”. Modena Trading Pty Ltd (“Modena”) distributed Caffe Molinari coffee products bearing the signs ORO and CINQUE STELLE.
In February 2011, Cantarella Bros commenced proceedings in the Federal Court against Modena to restrain Modena from importing for sale, selling, and offering for sale, coffee and coffee related products bearing the signs ORO and CINQUE STELLE. Modena denied infringement and sought, by way of cross-claim, to expunge Cantarella Bros’ registrations on the basis that they were not adapted to distinguish and did not in fact distinguish Cantarella Bros’ goods from the goods of other traders and did not qualify for registration under section 41 of the Trade Marks Act. A key issue was whether the trade marks were distinctive including to what degree they might be inherently adapted to distinguish.
Justice Emmett held that the appropriate question was whether the words themselves (“oro” and “cinque stelle”) would be commonly understood in Australia to mean “gold” and “five stars” respectively, not whether the words when translated would be distinctive, nor whether the Italian language is well understood by many people in Australia.
His Honour was not persuaded, on the material before him, that the Italian language was so widespread that “oro” and “cinque stelle” would be generally understood in Australia as meaning “gold” and “five stars” respectively. Accordingly, his Honour found that both ORO and CINQUE STELLE were inherently adapted to distinguish Cantarella Bros’ goods from the goods of other traders.
Full Court of the Federal Court of Australia – Modena Trading Pty Ltd v Cantarella Bros Pty Ltd (2013) 215 FCR 16
On appeal the Full Court (Mansfield, Jacobson and Gilmour JJ) overturned Justice Emmett’s decision and held that the words “oro” and “cinque stelle” were not inherently adapted to distinguish Cantarella’s goods.
The Full Court held that his Honour had erred by focussing on consumers as the relevant audience. Instead, the Full Court held that the focus of the inquiry should have been whether traders were likely to think of and want to use the words to describe their own goods. This conclusion was based on an understanding that section 41 “is directed to the distinctiveness of an applicant’s goods or services from those of other persons” where those “persons” are other traders. When the focus shifted from the consumers’ understanding of the Italian words to the need of other traders to describe their goods as having an Italian origin and of being high quality, the Court found it was clear that there were other traders who would seek to legitimately use “oro” and “cinque stelle” to describe their coffee products.
It was unnecessary to consider whether consumers know what the words mean in English, however, as Italian is a commonly spoken language in Australia, and it was accepted that many people in Australia would in fact know that “oro” meant “gold” and that “cinque stelle” meant “five stars”. The fact that these were common laudatory words in Italian rendered the trade marks descriptive, rather than distinctive.
High Court – Cantarella Bros Pty Ltd v Modena Trading Pty Ltd (2014) 254 CLR 337
A majority of the High Court (French CJ, Hayne, Crennan and Kiefel JJ (Gageler J dissenting)) overturned the Full Court’s decision and confirmed that the test for distinctiveness for trade marks composed of non-English words is the same as for trade marks composed of English words.
The majority adopted the reasoning of Kitto J in Clark Equipment Co v Registrar of Trade Marks (1964) 111 CLR 511, referring to Registrar of Trade Marks v W. and G. Du Cros Ltd (“Du Cros”) [1913] AC 624 where Lord Parker of Waddington said:
"The applicant’s chance of success in this respect (i.e. in distinguishing his goods by means of the mark, apart from the effects of registration) must, I think, largely depend upon whether other traders are likely, in the ordinary course of their businesses and without any improper motive, to desire to use the same mark, or some mark nearly resembling it, upon or in connection with their own goods. It is apparent from the history of trade marks in this country that both the Legislature and the Courts have always shown a natural disinclination to allow any person to obtain by registration under the Trade Marks Acts a monopoly in what others may legitimately desire to use."
Kitto J went on to say that:
"...[I]t is to insist that the question whether a mark is adapted to distinguish be tested by reference to the likelihood that other persons, trading in goods of the relevant kind and being actuated only by proper motives – in the exercise, that is to say, of the common right of the public to make honest use of words forming part of the common heritage, for the sake of the signification which they ordinarily possess – will think of the word and want to use it in connection with similar goods in any manner which would infringe a registered trade mark granted in respect of it."
