SingaporeLaw.sg
Singapore - Your Partner for Legal Solutions in Asia
Novelty Pte Ltd v Amanresorts Ltd and Another
Novelty Pte Ltd v Amanresorts Ltd and Another
[2009] 3 SLR 216; [2009] SGCA 13

Suit No:    CA 56/2007
Decision Date:    31 Mar 2009
Court:    Court of Appeal
Coram:    Andrew Phang Boon Leong JA, Chan Sek Keong CJ, V K Rajah JA
Counsel:    Tan Tee Jim SC, Christopher de Souza and Lim Ke Xiu (Lee & Lee) for the appellant, Alban Kang, Koh Chia Ling and Ang Kai Hsiang (Alban Tay Mahtani & de Silva LLP) for the respondents

Subject Area / Catchwords   
Trade Marks and Trade Names
Tort


Judgment

31 March 2009

Judgment reserved.

V K Rajah JA (delivering the judgment of the court):

Introduction

1       Should a modest cluster housing project in Yio Chu Kang be allowed to share the same name as an ultra-exclusive luxury resort in Bali where that name is not a registered trade mark in Singapore? In the present appeal against the decision of the High Court judge (“the Judge”) in Amanresorts Limited v Novelty Pte Ltd [2008] 2 SLR 32 (“the Judgment”), we were asked to consider this question from two angles: first, at common law under the tort of passing off, and, second, under s 55 of the Trade Marks Act (Cap 332, 2005 Rev Ed) (“the current TMA”), which provides for the protection of well-known trade marks. To facilitate the reading of this judgment, we set out below the outline of its contents:

Introduction...............................................................................................................................

221

The facts....................................................................................................................................

223

The respondents...............................................................................................................

223

The Respondents’ resort business............................................................................

223

The Respondents’ residential accommodation business.............................................

225

Other products and services offered by the Respondents..........................................

226

The Respondents’ marketing-cum-publicity efforts and the exposure received...........

226

The Respondents’ trade marks................................................................................

228

The appellant....................................................................................................................

228

The dispute between the parties.........................................................................................

230

The decision in the court below....................................................................................................

232

The passing off claim....................................................................................................................

232

Gooodwill..........................................................................................................................

234

The nature of goodwill...............................................................................................

234

Whether goodwill exists in the “Aman” names............................................................

235

(1) The specific sector of the public to be considered.......................................

235

(2) Whether the “Aman” names have an attractive force for custom among
the relevant sector of the public........................................................................

242

(3) The type of business in respect of which goodwill in the “Aman” names
exists...............................................................................................................

244

(4) Our conclusion on the extent of the goodwill attached to the “Aman” names.

244

The “foreign business problem”...................................................................................

244

Misrepresentation.................................................................................................................

245

The misrepresentation in the present case.....................................................................

245

(1) The content of the misrepresentation.............................................................

245

(2) The target audience of the misrepresentation.................................................

246

(3) Whether the misrepresentation resulted in confusion......................................

248

Damage................................................................................................................................

252

Analysis of the heads of damage claimed by the Respondents.......................................

253

(1) Inferiority of the Appellant’s residential accommodation................................

253

(2) Likelihood of damage should the Appellant get into financial, legal or other
trouble..............................................................................................................

255

(3) Loss of licensing opportunity or licensing income, and misappropriation of
goodwill and reputation......................................................................................

257

(4) Restriction on the Respondents’ expansion into the residential real estate
business.............................................................................................................

260

(5) Loss of exclusivity and erosion of the distinctiveness of the “Aman” names
as well as dilution of the goodwill attached to these names....................................

262

Our conclusion on the passing off claim...................................................................................

266

The claim under section 55 of the current TMA: Infringement of a well-known trade mark.................

266

What is a “well known trade mark”?.......................................................................................

267

The protection afforded by section 55(3)(a) of the current TMA.............................................

275

The meaning of “use … in the course of trade … of any trade mark” in section 55(3).....

275

The meaning of “would indicate a connection between those goods or services and the 
proprietor” in section 55(3)(a)......................................................................................


276

(1) International developments............................................................................

278

(2) The US law of dilution...................................................................................

282

(3) The EU Directive...........................................................................................

286

(4) The Joint Recommendation............................................................................

291

(5) The current TMA..........................................................................................

296

(6) The position in Australia..................................................................................

303

(7) Our view on whether confusion is required under section 55(3)(a) of the
current TMA........................................................................................................

305

Our decision on the Respondents’ claim for trade mark infringement..........................................

305

Conclusion.........................................................................................................................................

306

The facts

The respondents

2       The first respondent, Amanresorts Ltd (“the First Respondent”), was incorporated in Hong Kong in 1981 while the second respondent, Amanresorts International Pte Ltd (“the Second Respondent”), was incorporated in Singapore in 1999. The First Respondent and the Second Respondent (referred to collectively as “the Respondents”) operate under the umbrella of the Amanresorts group of companies (“the Amanresorts Group”), which comprises over 80 companies, all of which are either wholly or partially-owned subsidiaries of Silverlink Holdings Limited, a company incorporated in the British Virgin Islands. The Amanresorts Group was founded in the mid-1980s by Adrian Zecha, a Javanese-born Dutch national and businessman.

The Respondents’ resort business

3       According to the Respondents’ main witness, Mr Gregory Sirois (“Mr Sirois”), a director of the Respondents:

The Amanresorts Group’s vision was the creation of a string of exclusive, ultra-luxury [sic] boutique resorts situated at beautiful locations around the world, with the emphasis on providing impeccable service and privacy.

4       The first of the Respondents’ resorts was Amanpuri, which opened in Phuket, Thailand, in 1988. Since then, the Respondents have developed a string of 15 other resorts across the world, all with the prefix “Aman” in their names, namely: Amandari, Amankila, Amanusa, Amanwana, Amanpulo, Amanjiwo, Amangani, Amanjena, Amansara, Aman-i-Khás, Amankora, Amanbagh, Amangalla, Amanwella and Amanyara. The Respondents’ resorts (which we shall also refer to as “Aman resorts”) can be found in many different countries such as India, the US, Morocco, Cambodia, Sri Lanka and the Philippines. There are even resorts in more exotic locations such as the misty Himalayan kingdom of Bhutan and the Turks and Caicos Islands in the sunny Caribbean. Indonesia plays host to five Aman resorts (one on Moyo Island, one at Borobudur and three in Bali). The Respondents also own two hotels in France and French Polynesia named Le Mélézin and Hotel Bora Bora respectively.

5       It is undisputed that quality and exclusivity are the signatures of the Aman resorts. This is evinced by the numerous international awards showered upon the resorts as well as the high rankings which they have consistently garnered in prestigious guides such as The Gallivanter’s Guide (published in the UK) and Zagat (published in the US). One magazine gushed that the Aman resorts were “the most exclusive group of intimate resorts and hotels in the world” while an Australian newspaper proclaimed the resorts to be “the cocoon of luxury”. Despite the growing number of Aman resorts, the Respondents have chosen to keep the supply of rooms at these resorts very small. There are only 626 rooms worldwide between the 18 establishments run by the Respondents, not many more than the number of rooms usually found in a single large five-star hotel.

6       As befitting their high standards, the Aman resorts lay claim to be the playgrounds of the rich and famous – albeit with quiet understatement and elegant restraint. The following extracts from various reviews raving about the resorts illustrate this point:

a.    ‘It costs a bomb to stay there, but the rich and powerful seeking peace and privacy and people who want to treat themselves to a few days of romantic heaven are jetting in from all parts of the globe …’ (The Sydney Morning Herald, Australia, 1992).

c.    ‘… Targeting the top-end consumer who has more money than time, [the Aman resorts] reign supreme in their particular niche market. Most Aman aficionados wouldn’t even consider another brand.’ (Asian Style Hotels, Singapore, 2005).

d.    ‘Aman Resorts, those purveyors of super-exclusive, superb-luxury hotels, have opened two resorts in Sri Lanka … The understated, very perfect luxuries of an Aman resort are usually the preserve of a very wealthy few …’ … (Outlook Traveller, India, 2005).

e.    ‘… [The Aman resorts] are aimed at a more discerning class of [travellers] (never a ‘tourist’) who are cash-rich, style-conscious and wouldn’t be caught dead on a standard two-week package holiday to Thailand or Egypt’ (Serendipity, UK, 2005).

[emphasis in original omitted]

People from this “rarefied world of the luxury retreat”, as one UK newspaper enthused, often become repeat visitors, earning the affectionate soubriquet, “Amanjunkies”. According to another magazine, the Respondents can boast of “more than 130,000 self-proclaimed ‘Amanjunkies’”. Such devotees are not deterred by the stratospheric room rates charged at the Respondents’ resorts (a night’s stay at Amanyara in the Turks and Caicos Islands, for instance, can cost as much as US$10,900, excluding a 10% service charge and 10% government tax).

7       One particular Aman resort is of especial interest to us. This is Amanusa, which opened in Bali in 1992 (“the Amanusa Bali”). It is the latest of the Respondents’ three resorts on that island. Room rates at the Amanusa Bali range between US$700 and US$1,500 per night, excluding taxes. The word “aman” means “peace” or “peaceful” in Bahasa Indonesia, and the word “nusa” means “island” or “isle”. The phrase “peaceful island” in Bahasa Indonesia is “nusa aman”, but the Respondents deliberately created an ungrammatical portmanteau word, “Amanusa”, by inverting the order of the two component words (viz, by placing “aman” before “nusa”), combining the two words and dropping one of the ‘n’s.

The Respondents’ residential accommodation business

8       Apart from the resort business, the Respondents are also engaged in the residential accommodation business. They claim that they have received many requests from real estate developers for licences to use the word “Aman” as well as other words bearing the prefix “Aman” for real estate projects in return for a “branding fee”. Where the Respondents agree to grant such a licence, they typically charge a fee of 10% of the selling price of a unit in the project concerned. As part of the agreement, the Respondents also assume responsibility for the design of the development as well as for its management upon completion of construction. Such “branding agreements” have been entered into with respect to various residential projects in the Turks and Caicos Islands, Arizona, New York, Mexico and Greece. The development in Arizona comprises 31 residential units, while the one in New York consists of a hotel-cum-condominium complex. It is also claimed that the Respondents will be developing residential projects, to be owned exclusively by them, in New Delhi, Croatia and Turkey, all bearing the prefix “Aman”.

9       In addition, the Respondents develop and market long-term residential accommodation known as “Aman Villas”. The Respondents’ first real estate development of this nature consisted of 31 Aman Villas at Amanpuri sometime around 1990. Since then, the Respondents have, together with various partners, developed and marketed Aman Villas at Amanjena, Amangani, Amanpulo, Amanyara and Amandari as well. According to Mr Sirois:

These ‘AMAN Villas’ are managed by the respective AMAN resort[s] and the villa owners have the option of placing the villas in a rental pool so that they can be rented to hotel guests on an overflow basis or as complete villas. …

The ‘AMAN Villas’ … are referred to by appending the word ‘Villas’ or ‘Homes’ to the name of the respective adjacent AMAN resort, such as AMANPURI Villas, AMANJENA Villas, AMANGANI Homes, AMANPULO Villas and AMANDARI Villas, and the association with the [Respondents] and their … resorts and [accommodation] is strongly promoted.

The Respondents claim that they may in future proceed to develop Aman Villas adjacent to the Amanusa Bali as their joint venture partner for the resort has acquired the adjacent land specifically for the development of such villas.

Other products and services offered by the Respondents

10     The Respondents have associated luxury goods and services with the name “Aman”. Mr Sirois stated that:

… AMANPURI … has an accompanying ‘AMAN Spa’, while certain of the … AMAN resorts have ‘AMAN Villas’ as accommodation options. Further, the [Respondents] also offer a collection of high-value experiences for their guests called ‘AMAN Experiences’ that include, among others, ‘AMAN Cruises’, ‘AMAN Spa Season Retreat’, ‘AMANGANI in Winter’, ‘AMAN Weddings’, ‘AMAN Bali Experience’, ‘AMAN Bali Break’, ‘AMANWANA Komodo Expedition’, ‘AMANJIWO Cultural Trails’ … and ‘AMANPULO Underwater Experience’. From or about 2004 or thereabouts, the [Respondents] have also been selling various beauty spa/ wellness products under the AMAN name.

11     Mr Sirois also pointed out that:

[T]he world-renowned car manufacturer, Toyota, has approached the [Respondents] for a licence to use an ‘AMAN’-prefixed name for one of the models of [its] world-renowned Lexus luxury vehicles.

The Respondents’ marketing-cum-publicity efforts and the exposure received

12     The Respondents’ business is a successful one. Worldwide sales from the goods and services offered by the Aman resorts as well as from other accommodation related to the Respondents have risen from US$39,480,377 in 1995 to US$86,179,826 in 2006.

13     There are no Aman resorts or residential accommodation bearing the name “Aman” in Singapore. However, the Amanresorts Group shifted its corporate headquarters from Hong Kong to Singapore in 1999 with the setting up of the Second Respondent. Since 2001, the Respondents’ international corporate office and international reservations office have been located in Singapore. The latter handles approximately 30% to 40% of global reservations and inquiries which are not directed specifically to individual Aman resorts. In 2006, revenue from the Respondents’ international reservations office amounted to US$31,000,000 derived from room reservations, food and beverage services as well as spa services.

14     Between 1995 and 2005, more than 300,000 Singaporeans visited Bali, where the Amanusa Bali and two other Aman resorts are located. Of these Singaporeans, some 1,382 visited the Amanusa Bali between 1995 and 2005. During the same period, some travel agents in Singapore (as opposed to the Respondents themselves) also processed reservations for the various Aman resorts.

15     In 2006, the Respondents spent US$3,399,679 on promoting and marketing their goods and services globally. Over the years, they have also promoted their resorts regularly to travel agents and media personnel worldwide, participated in trade shows, sponsored prizes in charitable events as well as collaborated with airlines and credit card companies to advertise the resorts to individuals with high net worth.

16     The Respondents have been distributing newsletters at their resorts, including the Amanusa Bali, since 1994. These newsletters are distributed to actual customers and potential customers of the resorts who have signed up to be on the Respondents’ mailing list. One such newsletter, Amanews, is sent to over 122,000 recipients worldwide (including over 2,700 in Singapore) every quarter. The same group of recipients also receive via e-mail a newsletter, Aman Express, every two to three weeks. In addition, the Respondents send a newsletter, Aman Advisory, to travel agents, including some in Singapore, at the same interval (ie, every two to three weeks).

17     Brochures for the Respondents’ resorts are available in Singapore at the Respondents’ international corporate office and international reservations office. These brochures are also distributed to travel agents at trade shows and to guests at business or charitable functions.

18     One particular marketing strategy which the Respondents have employed is the targeted advertisement of their goods and services to two groups of American Express card holders – namely, holders of American Express’s “Platinum” card (“American Express Platinum members”) and holders of American Express’s “Centurion” card (“American Express Centurion members”) – via the “Fine Hotels and Resorts Programme” since 1988. A media pack from American Express describes such card holders in the following terms:

Platinum Cardmembers and Centurion members in Asia are super-affluent and high-net-worth individuals and are American Express’s highest spenders. They spend significantly more than other Cardmembers and they have real spending power, as their card must be paid in full every month. Centurion members spend 20–25 times more than Gold Cardmembers; Platinum Cardmembers spend 8–10 times more.

There are approximately 900 American Express Centurion members and 4,200 American Express Platinum members in Singapore. The Respondents have similar marketing arrangements with The Hongkong and Shanghai Banking Corporation for holders of its “Platinum” card.

19     As mentioned earlier (at [5][6] above), the Respondents’ resorts have been featured in various newspapers and periodicals around the world. These include international newspapers and periodicals such as The New York Times, Far Eastern Economic Review, Financial Times, Condé Nast Traveller, The Sydney Morning Herald, etc. Some local publications also contain references to the resorts: The Straits Times, Her World and Female, for instance.

20     Finally, the Respondents maintain a stable of websites under the name “Aman” which are hosted in Singapore (“the Respondents’ ‘Aman’ websites”). The website, , has been registered since 1996. Other websites are based on the names of the respective Aman resorts such as the website, , for the Amanusa Bali. According to the Respondents, in respect of searches for the word “Amanusa” done via the search engines, Google and Yahoo!, 28 out of the top 30 search results and all of the top 30 search results respectively contain references to this particular resort.

The Respondents’ trade marks

21     The Respondents own “Aman”-prefixed trade marks in various countries around the world for the names “Aman” and “Amanresorts” as well as for the individual names of all the resorts and hotels which they operate (see [4] above for the full list of these resorts and hotels). However, they have not registered as trade marks, in each of these countries, the full panoply of these “Aman”-based names. (The expression “the ‘Aman’ names” will hereafter be used to delineate, specifically, the name “Aman” as well as the names of the various resorts and hotels which the Respondents operate, while the term “‘Aman’-prefixed names” will be used to denote all other names which contain the prefix “Aman” but which fall outside the ambit of the “Aman” names.) In Singapore, only the names “Amanresorts”, “Amangani”, “Amanjena” and “Amanjiwo” are registered as trade marks. Notably, the word “Aman” is not, and has never been, registered as a trade mark in this country. The name “Amanusa” was registered as a trade mark in Singapore in 1998 under Class 3 (“cosmetics”), Class 16 (which includes “paper”, “goods made from [paper]” and “printed matter”), Class 25 (“clothing”) and Class 43 (“hotels”) of the International Classification of Goods and Services, but the registration was allowed to expire on 12 November 2001. According to the Respondents, this decision was made because of a shift in their commercial strategy and for reasons of cost-effectiveness.

The appellant

22     The appellant, Novelty Pte Ltd (“the Appellant”), was incorporated on 3 May 2005. It is a subsidiary of Novelty Dept Store Pte Ltd, which was incorporated on 27 September 1995. The two are part of the larger Novelty group of companies. The group is involved in the development of residential properties in Singapore, in particular, condominiums targeted at Singapore citizens and residents wishing to own private housing or to upgrade from public housing. According to the Appellant, its developments are successful because of their reasonable prices, their convenient locations as well as the quality finishes and furnishings which the Appellant claims it is well known for. The names of the Appellant’s developments often have a common thematic flavour: eg, the “Taipan” theme was used for the condominiums, The Taipan, Taipan Jade and Taipan Regency; and a “metal” theme was used for the developments, Platinum Edge, Iridium and Vanadium.

23     In July 2005, the Appellant completed the purchase of a 56,355 sq ft freehold site off Old Yio Chu Kang Road (“the Site”). The Site, a vacant plot of land surrounded by lush greenery, is situated in a quiet residential neighbourhood where established private housing estates such as Teachers’ Estate and Sembawang Hills Estate are located. After the purchase, the Appellant decided with its architect, JGP Architecture (S) Pte Ltd (“JGP”), to build a cluster housing project (“the Project”) comprising 36 three-storey terraced houses, each with a basement car park and a roof garden, at the Site. Various water-based features such as a lap pool and landscaped ponds were to be added to create a peaceful and soothing atmosphere. The Project would also feature modern and high-quality amenities and fittings. The Appellant decided to name the project “Amanusa”.

24     In the affidavit of evidence-in-chief of Mr Manoj Dharmadas Kalwani (a director of the Appellant) affirmed on 5 March 2007, it was claimed (at para 8) that “JGP suggested the name ‘Amanusa’ (meaning ‘peaceful island’ in Malay) for the [P]roject sometime in August 2005”. At that time, the Appellant did not know of the Respondents’ resorts or the Amanusa Bali. During cross-examination, Mr Chan Sze Chin (“Mr Chan”), the concept architect from JGP, testified that when he and his team were brainstorming for a name for the Project, they sought a name which would be evocative of the peaceful surroundings of the Site. They considered the words “aman” and “sentosa”, which mean “peaceful” in Malay and Bahasa Indonesia respectively, but rejected the latter because of its association with Sentosa Island. The word “nusantara”, meaning “archipelago”, was also considered, but it was perceived as being too long and was thus shortened to “nusa” (meaning “island”) to convey the image of “an oasis and sanctuary” (see the Judgment at [37]). The phrase “aman nusa” was chosen instead of “nusa aman” because the former “rolled off the tongue more easily” (id at [38]). It was also decided that the two words should be combined and one of the ‘n’s dropped for ease of pronunciation. While the brainstorming team had not heard of the Amanusa Bali, Mr Chan admitted that he himself had heard of Amanpuri (the Respondents’ resort in Phuket); he had not, however, foreseen any difficulty with using the name “Amanusa” for the Project as Amanpuri was, in his view, “completely different in many aspects from the [Project]” (ibid). Mr Chan also claimed that he had not checked the name “Amanusa” on the Internet.

25     On 8 September 2005, the Street and Building Names Board (“the Board”) approved the Appellant’s application for the name “Amanusa” to be used for the Project. This purportedly reinforced the Appellant’s view that the name was appropriate and acceptable because the Board would not have approved a proposed name that was identical or substantially similar to the name of an existing development, whether in Singapore or overseas. For instance, for another project, the Appellant had proposed the name “Twin Waters”, but that name had been rejected because there was an existing resort in Australia known as “Twin Waters”.

26     The marketing efforts for the Project began in January 2006 via newspapers (eg, The Straits Times), Internet advertisements as well as mailers distributed in Housing Development Board estates at, inter alia, Ang Mo Kio, Hougang and Yishun. A showflat (“the Showflat”) was also constructed to promote the Project. The Project was advertised as a “Balinese-inspired” retreat that included a “water-themed backdrop”. The Appellant’s brochures and advertisements also described the Project as an exclusive “private hideaway” with luxury fittings.

The dispute between the parties

27     The Respondents first became aware of the Project on 11 February 2006 after coming across advertisements placed in The Straits Times. A letter of demand was sent by the First Respondent’s then solicitors to the Appellant on 14 February 2006 demanding that the latter change the name used for the Project. The Appellant’s then solicitors replied by way of a letter dated 22 February 2006. In that letter, it was stated that the name “Amanusa”, a combination of Malay words meaning “peaceful motherland”, was appropriate for the Project given the quiet surroundings at the Site. The letter also stated that the Appellant had not had any knowledge of the Respondents or their resorts until the receipt of the letter of demand of 14 February 2006. This explanation (viz, that the word “Amanusa” meant “peaceful motherland” in Malay) was repeated in a letter dated 6 March 2006 from the Appellant’s then solicitors to the First Respondent’s then solicitors.