In Cantarella the majority observed:
"It is the “ordinary signification” of the word, in Australia, to persons who will purchase, consume or trade in the goods which permits a conclusion to be drawn as to whether the word contains a “direct reference” to the relevant goods (prima facie not registrable) or makes a “covert and skilful allusion” to the relevant goods (prima facie registrable). When the “other traders” test from Du Cros is applied to a word (other than a geographical name or a surname), the test refers to the legitimate desire of other traders to use a word which is directly descriptive in respect of the same or similar goods. The test does not encompass the desire of other traders to use words which in relation to the goods are allusive or metaphorical. In relation to a word mark, English or foreign, “inherent adaption to distinguish” requires examination of the word itself, in the context of its proposed application to particular goods in Australia."
For this purpose, the High Court propounded a two-stage test:
"In accordance with the principles established in Mark Foy’s and restated in Clark Equipment, Faulding and Burger King, determining whether a trade mark is “inherently adapted to distinguish”, as required by s 41(3), requires consideration of the “ordinary signification” of the words proposed as trade marks to any person in Australia concerned with the goods to which the proposed trade mark is to be applied.
As shown by the authorities in this Court, the consideration of the “ordinary signification” of any word or words (English or foreign) which constitute a trade mark is crucial, whether (as here) a trade mark consisting of such a word or words is alleged not to be registrable because it is not an invented word and it has “direct” reference to the character and quality of goods, or because it is a laudatory epithet or a geographical name, or because it is a surname, or because it has lost its distinctiveness, or because it never had the requisite distinctiveness to start with. Once the “ordinary signification” of a word, English or foreign, is established an enquiry can then be made into whether other traders might legitimately need to use the word in respect of their goods. If a foreign word contains an allusive reference to the relevant goods it is prima facie qualified for the grant of a monopoly. However, if the foreign word is understood by the target audience as having a directly descriptive meaning in relation to the relevant goods, then prima facie the proprietor is not entitled to a monopoly of it. Speaking generally, words which are prima facie entitled to a monopoly secured by registration are inherently adapted to distinguish." [emphasis added]
In practical terms, whether composed of English or non-English words, determining whether a mark is inherently adapted to distinguish therefore involves:
- first a consideration of the ordinary signification of the word or words for any person in Australia concerned with the relevant goods or services; and
- secondly whether other traders might legitimately need to use the word or words in respect of their goods or services for their ordinary signification.
The test is not concerned with asking what the word or words mean when translated into English (although this may be relevant). What is “critical” is how the word or words are understood by the target audience, being those who are concerned with the relevant goods.
After establishing that non-English words are not to be treated differently from English words, the majority discussed the principles that apply to trade marks that are descriptive of the goods for which they are to be registered. In this context, the majority held that what is relevant is the:
"…‘ordinary signification’ of the word, in Australia, to persons who will purchase, consume or trade in the goods which permits a conclusion to be drawn as to whether the word contains a ‘direct reference’ to the relevant goods (prima facie not registrable) or makes a ‘covert and skilful allusion’ to the relevant goods (prima facie registrable)."
If a trade mark composed of English or non-English words contains an allusive or indirect reference to the relevant goods, it is prima facie qualified for the grant of a monopoly because it is inherently distinctive. However, if a word is understood by the target audience as having a direct reference to some characteristic of the relevant goods, then prima facie it is not inherently distinctive and so not registrable, although evidence of acquired distinctiveness might raise it to the relevant threshold.
The majority held that ORO and CINQUE STELLE do not convey a meaning or idea sufficiently tangible to constitute a direct reference to the character or quality of the goods for people in Australia concerned with coffee goods.
Post-Cantarella
When the question of a trade mark’s distinctiveness is raised, determining the ordinary signification of the sign in question is therefore fundamental. However, there is no single way to determine that question. Ordinary signification has been equated with ordinary meaning but in the case of invented words there will be no such meaning. In the case of English words that are adopted to function as trade marks, the analysis is more involved and the complexity of that analysis will increase depending on the number of meanings that might be attributed to the words in issue in the context in which they appear when applied to the relevant goods or services. That complexity further increases where the trade marks in question are non-English words that might have multiple meanings (whether in one language or multiple languages) or might be understood by people from different ethnic or linguistics backgrounds or with different levels of education or in the different contexts in which the word or words appear.
Despite there being a very large number of decisions citing Cantarella, only a handful of those decisions have involved the registrability of non-English words and the question of distinctiveness. An analysis of decisions that have involved non-English words reveals the varying levels of complexity involved in assessing distinctiveness and in particular ordinary signification, as illustrated below.
Re Majans Pty Ltd [2016] ATMO 47
The applicant in this decision applied to register the trade mark BHUJA which it used in relation to savoury snack food mixes. The examiner objected on the basis that the word was commonly used to refer to a spicy dried snack food mix. The applicant submitted that “bhuja” is a foreign word and would have no meaning to customers whose first language is English. The Registrar’s delegate disagreed, holding that the word signified a type of snack food consisting of spices and dried ingredients akin to the way that “trail mix” is a type of snack food.
The applicant’s own promotional material also showed that “bhuja” ordinarily signified a type of spicy dried snack food. Additionally, there was evidence that other traders were already using the word as a descriptor of their goods.
The decision is unremarkable in its application of Cantarella and unsurprising in its outcome such that little can be gleaned from the approach taken to determining ordinary signification. (This application was subsequently accepted for registration without an appeal following minor amendments to the goods.)
Emilia v Vinedos Emiliana SA [2017] ATMO 67
This decision involved an application to register the trade mark EMILIANA for wines and alcoholic beverages. The application was opposed by a not-for-profit trade association for the Emilia Indicazione Geografica Tipica, who were responsible for various protective activities in respect of the Emilia region, a wine region recognised by the Government of Italy. In Australia, both “dell’Emilia” and “Emilia” are protected geographical indications under the Australian Grape and Wine Authority Act 2013 (Cth).
The Registrar’s delegate was satisfied that:
"Italian speaking Australians (of which there are many) would, on the balance of probabilities, understand the [EMILIANA] Trade Mark as identifying the Emilia wine region in Italy” and “that many Australian consumers of wine would be familiar with the word ‘Emilia’ denoting the Italian geographical region associated with wine production …"
Despite the Registrar’s delegate determining that the EMILIANA trade mark was a geographical indication and therefore unregistrable, the applicant was able to overcome this hurdle on the basis that it had been using EMILIANA as a trade mark for its wines since 1986, availing itself of the grandfathering provision in section 61(2)(c)(i) of the Trade Marks Act.
The Registrar’s delegate was also required to consider whether the sign was inherently adapted to distinguish for the purposes of section 41. Focusing largely on the legitimate need for other traders to use “Emilia” and “dell’Emilia” for their geographical indications, the opponent relied on the fact that the word “Emiliana” was a synonym of “dell’Emilia”, as both were Italian for “of/from Emilia.” The delegate also made reference to other Italian phrases in which “Emiliana” may appear, but determined that “this would not be understood by most Australians”.
Ultimately, the Registrar’s delegate found that the “ordinary signification of the Italian word ‘EMILIANA’ is a first name and a family name and not one other traders might legitimately need to use in respect of their goods”. The trade mark was therefore found to contain an allusive reference rather than a descriptive one.
The Registrar’s delegate also held that if they were wrong on prima facie registrability, other factors to be taken into account in dismissing the section 41 ground of opposition included:
(1) that the trade mark had been used by the applicant in respect of wine for a considerable time; and
(2) that the goods had been marketed within Australia in such a way that Australian consumers would not be caused to wonder if the goods originated in the Emilia region of Italy.
Emphasis was placed on the fact that the applicant’s products were produced in Chile, where there is no geographical indication for “Emiliana” and the Registrar’s delegate also noted that the inclusion of the suffix “na” in the trade mark had a significant impact on the ability of the trade mark as a whole to distinguish the goods of the applicant because:
"[t]he [EMILIANA] Trade Mark has no meaning in English in reference to the Applicant’s goods.” This last finding, however, appears to have involved what the majority in the High Court stated was not the critical enquiry – that is, “it is not the meaning of a foreign word as translated which is critical, although it might be relevant. What is critical is the meaning conveyed by a foreign word to those who will be concerned with the relevant goods."
Re Mantra IP Pty Ltd [2017] ATMO 10
This decision also involved questions of distinctiveness intertwined with questions of geographical indication. The applicant applied to register BALÉ for travel and tourism related services. The examiner issued an adverse report citing the fact that “Balé” is the name of geographical locations in Croatia and Ethiopia and was likely to be a word other traders would legitimately seek to use. The applicant sought to overcome this by limiting the proposed services as follows: “none of the foregoing related to services provided in respect of any places outside of Australia called Bale.” A second adverse report was issued and the applicant sought a hearing.