28     Legal proceedings against the Appellant were commenced on 3 May 2006. The Respondents asserted that they were the owners of the reputation and goodwill attached, both in Singapore and worldwide, to the name “Amanusa” as well as to the rest of the “Aman” names, which were used for (inter alia) hotels, resorts and residential accommodation that were “high-end, top quality, exclusive” (see the Judgment at [17]). It was alleged that the Appellant, by using the name “Amanusa” for the Project, had passed off accommodation which was not the Respondents’, or not associated or connected with the Respondents, as the Respondents’ accommodation. Furthermore, the Respondents claimed, the Appellant had made such use of the name “Amanusa” with the intention of confusing or deceiving members of the public into thinking that the Project was provided by the Respondents or, alternatively, associated or connected with them.

29     The Respondents later added a claim under s 55 of the current TMA. They asserted that the “Aman” names were well known in Singapore to one or more of the relevant sectors of the public, namely, actual consumers and potential consumers of goods and services provided under the “Aman” names as well as members of the hospitality, travel and/or property development industries. It was argued, relying on s 55(2), s 55(3)(a) and/or s 55(4)(a) of the current TMA, that the Appellant’s use of the name “Amanusa” for the Project was “likely to cause confusion” (see s 55(2)) or, in the alternative, would indicate a connection between the Project and the Respondents and would be likely to damage the Respondents’ interests (see s 55(3)(a)). (Eventually, the Respondents came to rely on only s 55(3)(a) of the current TMA.)

30     As a result of the First Respondent’s letter of demand, the Appellant issued questionnaires to actual and potential purchasers of units in the Project who visited the Showflat from July 2006 onwards. Some of the questions included:

Q5:      What were your reasons/factors for choosing the Amanusa development?

Q7:      Would you have chosen the Amanusa development if it were called by some other name? …

Q8:       Does the name/word ‘Amanusa’ mean anything to you? …

31     Fifty-one responses were collected by the Appellant. It appeared from these responses that the location and the design of the Project as well as the amenities which it featured were the main reasons why those who responded to the Appellant’s survey (“the survey participants”) were attracted by the Project; none of them cared if the Project had been called anything else. The name “Amanusa” generally did not mean anything to the survey participants, although some associated it with “peaceful” or “serenity”. Two responses in particular are of note, viz, that the name “sound [sic] like a Thai-spa kind of place”, and that the name was “very Balinese”, evocative of a “top of [the end] – Bali high end resort”.

32     Subsequently, the Appellant, probably out of caution, apparently told its sales agents not to link the Project with the Amanusa Bali. Further, all purchasers of units in the Project had to sign a declaration form acknowledging and agreeing that their request for an option to purchase and their subsequent exercise of such option did not confer any right in respect of the name “Amanusa”, and that that name did not constitute a consideration influencing their decision to exercise the option to purchase. Sales of units in the Project were initially not very brisk. By 5 March 2007, more than a year after marketing began, only 11 of the 36 units had been sold, unlike some of the Appellant’s other projects which had been sold out within six to eight months of their launch. However, by the time of the trial, only one unit in the Project remained unsold.

33     In the meantime, the Respondents engaged the services of two private investigators, who visited the Showflat in late March 2007. One of the investigators testified that a sales agent (“Jerry”) said that the name “Amanusa” was “like [the name of] a resort, there is a resort called Amanusa. It is a very famous resort …” (see the Judgment at [21]). Another sales agent (“Jason”) reportedly said that the Project had been named after a very well-known holiday resort, but was unable to say where that famous resort was located (id at [22]). Both Jerry and Jason gave evidence at the trial. Jerry denied making any reference to the Amanusa Bali and claimed that he had only referred to “resort living” (id at [42]), which was the Project’s selling point, while Jason said that he had only remarked that the Project was named after a resort, which he had thought was in some Western country (id at [43]).

34     By 30 March 2007, the Appellant had substantially changed the contents of its brochures advertising the Project. Some pictures of a thatched roof and a waterspout in the shape of a mythical creature were removed. The word “Balinese” was also omitted. This was the state of affairs when the matter went to trial in April 2007.

The decision in the court below

35     The Judge held that the Respondents’ passing off claim was made out. However, he declined to find that the Appellant had any fraudulent intention in using the name “Amanusa” (see the Judgment at [4] and [75]). He also held that the claim under s 55(3)(a) of the current TMA had been made out (id at [75]). In the circumstances, the Judge granted a declaration that the “Aman” names were well-known trade marks in Singapore and an injunction restraining the Appellant or anyone acting on its behalf from using those names (in particular, “Amanusa”) in relation to any form of accommodation (ibid). He expressly stated that the injunction was “confined to the [‘Aman’ names] and [did] not extend … to all words which include[d] the prefix ‘Aman’” (ibid). The Appellant was also ordered to stop distributing any material which described the Project as “Amanusa” (see the Judgment at [76]). Finally, concluding that there was only a likelihood of prospective damage rather than immediate damage to the Respondents, the Judge held that there was no need to order an inquiry as to damages or an account of profits (id at [4] and [77]).

The passing off claim

36     Two pronouncements of the House of Lords are generally considered to be the definitive statements on the requirements of an action for passing off. The first is from Lord Diplock’s speech in Erven Warnink Besloten Vennootschap v Townend & Sons (Hull) Ltd [1979] AC 731 (“the Advocaat case”) at 742, as follows:

A.GSpalding & Bros. v. A.W. Gamage Ltd. [(1915) 32 RPC 373] … and the later cases make it possible to identify five characteristics which must be present in order to create a valid cause of action for passing off: (1) a misrepresentation (2) made by a trader in the course of trade, (3) to prospective customers of his or ultimate consumers of goods or services supplied by him, (4) which is calculated to injure the business or goodwill of another trader (in the sense that this is a reasonably foreseeable consequence) and (5) which causes actual damage to a business or goodwill of the trader by whom the action is brought or (in a quia timet action) will probably do so.

The second is a reformulation by Lord Oliver of Aylmerton in Reckitt & Colman Products Ltd v Borden Inc [1990] 1 WLR 491 (“the Jif Lemon case”) at 499, as follows (this particular passage was also cited by the Judge at [48] of the Judgment as well as by this court in CDL Hotels International Ltd v Pontiac Marina Pte Ltd [1998] 2 SLR 550 (“the Millenia case”) at [86]):

The law of passing off can be summarised in one short general proposition – no man may pass off his goods as those of another. More specifically, it may be expressed in terms of the elements which the plaintiff in such an action has to prove in order to succeed. These are three in number. First, he must establish a goodwill or reputation attached to the goods or services which he supplies in the mind of the purchasing public by association with the identifying ‘get-up’ (whether it consists simply of a brand name or a trade description, or the individual features of labelling or packaging) under which his particular goods or services are offered to the public, such that the get-up is recognised by the public as distinctive specifically of the plaintiff’s goods or services. Secondly, he must demonstrate a misrepresentation by the defendant to the public (whether or not intentional) leading or likely to lead the public to believe that goods or services offered by [the defendant] are the goods or services of the plaintiff. Whether the public is aware of the plaintiff’s identity as the manufacturer or supplier of the goods or services is immaterial, as long as [the goods or services] are identified with a particular source which is in fact the plaintiff. For example, if the public is accustomed to rely upon a particular brand name in purchasing goods of a particular description, it matters not at all that there is little or no public awareness of the identity of the proprietor of the brand name. Thirdly, he must demonstrate that he suffers or, in a quia timet action, that he is likely to suffer damage by reason of the erroneous belief engendered by the defendant’s misrepresentation that the source of the defendant’s goods or services is the same as the source of those offered by the plaintiff.

37     The above dicta unequivocally articulate three clearly identifiable core concepts which are fundamental to a passing off action, namely, goodwill, misrepresentation and damage. Since the decision of the English Court of Appeal in Consorzio del Prosciutto di Parma v Marks & Spencer Plc [1991] RPC 351, these three elements have been known as the “classical trinity” (at 368, per Nourse LJ) of an action for passing off. We now proceed to examine whether each of these elements has been satisfied by the Respondents in the present case.

Goodwill

38     Passing off is a tort which protects goodwill and nothing else. Lord Diplock made this clear in Star Industrial Company Limited v Yap Kwee Kor [1976] FSR 256, where he stated (at 269):

A passing-off action is a remedy for the invasion of a right of property not in the mark, name or get-up improperly used, but in the business or goodwill likely to be injured by the misrepresentation made by passing-off one person’s goods as the goods of another.

As goodwill is the bedrock of this tort, it is unsurprising that one must properly define in each case what the plaintiff’s goodwill consists of in order to determine the boundaries of the protection which the plaintiff can claim. To do this, one must first have a proper understanding of what goodwill is.

The nature of goodwill

39     Like that other great force of attraction which we call “love”, “goodwill” is ephemeral and hard to define. To date, Lord Macnaghten’s speech in The Commissioners of Inland Revenue v Muller & Co’s Margarine, Limited [1901] AC 217 (“IRC v Muller & Co”) at 223–224 remains, in our view, the clearest exposition of what goodwill is:

What is goodwill? It is a thing very easy to describe, very difficult to define. It is the benefit and advantage of the good name, reputation, and connection of a business. It is the attractive force which brings in custom. It is the one thing which distinguishes an old-established business from a new business at its first start. The goodwill of a business must emanate from a particular centre or source. However widely extended or diffused its influence may be, goodwill is worth nothing unless it has a power of attraction sufficient to bring customers home to the source from which it emanates.

The above passage highlights two essential features of goodwill. First, it is the association of a good, service or business on which the plaintiff’s mark, name, labelling, etc (referred to generically as the plaintiff’s “get-up”) has been applied with a particular source. Second, this association is an “attractive force which brings in custom” (id at 224).

Whether goodwill exists in the “Aman” names

40     As a preliminary point, we note that it is not contested that the “Aman” names can be considered as distinctive indicia capable of being the vessels of goodwill. The evidence in respect of the name “Amanusa” shows that it is a distinctive and completely made-up name, even though its component parts, “aman” and “nusa”, have descriptive elements. We thus agree with the Judge’s finding that the “Aman” names are “fancy mark[s] when used in relation to hotels and resorts” (see the Judgment at [49]), as well as when used for projects that are associated with such developments. Needless to say, our views on this do not extend to the use of the “Aman” names in any other context.

41     The Judge found that the “Aman” names had “goodwill in Singapore” (id at [50]) before 2006, when the allegedly tortious conduct by the Appellant began. With respect, this is a somewhat ambiguous finding because it does not specify whether the goodwill existed in the minds of the general public in Singapore, or whether it was operative on only a small section of that public. In other words, the Judge did not identify who would make the requisite association of the “Aman” names with the ultra-luxurious hotels and resorts and other high-end residential accommodation offered by the Respondents. In our view, it is important to identify the specific section of the public which has goodwill towards the “Aman” names.

(1) The specific sector of the public to be considered

42     According to the Appellant, the “relevant sector of the public in Singapore” in this case ought to be “the actual [customers] and potential customers in the hotel, resort and travel [industries]” – a very wide group which covers more or less the entire public in Singapore. It was argued that the Judge ought to have considered whether this widely-defined group of consumers would have associated the “Aman” names with a specific source. The Appellant submitted that this group would not have made that association and, thus, the Respondents did not have any goodwill in Singapore at all vis-à-vis the “Aman” names.

43     In their submissions in the court below, the Respondents claimed that they had substantial reputation and goodwill, both in Singapore and worldwide, in the “Aman” names and “Aman”-prefixed names when those names were used in relation to hotels, resorts and residential accommodation. However, in their submissions in this appeal, the Respondents appear to have clarified that they are claiming goodwill only in respect of “the high-end hotel and resort market” [emphasis added]. It is argued that “the relevant purchasing public” consists of actual purchasers and prospective purchasers of the Respondents’ goods and services specifically, and not the wider group of consumers in the hotel, resort and travel industries.

44     In our view, goodwill is not an all-or-nothing attribute in that it is not the case that the plaintiff either has goodwill in Singapore or, otherwise, has no goodwill at all. Clearly, goodwill can be limited to particular sections of the public. These sections of the public can be small so long as they are not negligible. However, if goodwill is shown to exist only among a small section of the public in Singapore, it would mean that, while goodwill exists in this country, it exists only in relation to that small group and not to the entire public here at large.

45     In this case, we are of the view that goodwill exists in the “Aman” names among actual customers and potential customers in Singapore of the Aman resorts. As we shall explain more fully, this group is quite small, but, nonetheless, goodwill exists among it. We should also point out that the goodwill enjoyed by the Respondents covers the “Aman” names only and does not extend to “Aman”-prefixed names (as defined at [21] above).

46     In the instant case, the Judge noted that “the number of people patronising Aman resorts or who [were] familiar with [the Respondents’] goods and services [was] … relatively small” (see the Judgment at [57]). However, he also said that “a significant body of persons in Singapore” (id at [59]) had heard of the Aman resorts generally and the Amanusa Bali in particular.

47     The association of the plaintiff’s goods, services or business with a particular source may operate on only a small section of the public because of limited exposure of the plaintiff’s get-up in relation to those goods, services or business. Obviously, such an association can only form in the minds of those consumers who have been exposed to the plaintiff’s get-up. For instance, a small local business which is known only to the residents in its immediate vicinity will enjoy only geographically limited goodwill. In The Clock, Ld v The Clock House Hotel, Ld (1936) 53 RPC 269, for example, the plaintiff, which operated a small road-house (known locally as “The Clock”) offering its guests meals and facilities for bathing, tennis and golf, succeeded in obtaining a permanent injunction to restrain the defendant, which operated a hotel five miles away, from carrying on business under the name “Hotel Clock House”; this restriction, however, applied to the defendant’s premises only. Obviously, the goodwill which the plaintiff enjoyed in the name “The Clock” did not extend very far.

48     One of the more striking features about the present appeal is that the main part of the Respondents’ business is not based in Singapore. The Respondents’ resorts, including the Amanusa Bali, are located overseas. The Respondents’ physical presence in Singapore is limited to that established by their international corporate office and their international reservations office. This would suggest a much more limited exposure in Singapore of the “Aman” names than if there were an Aman resort located right in this country.

49     Furthermore, the Respondents’ resorts, apart from all being overseas, are located in exotic or relatively inaccessible locations not often frequented by mainstream tourists from Singapore. Indeed, the Respondents have taken great pains to keep their resorts low-key; understatement is a hallmark of their branding (see also [5][6] above). This can be seen from an article on Amanpuri entitled “Millionaires’ retreat”, published in South China Morning Post on 5 December 1994, which stated:

Perhaps one of the most unusual aspects of all about Amanpuri is that the resort does not even need to advertise.

Guests arriving at Amanpuri will vouch for its low profile stance. There is not even a sign at the reception declaring the name of the resort. And anyone venturing away from the retreat by hired jeep, for example, might find themselves at a loss on the return journey – for there is not even a sign post from the main road circling Phuket indicating the Amanpuri turn off.

Even the three Aman resorts located in Bali have been deliberately sited in out-of-the-way places in order to create a sense of exclusivity, seclusion and privacy. Indeed, in Tom Rao, “Honeymoon Home away from Home” Wedding & Travel (July 2001), an article on the Amanusa Bali adduced by the Respondents, the writer stated (id at 177):

If I were to drive, I never would have found the narrow road which turned into [the] Amanusa [Bali]. That’s how secretly-tucked away this luxury resort is, situated on a hilltop close to Bali’s southernmost point – Nusa Dua. [emphasis in original omitted]

50     Even though Bali welcomes many visitors from Singapore every year (see [14] above), the location of the three Aman resorts on that island means that most of these visitors are unlikely to come across those resorts in their peregrinations around the island. At the trial, there was no evidence of the Respondents’ advertising presence in Bali and the other countries where Aman resorts are located, but one might reasonably assume that any such advertising presence would be similarly limited.

51     Absent a prominent physical presence by the Respondents in Singapore and given the locations of the Respondents’ resorts, the only other means for the “Aman” names to gain exposure is through marketing and the media. Again, the evidence in this regard suggests that there has only been very limited, rather than extensive, exposure of these names thus far. The Respondents apparently have a policy of not advertising their resorts in the mainstream media. Instead, advertising is targeted at a select few: for instance, the small clutch of American Express Centurion members and American Express Platinum members in Singapore. There have been articles in various magazines on the Respondents’ resorts, but most of these periodicals are again targeted at rich individuals, both local and foreign, or other specialist consumers like those in the high-end travel industry. The Respondents highlighted three examples of mass media exposure in the form of articles in Female, Time and Her World, but this sort of publicity was sporadic at best and would have generated only trivial goodwill in the “Aman” names. This point is significant for, as Laddie J incisively noted in Sutherland v V2 Music Ltd [2002] EMLR 28 at [22]:

The law of passing off protects the goodwill of a small business as much as [the goodwill of] the large, but it will not intervene to protect the goodwill which any reasonable person would consider trivial.

52     With regard to the 20 domain names which have been registered for the Respondents’ “Aman” websites, no weight should be placed at all on the fact that those domain names were registered here – that is a purely technical matter which does not influence the extent of exposure of the “Aman” names. More importantly, it must be noted that although the Internet has the potential to be used for giving goods, services and businesses instant exposure across the entire world, it does not follow that any small, dinky business can set up a website and then claim that it has gained worldwide recognition. The Internet’s power of exposure is only a potential power. This issue was given some consideration in 800-FLOWERS Trade Mark [2000] FSR 697 (“the first instance decision in 800-FLOWERS”), which concerned a dispute over the registration of a service mark. In that case, an application to register a name as a service mark was objected to on, inter alia, the ground that the name had not been used in the UK. The applicant sought to counter this argument by relying on its use of that name for its website. Jacob J observed very sensibly (at 705):

Mr Hobbs [counsel for the applicant] submitted that any use of a trade mark on any website, wherever the owner of the site was, was potentially a trade mark infringement anywhere in the world because website use is in an omnipresent cyberspace: that placing a trade mark on a website was ‘putting a tentacle’ into the computer user’s premises. I questioned this with an example: a fishmonger in Bootle who puts his wares and prices on his own website, for instance, for local delivery can hardly be said to [be] trying to sell the fish to the whole world or even the whole country. And if any web surfer in some other country happens upon that website he will simply say ‘this is not for me’ and move on. For trade mark laws to intrude where a website owner is not intending to address the world but only a local clientele and where anyone seeing the site would so understand him would be absurd. So I think that the mere fact that websites can be accessed anywhere in the world does not mean, for trade mark purposes, that the law should regard them as being used everywhere in the world.

53     The requirement of “use” of a trade mark in a jurisdiction (in the context of trade mark law) is, of course, pegged at a higher standard than the standard which applies when determining whether goodwill exists in a name (in the latter context, exposure, as opposed to use, of the name in question is the relevant criterion). But, the principle that putting a name up on the Internet does not translate into instant global presence of that name is applicable vis-à-vis the analysis of goodwill in passing off actions.

54     It follows that evidence of the existence per se of domain names and websites featuring the name in which goodwill is said to exist will ordinarily be insufficient as proof of exposure of that name; something in addition to proof of the mere existence of such domain names and websites must be shown in order to establish that there has been sufficient exposure of the name in question. This approach is not dissimilar to that adopted by Buxton LJ in 800-FLOWERS Trade Mark [2002] FSR 12 (“the appellate court’s decision in 800-FLOWERS”), which was the English Court of Appeal’s decision on the appeal and the cross-appeal arising from the first instance decision in 800-FLOWERS. In the appellate court’s decision in 800-FLOWERS, Buxton LJ commented, in the context of use of a trade mark, that (at [138]):

[T]he very idea of ‘use’ within a certain area would seem to require some active step in that area on the part of the [trade mark owner] that goes beyond providing facilities that enable others to bring the mark into the area.

Such “active step[s]” (ibid), the learned judge suggested, could take the form of “direct encouragement or advertisement by the [trade mark] owner” (ibid) which led consumers to the website featuring the trade mark in question. In the present case, the Respondents must adduce evidence showing, among other matters, the following:

(a)     How many “hits” have the Respondents’ “Aman” websites received?

(b)     How many of these “hits” came from Internet users in Singapore?

(c)     If someone does a search for relevant topics (“resorts in Bali”, for instance) using any of the major search engines, are the Respondents’ “Aman” websites included among the top few search results?

(d)     How many other major websites (particularly those frequented by consumers in Singapore) are linked to the Respondents’ “Aman” websites?

In this regard, the only evidence of exposure is that the Respondents’ website, , has received “millions of ‘hits’” (see the Judgment at [11]) – but we do not know from where or from whom. Without such information, we would place little weight on the existence of the Respondents’ “Aman” websites in assessing the degree of exposure of the “Aman” names locally.

55     It is also unlikely that the fame of the Respondents’ resorts spread by word of mouth. As previously indicated (see [14] and [50] above), the number of visitors from Singapore who have patronised the Amanusa Bali is relatively small. Even if the total number of visitors from Singapore for all the Aman resorts were taken into account, this would not yield a very large figure because of the very limited number of rooms available at the various resorts; for instance, the Amanusa Bali itself has only 35 rooms. We acknowledge the Respondents’ point that many luxury brands are widely known among the general public even though the goods marketed under those brands are purchased by only a very limited number of consumers. However, such brands would usually have other means of exposure – whether through marketing or through the ubiquity of their names in popular culture, etc. In the present case, the “Aman” names did not have such exposure.

56     We are also aware of the impressive sales revenue derived by the Respondents from their resorts, more than 30% of which is attributable to the Respondents’ international reservations office in Singapore. However, there is no evidence as to what proportion of these sales was generated by customers based in Singapore. Sales generated by customers based overseas are entirely irrelevant for the purposes of establishing whether there has been exposure of the “Aman” names in this country and, if so, the extent of such exposure.

57     Finally, while a long-established business is, in general, likely to be more widely known than a recently-established one, this may not always hold true. The Respondents’ business has existed since 1988 when the first Aman resort, Amanpuri, was opened, but, apart from the two offices which the Respondents maintain in Singapore, the main part of their business is based overseas. The length of time for which the Respondents’ business has been operational is thus not of very great weight.

58     In our view, the exposure which the Respondents have succeeded in establishing is plainly limited to the well heeled. It is these people who visit the Respondents’ resorts; it is also this group which has been and continues to be the target of the Respondents’ selective marketing campaign. In contrast, exposure of the “Aman” names has not been established for other income groups in Singapore. Some indication of this can be found in the survey carried out by the Appellant among the visitors to the Showflat, the overwhelming majority of whom had not heard of the Amanusa Bali or any other Aman resort. However, we acknowledge that the survey was rather unscientific and do not place much weight on the results obtained. Further, the drafting of the questionnaires for the survey left a great deal to be desired. The use of surveys in trade mark disputes and passing off actions is a very technical matter and, if evidence based on surveys is to be relied on, it should be adduced in the form of expert evidence.

59     The uneven pattern of recognition of the “Aman” names among different income groups in the present case is no accident. It is the result of the deliberate policy of the Respondents to concentrate their marketing and advertising efforts on the wealthy. The image which the Respondents seek to project is one of understated class. A glimpse of this approach can be found in the following extract from an article in Time, which describes how the Aman resorts are designed and furnished:

Much of what makes [the Aman resorts] distinctive is how studiously they avoid most of the signifiers [that] the rest of the hospitality industry employs to trumpet their properties as ‘high class’. … The gift shops carry no logo-branded T shirts or beach towels. There are virtually no logos anywhere, except on matchbook covers and stationary [sic]. And there are certainly no single-serving plastic bottles of soap, shampoo and hair conditioner in the shower. [Adrian] Zecha literally shudders at the thought. ‘So tacky,’ he declares.

There is nothing objectionable about such an approach to marketing and business, but the Respondents must also be prepared for the consequence that whatever goodwill they may have in the “Aman” names may be limited in scope. The English High Court case of HFC Bank Plc v Midland Bank Plc  [2000] FSR 176 (“the HFC Bank case”) is instructive in this regard. The plaintiff in that case (“HFC”) was a bank which adopted the practice of dealing with its customers in indirect ways: for instance, through introductions from credit brokers, through retail finance agreements and through its credit card business conducted under other brand names. As a result, only a small proportion of its customers were familiar with the name “HFC”. When HFC tried to stop the defendant (“Midland Bank”) from using a similar name, HFC relied on the evidence of some of its customers, who claimed to have mistaken Midland Bank for HFC. Lloyd J held (at 198–199):

What this comes down to, in my judgment, is that in respect of none of these witnesses had HFC achieved brand recognition. This is not very surprising given the type of contact between each of [the witnesses] on the one hand and HFC on the other, the stage that it had got to, and also given HFC’s deliberate policy of not cultivating general awareness of its brand name among the public as a whole. …

I do not regard any of the six witnesses whom I treat as relevant for this purpose [ie, for the purpose of determining whether customers were confused between HFC and Midland Bank] as being people within the scope of HFC’s established goodwill at the time of [their] respective mistakes [in thinking that HFC and Midland Bank were the same].

(2) Whether the “Aman” names have an attractive force for custom among the relevant sector of the public

60     In the present case, even if the “Aman” names were generally known throughout the whole of Singapore, it would not necessarily mean that the Respondents have goodwill in those names to that extent. It must be recalled that goodwill, apart from being an association of a good, service or business with a particular source, must also be “[an] attractive force which brings in custom” [emphasis added] (see IRC v Muller & Co ([39] supra) at 224). In other words, whether or not there is goodwill attached to a name also depends on whether or not there are any actual and/or potential customers of the goods, services or business marketed under that name.

61     One example of a situation where an association with a source is not “[an] attractive force which brings in custom” (ibid) would be where a well-known good is unavailable for sale. This was the case in Anheuser-Busch Inc v Budejovicky Budvar NP [1984] FSR 413 (“the Budweiser case”), where the plaintiff manufactured the well-known “Budweiser” beer in the US. “Budweiser” beer was not available to the general market in the UK, although it was sold in American military bases and a limited number of other outlets in the UK on a duty-free basis. However, the beer was widely known throughout the UK because of the plaintiff’s publicity efforts as well as tourist and business traffic between Europe and the US. In view of this, the English Court of Appeal held (at 464, per Oliver LJ) that the plaintiff could “legitimately claim that … [it] had a reputation as the [brewer] of a beer, Budweiser, with a substantial section of the public”. The question for the court, then, was “whether this reputation associated with a beer which, for practical purposes, nobody could buy [in the UK], constituted … goodwill in any relevant sense” (ibid). Oliver LJ considered whether the plaintiff’s name could be said to be an attractive force for custom in the UK, and came to a negative conclusion. He reasoned (at 469 of the Budweiser case):

[O]ne asks oneself ‘what custom in this country in 1973 was brought in by the knowledge of members of the indigenous British public of the [plaintiff’s] Budweiser beer?’ And the answer must be that there was none, because however attractive [the British public] may have found the idea of drinking the [plaintiff’s] beer, they could not get it. In so far, therefore, as anyone was misled by the defendants’ use of the name ‘Budweiser,’ the [plaintiff] could suffer no damage either by loss of sales, for there were none at that time and none were contemplated, nor by loss of reputation, because if there was any such loss (which seems highly improbable) the reputation was quite unconnected with either an ability or a willingness to supply.

62     A desire to become a customer of the plaintiff, without the ability to actually be one, cannot ordinarily form the basis of goodwill. A person in such a position is not even a potential customer of the plaintiff because there is no way for the latter’s good, service or business to attract his custom. In such a case, the plaintiff can be described as having a good reputation among such persons (ie, persons who desire to be, but are unable to become, the plaintiff’s customers), but that is not the same as having goodwill. This court in the Millenia case ([36] supra) cited the Budweiser case with approval, noting that “having a widespread reputation in a particular place [did] not necessarily mean that the trader [had] … goodwill in his business” (see the Millenia case at [50]).

63     Another situation in which there may be a good reputation attached to a get-up without corresponding goodwill in that get-up is when the good or service marketed under the get-up is priced out of the reach of those who desire it. People who cannot afford a good or service cannot be considered to be part of the potential customer base of the plaintiff who provides that good or service. In the present case, the Respondents’ resorts cater for the wealthy and the room rates there reflect this. As noted at [65] of the Judgment, visitors to the Respondents’ resorts “pay a premium” to stay there. The Respondents’ high room charges were described by Mr Sirois as an “economic barrier”. The remarkable prices contribute to the overall image of exclusivity which the Respondents have chosen to imbue their business with. The Respondents’ resorts cannot even be put in the same class as other five-star hotels, which, although pricey, often still receive guests from middle-income groups. Even if the “Aman” names are well known throughout Singapore such that they have widespread reputation, the high room rates at the Aman resorts mean that these names carry no attractive force for custom among those who are less well off; such individuals simply cannot afford to stay at the Aman resorts. In the Budweiser case, the plaintiff’s goodwill was limited by the restricted availability of “Budweiser” beer in the UK. In the instant case, the limitation stems not from the availability of the product in question but from its price. However, both types of limitation (ie, limitation due to availability and limitation due to price) effectively mean that the sphere of actual customers and potential customers of the product in question is limited.

64     We should, however, point out that we accept that there are many people who, although unable to go to the Aman resorts on a regular basis, may nonetheless save up for a once-in-a-lifetime visit, such as for a honeymoon or a golden wedding anniversary, etc. Exposure of the Aman resorts in magazines like Wedding & Travel (see [49] above) illustrates that there may be goodwill in relation to this group of persons even though they may not fall into the high-income category. We also accept that goodwill may subsist in potential customers – a person may be unable to afford to stay at an Aman resort today, but he may nonetheless have been exposed to the “Aman” names and thus aspire to visit one of the Aman resorts some day should his financial position improve. Thus, the goodwill attached to the “Aman” names, apart from existing among the core group of wealthy persons who are presently able to patronise the Aman resorts on a regular basis, also subsists among once-in-a-lifetime guests as well as the penumbra of aspirants who may one day be able to realise their dreams of staying at an Aman resort if their financial status improves.

(3) The type of business in respect of which goodwill in the “Aman” names exists

65     The Judge was not clear about whether the Respondents had goodwill in the “Aman” names in relation to hotels and resorts only, or whether the goodwill extended to residential accommodation. The Appellant argued – correctly, in our view – that the goodwill proved did not extend so far. While there is evidence that the Respondents are engaged in tie-ups and licensing agreements for the use of the “Aman” names vis-à-vis luxury residential developments (which will be discussed in greater detail later), there is no evidence that this aspect of the Respondents’ business is known to the general public in Singapore or even the more limited wealthy market, which forms the target clientele for the Respondents’ resorts. In any event, as mentioned earlier (at [43] above), the Respondents’ case on appeal was that the alleged goodwill existed only in respect of the high-end hotel and resort market.

(4) Our conclusion on the extent of the goodwill attached to the “Aman” names

66     Given the limitations outlined above, we are of the view that the Respondents’ goodwill in the “Aman” names in Singapore is limited largely to those of a high income level who would have been exposed to the Respondents’ advertising. To this section of the public, we would add the category of once-in-a-lifetime guests and aspirants which we alluded to earlier (at [64] above). Finally, we also note that there is ample evidence that the “Aman” names have goodwill among those in the high-end travel and resort industry as well.

The “foreign business problem”

67     There is considerable academic debate in both the UK and Singapore over what is sometimes called the “foreign business problem” in the law of passing off. In a nutshell, this issue concerns the circumstances under which foreign businesses can bring a claim for passing off. The nub of the problem lies in whether a foreign entity must actually do business in a particular jurisdiction before it can bring an action for passing off in that jurisdiction, or whether it can simply rely on its reputation in the jurisdiction concerned without doing any business there. The former position was dubbed “the hard line” in Mechanical Handling Engineering (S) Pte Ltd v Material Handling Engineering Pte Ltd [1993] 2 SLR 205 at 210, [22], and some suggest that it is the position which prevails in the UK and Singapore. The latter is the less traditional approach exemplified by cases like Conagra Inc v McCain Foods (Aust) Pty Ltd (1992) 33 FCR 302 (a decision of the Federal Court of Australia) and Orkin Exterminating Co Inc v Pestco Co of Canada Ltd (1985) 19 DLR (4th) 90 (a decision of the Ontario Court of Appeal).

68     We do not wish to wade into this controversy at present, but there are nevertheless two points pertinent to this appeal which we wish to make. First, although the Respondents do not have any resorts in Singapore, they do have two offices here. Indeed, the Amanresorts Group’s headquarters are in Singapore (see [13] above) and the Amanresorts Group has an actual customer base here. As such, it is not strictly correct to think of the Respondents’ business as a foreign one. The second point is that whether or not Singapore adopts the hard-line approach to the “foreign business problem” may not be of much significance, considering that foreign entities which have no goodwill and/or do not carry on any business in Singapore may still get protection for the names which they use to market their goods, services or business if those names are “well known trade marks” for the purposes of s 55 of the current TMA.

Misrepresentation

69     In order to establish an actionable tort of passing off, the plaintiff, apart from proving that it has goodwill in the goods, services or business in question, must also show that:

(a)     the defendant has made a misrepresentation to the relevant sector of the public; and

(b)     such misrepresentation has resulted in or is likely to result in damage to the plaintiff’s goodwill.

The elements of misrepresentation and damage are very closely linked, and it is crucial to appreciate that both the misrepresentation and the damage must relate to the plaintiff’s goodwill. A misrepresentation is actionable as a tort of passing off only if it causes (or is likely to cause) damage to the plaintiff’s goodwill.

The misrepresentation in the present case

(1) The content of the misrepresentation

70     In the present case, the Respondents allege that the Appellant’s use of the name “Amanusa” constituted a misrepresentation that the Project had the same source as the Respondents’ resorts or that the two (ie, the Project and the Respondents’ resorts) were somehow connected.

(2) The target audience of the misrepresentation

71     The Judge described the relevant section of the public to whom the misrepresentation was made in the following terms (see the Judgment at [57]):

The relevant section of the public in the present case would encompass Singapore citizens or residents who have visited the Aman resorts one or more times or [who] are prospective visitors to the resorts, foreign individuals and corporations here intending to visit the resorts or seeking to invest in high-end properties (and there [are] a substantial number of foreign entities here) and developers and hoteliers interested in becoming licensees of the ‘Aman’ brand names. Conceivably, there would also be a portion of the more than 300,000 Singapore visitors to Bali since 1995 who have read about or heard of the three Aman resorts there.

The Appellant argued that the Judge erred in delineating the relevant sector of the public in the above manner because, in effect, he was confining the relevant sector to the Respondents’ customers when the question was whether the Appellant’s customers were confused by the misrepresentation.

72     The established cases take inconsistent positions on the issue of whether the target audience of the misrepresentation in a passing off action consists of the plaintiff’s customers or the defendant’s customers. In the Advocaat case ([36] supra), Lord Diplock stated that the misrepresentation must be made to the defendant’s customers (see the passage reproduced at [36] above). However, according to Lord Oliver’s formulation in the Jif Lemon case ([36] supra), the misrepresentation must be made to “the public” (id at 499 (reproduced at [36] above)). And, in yet another case, Harrods Limited v Harrodian School Limited [1996] RPC 697, which concerned a passing off action brought by the proprietor of the world-famous department store, Harrods, to restrain the defendant from carrying on a private preparatory school under the name “The Harrodian School”, the English High Court considered whether confusion existed among those parents who had inquired about sending their children to the defendant’s school. The English Court of Appeal disagreed and held instead that the misrepresentation must be considered from the perspective of “a common section of the public” (at 716) which comprised both the plaintiff’s customers and the defendant’s customers. Millett LJ stated (ibid):

[T]he relevant public is not confined to parents who take up places at the [defendant’s] school. The [plaintiff] and the [defendant] appeal to a common section of the public – affluent members of the middle class who live in London, shop at Harrods and wish to send their children to fee-paying schools.

73     With the greatest respect to the judges in the above-mentioned cases, these formulations of the relevant sector of the public do not link the element of misrepresentation back to the foundation of a claim for passing off – viz, goodwill. These formulations include people who may not have any goodwill towards the plaintiff’s get-up at all. Logically, misrepresentations made to such people cannot affect the plaintiff’s goodwill in any way. Christopher Wadlow perceptively notes in The Law of Passing-Off: Unfair Competition by Misrepresentation (Sweet & Maxwell, 3rd Ed, 2004) (“The Law of Passing-Off”) that (at para 5-92):

The relevant public is normally stated to consist of the actual or potential customers of the claimant, and this is consistent with the fact that only misrepresentations directed to them are likely to damage the claimant’s goodwill. The claimant is much less likely to suffer from a misrepresentation made to a person who would never be likely to deal with him. [emphasis added]

To this comment, we would add that not all actual customers and potential customers of the plaintiff necessarily have goodwill vis-à-vis the plaintiff’s get-up. For instance, if a customer buys the plaintiff’s goods solely because of reasons unrelated to the source of those goods (for instance, because of the price, the availability or the functionality, etc, of the goods) and not because of positive association of the goods with the source indicated by the plaintiff’s get-up, then such customers do not ordinarily count for the purposes of assessing the impact of the defendant’s misrepresentation.

74     We find support for this approach in the HFC Bank case ([59] supra). One might recall that, in that case, HFC was held not to have achieved brand recognition among many of its customers, including a number who were called by HFC as witnesses to testify on the issue of confusion. Lloyd J held that those witnesses did not have goodwill towards HFC’s name, and concluded (at 199):

I accept that persons as respects whom HFC do not have goodwill, because they have no established brand recognition, have been confused as between the two entities [ie, HFC and Midland Bank], but this is not enough to make out a case of passing off; because even if in some sense it may be regarded as a misrepresentation, it is not made to the relevant people.

75     We agree with the Respondents’ submission that the approach taken by Lloyd J in the HFC Bank case is the right approach to adopt. The alleged misrepresentation must be analysed from the perspective of those who have goodwill in the plaintiff’s get-up. We further note that the definition of the “relevant section of the public” (see the Judgment at [57]) which the Judge adopted in the context of the misrepresentation in the present case is largely consonant with the group of people who, we have held, have goodwill towards the “Aman” names (see [66] above).

76     The question of whether those in Singapore with goodwill towards the “Aman” names have received the Appellant’s misrepresentation depends on how widely the misrepresentation has been broadcast. This is a matter of fact. Most misrepresentations are broadcast to the public at large in Singapore. In this case, the name “Amanusa” was apparently prominently displayed outside the Showflat itself; it also appeared in promotional brochures and advertisements of the Project in the mass media, which many people in Singapore would have been exposed to. The misrepresentation has thus been made to the public at large, which would include those in Singapore with goodwill towards the “Aman” names.

(3) Whether the misrepresentation resulted in confusion

77     A misrepresentation is actionable under the law of passing off only if it gives rise to confusion. The two elements – misrepresentation and confusion – together form the “deception” referred to by Lord Diplock in the Advocaat case ([36] supra) at 740. The Respondents raised the remarkable suggestion that confusion was not an element of an action for passing off in view of the following dicta in the Millenia case ([36] supra) at [71]:

It is not necessary that there should be any confusion occasioned by the misrepresentation, although confusion is of great evidential value in determining whether or not there has been any misrepresentation. If the names of two entities are confusingly similar, there is a great likelihood that the public will be misled into thinking that they or their businesses are connected with each other.

In our view, the above passage clearly only means that evidence of actual confusion is not required before a passing off action can succeed. It is open to the court to infer a likelihood of confusion from the surrounding facts. Confusion is still an essential element of the tort of passing off and we hope that we have given any misunderstanding of the court’s dicta in the Millenia case (at [71]) its quietus.

78     The question to be asked at this stage is actually fairly simple: Did those in Singapore with goodwill towards the “Aman” names believe that the Project had the same source as the Respondents’ resorts or was somehow connected with the source of those resorts?

79     Before we continue, we pause to note that there was scant evidence from the Respondents that persons in Singapore with goodwill towards the “Aman” names were confused, in actual fact, by the Appellant’s use of the name “Amanusa” for the Project. The only evidence of actual confusion came from the husband of one of the Appellant’s witnesses (a purchaser of a unit in the Project), who said that he had asked the marketing agent whether the Project had anything to do with the Amanusa Bali (see the Judgment at [59]). There was, however, no evidence that high-income individuals (who form the bulk of the group of persons with goodwill towards the “Aman” names) were actually confused.

80     Evidence of actual confusion between the business, goods or services of the plaintiff and those of the defendant may be helpful, but the lack of such evidence is not fatal to the plaintiff’s claim. Instead, the court is entitled to consider whether the average reasonable person, with characteristics reflective of the relevant section of the public as identified under the examination of goodwill, is likely to be confused by the defendant’s misrepresentation. In Parker-Knoll Limited v Knoll International Limited [1962] RPC 265, Lord Devlin said (at 291–292):

Instances of actual deception may be useful as examples, and evidence of persons experienced in the ways of purchasers of a particular class of goods will assist the judge. But his decision does not depend solely or even primarily upon the evaluation of such evidence. The court must in the end trust to its own perception into the mind of the reasonable man.

81     With this in mind, we turn to consider whether the factual matrix of the present case supports the conclusion that the reasonable member of the relevant section of the public identified earlier (at [66] above) was likely to have been confused by the Appellant’s misrepresentation.

82     The Judge quite rightly considered that the Appellant had used for the Project a name which was identical to the name used by the Respondents for the Amanusa Bali. It was also a relevant consideration that the Project was originally heavily touted as having a Balinese atmosphere, just like the Amanusa Bali. These are all factors which increase the likelihood of confusion. On the Balinese theme point, we reject the Appellant’s suggestion that the Judge was somehow granting the Respondents a monopoly over that particular architectural style. The Balinese theme is only relevant in so far as the name of the Project and that of the Amanusa Bali are identical – a factor which increases the likelihood of confusion.

83     Notably, it would have been a relevant consideration increasing the likelihood of confusion if the Appellant had been found to have had fraudulent intentions in using the name “Amanusa” for the Project, ie, if it had set out to deceive those with goodwill towards the “Aman” names (see Office Cleaning Services, Ld v Westminster Window and General Cleaners, Ld (1946) 63 RPC 39 at 42, per Lord Simonds). Although the account by the Appellant of how it came up with the name “Amanusa” for the Project lacked an air of reality and although there certainly was some patent plagiarism of certain laudatory phrases from the Respondents’ promotional materials, the Judge was satisfied that there was no fraudulent intention on the Appellant’s part (see the Judgment at [4] and [75]). This finding has not been challenged by the Respondents. It should be noted, however, that a finding of innocent copying would not in itself exonerate the defendant in a passing off action.

84     Another factor to consider in assessing whether the misrepresentation in question has led to confusion between the business, goods or services of the plaintiff and those of the defendant is whether the parties are in the same field or closely related fields of business – or, to be more precise, whether the business, goods or services to which the plaintiff’s get-up is applied are in the same or a closely related field as that to which the defendant’s business, goods or services belong.

85     We note that the Respondents use the “Aman” names in the high-end hotel and resort business while the Appellant seeks to use the name “Amanusa” for middle-class residential accommodation (specifically, the Project). Nevertheless, we accept the Judge’s finding that “the dividing line between purely residential developments and luxury hotel or resort developments is no longer pronounced” (see the Judgment at [59]). As we noted earlier, the Respondents are involved in the high-end residential accommodation industry as well (see [8][9] above). The Respondents have also put forward cogent evidence that the luxury hotel and resort market as a whole is “converg[ing]” with the residential accommodation market. The Judge summarised the Respondents’ position in relation to the development of Aman Villas thus (see the Judgment at [9]):

The development of such branded real estate projects [is] part of a worldwide trend whereby luxury hotel groups [will] develop and manage branded long-term residential projects. Conversely, developers of houses and condominiums have also started building hotels and resorts. There [is] therefore an ongoing convergence of the two erstwhile separate industries. In Singapore, for example, the Shangri-La Hotels and Resorts manage the Shangri-La Residences, a branded residential condominium development, and the Shangri-La Apartments, a project involving branded serviced apartments. Similarly, Starwood Hotels & Resorts Worldwide [will] be managing the six-star St Regis Hotel together with the St Regis Residences.