On the question of ordinary signification, the Registrar’s delegate noted that the word is Portuguese for “ballet”, as well as the name of 45 provinces in Burkina Faso, which is known for its Deux Balé Forest and savannah elephants. The Registrar’s delegate also noted that:
"In assessing the ordinary signification of a word it is important that the persons who will purchase, consume or trade in the goods or services are identified. Asking ‘the man on the Clapham omnibus’ will not necessarily yield useful results. The word ‘EUTECTIC’, for example, is unlikely to have any signification at all for nearly the entire population of Australia, yet that term entirely lacks inherent adaptation to distinguish certain goods. Equally, that to the man on the Clapham omnibus ‘BALÉ’ has no ordinary signification is not particularly instructive."
As Africa is a popular tourist destination for those seeking to see “the Big Five” African animals, the Registrar’s delegate concluded it was reasonable to suppose that if not already, Balé is likely to become a recognised destination for those seeking to see the elephants that live there. If this occurs, the ordinary signification of “Balé” to persons who wish to travel to Africa to see elephants would be the geographical location where elephants can be found.
When considering the second limb of Cantarella, the delegate noted that Balé was fairly unknown as a tourist destination and accepted as persuasive the submission that it would be absurd if every geographical indication was held to lack inherent adaption to distinguish tourism services, regardless of its popularity. However, the Registrar’s delegate considered it reasonable to suppose that tourism services might be provided in connection with Balé at some point in the future and that on that basis the trade mark was one which other traders may legitimately need to use.
The applicant also failed to establish that the trade mark had acquired distinctiveness. The examples of use of BALÉ as a trade mark provided by the applicant were held to fall well short of the threshold for proving that the trade mark did in fact distinguish the applicant’s services from those of other traders.
The outcome of this decision reflects a strict application of the test propounded and the Australian Trade Mark Office’s strict approach to geographical indications. A significant focus appears to have been placed on the second limb and future circumstances. While it is true that regard ought to be had to what other traders may legitimately wish to do in future, the relevant analysis must be by reference to the potential future use of the words as signifying the meaning that they ordinarily possess. A qualification of the kind that was volunteered during examination, limiting the use of the trade mark so as not to be in connection with services provided in respect of any place outside of Australia called Balé, would arguably have no effect on the ability of other traders to use the word to describe services related to places called Balé.
Goodman Fielder Pte Ltd v Conga Foods Pty Ltd [2020] FCA 1808
As in Cantarella, the section 41 issue in this case arose as part of a cross-claim filed by the respondents to an infringement claim. Again, the trade marks were composed of two Italian words “La Famiglia”, applied to common foodstuffs (garlic and other bread products) sold in major supermarkets across the country, mirroring to some extent the factual background in Cantarella. The words in issue are, however, arguably of broader import given the implications that this decision may have for other traders.
The applicants are the registered owners of the following trade marks:
The infringement proceedings were commenced to restrain the respondents from selling pasta products bearing the sign:
The applicants alleged infringement of each of the word mark registrations above that included the words “la famiglia”, namely LA FAMIGLIA and LA FAMIGLIA KITCHEN, registered for broad class 30 goods that included bread products.
Aside from denying infringement, the respondents cross-claimed arguing that LA FAMILGIA and LA FAMIGLIA KITCHEN were not distinctive at their respective filing dates and were not capable of distinguishing within the meaning of section 41. They sought to distinguish the outcome in Cantarella on the basis that the evidence indicated that the words “la famiglia” are much simpler, more readily understood and more familiar in Australia, and needed to be used by Italian family-run businesses. That is, “la famiglia” clearly means “the family” and “la famiglia kitchen” clearly means “the family kitchen” and this would be obvious to Australian traders and the public.
The respondents relied on the similarities in the pronunciation and spelling of the Italian words and the English translation, as well as the similarity to words for “the family” in 20 languages including Albanian, Basque, Danish, Esperanto, French, German, Italian, Latin, Swahili and Swedish. Evidence was also led of other traders using “la famiglia” in relation to restaurants, as well as evidence of “famiglia” appearing on graves, all with the intention of establishing the ubiquity of the sign in Australia.