86     We accept the above evidence. This convergence between the hotel and resort industry on the one hand and the residential accommodation industry on the other makes it more likely that confusion will occur in the present case. Furthermore, it is well established that “the parties in a passing off action need not be in mutual competition” (see the Millenia case ([36] supra) at [63]) and may even be engaged in “different fields of business activities” (ibid) (see also generally the Millenia case at [63]–[66], citing English and Australian authorities on this point).

87     The Appellant argued that there were such great differences between the Project and the Amanusa Bali that no one would think that they were from the same source or were somehow connected. We are aware that the Respondents’ resorts are aimed at the high-end market while the Project is aimed at the middle class in Singapore, and that the quality of the two kinds of accommodation differs accordingly. In this regard, the Judge found (see the Judgment at [65]):

Visitors to Aman resorts, like the guests at RCMS (The Ritz-Carlton, Millenia Singapore), pay a premium to stay at the resorts. On the other hand, as emphasised by the [Appellant] and its counsel, the [Project’s] target buyers were people looking for reasonably-priced landed property (the listed prices for most of the units [in the Project] ranged from $1.3m to $1.45m, with the most expensive unit priced at $1.8m). Although the [Appellant] did assert that its projects were well known for their high-quality finishes and imported fixtures, I doubt the [Project] would be capable of replicating the ambience and spaciousness of the [Respondents’] resorts. I quote the words of one of the visitors to the [Appellant’s] sales office, who responded to the [Appellant’s] questionnaire asking whether the word ‘Amanusa’ meant anything to him by answering, ‘means spa; alone; serene – but the development is totally opposite to its meaning. Very crowded[.]’ [emphasis in original]

88     We do accept that in some cases, the differences between the business, goods or services of the plaintiff and those of the defendant may be so great and so obvious that no one would mistake the former for the latter and vice versa. For instance, in Miss World (Jersey) Ltd v James Street Productions Ltd [1981] FSR 309 (“the Miss World case”), the plaintiff’s “Miss World” beauty contest was held to be too well known and respectable to be confused with the defendant’s film entitled “Miss Alternative World”, which Lord Denning MR described as “not a film showing beautiful girls – as in the Miss World contest – but a film of the grotesque” (at 310).

89     The decision in the Miss World case may be contrasted with that in the Millenia case ([36] supra). In the latter, the plaintiff alleged that the defendant’s “Millennium” hotels were being passed off as the plaintiff’s “Millenia” complex, which comprised a hotel (“The Ritz-Carlton, Millenia Singapore”), a shopping mall and two office towers. This court noted at [84]:

In so far as the hotel industry is concerned, [The Ritz-Carlton, Millenia Singapore] symbolises world class quality and prestige with top-of-the-range services and amenities. The [defendant’s “Millennium”] hotels in Singapore are of a lower class and quality and are more economically priced.

90     Despite these differences, this court held that the public would very likely be deceived into thinking that the defendant’s “Millennium” hotels were associated with or related to The Ritz-Carlton, Millenia Singapore, or that the former and the latter belonged to the same chain. The court was influenced by (inter alia) the similarity of the words “Millenia” and “Millennium”.

91     We are of the view that the facts in the instant case are closer to those in the Millenia case than those in the Miss World case ([88] supra). We find that the Appellant presented the Project as high-quality accommodation (see [23] above) which was being sold at affordable prices. This manner of presentation would have increased the likelihood that someone who, for example, came across an advertisement of the Project in the newspapers might confuse the Project with the Amanusa Bali.

92     We have said that the bulk of those who have goodwill towards the “Aman” names are high-income individuals (see [66] above). These individuals are not likely to be actual or potential purchasers of units in the Project. Hence, if an initial confusion as to the source of the Project or its connection with the Aman resorts sets in (for example, from a cursory glance at one of the Appellant’s advertisements featuring the Project in the newspapers), there is not likely to be further investigation by the aforementioned individuals into the nature of the Project out of an interest in purchasing units in that project, which investigation (if carried out) would reveal the differences between the Project and the Amanusa Bali so as to dispel such confusion. Further, it is also difficult to dispel confusion where news of the Project gets passed around by word of mouth. Thus, we agree with the Judge’s view that (see the Judgment at [60]):

It is no answer to an allegation of misrepresentation in such a case to say that one could easily check the facts with the [Appellant] or its agents or through its sales brochures (which would indicate the name of the developers for the [P]roject) or otherwise check with the [Respondents] or other independent sources. The [Appellant] should also not be allowed to say that the relevant section of the public in this case would tend to be more intelligent and would verify matters which they are not certain about. The fact that such checking or verification is required is itself evidence of confusion.

93     In conclusion, we accept the Judge’s finding (id at [59]) that there was a likelihood of confusion arising from the misrepresentation in the present case in that the use of the name “Amanusa” for the Project:

… would [create] … the impression that [the Project] is somehow related or connected to the [Respondents], perhaps by way of licence to use the name or that the [Project] has the [Respondents’] endorsement and will in future be managed by [the Respondents]. It may even convey the wrong impression that the [Respondents] are involved as developers.

Damage

94     The tort of passing off protects the plaintiff against damage caused to the goodwill attached to its business, goods or services by the defendant’s misrepresentation. Thus, even if the defendant’s misrepresentation is shown to have caused or to be likely to cause confusion between the plaintiff’s business, goods or services and those of the defendant, such misrepresentation is not in itself actionable under the law of passing off unless it has caused (or is likely to cause) damage to the plaintiff’s goodwill. As correctly identified by the Judge, the test for damage in passing off cases is either “actual or probable damage” (id at [63]) to the plaintiff’s goodwill.

95     In the court below, the Respondents claimed the following heads of damage:

(a)     “inferiority of the [Appellant’s] goods or services” (ibid);

(b)     “loss of licensing opportunity or income” (ibid);

(c)     “likelihood of damage should the [Appellant] get into financial, legal or other trouble” (ibid);

(d)     “misappropriation of the [Respondents’] goodwill and reputation” (ibid);

(e)     “restriction on the [Respondents’] natural expansion into residential real estate” (ibid);

(f)     “loss of exclusivity and erosion of distinctiveness” (ibid) of the “Aman” names; and

(g)     “dilution of [the goodwill attached to] the [‘Aman’ names]” (ibid).

96     The law relating to which heads of damage are (or should be) judicially recognised under the tort of passing off is notoriously difficult. Not only are the pleaded heads of damage often overlapping, but there is also substantial judicial and academic disagreement over which heads of damage should be recognised and which should not. A number of the heads of damage claimed by the Respondents are of an inchoate nature – they are still very much in development at common law and their universal acceptance cannot be taken for granted. As stated by Wadlow at para 4-24 of The Law of Passing-Off ([73] supra):

It is suggested that there may be a distinction to be drawn between heads of damage which are sufficient to support a passing-off action in their own right, and those which may be taken into account in some circumstances, but only if the damage under one of the heads in the first category can be proven or presumed. … One has, so to speak, a category of recoverable parasitic damages.

Analysis of the heads of damage claimed by the Respondents

(1) Inferiority of the Appellant’s residential accommodation

97     There are two primary and very well-established means by which goodwill can be damaged, namely, by “blurring” and by “tarnishment”. Blurring occurs when the plaintiff’s get-up, instead of being indicative of only the plaintiff’s goods, services or business, also becomes indicative of the defendant’s goods, services or business. While customers may still be drawn by the attractive force of the plaintiff’s get-up, they may be drawn to the business, goods or services of the defendant instead of those of the plaintiff. In other words, the goodwill attached to the plaintiff’s business, goods or services becomes spread out over business, goods or services which are not the plaintiff’s. This phenomenon occurs only when the business, goods or services of the plaintiff and those of the defendant are in competition with or are at least substitutes for each other. The damage manifests itself in sales being diverted from the plaintiff to the defendant.

98     Tarnishment occurs when the business, goods or services of the defendant are of a worse quality than those of the plaintiff or have some other undesirable characteristic. Customers think that the plaintiff is now the source of such poor quality or undesirable business, goods or services. The goodwill previously attached to the plaintiff’s business, goods or services loses its attractive quality and may even become a liability, driving away custom rather than attracting it. Where it is alleged that goodwill has been damaged by tarnishment, there is no need for the plaintiff’s business, goods or services and those of the defendant to be in competition with each other (contra the position where damage to goodwill by blurring is concerned). The plaintiff and the defendant can be engaged in entirely different fields of business so long as it is shown that the poor quality or undesirability of the defendant’s business, goods or services rebounds on the plaintiff.

99     In the instant case, the Respondents did not make any claim that the goodwill attached to the “Aman” names had been damaged by blurring. This was a sensible course of action, seeing that those with goodwill in the “Aman” names, being in the main individuals with significantly higher income than the targeted clientele for the Project, are not likely to be purchasers of units in the Project. As for those persons in Singapore with goodwill towards the “Aman” names who may not necessarily be in the high-income group (ie, the once-in-a-lifetime guests and the aspirants mentioned at [64] above), we agree with the Judge’s finding (at [64] of the Judgment) that if such persons had at first mistakenly thought that the Project was related to the Respondents’ resorts generally and/or the Amanusa Bali in particular and had in turn sought to purchase units in the Project on that erroneous basis, the purchasing process itself would have dispelled such misunderstanding. There has thus been no damage by blurring in the present case.

100    Although the Respondents did not assert damage to their goodwill by blurring, they did claim that the Appellant’s use of the name “Amanusa” for the Project would tarnish the goodwill attached to the “Aman” names. In the classic case of tarnishment, the defendant’s business, goods or services have an unsavoury reputation (see, for instance, Annabel’s (Berkeley Square) Limited v G Schock [1972] RPC 838, where the defendant operated an escort agency using the name of the plaintiff’s upscale night club, and C A Sheimer (M) Sdn Bhd’s Trade Mark Application [2000] RPC 484, where it was sought to register the word “Visa” as a trade mark for use in relation to, inter alia, condoms and contraceptive devices). Differences in quality may also be sufficient to establish damage to goodwill by tarnishment (see, eg, the Millenia case ([36] supra)).

101    Obviously, the Project has no immoral or unsavoury associations. The only relevant factor for the purposes of assessing whether there has been damage to goodwill by tarnishment lies in the quality of the Project. In this respect, we find it hard to resist the Judge’s conclusion that if the name “Amanusa” were allowed to be used for the Project (see the Judgment at [65]):

[T]he name which inspired hushed awe could in future evoke suppressed laughter because of the association with [a project] that [does] not quite exude the same luxury and class or quality of upkeep, especially when [the Project has] begun to show [its] age.

(2) Likelihood of damage should the Appellant get into financial, legal or other trouble

102    The Respondents argued that if the Appellant were to get into any sort of financial, legal or other trouble, persons who had confused the Project with the Respondents’ resorts might assume that it was the Respondents which were having difficulties; in effect, the Respondents would have lost control over the goodwill or reputation attached to the “Aman” names. It was submitted that if, for instance, a newspaper article headed “Fire at Amanusa Due To Faulty Wiring” were published, the public might assume that the article was referring to the Amanusa Bali as opposed to the Project. People who made that assumption might decide not to stay at the Amanusa Bali and might also assume that the quality of the other Aman resorts had likewise deteriorated.

103    Some case authorities indicate that the threshold which the plaintiff has to cross for the purposes of proving this particular head of damage (ie, a likelihood of damage to the plaintiff’s goodwill should the defendant encounter financial, legal or other trouble) is relatively low. British Legion v British Legion Club (Street) Ld (1931) 48 RPC 555 (“the British Legion case”) is one such example. The plaintiff there, a charitable association known as the “British Legion”, was formed at the close of World War I with the aim of “[inaugurating] and [obtaining] united democratic comradeship between those who had served in [World War I] and [of making] provision for their welfare and that of their dependants” (at 560). The defendant was a small social club in a quiet corner of Somerset county which styled itself the “British Legion Club (Street) Ld”. A considerable number of its members were ex-soldiers and/or members of the British Legion, but there were also a significant number of its members who were not ex-soldiers or members of the British Legion. The plaintiff brought an action to restrain the defendant from using the words “British Legion” in its name. Farwell J ruled in favour of the plaintiff, holding (at 564):

If the use of the [defendant’s] name, as it is now used, is to suggest to any ordinary person that the [plaintiff] is in some way connected with the [defendant] or is in some way, as it were, its sponsor, then if evil befalls the [defendant] and it finds itself (I am not suggesting [that] it will) in trouble either under the [l]icensing laws or in financial trouble or in some other way of discredit, the result will be, in my judgment, that many of the persons who knew of the existence of the [defendant might] think that the [plaintiff] had been very ill-advised and very unfortunate in having any connection with such a [c]ompany, and that might well tend to prevent persons who otherwise would have supported the [plaintiff] by subscriptions or otherwise from continuing to do so.

We note that there was no evidence in that case that the evils envisaged by Farwell J were about to befall the defendant club.

104    A similarly lenient approach vis-à-vis proof of the likelihood of damage was adopted by the English Court of Appeal in Dawnay Day & Co Limited v Cantor Fitzgerald International [2000] RPC 674 (“the Dawnay Day case”). There, Sir Richard Scott VC held that if the defendant (a securities trading company) were allowed to use the name “Dawnay Day” for its business, one form of damage likely to be suffered by the plaintiffs (securities trading companies belonging to the Dawnay Day group of companies) was the damage to their reputation should the defendant engage in activities “in any respect reprehensible” (at 705–706), which activities the plaintiffs would be unable to prevent. As in the British Legion case, there was no suggestion that the defendant was at the time of the trial engaged in or might in future engage in such activities.

105    With respect, we do not think that it is sufficient for the plaintiff, for the purposes of establishing a likelihood of damage to its goodwill, to merely allege that evil days may befall the defendant. If such allegations were accepted as sufficient proof, the requirement of damage would be rendered otiose. Wadlow in The Law of Passing-Off ([73] supra) argues at para 4-40 that there must be “a real tangible risk of substantial damage”; we agree. In this regard, we would also draw attention to Habib Bank Ltd v Habib Bank AG Zurich [1981] 1 WLR 1265, where the English Court of Appeal dismissed the argument of the plaintiff, Habib Bank Ltd, that there was a danger of some malpractice by the defendant, a bank whose name contained the words “Habib Bank”, rubbing off on the plaintiff. It was held that the plaintiff’s contention was “two-way traffic” (at 1282) and “pure speculation” (ibid).

106    In the instant case, while there is an obvious difference in quality between the Aman resorts and the Project (which, we have found, gives rise to damage in the form of tarnishment (see [101] above)), it is too much of a stretch to allege, without more, that there is a real tangible risk that the Appellant may go into liquidation or that some disaster like a fire will befall the Project.

(3) Loss of licensing opportunity or licensing income, and misappropriation of goodwill and reputation

107    The Respondents stated that they were in the business of granting licences for the use of the “Aman” names to residential property developers in other countries who wanted to adorn their developments with a certain classy flavour. They pointed out that the Appellant had a history of building condominiums and giving them thematic names, and was thus was likely to develop its own “Aman” series of condominiums if it were not prevented from doing so. The Respondents further argued that if the Appellant’s use of the name “Amanusa” for the Project were allowed to pass unhindered, other real estate developers in Singapore could also begin using the “Aman” names without seeking licences from the Respondents.

108    In the UK, loss of licensing opportunity or licensing income has been held to be a distinct head of damage in itself in passing off actions. For instance, in Irvine v Talksport Ltd [2002] 1 WLR 2355, Laddie J thought that the law of passing off should afford a remedy to the plaintiff (“Irvine”), a well-known Formula One racing driver, against the defendant company, which had used his photograph in a brochure promoting its radio station. Irvine was already in the business of endorsement, and it was common knowledge that many sportsmen earned a lot of money through endorsements. Laddie J said at [39]:

Not only has the law of passing off expanded over the years, but the commercial environment in which it operates is in a constant state of flux. Even without the evidence given at the trial in this action, the court can take judicial notice of the fact that it is common for famous people to exploit their names and images by way of endorsement. They do it not only in their own field of expertise but, depending on the extent of their fame or notoriety, wider afield also. It is common knowledge that for many sportsmen, for example, income received from endorsing a variety of products and services represents a very substantial part of their total income. The reason large sums are paid for endorsement is because, no matter how irrational it may seem to a lawyer, those in business have reason to believe that the lustre of a famous personality, if attached to their goods or services, will enhance the attractiveness of those goods or services to their target market. In this respect, the endorsee is taking the benefit of the attractive force which is the reputation or goodwill of the famous person.

109    Another striking example where loss of licensing opportunity was recognised as a recoverable head of damage is the English High Court case of Lego System Aktieselskab v Lego M Lemelstrich Ltd [1983] FSR 155 (“the LEGO case”). There, the famous toy-maker, Lego, and its UK distributor (referred to collectively as “the plaintiffs” in our discussion of this particular case) successfully argued that the use of the name “Lego” by the defendant on its garden sprinklers and other irrigation equipment amounted to passing off, even though there was no evidence that the plaintiffs were going to expand into the defendant’s unrelated field of business (viz, the manufacture of irrigation equipment). Falconer J held (at 194):

… I am satisfied … that the reputation of the plaintiffs’ mark LEGO and their goodwill is wide enough to extend to goods such as garden sprinklers made of coloured plastics [sic] material, [and that] for the [defendant] to market [its] garden sprinklers and other irrigation equipment products would cause damage to the plaintiffs. The plaintiffs now have the potentiality of using the mark and the attractive force of its reputation themselves to market their own goods in the garden equipment field or to license or franchise another trader to use it in that field. As to using it themselves, the evidence … was to the effect that the plaintiffs would be prevented from entering that field if the [defendant were] allowed to continue to use the trade mark LEGO in this country on [its] products. Another course open to the plaintiffs would be to license or franchise another trader to use the mark LEGO in that field. The evidence of the plaintiffs’ expert witness, Mr. Weisz, was important in this respect. He is a marketing expert with very wide and successful experience in advertising and marketing, particularly in the launching of new products and new enterprises in this country, and in franchising operations; he is now Managing Director of A.T.V. Licensing Limited which is concerned with the commercial exploitation by franchising of characters associated with television programmes. The effect of his evidence was that, because of the reputation of Lego (meaning the plaintiffs’ toy construction products), there would be an opportunity for licensing or franchising the mark LEGO in other fields, that, because of the nature of the Lego products, primarily plastic bricks, the plastics area would be a likely one to exploit and that garden implements would be an ideal market for franchising LEGO, because the purchasers of Lego toys are parents and grandparents; as he put it, the very same people who are likely to be purchasers of garden equipment. Obviously, the possibility of licensing or franchising another trader to use LEGO in the gardening equipment area would be lost if the [defendant is] allowed to continue using LEGO in this country in relation to [its] products. The effect, therefore, of the [defendant] continuing to use LEGO in this country in relation to [its] products would be to destroy that part of the plaintiffs’ reputation in their mark LEGO and [the] goodwill attached to it which extends to such goods. [emphasis added]

110    On the other hand, it has been argued that accepting the loss of licensing opportunity or licensing revenue as a recoverable head of damage in passing off actions will lead to a circularity in reasoning. In the New Zealand case of Tot Toys Ltd v Mitchell [1993] 1 NZLR 325, Fisher J said (at 362):

It is axiomatic that damage must be proved or presumed as one of the ingredients of passing off. It has never been sufficient for plaintiffs to fill this gap by arguing that their loss is loss of the right to charge the defendant a fee for continuing conduct the lawfulness of which is the subject currently under inquiry. To accept that proposition would be to deny that damage is an essential and independent ingredient of the tort. And if the defendant’s conduct is otherwise lawful, it cannot be rendered unlawful upon the ground that it might induce others to act in the same way without the plaintiff’s permission.

111    In our view, there is much force in Fisher J’s analysis. Even if the plaintiff does have an existing business of licensing its get-up for use by others, this might only mean that its licensees mistakenly think that a licence is necessary before they can use the plaintiff’s get-up. In fact, it is only if passing off is first proved that the plaintiff then has the right to demand that those who wish to use its distinctive get-up must first obtain a licence from it. We agree with Wadlow’s argument in The Law of Passing-Off ([73] supra) that (at para 4-38):

It might be said that the misrepresentation in passing-off could always have been sanctioned by the claimant for payment; and, therefore, that the claimant has automatically been damaged by the unauthorised making of the misrepresentation. This does not represent English law. If right, it would mean that there would never be any independent need to prove likelihood of actual damage in passing-off.

112    It has also been noted that developments in this area of law (ie, recognising loss of licensing opportunity or licensing income as a recoverable head of damage) appear to be extensions of the classical trinity of passing off delineated at [37] above, such that what appears to be protected is really the plaintiff’s fame rather than the goodwill attached to the plaintiff’s business, goods or services (see, for instance, Hazel Carty, “Advertising, Publicity Rights and English Law” [2004] IPQ 209).

113    In our view, the recognition of loss of licensing opportunity or licensing income as a distinct head of damage should be approached conservatively. In the present case, even if we accept that this head of damage is recoverable in passing off actions under our law, it is not clear that the evidence adduced by the Respondents of their licensing agreements with real estate developers overseas really proves damage arising from misrepresentation and confusion in Singapore – the Respondents may be entitled to license the “Aman” names overseas because they have been registered as trade marks overseas or for other reasons, but this does not mean that the “Aman” names have a natural capacity for generating licensing opportunity or licensing revenue in Singapore for the Respondents.

114    That said, the loss of licensing opportunity or licensing revenue is not entirely irrelevant to the equation. If passing off is proved (based on some other head of damage which is well established as being recoverable), the plaintiff, in addition to claiming for any loss of sales occasioned by such damage, can also claim for loss of licensing revenue. Thus, the loss of licensing opportunity or licensing revenue goes towards quantifying rather than proving damage.

115    The Respondents’ claim for “misappropriation of … goodwill and reputation” (see the Judgment at [63]) is closely linked to the claim for loss of licensing opportunity or licensing income. The Respondents argued that the Appellant’s use of the name “Amanusa” amounted to “an unlicensed appropriation of the Respondents’ goodwill and reputation.” Our views stated earlier (at [110][114] above) also apply to this argument.

(4) Restriction on the Respondents’ expansion into the residential real estate business

116    The Respondents claimed that if the Appellant were permitted to use the name “Amanusa” for the Project, they would be prejudiced in any future commercial plans to expand into residential development in Singapore. It was submitted that damage should be inferred in this case because the Respondents not only had every intention of expanding into the field of residential accommodation, but were in fact already involved in that field.

117    We accept the principle that a plaintiff who has established goodwill in one form of commercial activity (“the established activity”) may be entitled to protection from passing off vis-à-vis another form of commercial activity which is a natural expansion of the first (“the extended activity”). Thus, in Alfred Dunhill Limited v Sunoptic SA [1979] FSR 337, it was understandable why the English Court of Appeal prevented the defendant from marketing sunglasses under the name “Dunhill” given that the plaintiff, a well-known producer of tobacco goods under the same name, had expanded its business to include the sale of luxury goods for men, likewise under the name “Dunhill”. The plaintiff had also showed that although it had not sold sunglasses in England, it was planning to produce sunglasses itself.

118    It is, however, important to stress the close connection which must exist between the established activity and the extended activity for that connection forms the foundation for this head of damage. Wadlow in The Law of Passing-Off ([73] supra) cites (at para 4-42) the following extract from the American case of S C Johnson & Son, Inc v Johnson 116 F 2d 427 (2nd Cir, 1940) at 429 (per Hand J) as being representative of the English position in this regard:

It is true that a merchant who has sold one kind of goods … sometimes finds himself driven to add other ‘lines’ in order to hold or develop his existing market; in such cases he has a legitimate present interest in preserving his identity in the ancillary market, which he cannot do … if others make his name equivocal there. But if the new goods have no such relation to the old, and if the first user’s interest in maintaining the significance of his name when applied to the new goods is nothing more than the desire to post the new market as a possible preserve which he may later choose to exploit, it is hard to see any basis for its protection [ie, for the protection of the merchant’s name in the new market]. The public may be deceived, but [the merchant] has no claim to be its vicarious champion; his remedy must be limited to his injury and by hypothesis he has none.

In our view, this passage is representative of the position in Singapore as well.

119    Some of the English cases in which claims for passing off have succeeded despite the apparent gulf between the established activity and the extended activity of the plaintiff can be understood on their particular facts. One such case is the English High Court’s decision in The Eastman Photographic Materials Company, Ld v The John Griffiths Cycle Corporation, Ld and The Kodak Cycle Company, Ld (1898) 15 RPC 105 (“the KODAK case”). In that case, the plaintiff carried on a business involving cameras and other related goods under the trade mark “Kodak”. It also made cameras, called “Cycle Kodaks” or “Bicycle Kodaks”, which were suitable for use on bicycles. The plaintiff sought an injunction to restrain the defendant from carrying on a bicycle business under the name “The Kodak Cycle Company, Ld”. This application was successful. Romer J noted (at 110):

[S]o great is the connection between the two classes of business [ie, the plaintiff’s camera business and the defendant’s bicycle business], that in all probability, I may say, the [p]laintiff … may wish hereafter to manufacture and sell cycles specially adapted to carry [its] ‘Kodaks.’

It is important to note that not only were the plaintiff’s “Kodak” cameras “very well known indeed throughout the length and breadth of [the UK]” (ibid), but the defendant’s fraudulent intention, so Romer J found, “was really to try and get a monopoly of the word ‘Kodak’ as connected with cycles … in the hope and intention of, in some shape or other, identifying [the defendant company] with the [p]laintiff” (ibid). It also appears from the judgment that at the time the case was decided (viz, in the late 19th century), the camera industry and the bicycle industry were not quite as distinct as they are today. Apparently, many shops sold both cameras and bicycles.

120    In the LEGO case ([109] supra), it was held that the toy brick industry was close enough to the garden sprinkler industry to warrant the protection of the “Lego” mark against passing off in the latter industry. It should be noted that Falconer J was assisted in this conclusion by the fact that the plaintiffs in that case had established “a high reputation in the mark, which [had] become very widely known in [the UK] in relation to [toy construction sets made of moulded coloured plastic materials] as denoting the goods of the plaintiffs” (at 157–158), and by the fact that the word “Lego” was “[a] household word” (at 160). Also, there were similarities in the plaintiffs’ products and the defendant’s garden sprinklers as both consisted of brightly-coloured plastic goods.

121    In our view, in the instant case, the Respondents’ field of business (which, in the context of this appeal, is the high-end hotel and resort business (see [65] above)) and the Appellant’s field of business (viz, the residential accommodation business) are closely connected. Both fields concern accommodation. The fact that the Respondents have already expanded into the residential accommodation business overseas (see [8][9] above) buttresses this conclusion. The Respondents’ and the Appellant’s respective fields of business are certainly much closer than the camera and the bicycle industries (see the KODAK case) or the toy brick and the garden sprinkler industries (see the LEGO case). The use by the Appellant of the name “Amanusa” or of other names similar to the “Aman” names in the field of residential accommodation in Singapore would prevent the Respondents from expanding into the residential accommodation business in this country. For this reason, we hold that this head of damage (viz, restriction on the Respondents’ expansion into the residential accommodation business) has been proved.

(5) Loss of exclusivity and erosion of the distinctiveness of the “Aman” names as well as dilution of the goodwill attached to these names

122    The Respondents argued that that the use of the name “Amanusa” by the Appellant for the Project would lead to the gradual whittling away and erosion of the uniqueness and exclusivity associated with the “Aman” names, and would lessen the capacity of these names to identify and distinguish the Respondents’ goods and services from the goods and services of others. The Respondents contended that:

[T]he distinctiveness of the ‘AMAN’ brand would be eroded by the haphazard and random use of the brand, for example, where geographical exclusivity is compromised, or where there is indiscriminate and widespread use and the use is not protected by the involvement of the [Respondents] ...

The Judge agreed with the Respondents, holding that “the most potent form of likely damage is the intangible loss that may be occasioned by the insidious process of dilution such that the famous name [in question] loses its uniqueness” (see the Judgment at [65]).

123    The Respondents cited Taylor Bros Ltd v Taylors Group Ltd [1988] 2 NZLR 1 (“the Taylor Bros case”) in support of the proposition that these heads of damage – viz, loss of exclusivity and erosion of the distinctiveness of the plaintiff’s get-up, as well as dilution of the goodwill attached to that get-up – ought to be recognised. The plaintiff in that case operated, inter alia, a dry-cleaning business under the style “Taylors”, “Taylors Drycleaners” and “Taylors Drycleaning”. It sought an injunction to restrain the defendant, which likewise ran (inter alia) a dry-cleaning business, from trading in the Wellington region under the name “Taylors”. The question before the court was whether damage had been caused in the form of the loss of the distinctiveness of the name “Taylors” or a risk of inundation or dilution of the plaintiff’s goodwill through the incorrect association of the plaintiff with the defendant. The Wellington Court of Appeal upheld the High Court of Wellington’s decision to grant the plaintiff the injunction sought, stating (at 37–38):

[I]n some cases it is legitimate to infer damage from a tendency to impair distinctiveness. … Here there is ample evidence that Taylors is at present distinctive of the [plaintiff] in Wellington in the particular field of businesses associated with textile cleaning. If the [defendant is] permitted to use the name in Wellington, albeit in a special branch of cleaning business in which [the plaintiff] do [sic] not engage, there will be not only the natural tendency to treat the parties as associated but also a loss by [the plaintiff] of control and licensing capacity (and therefore revenue) of the same kind as [that] illustrated by the [LEGO] case.

124    We find it difficult to determine if the Taylor Bros case really established a separate head of damage. It is clear that the Wellington Court of Appeal thought that damage had been caused to the plaintiff in two ways, namely, due to: (a) a “natural tendency to treat the parties as associated” (id at 38); and (b) “a loss by [the plaintiff] of control and licensing capacity (and therefore revenue)” (ibid). With regard to the first type of damage, it is hard to see how this constitutes damage as opposed to mere proof that there has been a misrepresentation by the defendant which has led to confusion between the business, goods or services of the plaintiff and those of the defendant. As for the second form of damage, we have already dealt with it earlier (at [107][114] above).

125    The Respondents also cited the Dawnay Day case ([104] supra), in which it was held by Sir Scott VC (at 706) that “the use of Dawnay Day as a trading style by a company that [was] not a member of the Dawnay Day group [would] dilute and, potentially, [might] destroy the distinctiveness of the name”. No further elaboration on what this meant was given by Sir Scott VC, although he referred to Taittinger SA v Allbev Ltd [1993] FSR 641 (“the Elderflower Champagne case”). In that case, the issue was whether it was permissible for the defendants to sell a non-alcoholic, fruit-flavoured soft drink called “Elderflower Champagne” in a bottle and a get-up that looked much like the bottle and the get-up used for “champagne”, which term was used only in respect of sparkling wines from the Champagne region in France. Peter Gibson LJ said (at 670) that the “erosion of the distinctiveness of the name champagne in [the UK was] a form of damage to the goodwill of the business of the champagne houses”.

126    Sir Thomas Bingham MR came to the same conclusion. He said (at 678 of the Elderflower Champagne case):

The first [plaintiff’s] reputation and goodwill in the description Champagne derive not only from the quality of [its] wine and its glamorous associations, but also from the very singularity and exclusiveness of the description, the absence of qualifying epithets and imitative descriptions. Any product which is not Champagne but is allowed to describe itself as such must inevitably, in my view, erode the singularity and exclusiveness of the description Champagne and so cause the first [plaintiff] damage of an insidious but serious kind. … I cannot see, despite the defendants’ argument to the contrary, any rational basis upon which, if the defendants’ product were allowed to be marketed under its present description, any other fruit cordial diluted with carbonated water could not be similarly marketed so as to incorporate the description champagne. The damage to the first [plaintiff] would then be incalculable but severe.

127    The concept of loss of exclusivity or erosion of the distinctiveness of the plaintiff’s get-up as a separate head of damage in passing off actions is not an immediately obvious one. The danger of recognising such a head of damages lies in truism: In almost all cases where the defendant is alleged to have applied the plaintiff’s distinctive get-up not on the same or similar goods or services as those of the plaintiff but, rather, on different goods or services, there will likely be a loss of exclusivity by definition. On this matter, Wadlow in The Law of Passing-Off ([73] supra) warns that (at para 4-44):

Considered as a head of damage in its own right, loss of exclusivity in a name, mark or get-up needs to be treated with caution. It lends itself to circular arguments, thereby undermining the importance of damage as an essential element of passing-off, and it leads easily to the fallacy that the property which is being protected is the name or mark or get-up itself, rather than the goodwill in the claimant’s business. In the normal course of events there is little to be gained by enumerating loss of exclusivity as a separate head of damage, since it overlaps with, or is an intermediate cause of, several other heads of damage which are better established.

128    This is certainly the case on the facts of the present appeal. We have found that the application of the name “Amanusa” to the Project is likely to tarnish the goodwill attached to the “Aman” names because of the difference between the quality of the accommodation provided by the Appellant and that of the accommodation provided by the Respondents. One can also recast this by saying that the “Aman” names, which have hitherto been used exclusively in respect of high-end hotels, resorts and residential accommodation, would be in danger of losing their distinctiveness if the Appellant were allowed to continue using the name “Amanusa” for the Project as consumers might mistakenly think that the goodwill in the “Aman” names extends to middle-class residential accommodation in suburban Singapore. Thus, the inclusion of a claim for loss of exclusivity or erosion of distinctiveness does not, in our view, add anything to the Respondents’ claim.

129    The same is to be said of the Respondents’ claim for dilution of the goodwill attached to the “Aman” names. The Respondents cited the Millenia case ([36] supra) in support of the proposition that this head of damage is recognised in Singapore. In the Millenia case, this court stated (at [84]):

It would be commercially damaging for [the] business of a super deluxe ‘five-star’ hotel, as [The Ritz-Carlton, Millenia Singapore] is, to be associated with that of a ‘four-star’ hotel, which the [defendant’s “Millennium”] hotels admittedly are. The reason for this is simple and is best illustrated in the sale or provision of goods or services. If there is a misrepresentation induced by the defendant, such that the public is led to believe erroneously that the plaintiff’s and the defendant’s goods or services are the same or that their businesses are connected, the plaintiff’s goodwill in his business would be likely to be diluted … if the defendant’s goods or services are of a quality or class lower than that of the plaintiff’s. Deceived into thinking that the plaintiff’s and the defendant’s goods or services are the same or that their businesses are connected, the public would mistakenly perceive the plaintiff as having a range of lower quality goods or services which in reality belong not to him but to the defendant. The public would not perceive the plaintiff as supplying solely top-of-the-range goods or services but [goods or services] of varying qualities. The prestigious image of the plaintiff would in these circumstances be likely to be undermined. [emphasis added]

130    We do not see this passage as holding anything more than that the association of The Ritz-Carlton, Millenia Singapore (the plaintiff’s five-star establishment) with the defendant’s four-star “Millennium” hotels would cause tarnishment of the goodwill attached to the name “Millenia”. This, as we have seen (at [97][98] above), is a conventional head of damage.

131    We shall have occasion to consider the technical and statutory definitions of “dilution” in due course in the context of the Respondents’ claim under s 55 of the current TMA. It suffices for the moment to say that, to avoid unnecessary confusion, it is better to avoid using the word “dilution” in a passing off action when what is meant is simply tarnishment or blurring of the goodwill in the plaintiff’s get-up.

Our conclusion on the passing off claim

132    We agree with the Judge’s finding that the Respondents’ claim for passing off has been made out. Goodwill exists in the “Aman” names (but not in “Aman”-prefixed names). Although the goodwill in the “Aman” names in Singapore may not have reached across the entire spectrum of the public here, it certainly does exist in respect of high-income individuals (who form the Respondents’ core clientele), as well as once-in-a-lifetime visitors to the Respondents’ resorts and aspiring visitors who have somewhat less means but who are nonetheless attracted by the exclusivity and luxury embodied in the “Aman” names in the field of hotel and resort accommodation. This section of the public would have been exposed to the Appellant’s use of the name “Amanusa” for the Project – ie, to the Appellant’s misrepresentation that the Project had the same source as or was in some way connected with the Respondents’ resorts. They would have been likely to have been confused by this misrepresentation into assuming that the Project was either developed by the Respondents or somehow connected with the latter’s resorts. The Respondents have not succeeded in convincing us on all their pleaded heads of damage, but this is not significant as the Respondents have successfully demonstrated a likelihood of damage in the form of: (a) tarnishment of the goodwill attached to the “Aman” names due to the difference in quality between the Aman resorts and the Project (see [101] above), and (b) restriction on the Respondents’ expansion into the residential accommodation business in Singapore (see [121] above).

The claim under section 55 of the current TMA: Infringement of a well-known trade mark

133    Apart from pursuing a claim for passing off, the Respondents also claimed that the “Aman” names were “well known trade mark[s]” within the meaning of s 2(1) of the current TMA and were entitled to the protection set out in s 55 of that Act. Specifically, the Respondents alleged that the use of the name “Amanusa” by the Appellant in the course of trade, without the Respondents’ consent, on goods or services related to residential accommodation would indicate a connection between the Appellant’s residential accommodation-related goods or services (specifically, the Project) and the Respondents, and was likely to damage the Respondents’ interests (see s 55(3)(a) of the current TMA).

What is a “well known trade mark”?

134    The first question to ask in considering a claim under s 55 of the current TMA is what the term “well known trade mark” in this section means. According to the definition set out in s 2(1) of the current TMA:

‘well known trade mark’ means —

(a)    any registered trade mark that is well known in Singapore; or

(b)    any unregistered trade mark that is well known in Singapore and that belongs to a person who —

(i)    is a national of a Convention country [ie, a country which is a party to the Paris Convention for the Protection of Industrial Property of 20 March 1883 or a member of the World Trade Organization]; or

(ii)   is domiciled in, or has a real and effective industrial or commercial establishment in, a Convention country,

whether or not that person carries on business, or has any goodwill, in Singapore …

135    This definition was adopted in order for Singapore to comply with its obligations under the Paris Convention for the Protection of Industrial Property of 20 March 1883 (“the Paris Convention”) and the Agreement on Trade-Related Aspects of Intellectual Property Rights (“the TRIPS Agreement”), which is set out in Annex 1C of the Agreement Establishing the World Trade Organization signed in Marrakesh on 15 April 1994 (we shall have occasion to refer to both the Paris Convention and the TRIPS Agreement in greater detail at a later stage). The important development which underlies the current TMA is the recognition that some unregistered trade marks should be entitled to protection, and that such protection does not depend on whether the proprietor of the trade mark in question carries on business or has any goodwill in Singapore. (For ease of discussion, we shall, in the context of the Respondents’ claim under the current TMA, refer to the proprietor of a trade mark as “the plaintiff” and the party whose trade mark allegedly infringes or is in conflict with the plaintiff’s trade mark as “the defendant”; we shall also use the terms “the plaintiff’s goods or services” and “the defendant’s goods or services” to denote the goods or services to which the plaintiff’s trade mark and the defendant’s trade mark, respectively, are applied.) This constitutes a workaround the hard-line stance taken towards the “foreign business problem” discussed earlier (at [67][68] above) in the context of the tort of passing off. This development also to some extent represents the erosion of the principle of the territoriality of trade marks (viz, that trade marks are entitled to protection only within the national borders of the countries in which they are registered), but, at the same time, it must be stressed that s 55 of the current TMA does not dispense with the territoriality element entirely. A trade mark must still be “well known in Singapore” [emphasis added] (see s 2(1) of the current TMA) before it is entitled to enjoy the protection conferred by s 55.

136    In assessing whether a trade mark is “well known in Singapore” within the meaning of s 2(1) of the current TMA, regard must be had to s 2(7), which states:

Subject to [s 2(8)], in deciding, for the purposes of this Act [ie, the current TMA], whether a trade mark is well known in Singapore, it shall be relevant to take into account any matter from which it may be inferred that the trade mark is well known, including such of the following matters as may be relevant:

(a)    the degree to which the trade mark is known to or recognised by any relevant sector of the public in Singapore;

(b)    the duration, extent and geographical area of —

(i)    any use of the trade mark; or

(ii)   any promotion of the trade mark, including any advertising of, any publicity given to, or any presentation at any fair or exhibition of, the goods or services to which the trade mark is applied;

(c)    any registration or application for the registration of the trade mark in any country or territory in which the trade mark is used or recognised, and the duration of such registration or application;

(d)    any successful enforcement of any right in the trade mark in any country or territory, and the extent to which the trade mark was recognised as well known by the competent authorities of that country or territory;

(e)    any value associated with the trade mark.

137    This definition is based on Art 2(1) of another document which we shall examine in greater detail in due course, viz, the Joint Recommendation Concerning Provisions on the Protection of Well-Known Marks (“the Joint Recommendation”), which was adopted at the Thirty-Fourth Series of Meetings of the Assemblies of the Member States of the World Intellectual Property Organization (“WIPO”) held from 20 to 29 September 1999 (“the 34th WIPO Meeting”). The factors laid out in ss 2(7)(a)–2(7)(e) of the current TMA (which collectively correspond to Art 2(1)(b) of the Joint Recommendation) are not an exhaustive list in that the court is obliged to take into account “any matter from which it may be inferred that the trade mark is well known” [emphasis added] (per s 2(7) of the current TMA). It appears that the court is ordinarily free to disregard any or all of the factors listed in s 2(7) as the case requires (subject to one particular factor which we shall discuss at [139][140] below), and to take additional factors into consideration. The list of factors in s 2(7) can thus be more accurately described as a set of guidelines. This conclusion is supported by Art 2(1)(c) of the Joint Recommendation, which refers to the factors listed in Art 2(1)(b) (ie, the equivalent of our ss 2(7)(a)–2(7)(e)) in the following manner:

The above factors, which are guidelines to assist the competent authority to determine whether the mark is a well-known mark, are not pre-conditions for reaching that determination. Rather, the determination in each case will depend upon the particular circumstances of that case. In some cases all of the factors may be relevant. In other cases some of the factors may be relevant. In still other cases none of the factors may be relevant, and the decision may be based on additional factors that are not listed in subparagraph (b), above [ie, Art 2(1)(b)]. Such additional factors may be relevant, alone, or in combination with one or more of the factors listed in subparagraph (b), above. [emphasis added]

138    This somewhat ambivalent passage gives an inkling of how difficult it is to determine whether a trade mark is well known in Singapore. There is no single universally-applicable test, but the guidelines in s 2(7) of the current TMA strive toward some semblance of objectivity.

139    There has been some significant and interesting debate among legal commentators on the actual scope of ss 2(7)(c)–2(7)(e) of the current TMA. However, as the instant case does not turn on the consideration of these factors, we shall limit our discussion to s 2(7)(a) (primarily) and s 2(7)(b) of the current TMA. Despite what has been said earlier, it can be persuasively said that s 2(7)(a) is arguably the most crucial factor when determining whether a trade mark is well known in Singapore. This is because, by virtue of s 2(8) of the current TMA:

Where it is determined that a trade mark is well known to any relevant sector of the public in Singapore, the trade mark shall be deemed to be well known in Singapore.

140    This deeming provision suggests that the court cannot disregard s 2(7)(a) (which looks at “the degree to which the trade mark is known to or recognised by any relevant sector of the public in Singapore” [emphasis added]) in its deliberations. Indeed, it appears that all the other factors listed in s 2(7) are irrelevant once it is determined that the trade mark in question is well known to “any relevant sector of the public in Singapore” (per s 2(7)(a)).

141    Section 2(9) of the current TMA elaborates on who constitutes the “relevant sector of the public in Singapore” as follows:

In [ss 2(7) and 2(8)], ‘relevant sector of the public in Singapore’ includes any of the following:

(a)    all actual consumers and potential consumers in Singapore of the goods or services to which the trade mark is applied;

(b)    all persons in Singapore involved in the distribution of the goods or services to which the trade mark is applied;

(c)    all businesses and companies in Singapore dealing in the goods or services to which the trade mark is applied.

142    The ambit of these categories, in particular s 2(9)(a), is in dispute in this case. The Appellant argues that s 2(9)(a) should be read as covering all actual consumers and potential consumers in Singapore of the type of goods or services to which the plaintiff’s trade mark is applied, while the Respondents contend that the provision should be taken to mean what it says – ie, one considers actual consumers and potential consumers of, specifically, the plaintiff’s goods or services only. The former interpretation, which encompasses a wider group of persons within the “relevant sector of the public in Singapore”, makes it harder for the plaintiff to show that its trade mark is well known in Singapore and this is, presumably, the Appellant’s aim in advancing this particular interpretation in the present appeal.