The respondents argued that the ordinary signification of “la famiglia” when used in relation to bread products was “that it’s a family product intended to be consumed with your family on an occasion that will bring the people together around the table with shared food. It’s a sharing food.”
The applicants submitted that the words “la famiglia” or even “the family” were not directly descriptive of the goods in issue and that even based on the translation “the family”, when used in relation to bread products those words are no more than an allusive or metaphorical reference to a family.
(a) “La Famiglia”
Justice Burley accepted that “a substantial number of consumers and traders in Australia would understand that the words LA FAMIGLIA are foreign words that mean ‘the family’”. The words are not within the class of foreign words for which his Honour considered the original meaning to be opaque to English speakers in Australia, nor are they, however, likely to be understood by all English speaking Australians as having this connection. Nevertheless, his Honour accepted that “one signification of the trade mark LA FAMIGLIA is likely to be that it is a foreign word, probably Italian (or another Romance language, such as Spanish) that means ‘the family’”.
It did not follow, however, that the words would be understood as describing a family sized or a family made product. Rather the words evoke “for those who recognise the language an Italian feel, with whatever connotations or associations that supplies”. This was contrasted with “la deliziosa” which his Honour considered more directly referred to a quality of a product. The words “la famiglia” would not be understood as any more than an allusive reference not conveying any meaning or idea sufficiently tangible to amount to a direct reference to the character or quality of the goods:
"In the present case, and in the context of the goods to which they are applied, the words “la famiglia” have vague, if any definite meaning, and connote, in relation to the designated products, no more than an emotive tendency that the designated goods (like garlic bread) perhaps may ultimately be consumed in a convivial dining setting a bit like that in which Italian families are imagined to dine when together, or perhaps may be concerned with family. To institute a search for a meaning almost necessarily implies that in ordinary English speech the words do not possess a connotation sufficiently definite to amount to a direct reference to the character or quality of the goods …"
While his Honour’s reasoning as to the limited value of translating “la famiglia” into English is consistent with the majority’s reasoning in Cantarella, it is difficult to reconcile how a search for meaning of the kind to which his Honour refers would have been required in circumstances where it was accepted that “a substantial number of consumers and traders in Australia would understand that the words LA FAMIGLIA are foreign words that mean ‘the family’”.
Having concluded in effect that the words did not have an ordinary signification that was directly descriptive for the relevant goods, for the purposes of the first limb in Cantarella, his Honour considered that the question of whether other traders would desire to use the mark for the purposes of the second limb did not arise. Nonetheless, for completeness, his Honour noted that the evidence did not lead to the conclusion that other traders would wish to use the words “la famiglia” in relation to the registered goods or goods of the same description. This was notwithstanding that one such other trader, the second respondent, was using the words “la famiglia” in a manner which, on his Honour’s earlier reasoning, would arguably be understood to mean “The Family Rana” or “The Rana Family” (ie. descriptively).
His Honour said that there was no uncertainty that LA FAMIGLIA was inherently adapted to distinguish, and so it was unnecessary to consider acquired distinctiveness arising from use or other circumstances under section 41(5) (pre-Raising the Bar) to render it “capable of distinguishing”. Despite this, his Honour observed that the evidence of significant use of the mark demonstrated that the requirements would have been satisfied regardless.
(b) “La Famiglia Kitchen”
With respect to the trade mark LA FAMIGLIA KITCHEN, his Honour found that the addition of the word “kitchen” had the effect of moving the trade mark into the territory of being a trade mark that is less allusive, and more literal. Nevertheless, his Honour found that the reference in the trade mark was indirect, and that the combination of Italian and English was sufficient to meet the distinctiveness requirements of section 41 despite it sitting near the borderline.
For completeness, his Honour also considered whether, if LA FAMIGLIA KITCHEN was to some extent, but not sufficiently inherently adapted to distinguish, the evidence was sufficient to satisfy section 41(4) (post-Raising the Bar). His Honour concluded that the mark had actually been used and had acquired distinctiveness in relation to all but the general foodstuffs category for which it was registered.
The decision arguably demonstrates the practical difficulties associated with applying the two-step test in Cantarella and the attendant risk of conflating the first and second limb. All that the first limb requires is to identify the “ordinary signification” of the mark, bearing in mind that there may be more than one. Whether or not the ordinary signification is allusive is one of the issues that bears upon whether the impairment of other’s rights to use the sign is justified.
The authorities indicate that the needs of other traders must be tested by reference to the ordinary signification that the words in question possess once that ordinary signification has been identified. Just because a sign has a level descriptiveness does not mean that other traders will need to use it to describe their goods or services (see for example SA Brewing Co Pty Ltd v Carlton & United Breweries [2001] FCA 902; (2001) 185 ALR 719 where the mark in issue was SHOWDOWN). This is the evaluative assessment of direct versus indirect description. Equally, it may not always follow that where the ordinary signification is not prima facie descriptive, other traders may still need to use those words for the sake of their ordinary signification (see for example Du Cros).
Are the words “la famiglia” any more distinctive than the initials W & G? In Du Cros, the issue was the rights of other to use the same initials. Even if it is accepted that words such as “la famiglia” on their own are not directly descriptive of the goods in issue, what of the need for other traders to use the words “la famiglia” to indicate the name of their family? The upshot of this decision appears to be that no Italian family can indicate that its goods come from them – for example, “La Famiglia Montini”, “La Famiglia Borghese”, “La Famiglia Lucchese”. On one view this may be an unwarranted impairment of the rights of Italian family traders at least insofar as the same goods the subject of the registrations or goods of the same description as the goods the subject of the registrations, even if it is accepted that the words “la famiglia” are not directly descriptive.
Practical considerations for non-English trade marks
As illustrated above, there are a number of different approaches taken to establishing whether non-English words have an ordinary signification when their distinctiveness is called into question under section 41. The decisions from both the Australian Trade Marks Office and the Federal Court provide some interesting insights into which aspects of the often voluminous evidence decision makers will and will not consider and/or rely upon.
One of the key issues is clearly that of establishing what the ordinary signification of a non-English word is for consumers and traders who will be concerned with the relevant goods or services. This undoubtedly raises questions of translation, the proportion of the target audience that speak the language in which the trade mark is presented and the ways in which the words will be understood in context.
Evidence might, for example, be adduced of the literal meaning of the words in issue in the native language from which the mark has been adopted, and applied for or registered, together with a translation into English. It might also be necessary to consider whether there are other languages that might share a common heritage or morphology that may be informative. Evidence might also be adduced of the way the words in issue appear in everyday contexts as a means of demonstrating their ubiquity. Evidence might also be adduced as to the number of persons who speak or comprehend the language in question in Australia.
Given the nature of the inquiry and the focus on “ordinary meaning” to any persons concerned with the relevant goods, it remains to be seen whether expert evidence can help with the inquiry. Such evidence may be considered useful, for example, where there are competing translations at issue or even multiple dialects within the same language.
Taking this analysis a step further, one can imagine a situation where a single word may have multifaceted meanings, which are informed by the context in which the word appears. Still more complex would be a situation in which the relevant language is predominantly oral, rather than written, or where a written word may have multiple pronunciations, each with a different meaning.
As a practical matter, the assistance of a translator may also be required, for the purposes of making statutory declarations, taking affidavits or cross-examination or where reliance is placed on the appearance of the word in issue in non-English materials from which examples of use might be derived, such as non-English publications or broadcasts.
These enquiries are often closely intertwined with the second limb of Cantarella which requires consideration of whether other traders may legitimately wish to use the words in issue in respect of their goods. As the Full Federal Court noted in Telstra Corp Ltd v Phone Directories Company Australia Pty Ltd (2015) 237 FCR 388 the same evidence might be called for the purposes of either limb:
"It is important to note the possible dual significance of such evidence. Evidence of other traders using the word yellow or the colour yellow in connection with their print and online directories could be evidence of the ordinary signification of the YELLOW mark and of the fact that other traders might want to use the word to describe their goods or services."
In the case of the second limb it is, however, important to note that not every use of the word in question will be relevant. It is not enough to show only that there will be some impairment to the rights of other traders. The monopoly secured by registration will, as a practical matter always impair the rights of other traders to some degree, irrespective of whether there are pre-existing common law rights in the trade mark in question or not. What must be shown is that the impairment is unreasonable – that is, will registration impair other traders without improper motive from using a word for its “ordinary signification”? As a practical matter the challenge is showing not only that other traders are in fact using the word but how those other traders are using the word.