143    There is an overarching policy question, applicable to the general construction of all provisions on the protection of well-known trade marks, which reflects a tension between two key positions. On the one hand, well-known trade marks are usually the result of careful planning, hard work, large investments in branding and marketing as well as an established presence in a market. From this perspective, well-known trade marks deserve to be protected because of the time, effort and money which have been expended on their development and promotion. Alternatively, one could justify protection of well-known trade marks from a consumer welfare perspective, in that such trade marks perform crucial signalling, symbolic and even myth-making roles in society. These views may lead one to conclude that well-known trade marks should be given greater protection than what may loosely be called “ordinary” trade marks (ie, trade marks which do not satisfy the definition of “well known trade mark[s]” in s 2(1) of the current TMA).

144    On the other hand, one may legitimately ask whether well-known trade marks deserve any better protection than ordinary trade marks. There are strong economic and social reasons why the answer to this question may well be “no”. Monopolies and barriers to entry created by strong trade mark protection are disincentives to competition and distort the proper functioning of the free market economy. Socially, the commercial monopolisation of words and symbols is detrimental to free enterprise, whether of a commercial or a private nature, in so far as well-known trade marks are likely to be important subjects of everyday commentary, comparison and critique. Even if strong protection of well-known trade marks is justified, some argue that the extra-special rights which accompany such protection should be strictly limited to an extremely small group of particularly well-known trade marks.

145    On the present point of how s 2(9)(a) of the current TMA should be construed (see [142] above), we are of the view that the Respondents have a stronger case. The point made by the Appellant is that the Respondents’ interpretation of s 2(9)(a) (ie, that the consumers referred to in that provision are limited to actual consumers and potential consumers of, specifically, the plaintiff’s goods or services only) would make it too easy for a trade mark to be deemed to be well known in Singapore. It is argued that the nature of the test itself – which is based on the viewpoint of “any relevant sector of the public” [emphasis added] (see s 2(7)(a) and s 2(8) of the current TMA) – already means that any trade mark which is known to even the smallest and most exclusive of niche markets can become well known in Singapore. Furthermore, according to the Appellant, the miniscule size of the overall market in Singapore makes it very easy for even a small marketing campaign to establish awareness of a trade mark among “any relevant sector of the public”.

146    Notwithstanding the Appellant’s arguments, there are, in our view, sound reasons why it is inappropriate to demand that a trade mark must be well known across an entire country before it is regarded as a well-known trade mark. The explanatory notes which accompany the Joint Recommendation (“the Explanatory Notes”) explain this well at para 2.15 vis-à-vis Art 2(2)(b) of the Joint Recommendation (which is the equivalent of s 2(8) of the current TMA):

It is not permitted to apply a more stringent test such as, for example, that the mark be well known by the public at large. The reason for this is that marks are often used in relation to goods or services which are directed to certain sectors of the public such as, for example, customers belonging to a certain group of income, age or sex. An extensive definition of the sector of the public which should have knowledge of the mark would not further the purpose of international protection of well-known marks, namely to prohibit use or registration of such marks by unauthorized parties with the intention of either passing off their goods or services as those of the real owner of the mark, or selling the right to the owner of the well-known mark. [emphasis added]

147    The power of the deeming provision in s 2(8) of the current TMA (see [139] above) should also not be overstated. It must be noted that before a well-known trade mark qualifies to be protected under s 55, it must still be found to be (broadly speaking) in conflict with the defendant’s trade mark. This area will be more fully explored in subsequent sections of this judgment; it suffices for present purposes to note three points. First, the additional protection granted to a trade mark that is well known in Singapore (as compared to the protection granted to ordinary trade marks) is not very extensive. By comparison, far greater protection is granted to trade marks which are “well known to the public at large in Singapore” [emphasis added] (see, inter alia, s 55(3)(b) and s 55(4)(b)).

148    Second, the deeming provision in s 2(8) does not (and should not) mean that every person in Singapore is deemed to actually know of the plaintiff’s (well-known) trade mark. The likelihood of confusion between the plaintiff on the one hand and the defendant’s goods or services on the other must still be proved (at least where the plaintiff seeks to avail itself of certain subsections of s 55), and, for this purpose, the fact that a well-known trade mark is known only to a niche market may make it harder to show a likelihood of confusion.

149    Third, the deeming provision in s 2(8) still requires the trade mark in question to be “well known” [emphasis added] to the relevant sector of the public in Singapore. In this regard, it is interesting that Singapore did not implement Art 2(2)(c) of the Joint Recommendation, which provides that a mark which is merely “known” to the relevant sector of the public “may” be considered to be well known. All this suggests that there are different degrees of public knowledge of a trade mark, and the requisite level of knowledge required under Singapore’s legislation tends towards the higher end of the scale: ie, to qualify for protection under s 55 of the current TMA, the trade mark concerned must be more than merely “known” to the relevant sector of the public in Singapore. This understanding is borne out by our legislature’s intention in enacting s 55 of the current TMA (and the related provisions therein pertaining to well-known trade marks). At the second reading of the Trade Marks (Amendment) Bill 2004 (Bill 18 of 2004), ie, the Bill which led to the enactment of the current s 55 and the other relevant accompanying provisions, Prof S Jayakumar, the Minister for Law, stated (see Singapore Parliamentary Debates, Official Report (15 June 2004) vol 78 at col 111):

[T]he Bill will give better protection [to] well-known marks. These are marks which are famous and known to many people. Their owners would have invested substantial amounts of time and money … to generate the high levels of recognition and goodwill associated with these marks.

150    The current TMA, when read in conjunction with the Joint Recommendation, also does not lend support to the Appellant’s contention (viz, that s 2(9)(a) of the current TMA should be interpreted as referring to actual consumers and potential consumers of the type of goods or services to which the plaintiff’s trade mark is applied, regardless of whether or not such consumers are the plaintiff’s actual or potential customers). Article 2(2)(a)(i) of the Joint Recommendation (which formed the basis for s 2(9)(a) of the current TMA), in referring to “actual and/or potential consumers of the type of goods and/or services to which the mark [is applied]” [emphasis added], differs crucially from s 2(9)(a). This departure (in s 2(9)(a) of the current TMA) from the wording used in Art 2(2)(a)(i) of the Joint Recommendation may in itself signal our legislature’s intentions. Besides, even the commentary on Art 2(2)(a)(i) of the Joint Recommendation appears to focus on consumers of the plaintiff’s goods or services specifically rather than the more general category of consumers of a particular type of good or service (see the Explanatory Notes at para 2.12):

The expression ‘consumers’ is to be understood in the wide sense of the term, and should not be restricted to those persons who actually and physically consume the product. In that respect, reference can be made to the term ‘consumer protection’ which covers all parts of the consuming public. Because the nature of the goods or services to which a mark is applied can vary considerably, actual and/or potential consumers can be different in each case. Groups of actual and/or potential consumers may be identified with the help of parameters such as the target group for the goods and services in relation to which the mark is used or the group of actual purchasers. [emphasis added]

151    Similarly, in respect of Art 2(2)(a)(iii) of the Joint Recommendation (which corresponds to our s 2(9)(c)), the Explanatory Notes state (at para 2.14):

The business circles which deal with the goods and/or services to which a mark [is applied] are in general constituted by importers, wholesalers, licensees or franchisees interested and prepared to deal in the goods or services to which the mark [is applied]. [emphasis added]

This position (viz, that s 2(9)(a) of the current TMA refers to only actual consumers and potential consumers of the plaintiff’s goods or services) is also supported by authority. The second sentence of Art 16(2) of the TRIPS Agreement, which (broadly speaking) corresponds to s 2(7)(a) of the current TMA in so far as the assessment of whether a trade mark is well known is concerned, states that “[i]n determining whether a trademark is well-known … account [shall be taken] of the knowledge of the trademark in the relevant sector of the public”. Frederick W Mostert notes in Famous and Well-known Marks: An International Analysis (International Trademark Association, 2nd Ed, 2004) at p 1-36 that:

The probable interpretation [of Art 16(2) of the TRIPS Agreement] is that it must be shown that the target audience – the relevant trade circle or the relevant class of consumers for which the particular goods or services are destined – has knowledge of the mark. In other words, the specific sector of the public within which the particular goods or services are purchased or used, or where the mark is exposed through advertising, should become the focus of the factual enquiry. The focus will very much depend on the type of goods or the nature of the services involved. [emphasis added]

152    We also agree with the Respondents that expanding the realm of the “relevant sector of the public in Singapore” mentioned in s 2(7)(a) and s 2(8) of the current TMA to include actual consumers and potential consumers of the type of goods or services to which the plaintiff’s trade mark is applied does not really assist the Appellant. The concept of a “type” of goods or services is not useful. The problem is one of definition – for instance, in the present case, would the relevant type of goods or services encompass the entire hotel and resort market as a whole, or only the high-end hotel and resort market? Or, could the relevant type of goods or services be even more limited (eg, the market for high-end hotels and resorts in Bali specifically)? In contrast, the inquiry is much more focused and manageable if one looks only at the specific goods or services to which the plaintiff’s trade mark has been applied (ie, if one considers only the plaintiff’s goods or services).

153    The Appellant argues that such an approach is tautological in that the plaintiff’s actual customers and potential customers would naturally know of the plaintiff’s trade mark. This is not, however, always the case. The plaintiff could be but one of many producers or suppliers of similar goods or services, and its customers may equally happily purchase those goods or services from other sources without paying attention to what trade mark is attached to the goods or services which they buy. The plaintiff’s advertising could also have been unsuccessful, thus leaving no imprint of the plaintiff’s trade mark on the target market.

154    Applying the law to the facts, we find that the Judge correctly identified the relevant sector of the public in Singapore as the actual and/or potential consumers of the Respondents’ good and services. We also find that there has been a great deal of promotion of the “Aman” names targeted at this group of consumers, and that the “Aman” names enjoy an established reputation among this group (see in this regard, respectively, s 2(7)(b)(ii) and s 2(7)(a) of the current TMA). As we indicated earlier (at [139] above), there is some ambiguity as to the application of some of the other factors listed in s 2(7) of the current TMA. Since it is clear to us that the “Aman” names are well known in Singapore when considered from the viewpoint of the factors listed in s 2(7)(a) and s 2(7)(b), we decline to express any opinion as to whether and to what extent the factors set out in ss 2(7)(c)–2(7)(e) are applicable to the facts of the instant case. Further, since the “Aman” names are “known to or recognised by any relevant sector of the public in Singapore” (see s 2(7)(a)), it follows that, by virtue of s 2(8), they are to be deemed to be well known in Singapore.

The protection afforded by section 55(3)(a) of the current TMA

155    The Respondents argued that the “Aman” names, being trade marks which were well known in Singapore, were entitled to protection under s 55(3) of the current TMA from the Appellant’s use. Section 55(3) states:

Subject to [ss 55(6) and 55(7)], the proprietor of a well known trade mark shall be entitled to restrain by injunction the use in Singapore, in the course of trade and without the proprietor’s consent, of any trade mark which, or an essential part of which, is identical with or similar to the proprietor’s trade mark, in relation to any goods or services, where the use of the trade mark —

(a)    would indicate a connection between those goods or services and the proprietor, and is likely to damage the interests of the proprietor; or

(b)    if the proprietor’s trade mark is well known to the public at large in Singapore —

(i)    would cause dilution in an unfair manner of the distinctive character of the proprietor’s trade mark; or

(ii)   would take unfair advantage of the distinctive character of the proprietor’s trade mark.

156    As mentioned earlier (see [29] above), the Respondents eventually relied on only s 55(3)(a) of the current TMA in the court below, and the parties’ submissions were largely confined to that subsection. It will be seen, however, that s 55(3)(a) must be construed and applied as one part of the entire scheme relating to the protection of well-known trade marks.

The meaning of “use … in the course of trade … of any trade mark” in section 55(3)

157    There was a preliminary issue as to whether the Appellant’s use of the name “Amanusa” at the Showflat and in various promotional materials relating to the Project was “use … in the course of trade … of any trade mark” (see s 55(3) of the current TMA). Section 2(1) of the current TMA defines a “trade mark” as:

… any sign capable of being represented graphically and which is capable of distinguishing goods or services dealt with or provided in the course of trade by a person from goods or services so dealt with or provided by any other person …

158    The Appellant argued that its use of the name “Amanusa” was not use of that name as a trade mark because the name was not used as “a badge of trade origin” (per Andrew Phang Boon Leong J in Nation Fittings (M) Sdn Bhd v Oystertec Plc [2006] 1 SLR 712 (“the Nation Fittings case”) at [136]).

159    We do not think this submission could have been put forward with much conviction. The Nation Fittings case dealt with trade marks consisting of two-dimensional diagrammatic representations of pipe-fittings, which trade marks were allegedly infringed by real-life three-dimensional pipe-fittings made to those specifications. It was held that the trade marks had not in fact been used as trade marks as there was no evidence that they had been perceived as a badge of trade origin (id at [136]).

160    The situation in the present case is obviously quite different. For one, the trade marks concerned are straightforward names. The Appellant has applied the name “Amanusa” to what one may reasonably call its “goods”– ie, the Project. We appreciate that not all names of residential developments are source-indicators. Many of them are more akin to personal names of people in that they are “identifiers” (ie, each name identifies the particular residential development to which it has been applied) rather than indicators of source. But, it is also possible for the name of a residential development to be seen as an indicator of source, particularly where it is the same as or similar to another name which has been used as an indicator of source. The “Aman” names have clearly been used in the latter sense (ie, as indicators of source) primarily because of the “Aman” prefix which is common to all of these names.

The meaning of “would indicate a connection between those goods or services and the proprietor” in section 55(3)(a)

161    One area of serious dispute between the parties was whether the requirement in s 55(3)(a) of the current TMA of “a connection between [the defendant’s] goods or services and the [plaintiff]” entailed that there had to be a likelihood of confusion that the plaintiff: (a) was the source of the defendant’s goods or services; or (b) had licensed the use of its (ie, the plaintiff’s) well-known trade mark on the defendant’s goods or services; or (c) had endorsed the defendant’s goods or services (the word “confusion” as used in the rest of this judgment should be read as denoting these particular forms of confusion, which, in essence, involve confusion between the plaintiff on the one hand and the defendant’s goods or services on the other). Given our finding in the context of the Respondents’ passing off claim that a likelihood of confusion (in the sense just described) exists (see [93] above), the question is actually academic for the purposes of this appeal. However, as this question – viz, whether s 55(3)(a) incorporates a requirement that there must be a likelihood of confusion (“the confusion requirement”) – touches on a crucial and difficult area of law which requires clarification, we shall nevertheless set out our views on it.

162    The scope of the problem can be appreciated from a brief overview of the somewhat byzantine history of trade mark legislation in Singapore. Section 55 of the current TMA and the other sections therein dealing with well-known trade marks are the product of two stages of legislative development. The first stage consisted of the enactment of the Trade Marks Act 1998 (Act 46 of 1998) (“the 1998 TMA”), which came into effect on 15 January 1999. This was necessary in order for Singapore to comply with Art 16 of the TRIPS Agreement, which was in the main an elaboration of the rights given to well-known trade marks by Art 6bis of the Paris Convention. The provisions of the 1998 TMA were largely Singapore’s own work based on the wording of Art 16 of the TRIPS Agreement, but Singapore also took the opportunity to incorporate s 56(2) of the Trade Marks Act 1994 (c 26) (UK) (“the English Trade Marks Act”), which our parliament enacted as s 55(2) of the 1998 TMA.

163    Amendments were made to the 1998 TMA in 2004 via the Trade Marks (Amendment) Act 2004 (Act 20 of 2004). These amendments, which we shall refer to as “the 2004 TMA amendments”, constituted the second stage of the legislative development which culminated in the enactment of s 55 (and the relevant accompanying provisions) of the current TMA. The impetus for the 2004 TMA amendments came from the signing of the United States-Singapore Free Trade Agreement (“the FTA”). The FTA obliged Singapore to adopt the standards for the protection of well-known trade marks set out in the Joint Recommendation, which standards are even higher than those contained in the TRIPS Agreement (the standards set by the Joint Recommendation are the so-called “TRIPS-plus” standards). The Joint Recommendation is itself what some might call a legal platypus. It is an amalgam of Art 16 of the TRIPS Agreement, the law of dilution in American jurisprudence, as well as the law relating to trade marks with a reputation in the European Union (“the EU”) as set out in the First Directive 89/104/EEC of the Council, of 21 December 1988, to Approximate the Laws of the Member States Relating to Trade Marks (“the EU Directive”) – all of these condensed into a single section, namely, Art 4(1)(b) of the Joint Recommendation.

164    The 2004 TMA amendments are largely based on the wording of the Joint Recommendation, and Singapore remains the only country in the world to have adopted Art 4(1)(b) of the Joint Recommendation in its entirety. Even so, small but important changes were made by our legislature, so the grafting exercise was not perfect. At the same time, Singapore directly imported the definition of “dilution” set out in the US Trademark Act of 1946 (or, as it is commonly known, “the Lanham Act”) – specifically, the definition in s 45 of the Lanham Act (15 USC (US) § 1127), as amended by s 4 of the Federal Trademark Dilution Act of 1995, Pub L No 104-98, 109 Stat 985 (“the FTDA”) – into our local trade mark legislation, even though this definition was not contained in the Joint Recommendation.

165    Trying to decipher the law on well-known trade marks in Singapore can be akin to engaging in a game of Chinese whispers. We propose to adopt a more reliable route by following history’s trail. The story of the protection of well-known trade marks actually has a number of themes, the chief of which is probably the erosion of the territoriality principle of trade mark law. However, that issue is not our concern in the present appeal; instead, our focus will be on the issue of confusion and its cousin, dilution. These issues are related because the removal of the confusion requirement (as defined at [161] above) from s 55(3)(a) of the current TMA would in effect mean that this provision protects a well-known trade mark against dilution. In analysing whether or not s 55(3)(a) of the current TMA should be interpreted thus (ie, as not incorporating the confusion requirement), we shall first examine the international movement towards granting greater protection to well-known trade marks, which culminated in the TRIPS Agreement. An overview of the law of dilution in the US follows. We will then look at developments in Europe arising from the EU Directive. These three legal systems will form the basis for analysing whether the confusion requirement is part and parcel of the regime under the Joint Recommendation. We shall also touch briefly on the position in Australia. Finally, the conclusions reached will be linked back to the current TMA.

(1) International developments

166    Article 6bis of the Paris Convention is the grandfather of provisions on the international protection of well-known trade marks. Well-known trade marks were first mentioned at the Diplomatic Conference for the Revision of the Paris Convention (referred to hereafter as the “Diplomatic Conference” for short) held at The Hague in 1925, but there were already references to the protection of a trade mark in countries other than the mark’s country of origin at the earlier Diplomatic Conference held in Washington in 1911. The essential elements of the scope of protection for well-known trade marks were outlined at The Hague in 1925. Article 6bis, as it stood then, related to trade marks which were identical or similar to a well-known trade mark or, to use the language of the provision, “which constitute[d] a reproduction, an imitation, or a translation” of that mark. The scope of protection was limited to situations in which the trade mark which reproduced, imitated or translated the plaintiff’s well-known trade mark was applied to identical or similar goods (but not services). Importantly, the provision also required that the defendant’s trade mark had to be “liable to create confusion” – ie, Art 6bis expressly incorporated the confusion requirement.

167    Subsequent changes were made to the 1925 version of Art 6bis of the Paris Convention at the Diplomatic Conference held in London in 1934 and the Diplomatic Conference held in Lisbon in 1958, but those changes did not affect the portions of Art 6bis that are material to us in the present appeal. The Diplomatic Conference held in Stockholm in 1967 adopted the version of Art 6bis propounded at Lisbon without any changes.

168    The next major development vis-à-vis the protection of well-known trade marks came in the form of the TRIPS Agreement, which was concluded under the aegis of the World Trade Organization (“WTO”). Under Art 2(1) of the TRIPS Agreement, WTO member states must comply with various provisions of the Paris Convention, including Art 6bis. The TRIPS Agreement, which came into force on 1 January 1996, expanded the scope of the protection provided by Art 6bis of the Paris Convention in a number of ways. First, under Art 16(2) of the TRIPS Agreement, the protection set out in Art 6bis was extended to services.

169    Second, under Art 16(3) of the TRIPS Agreement, Art 6bis of the Paris Convention was to apply mutatis mutandis to situations where the defendant’s goods or services were dissimilar to the plaintiff’s goods or services provided that the defendant’s use of the plaintiff’s trade mark in relation to the defendant’s goods or services would indicate “a connection” (per Art 16(3) of the TRIPS Agreement) between those goods or services and the plaintiff and provided that the interests of the plaintiff would be “likely to be damaged by such use” (ibid). Notably, Art 16(3) of the TRIPS Agreement did not mention the “liable to create confusion” requirement (ie, the confusion requirement) contained in Art 6bis of the Paris Convention.

170    There is thus some debate as to whether Art 16(3) of the TRIPS Agreement did away with the confusion requirement or whether the “connection” and the “likely to be damaged” conditions in that provision were in addition to (rather than in lieu of) the confusion requirement. David Vaver, in “Unconventional and Well-Known Trade Marks” [2005] Sing JLS 1, argues (at 14, n 49) that if confusion were a requirement under Art 16(3) of the TRIPS Agreement, it would have been explicitly stated and the drafters of Art 16(3) would not merely have used the words “indicate a connection”, particularly because the requirement of “a likelihood of confusion” is expressly stated in Art 16(1) of the TRIPS Agreement. In contrast, Ng-Loy Wee Loon, in “The IP Chapter in the US-Singapore Free Trade Agreement” (2004) 16 SAcLJ 42 at 52, takes the alternative view (viz, that the confusion requirement remains part and parcel of Art 16(3) of the TRIPS Agreement), which we agree with. Article 16(3) of the TRIPS Agreement explicitly uses the words “mutatis mutandis”, which suggest that only obvious and necessary changes should be effected to Art 6bis of the Paris Convention. As such, in our view, the “liable to create confusion” requirement in Art 6bis likewise applies vis-à-vis Art 16(3) of the TRIPS Agreement.

171    Indeed, nothing in Art 16(3) of the TRIPS Agreement necessitates the removal of the confusion requirement. If the confusion requirement were not preserved, it is hard to see what is left of Art 6bis. Furthermore, comparisons between Art 16(1) and Art 16(3) of the TRIPS Agreement are inapt. The former is an independent provision which does not seek to refer to Art 6bis in any way.

172    In support of their contention that the confusion requirement was not part and parcel of Art 16(3) of the TRIPS Agreement, the Respondents cited two authorities (apart from Art 16(1) and Art 16(3) of the TRIPS Agreement), but we think neither is clearly supportive of their position. The first is Daniel Gervais, The TRIPS Agreement: Drafting History and Analysis (Sweet & Maxwell, 2nd Ed, 2003), where the learned author comments (at p 174):

Art. 16(3) [of the TRIPS Agreement] protects against ‘dilution’ of a mark, including use detrimental to a mark’s reputation (not necessarily a well-known mark), its quality indication, usually leading to depreciation of the goodwill attached to it, even in cases where there is no likelihood of confusion. [emphasis in original omitted]

But, oddly, Gervais’s definition of “dilution” seems to include an element of confusion for Gervais goes on to say (ibid):

Through the reference to Art. 6bis, Art. 16(3) also protects the public against the confusing effects of such use. [emphasis added]

In so far as Gervais is advocating the view that there are two different tests – one involving confusion and the other, dilution – to be satisfied before the protection under Art 16(3) of the TRIPS Agreement can be invoked, this conflicts with the aim of Art 16(3), which was intended to prescribe a single minimum standard for WTO member states to comply with.

173    The same ambivalence is to be found in a joint publication by the United Nations Conference on Trade and Development and the International Centre for Trade and Sustainable Development, Resource Book on TRIPS and Development (Cambridge University Press, 2005) (“the TRIPS Resource Book”), which the Respondents also referred to. Paragraph 3.2.3 of the TRIPS Resource Book states:

[I]f there is no likelihood of confusion in the ordinary trademark sense, Article 16.3 indicates that the finder of fact should proceed to ask whether a consumer would consider there to be a connection between the goods, even if not part of a natural product line expansion … [emphasis in original]

In our view, the above extract still appears to be referring to a scenario involving confusion, albeit not of the straightforward kind where the consumer mistakenly assumes that the plaintiff itself made, provided or endorsed the defendant’s goods or services (described in the TRIPS Resource Book (at para 3.2.3) as “confusion in the ordinary trademark sense” [emphasis in original]). The TRIPS Resource Book describes the ambit of Art 16(3) of the TRIPS Agreement as covering the situation where (ibid):

… the goods or services for which the confusingly similar trademark is used are different from those goods or services that are covered by the well-known mark.

174    The passage cited by the Respondents also goes on to describe what constitutes a situation with “no likelihood of confusion in the ordinary trademark sense” [emphasis in original omitted] (ibid) as follows (ibid):

Would a consumer seeing the term ‘AUDI’ on a television set think that there was a connection with the automobile company? In recent years there has been an increasing tendency for producers well known in one area of commerce to market [sic] into un-related lines of commerce. Would it have been anticipated, for example, that the ‘Marlboro’ and ‘Camel’ cigarette marks would be used on clothing and shoes?

With respect, all of these examples are nothing more than examples of confusion, albeit arising from the consumer’s mistaken impression that the plaintiff has extended the use of its trade mark to products unrelated to the goods or services in respect of which its trade mark has hitherto been applied, or has licensed such use of its trade mark. There is nothing to suggest that the learned authors of the TRIPS Resource Book were considering examples where there was truly no likelihood of confusion at all but Art 16(3) of the TRIPS Agreement still applied.

175    Indeed, it is arguable that the reference in Art 16(3) of the TRIPS Agreement to “a connection between [the defendant’s] goods or services and the owner of the registered trade mark [ie, the plaintiff]” [emphasis added] means that there must be confusion. How else would the owner of the well-known trade mark (ie, the plaintiff) be connected with the defendant’s goods or services except through the owner being assumed to have either: (a) produced, provided or endorsed the defendant’s goods or services, or (b) licensed the use of its (ie, the owner’s) trade mark on the defendant’s goods or services? It would have been different if Art 16(3) had spoken of a connection between the defendant’s and the plaintiff’s respective trade marks, or between the defendant’s goods or services and the plaintiff’s trade mark. The provision might then have covered dilution in the sense that the plaintiff’s trade mark now represents not just its own goods or services, but the defendant’s goods or services too. A similar point was made in Michael Handler, “Trade Mark Dilution in Australia?” [2007] EIPR 307 (“Handler’s article”) at 309:

The focus of anti-dilution law is on conduct that causes harm to the mark itself: the aim is to protect the mark’s inherent ‘selling power’ as distinct from its ability to guarantee the trade origin of particular goods or services. Thus, not only does anti-dilution law not depend on the existence of consumer confusion, but it does not even look to consumers’ responses as to the existence or otherwise of any connection between the defendant’s goods or services and the owner of the mark in order to determine liability. Yet Art. 16.3 [of the TRIPS Agreement] requires that, as a minimum standard, [WTO] members base liability for [the] defendant’s use of a well-known mark on dissimilar goods or services on a finding of such a connection.

176    Handler also points out that WTO member states do not seem to have interpreted Art 16(3) of the TRIPS Agreement as prescribing a minimum standard – which is not based on confusion – for the protection of well-known trade marks. For instance, Singapore made the confusion requirement explicit, at least in the 1998 TMA (see s 55(2) thereof). Likewise, the Trademark Law of the People’s Republic of China, as adopted on 23 August 1982 and subsequently amended (for the second time) on 27 October 2001 (available at (accessed 14 February 2009)), states at Art 13:

Where a trademark in respect of which the application for registration is filed for use for identical or similar goods is a reproduction, [an] imitation or [a] translation of another person’s trademark not registered in China and likely to cause confusion, it shall be rejected for registration and prohibited from use.

Where a trademark in respect of which the application for registration is filed [for] use for non-identical or dissimilar goods is a reproduction, [an] imitation or [a] translation of the well-known mark of another person that has been registered in China, misleads the [public] and is likely to create prejudice to the interests of the well-known [mark’s] registrant, it shall be rejected for registration and prohibited from use.

[emphasis added]

177    Similarly, the EU Directive makes protection against dilution (as set out in Art 5(2) of that directive) merely an optional measure. In contrast, if Art 16(3) of the TRIPS Agreement indeed did not incorporate the confusion requirement – ie, if Art 16(3) were indeed an anti-dilution provision – then protection against dilution would have been mandatory under the EU Directive. That said, the US has taken the opposite view and considers Art 16(3) to be an anti-dilution provision. We shall discuss the corresponding provisions in the EU Directive and the Joint Recommendation in due course; for now, we shall turn our attention to the position under US law.

(2) The US law of dilution

178    Dilution, as it presently exists in American law, is described in Restatement (Third) of Unfair Competition (1995) (“the Restatement”) at § 25, as follows:

(1)    One may be subject to liability under the law of trademarks for the use of a designation that resembles the trademark, trade name, collective mark, or certification mark of another without proof of a likelihood of confusion only under an applicable antidilution statute. An actor is subject to liability under an antidilution statute if the actor uses such a designation in a manner that is likely to associate the other’s mark with the goods, services, or business of the actor and:

(a)    the other’s mark is highly distinctive and the association of the mark with the actor’s goods, services, or business is likely to cause a reduction in that distinctiveness; or

(b)    the association of the other’s mark with the actor’s goods, services, or business, or the nature of the actor’s use, is likely to disparage the other’s goods, services, or business or tarnish the images associated with the other’s mark.

[emphasis added]

179    The concept of dilution was first articulated by the American academic, Frank I Schechter, in his landmark article, “The Rational Basis of Trademark Protection” (1926–1927) 40 Harv L Rev 813, which was in turn based on the German concept of Verwässerung. Schechter’s innovation was to suggest that the protection of the distinctiveness of a trade mark was “the only rational basis” (at 831) for trade mark protection, and that a trade mark should be protected from “the gradual whittling away or dispersion of [its] identity and hold upon the public mind … by its use upon non-competing goods” (at 825). This concept became known as the “blurring” aspect of dilution (see § 25(1)(a) of the Restatement (reproduced at [178] above)). Crucially, it is not necessary to show confusion for the purposes of establishing dilution by blurring – as distinct from blurring in the context of damage to goodwill under the tort of passing off (as we have seen (at [97] above), damage to goodwill by blurring is made out only if there is a likelihood of confusion). Schechter, in Hearings Before the House Committee on Patents, 72d Cong 15 (1932), gave the following example of dilution by blurring (see Miles J Alexander & Michael K Heilbronner, “Dilution under section 43(c) of the Lanham Act” (1996) 59 Law & Contemp Probs 93 at 98, n 27):

If you allow Rolls Royce restaurants and Rolls Royce cafeterias, and Rolls Royce candy, in ten years you will not have the Rolls Royce marks any more.

In Schechter’s example, although consumers may not think that the other “Rolls Royce”-labelled goods or services really originate from the same source as “Rolls Royce” cars, it may become more difficult for consumers to associate the “Rolls Royce” name with hyper-luxurious cars specifically as the use of that name for non-automobile related goods or services becomes established.

180    Although Schechter only referred to dilution by blurring, it has become established that there is another form of dilution, viz, by tarnishment. There is no need to show confusion either for the purposes of proving dilution by tarnishment. Dilution by tarnishment occurs when a trade mark is used in a negative way which affects the positive associations that consumers had previously made with the mark. The Restatement gives the following rather vivid example in Illustration 3 to § 25(1)(b):

A, a bank, uses the designation ‘Cookie Jar’ to identify its automatic teller machines. B opens a topless bar across the street from A under the trade name ‘Cookie Jar.’ Although prospective customers of A are unlikely to believe that A operates or sponsors the bar, B is subject to liability to A for tarnishment under an applicable antidilution statute if the customers are likely to associate A’s mark or A’s business with the images evoked by B’s use.

In this example, although consumers are not confused into thinking that A (the bank) operates B’s topless bar, whenever they hear the name “Cookie Jar”, they cannot help but think of topless bars because of B’s diluting usage of that name.

181    It was not always the case that both forms of dilution were recognised, and the history of this area of the law of dilution is important in the present case. The first state anti-dilution statute was passed by Massachusetts in 1947. That statute became the basis of the Model State Trademark Bill of 1964, which in turn became the basis of anti-dilution legislation in several more states. However, it was only much later that dilution became recognised at the federal level with the implementation of the FTDA in 1996. The FTDA was introduced to effect amendments to the Lanham Act in an effort to harmonise state dilution laws, which had initially faced substantial suspicion from the courts.

182    It is important to note that the FTDA was intended to cover both blurring and tarnishment as forms of protectable dilution. This can be seen from the House of Representatives’ statement in H R Rep No 104-374 (1995) (“the 1995 House Report”) that the objective of the FTDA was “to protect famous trademarks from subsequent uses that [blurred] the distinctiveness of the mark or tarnish[ed] or disparage[d] it, even in the absence of a likelihood of confusion” (at 2).

183    Section 3 of the FTDA added a new s 43(c) to the Lanham Act (15 USC (US) § 1125) (“the 1996 version of s 43(c) of the Lanham Act”), which provided a cause of action for dilution of famous trade marks. Subsection (1) of the 1996 version of s 43(c) of the Lanham Act read as follows:

The owner of a famous mark shall be entitled, subject to the principles of equity and upon such terms as the court deems reasonable, to an injunction against another person’s commercial use in commerce of a mark or trade name, if such use begins after the mark has become famous and causes dilution of the distinctive quality of the mark, and to obtain such other relief as is provided in this subsection. [emphasis added]

The 1996 version of s 43(c) of the Lanham Act also set out certain criteria for determining whether a trade mark was distinctive or famous. (The 1996 version of s 43(c) of the Lanham Act was subsequently amended in 2006 (see [186] below).)

184    Important for present purposes is the amendment of s 45 of the Lanham Act (15 USC (US) § 1127), via s 4 of the FTDA, to include a definition of “dilution”, as follows:

The term ‘dilution’ means the lessening of the capacity of a famous mark to identify and distinguish goods or services, regardless of the presence or absence of —

(1)    competition between the owner of the famous mark and other parties, or

(2)    likelihood of confusion, mistake, or deception.

(We shall hereafter refer to the amended s 45 as “the 1996 version of s 45 of the Lanham Act”.) The above is, of course, the definition of “dilution” that has been incorporated into the current TMA (see s 2(1) thereof).

185    It should be noted that the 1995 House Report ([182] supra) specifically stated in relation to the new definition of “dilution” that (at 8):

The definition is designed to encompass all forms of dilution recognized by the courts, including dilution by blurring, by tarnishment and disparagement, and by diminishment.

Despite its noble aims, the FTDA engendered even more disputes than it solved. These disputes need not detain us, except for one, viz, that arising from the important US Supreme Court decision of Victor Moseley v V Secret Catalogue, Inc 537 US 418 (2003) (“the Victoria’s Secret case”). The plaintiff in that case, which owned the famous “Victoria’s Secret” trade mark and operated a chain of lingerie stores under that name, alleged that the defendants’ use of the name “Victor’s Little Secret” for the defendants’ sex shop was likely to both “blur and erode the distinctiveness” (at 424) as well as “tarnish the reputation” (ibid) of the plaintiff’s name. This case is tremendously important for all sorts of reasons, but the only point of concern for us for present purposes is the following passage (see the Victoria’s Secret case at 432):

[The defendants] have not disputed the relevance of tarnishment … presumably because that concept was prominent in litigation brought under state antidilution statutes and because it was mentioned in the legislative history. Whether it is actually embraced by the statutory text, however, is another matter. Indeed, the contrast between the state statutes, which expressly refer to both ‘injury to business reputation’ and to ‘dilution of the distinctive quality of a trade name or trademark,’ and the federal statute [ie, the FTDA] which refers only to the latter, arguably supports a narrower reading of the FTDA.

Essentially, the US Supreme Court confined dilution under the FTDA to dilution by blurring only, and declined to interpret the Act as encompassing dilution by tarnishment as well.

186    The legislative reaction to the decision in the Victoria’s Secret case came in the form of the Trademark Dilution Revision Act of 2006, Pub L No 109-312, 120 Stat 1730 (“the TDRA”), which was enacted on 6 October 2006 in order to further amend the Lanham Act. The TDRA cleared up many of the issues raised by the Victoria’s Secret case. In particular, the issue of whether the FTDA had made dilution by tarnishment actionable was resolved. Section 3(e) of the TDRA removed the definition of “dilution” set out in the 1996 version of s 45 of the Lanham Act. At the same time, s 2 of the TDRA amended the 1996 version of s 43(c) of the Lanham Act so as to make it clear that injunctive relief was available in cases of “dilution by blurring or dilution by tarnishment” [emphasis added] (see the new s 43(c)(1) of the Lanham Act, as enacted by s 2 of the TDRA). Further, the term “dilution by tarnishment” was defined in the new s 43(c)(2)(C) of the Lanham Act (as enacted by s 2 of the TDRA) as “association arising from the similarity between a mark or trade name and a famous mark that harm[ed] the reputation of the famous mark”.

187    The importance of the preceding discussion on the US law of dilution lies in the fact that Singapore has imported elements of this area of US law both indirectly through the adoption of (inter alia) Art 4(1)(b) of the Joint Recommendation (which has been translated into s 55(3) of the current TMA) and directly through the incorporation in s 2(1) of the current TMA of the definition of “dilution” set out in the 1996 version of s 45 of the Lanham Act. However, our importation of the US law of dilution took place in 2004, before the TDRA was enacted and under the shadow of the US Supreme Court’s view in the Victoria’s Secret case that dilution under the FTDA included dilution by blurring only and not dilution by tarnishment. As the amendments effected by the TDRA show, Congress had in fact intended for dilution to include both forms, just as the 1995 House Report ([182] supra) had stated. We shall return to this point in due course.

(3) The EU Directive

188    European trade mark law is marked by two major developments – the adoption of the EU Directive and the introduction of the Community Trade Mark (“CTM”) regime. The CTM regime was conceived as a parallel, supranational system of trade mark registration that would exist alongside national trade mark systems, which were to be preserved. However, it was recognised that there was also a need for some degree of compatibility and uniformity among the various trade mark regimes of EU member states. Harmonisation efforts were focused on substantive issues such as the rights of a trade mark holder and the conditions for obtaining and holding a trade mark. To this end, the EU Directive was adopted on 21 December 1988.

189    The EU Directive protects a trade mark with a reputation in an EU member state (ie, the plaintiff’s trade mark) against the use of a mark or sign (ie, the defendant’s trade mark) on dissimilar goods or services where “use of [the defendant’s trade mark] without due cause takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the [plaintiff’s] trade mark” (see Art 5(2) of the EU Directive). The scope of protection extends to the refusal of registration or the invalidation of the defendant’s trade mark (see Art 4(4)(a) of the EU Directive), and to the grant of an injunction against the use of the defendant’s trade mark (see Art 5(2) of the EU Directive). The EU Directive also gives the plaintiff’s well-known trade mark protection similar to that provided by Art 6bis of the Paris Convention in that, under Art 4(1)(b) of the EU Directive, the defendant’s trade mark is not to be registered or, if registered, is liable to be declared invalid if:

… because of its identity with, or similarity to, the [plaintiff’s] trade mark and the identity or similarity of the goods or services covered by the [plaintiff’s and the defendant’s respective] trade marks, there exists a likelihood of confusion on the part of the public, which includes the likelihood of association with the [plaintiff’s] trade mark.

Except where the defendant’s trade mark is identical to the plaintiff’s trade mark and also covers goods or services identical to the plaintiff’s goods or services (see Art 4(1)(a) of the EU Directive), “a likelihood of confusion on the part of the public, which includes the likelihood of association with the [plaintiff’s] trade mark” (see Art 4(1)(b) of the EU Directive) must be shown. In the UK, Art 4(4)(a) and Art 5(2) of the EU Directive have been translated into s 5(3) and s 10(3) respectively of the English Trade Marks Act.

190    The unusual structure and syntax of the EU Directive raises a number of interesting issues. Of concern to us in the present appeal is the meaning of the phrase “without due cause takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the [plaintiff’s] trade mark” in Art 5(2) (a similar phrase, albeit worded in a slightly different manner, also appears in Art 4(4)(a)). The phrase “without due cause” in s 10(3) of the English Trade Marks Act (ie, the English equivalent of Art 5(2) of the EU Directive) was interpreted by the English High Court in Premier Brands UK Ltd v Typhoon Europe Ltd [2000] FSR 767 (“Premier Brands”). There, Neuberger J rejected the defendant’s submission that the phrase meant “[not] in good faith” (at 789) or “[not] for good and honest commercial reasons” (at 789–790) because such a conclusion could have unfair and capricious results, depending on the subjective intention or even unconscious desires of the defendant. Instead, he preferred an interpretation which required the defendant to show that its use of its trade mark was “with due cause” (at 791). Neuberger J also adopted the Benelux Court of Justice’s interpretation in Lucas Bols v Colgate-Palmolive (1976) 7 IIC 420 (“Lucas Bols”) of the phrase “without justifiable reason”, which appears in the Benelux equivalent of s 10(3) of the English Trade Marks Act. In Lucas Bols, it was held that there would be “justifiable reason” for the defendant to use its trade mark if (see Premier Brands at 791):

… the [defendant] is under such a compulsion to use this very mark that [it] cannot honestly be asked to refrain from doing so regardless of the damage the owner of the mark would suffer from such use, or … the [defendant] is entitled to the use of the mark in [its] own right and does not have to yield this right to that of the owner of the mark …

191    Notably for present purposes, Neuberger J observed in Premier Brands that the phrase “without due cause” might seem superfluous in the context of the stipulation in s 10(3) of the English Trade Marks Act that the use of the defendant’s trade mark must also “[take] unfair advantage” of the plaintiff’s reputable trade mark. He concluded, however (at 791–792 of Premier Brands):

[I]t is fair to say that it is not easy to see how the use of a sign could take ‘unfair advantage’ of a registered mark and yet be not ‘without due cause’. In my judgment, however, there are two answers to that point. First, it is conceivable that, in certain circumstances, the court might conclude that, notwithstanding the fact that the alleged infringer’s use of a sign took ‘unfair advantage’ of a mark, this was outweighed by the use being with ‘due cause’, within the limited meaning of that expression as explained in Lucas Bols [1976] 7 I.I.C. 420. Secondly, the words ‘being without due cause’ apply not only to a case of alleged unfair advantage, but also to a case where the use of the sign is allegedly ‘detrimental’ to the mark: there is no difficulty in envisaging circumstances where a ‘detrimental’ use could not be ‘without due cause’ within the sense I have suggested. [emphasis in original]

192    The trickier limb of Art 5(2) of the EU Directive consists of the phrase “takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the [plaintiff’s] trade mark”. In Adidas-Salomon, Adidas Benelux v Fitnessworld Trading Ltd [2003] ETMR 91 (“Adidas-Salomon (AG)”), Advocate General Jacobs (“Jacobs AG”) noted in his opinion (at [36]):

Article 5(2) protects the proprietor of a mark with a reputation against use of an identical or [a] similar sign where use of that sign without due cause takes unfair advantage of, or is detrimental to, the distinctive character or the repute of the trade mark. There are thus in principle four types of use which may be caught: use which takes unfair advantage of the mark’s distinctive character, use which takes unfair advantage of [the mark’s] repute, use which is detrimental to the mark’s distinctive character and use which is detrimental to [the mark’s] repute.

193    Jacobs AG opined that the concept of detriment to the distinctive character of a trade mark reflected dilution in the form of blurring (id at [37]), while the concept of detriment to the repute of a trade mark corresponded to dilution in the form of tarnishment (id at [38]). In contrast, the concept of taking unfair advantage of the distinctive character of a trade mark and that of taking unfair advantage of the repute of a trade mark encompassed “instances where there [was] clear exploitation and free-riding on the coat-tails of a famous mark or an attempt to trade upon its reputation” (id at [39]). Jacobs AG also made a specific reference (ibid) to the explanation given in the Explanatory Notes of Art 4(1)(b)(iii) of the Joint Recommendation, which we shall refer to later (see [210] below).