ASIC searches, ABN searches, Wayback searches, Google searches and trade mark searches may obviously be a good start but may not in fact necessarily establish use in the requisite manner, if at all. Where such evidence can be marshalled it must demonstrate use of the relevant sign in a descriptive fashion. It is also not enough to rely upon uses that might be able to be readily distinguished as aurally, visually and semantically different. Similarly, the presence of other proprietor’s trade marks comprising similar signs (whether in whole or in part) may not be sufficient to show that the words in issue, standing alone, would be understood in Australia by persons concerned with the relevant goods or services to be directly descriptive of character or quality.
Even then, the fact that a sign is descriptive may not be the end of the road. Courts have accepted that identifying a mark as descriptive is not fatal to registration, because descriptiveness and distinctiveness are not mutually exclusive. As Hill J noted in Stores v Frito-Lay Trading Co Gmbh (2001) 190 ALR 185, citing Johnson & Johnson Australia Pty Ltd v Sterling Pharmaceuticals Pty Ltd :
"It is well-established that there is not a true dichotomy between words capable of being used as a badge of origin and words that are descriptive: see Johnson & Johnson at FCR 347; IPR 24 per Gummow J and at FCR 339; IPR 16-17 per Lockhart J."
The same considerations ought to also apply to the other uses for which one may seek to utilise the word or words. These uses need to be carefully considered in the context in which they appear to determine whether they are in fact sufficient to demonstrate that other traders may need to use the sign in question without improper motive for their ordinary signification.
The examples above demonstrate that mechanistically applying the two-step test, particularly to trade marks containing non-English words, risks muddying the analysis required for the purposes of considering distinctiveness. This has the potential to result in the test devolving into a binary equation of the kind Professor Michael Handler portended in his early commentaries on the Cantarella decision.
It is important to bear in mind that while inherent adaptability depends upon the nature of the trade mark itself and is not something that can be acquired or changed by use, proprietors are able to rely upon use to secure or maintain a registration even where the sign in question might not itself be inherently adapted to distinguish or is only to some extent inherently adapted to distinguish. As with a trade mark containing English words, the extent and manner of such use will be important.
What the future holds
As the demographic of the average Australian consumer, and their understanding of non-English words, continue to grow, it is likely that there will likewise be a similar growth in both registrations and cases involving challenges to the distinctiveness of trade marks composed of non-English words. There is also the potential for further examples of hybrid word marks, which contain a mixture of English and non-English words, to emerge, which will likely engender their own nuances when applying the test for distinctiveness under section 41.
As the decisions to date have highlighted, interesting additional questions have arisen in this context following the decision in Cantarella, including, for example, the scope of the monopoly granted by registration of a non-English word mark that might be commonly understood but be deemed sufficiently allusive to warrant registration. Would a finding that a trade mark was inherently adapted to distinguish grant the proprietor a monopoly that extends to translations that might have the same obvious meaning in other languages? While it may be more difficult to establish substantial identity, one could conceive of examples where multi-lingual equivalents could be deceptively similar.
Further, the focus to date has largely been on goods and services in relation to which the relevant persons concerned with the goods or services and the other traders includes a substantial proportion of the Australian population that speak English. It remains to be seen how the two-step test operates where the word in question is more squarely directed to a confined audience of non-English speakers who may be the relevant persons concerned with the goods for the purposes of the first limb in Cantarella. For example, presumably “dhahabu” (Swahili for “gold”) would be registrable for coffee, notwithstanding Kenyan coffee is considered some of the world’s best. What of the word “familia” (Swahili for “family”) when used on Chapati?
It is likely that the same questions will continue to arise in relation to trade marks composed of non-English words as presently arise in relation to trade marks composed of English words. For example, how does one establish what the words would mean to the relevant persons in Australia concerned with the relevant goods? How many of those persons needs to be shown to understand the words to have that ordinary signification? How does one go about establishing that other traders might legitimately need to use the words in respect of their goods?
A smorgasbord of options is set out above but each case will, of course, turn on its own facts. Much will depend on the nature of the language in question, the nature of the goods and the sign in issue. While the decisions outlined above have assisted in some ways to illustrate aspects of the practical difficulties in applying Cantarella to trade marks composed of non-English words, they also demonstrate that there are still a significant number of questions that remain unanswered.