194    On reference of the case to the European Court of Justice (“the ECJ”), the court held, in Adidas-Salomon AG, Adidas Benelux BV v Fitnessworld Trading Ltd [2004] ETMR 10 (“Adidas-Salomon (ECJ)”), that proof that the public was likely to confuse the plaintiff’s well-known trade mark with the defendant’s trade mark was not necessary for actions under Art 5(2) of the EU Directive. The ECJ characterised the scope of that provision as follows (at [29]):

The infringements referred to in Art. 5(2) of the [EU] Directive, where they occur, are the consequence of a certain degree of similarity between the [plaintiff’s] mark and the [defendant’s] sign, by virtue of which the relevant section of the public makes a connection between the [defendant’s] sign and the [plaintiff’s] mark, that is to say, establishes a link between them even though it does not confuse them …

195    The fact that there is no requirement of a likelihood of the public confusing the plaintiff’s trade mark with the defendant’s trade mark vis-à-vis Art 5(2) of the EU Directive is balanced against the relatively strict standards which apply to the other components of this provision. The concept of a link between the plaintiff’s trade mark and the defendant’s trade mark as enunciated in Adidas-Salomon (ECJ) at [29] (“the ‘link’ requirement in Adidas-Salomon (ECJ)”) was recently interpreted by Jacob LJ in the English Court of Appeal case of Intel Corp Inc v CPM United Kingdom Ltd [2007] ETMR 59 (“the English Intel Corp case”) in relation to Art 4(4)(a) of the EU Directive, which contains a phrase similar to that in Art 5(2) (namely, “without due cause would take unfair advantage of, or be detrimental to, the distinctive character or the repute of the [plaintiff’s] trade mark”). Jacob LJ commented at [30] that “a mere calling to mind” of the plaintiff’s reputable trade mark was insufficient to establish the requisite connection. Instead, such connection was established only in the following situations (ibid):

[T]he facts … go beyond a mere calling to mind. They may cause the consumer to think [that] there is a trade connection between the owner of the [reputable] mark [ie, the plaintiff] and the user of the later [allegedly conflicting] mark [ie, the defendant] or at least wonder whether there is such a connection. Clearly the former case ought to be protected, and perhaps the latter too if the wonder is substantial as opposed to fleeting.

Jacob LJ also posited that speculation by the public about “some sort of licensing arrangement” (see the English Intel Corp case at [31]) between the plaintiff and the defendant might suffice for the purposes of establishing the requisite trade connection.

196    The sort of strictness advocated by Jacob LJ has not been seen in other jurisdictions. For instance, the Benelux courts have been quite happy to accept “mere association” (see Seiko Hidaka et al, “A Sign of the Times? A Review of Key Trade Mark Decisions of the European Court of Justice and Their Impact Upon National Trade Mark Jurisprudence in the EU” (2004) 94 Trademark Rep 1105 at 1142) as sufficient. Even more pertinent in this regard is the decision of the ECJ on the question which the English Court of Appeal in the English Intel Corp case referred to it, viz, the question of what would suffice to satisfy the “link” requirement in Adidas-Salomon (ECJ). In Intel Corporation Inc v CPM United Kingdom Ltd Case C-252/07, 27 November 2008 (unreported) (“the European Intel Corp case”), the ECJ held (contra Jacob LJ’s view in the English Intel Corp case) that this “link” requirement would be satisfied by the mere calling to mind of the plaintiff’s trade mark on the part of “the average consumer, who [was] reasonably well informed and reasonably observant and circumspect” (see the European Intel Corp case at [60] and [63]). It should be noted, however, that the ECJ’s more lenient approach to the “link” requirement in Adidas-Salomon (ECJ) was counterbalanced by the ECJ’s holding that the plaintiff, apart from showing that that requirement had been satisfied, also had to prove “actual and present injury to its mark … or a serious likelihood that such an injury [would] occur in the future” (see the European Intel Corp case at [71]). With regard to damage in the form of, specifically, “[detriment] to … the distinctive character … of the [plaintiff’s] trade mark” (per Art 4(4)(a) of the EU Directive), the ECJ stated that such detriment had to be proved by “evidence of a change in the economic behaviour of the average consumer of the [plaintiff’s] goods or services … or a serious likelihood that such a change [would] occur in the future” (see the European Intel Corp case at [77]).

197    Jacob LJ also interpreted the term “unfair” (in the context of the term “unfair advantage” in Art 5(2) of the EU Directive) in L’Oreal SA v Bellure NV [2008] ETMR 1 (“the English L’Oreal case”). He opined (id at [91]) that evidence of “free-riding” by the defendant per seviz, of the defendant capitalising on the distinctive character or repute of the plaintiff’s trade mark to market and promote its (ie, the defendant’s) own goods or services (id at [88]) – was not sufficient to prove that unfair advantage had been taken of the plaintiff’s well-known trade mark. Instead, there had to be evidence of “harm, present or prospective, caused to the [plaintiff’s trade] mark … [or] its distinctive character or to the [plaintiff] or [its] business, present or reasonably prospective” (id at [94]). Blackburne J, on the other hand, appeared to take the view that “free-riding” could in certain cases constitute an “unfair advantage” for the purposes of Art 5(2) of the EU Directive even if there was no “harm, present or prospective” (see the English L’Oreal case at [94]) as mentioned by Jacob LJ. Blackburne J commented (id at [166]):

[W]here [the defendant] is able to derive a commercial advantage through the ‘wink’ which [its] product makes at the [plaintiff’s] registered mark but no harm, present or prospective, can be shown to that mark … [or] its distinctive character, or to the [plaintiff] or [its] business … such conduct, assuming that it gives to the [defendant] a commercial advantage, should be treated as ‘unfair’.

The English L’Oreal case was likewise referred by the English Court of Appeal to the ECJ for a ruling on the meaning of “unfair advantage” in Art 5(2) of the EU Directive. In Advocate General Mengozzi’s non-binding opinion on this matter (see L’Oréal SA, Lancôme parfums et beauté & Cie and Laboratoire Garnier & Cie v Bellure NV, Malaika Investments Ltd and Starion International Ltd Case C-487/07, 10 February 2009 (unreported)), it was stated that whether or not the advantage taken of the plaintiff’s well-known trade mark was “unfair” was “for the national court to verify … taking … all the relevant circumstances of the case into account” (at [110]). The ECJ’s decision is expected later this year.

(4) The Joint Recommendation

198    The Joint Recommendation was the result of further developments following the TRIPS Agreement. Work was initially done by WIPO’s Committee of Experts on Well-known Marks (“the Committee of Experts”), and later continued by the Standing Committee on the Law of Trademarks, Industrial Designs and Geographical Indications (“the SCT”). Various draft proposals were presented over the course of six meetings of the Committee of Experts and the SCT. The proposals by the SCT were eventually adopted at a joint session of the Assembly of the Paris Union for the Protection of Industrial Property and the General Assembly of WIPO held at the 34th WIPO Meeting (see [137] above). The Joint Recommendation does not have any binding force and its provisions are merely guidelines for WIPO member states to consider.

199    Article 2 of the Joint Recommendation delineates the type of trade marks which are deemed to be sufficiently well known to enjoy the protection set out in the Joint Recommendation. We have considered this provision earlier in our discussion (at [134][153] above) of what constitutes a well-known trade mark in the context of the current TMA.

200    Article 3(1) of the Joint Recommendation requires each WIPO member state to “protect a well-known mark against conflicting marks, business identifiers and domain names, at least with effect from the time when the mark has become well known in the [m]ember [s]tate”. In this regard, the defendant’s trade mark is deemed to be a “conflicting” mark in the various scenarios set out in Art 4(1) of the Joint Recommendation. Under Art 4(1)(a), the defendant’s trade mark :

… shall be deemed to be in conflict with a well-known mark [ie, the plaintiff’s mark] where [the defendant’s] mark, or an essential part thereof, constitutes a reproduction, an imitation, a translation, or a transliteration, liable to create confusion, of the well-known mark … if the [defendant’s] mark, or an essential part thereof, is used, is the subject of an application for registration, or is registered … in respect of goods and/or services which are identical or similar to the goods and/or services to which the well-known mark applies.

201    The more difficult provision as to when the defendant’s trade mark constitutes a “conflicting” mark is Art 4(1)(b) of the Joint Recommendation. This provision applies regardless of the nature of the goods or services on which the defendant’s trade mark is used; ie, it is immaterial whether or not those goods or services are identical or similar to the plaintiff’s goods or services. Article 4(1)(b) sets out three different situations in which the defendant’s trade mark is deemed to be in conflict with the plaintiff’s trade mark. The first, which is to be found in Art 4(1)(b)(i), is based on Art 16(3) of the TRIPS Agreement. It states that the defendant’s trade mark will be deemed to be a “conflicting” mark where the use of that mark would indicate a “connection” (see Art 4(1)(b)(i) of the Joint Recommendation) between the defendant’s goods or services and the plaintiff, and would also be “likely to damage” (ibid) the plaintiff’s interest.

202    The second situation in which the defendant’s trade mark is deemed to be a “conflicting” mark under Art 4(1)(b) of the Joint Recommendation is that delineated in Art 4(1)(b)(ii), which appears to be drawn from both the American law of dilution as well as Art 4(4)(a) and Art 5(2) of the EU Directive. Article 4(1)(b)(ii) of the Joint Recommendation states that the defendant’s trade mark will be deemed to be in conflict with the plaintiff’s trade mark if the use of the defendant’s mark “is likely to impair or dilute in an unfair manner the distinctive character of the [plaintiff’s] well-known mark” (cf the phrase “detrimental to … the distinctive character … of the [plaintiff’s] trade mark” in Art 4(4)(a) and Art 5(2) of the EU Directive).

203    Finally, Art 4(1)(b)(iii) of the Joint Recommendation covers the situation where “the use of [the defendant’s] mark would take unfair advantage of the distinctive character of the [plaintiff’s] well-known mark”. The use of the phrase “distinctive character” in both this provision and Art 4(1)(b)(ii) of the Joint Recommendation seems to have been inspired by the wording used in Art 4(4)(a) and Art 5(2) of the EU Directive. The correlation is particularly strong in the case of Art 4(1)(b)(iii) of the Joint Recommendation because this provision also imports the concept of “unfair advantage” from Art 4(4)(a) and Art 5(2) of the EU Directive.

204    Article 4(1)(c) of the Joint Recommendation is important as well. It states that a WIPO member state may require the plaintiff’s well-known trade mark to be “well known by the public at large” before Art 4(1)(b)(ii) and Art 4(1)(b)(iii) are applicable.

205    The first thing to note is that Art 4(1)(b) of the Joint Recommendation does not mention that there must be a likelihood of confusion before the defendant’s trade mark will be deemed to be in conflict with the plaintiff’s trade mark, unlike Art 4(1)(a), which expressly incorporates the confusion requirement (see the words “liable to create confusion” in Art 4(1)(a)). It may thus be thought that the confusion requirement is absent from Art 4(1)(b) generally and Art 4(1)(b)(i) in particular. We disagree, however, with this interpretation for four reasons.

206    First, as we have seen (at [201] above), Art 4(1)(b)(i) of the Joint Recommendation is derived directly from Art 16(3) of the TRIPS Agreement. It has been shown earlier (at [170][177] above) that Art 16(3) of the TRIPS Agreement did not dispense with the confusion requirement contained in its originating provision, Art 6bis of the Paris Convention.

207    Second, the commentary in the Explanatory Notes on Art 4(1)(b)(i) of the Joint Recommendation indicates that the drafters of the Joint Recommendation considered the likelihood of confusion to be a requirement. The Explanatory Notes state (at para 4.3):

Under [Art 4(1)(b)(i)], a connection between a well-known mark and a third party’s goods or services may be indicated, for example, if the impression is created that the owner of the well-known mark is involved in the production of those goods, or the offering of those services, or that such production or offering was licensed or sponsored by him. The interests of the owner of the well-known mark could be damaged if the goods and/or services with which the connection is established have a down-market image, thereby reflecting negatively on the goodwill of the well-known mark. [emphasis added]

208    Burton Ong, in “Protecting Well-known Trade Marks: Perspectives from Singapore” (2005) 95 TMR 1221 (“Ong’s article”), disagrees with this view. Instead, he suggests (at 1248) that Art 4(1)(b)(i) of the Joint Recommendation is meant to “extend protection for well-known trade marks to all situations where damage to the [plaintiff] arises from any connection made by the public between the [plaintiff] and the goods or services of the unauthorised user of the mark” [emphasis added in bold italics]. He further suggests that the two sentences in para 4.3 of the Explanatory Notes actually delineate two different instances in which the plaintiff’s well-known trade mark is protected, namely (see Ong’s article at 1249):

The first sentence makes it clear that the [plaintiff’s] well-known mark is protected from origin-related confusion … However, this is merely one example of the type of misconduct that [Art 4(1)(b)(i)] is seeking to address. The second sentence expands the scope of [Art 4(1)(b)(i)] to include situations where the unauthorised application of the [plaintiff’s] mark on down-market goods might have an adverse impact on the attractive force of the well-known mark itself, indeed suggesting that instances of tarnishment [which, as stated at [180] above, does not require proof of confusion] fall within this category of misconduct as well.

With respect, this is, in our view, an incorrect reading of Art 4(1)(b)(i) of the Joint Recommendation. One recalls that Art 4(1)(b)(i), besides requiring a “connection” between the plaintiff and the defendant’s goods or services (in essence, confusion) to be shown, also requires the connection to be “likely to damage [the plaintiff’s] interests”. The second sentence of para 4.3 of the Explanatory Notes merely makes the point that confusion could result in tarnishment as a form of damage; that sentence does not suggest that confusion and tarnishment delineate two different scenarios in which the defendant’s trade mark is deemed to be a “conflicting” mark for the purposes of Art 4(1)(b)(i).

209    Third, it is crucial to note that the “connection” required in Art 4(1)(b)(i) of the Joint Recommendation is that between the defendant’s good or services and the plaintiff itself. This is distinguishable from the connection or link required under Art 5(2) of the EU Directive, which was interpreted in Adidas-Salomon (ECJ) ([194] supra) at [29] as “a connection between the [defendant’s] sign and the [plaintiff’s] mark” [emphasis added] – ie, the connection required under Art 5(2) of the EU Directive is that between two trade marks (viz, the plaintiff’s trade mark and the defendant’s trade mark), although such connection need not amount to confusion. In respect of the connection between the defendant’s goods or services and the plaintiff (ie, the connection required under Art 4(1)(b)(i) of the Joint Recommendation), Handler’s article ([175] supra) notes (at 314):

This condition cannot be satisfied simply by looking to whether consumers might make a conceptual link between the two marks. A quite different type of assessment is needed. To use an example from US dilution law, consumers who see ‘The Greatest Snow on Earth’ on licence plates in the context of an organisation’s promotion of winter sports tourism might well call to mind Ringling Bros circus’ well-known mark THE GREATEST SHOW ON EARTH, and thus form a mental association or connection between the two marks. However, this does not address the question … [of] whether such consumers would make a connection between the tourism promotion services and Ringling Bros. Whether this sort of connection exists will depend on factors going beyond the similarity of the two marks and the degree of fame of the registered mark. [emphasis added]

210    Finally, our view that the likelihood of confusion is an essential element of Art 4(1)(b)(i) of the Joint Recommendation is supported by an examination of this particular provision in the wider context of all three situations in which the defendant’s trade mark is deemed to be a “conflicting” mark under Art 4(1)(b). It appears from the reference in Art 4(1)(b)(ii) to “the distinctive character of the well-known mark” that this provision is grafted from Art 4(4)(a) and Art 5(2) of the EU Directive, which (as we have seen at [194] above) do not require a likelihood of confusion between the plaintiff’s trade mark and the defendant’s trade mark to be shown. More importantly, as noted earlier (at [193] above), Jacobs AG regarded the words “detrimental to … the distinctive character … of the trade mark” in Art 5(2) of the EU Directive (and, presumably, the corresponding words in Art 4(4)(a) thereof as well) as encompassing the concept of dilution by blurring (see Adidas-Salomon (AG) ([192] supra) at [37]). This indicates that Art 4(1)(b)(ii) of the Joint Recommendation is an explicit anti-dilution provision, and anti-dilution legislation, as we have seen from our earlier analysis of US law, does not require confusion to be proved before the trade mark in question can be protected (see [179] above). It appears from the wording of Art 4(1)(b)(ii) of the Joint Recommendation, read together with Art 4(1)(b)(iii), that the “[detriment] to … the distinctive character … of the trade mark” limbs of Art 4(4)(a) and Art 5(2) of the EU Directive (which, as just mentioned, were regarded by Jacobs AG in Adidas-Salomon (AG) as providing protection against dilution by blurring) were removed and transplanted specifically onto Art 4(1)(b)(ii) of the Joint Recommendation. What essentially remains in Art 4(1)(b)(iii) is “free-riding”. The commentary on Art 4(1)(b)(iii) in the Explanatory Notes (at para 4.5) confirms that proof of confusion is not required for the purposes of this provision:

The case referred to in [Art 4(1)(b)(iii) differs from the cases covered by [Art 4(1)(b)(i)] and [Art 4(1)(b)(ii)] in that no wrong connection concerning the real source of the goods and/or services is indicated (as in [Art 4(1)(b)(i)]), and the value of the well-known mark has not diminished in the eyes of the public (as in [Art 4(1)(b)(ii)]), but rather the use in question would, for example, amount to a free ride on the goodwill of the well-known mark for the person who uses a conflicting mark. [emphasis added]

211    In short, both Art 4(1)(b)(ii) and Art 4(1)(b)(iii) of the Joint Recommendation do not require proof of confusion before the defendant’s trade mark is deemed to be in conflict with the plaintiff’s trade mark. If Art 4(1)(b)(i) similarly does not require confusion to be shown in this regard, it is very likely to swallow up Art 4(1)(b)(ii) and Art 4(1)(b)(iii). Article 4(1)(b)(i), if interpreted in this manner, would essentially incorporate the “link” requirement in Adidas-Salomon (ECJ), which applies to both dilution by blurring and “free-riding”. However, the presence of Art 4(1)(b)(ii) and Art 4(1)(b)(iii) in Art 4(1) of the Joint Recommendation suggests that Art 4(1)(b)(i) must serve an entirely different purpose from that of providing protection against dilution by blurring (which is covered by Art 4(1)(b)(ii)) and protection against “free-riding” (which is covered by Art 4(1)(b)(iii)). This interpretation is reinforced by Art 4(1)(c) of the Joint Recommendation, which demonstrates the recognition that Art 4(1)(b)(ii) and Art 4(1)(b)(iii) confer a higher level of protection than that provided by Art 4(1)(b)(i) and may thus apply only to commensurately more well-known trade marks.

212    We conclude from the foregoing that Art 4(1)(b)(i) of the Joint Recommendation requires the likelihood of confusion to be shown before the defendant’s trade mark will be deemed to be in conflict with the plaintiff’s trade mark, unlike Art 4(1)(b)(ii) and Art 4(1)(b)(iii).

(5) The current TMA

213    As mentioned earlier (at [162] above), our law on well-known trade marks has undergone two distinct stages of development. Protection for well-known trade marks was first introduced into our law with the advent of the 1998 TMA. Under the 1998 TMA, all registered trade marks were protected from the registration and the use of identical trade marks which covered identical goods or services (see, respectively, s 8(1) and s 27(1) of the 1998 TMA). The protection afforded by s 8(1) and s 27(1) of the 1998 TMA was not dependent on proof of confusion, in accordance with Art 16(1) of the TRIPS Agreement (under Art 16(1) of the TRIPS Agreement, “[i]n [the] case of the use of an identical sign for identical goods or services, a likelihood of confusion shall be presumed” [emphasis added]). In contrast, “a likelihood of confusion on the part of the public” (per s 8(2) and s 27(2) of the 1998 TMA) had to be shown where the defendant’s trade mark:

(a)     was identical to the plaintiff’s trade mark and covered goods or services which were similar to the plaintiff’s goods or services (see s 8(2)(a) and s 27(2)(a) of the 1998 TMA); or

(b)     was similar to the plaintiff’s trade mark and covered goods or services which were identical or similar to the plaintiff’s goods or services (see s 8(2)(b) and s 27(2)(b) of the 1998 TMA).

For trade marks which were well known in Singapore, additional protection was granted under s 8(3) and s 27(3) of the 1998 TMA even where the defendant’s goods or services were dissimilar, provided that the following conditions were fulfilled:

(a)     the use of the defendant’s trade mark in relation to the defendant’s goods or services would indicate a connection between those goods or services and the plaintiff (see s 8(3)(ii) and s 27(3)(b) of the 1998 TMA);

(b)     there was a likelihood of confusion on the part of the public because of such use (see s 8(3)(iii) and s 27(3)(c) of the 1998 TMA); and

(c)     the plaintiff’s interests in its well-known trade mark were likely to be damaged by such use (see s 8(3)(iv) and s 27(3)(d) of the 1998 TMA).

214    In addition, under s 55(2) read with s 55(1) of the 1998 TMA, where the plaintiff owned a trade mark which was entitled to protection under either the Paris Convention or the TRIPS Agreement as a well-known trade mark, the plaintiff could obtain an injunction to restrain the defendant from using an identical or a similar trade mark on identical or similar goods or services if such use was “likely to cause confusion” (see s 55(2)), regardless of whether or not the plaintiff carried on business or had any goodwill in Singapore.

215    The 2004 TMA amendments were brought about by the FTA, which obliged both the US and Singapore to implement the Joint Recommendation (see [163] above). Section 8(3) of the 1998 TMA vis-à-vis protection from the registration of identical or similar trade marks for goods or services dissimilar to those of the plaintiff was altered, and now, under s 8(4) of the current TMA, where the defendant applies on or after 1 July 2004 (viz, the date on which the 2004 TMA amendments came into effect) to register a trade mark which is identical or similar to the plaintiff’s well-known trade mark, the defendant’s trade mark – regardless of whether the goods or services which it covers are similar or dissimilar to the plaintiff’s goods or services – will be refused registration only if:

(a)     use of the defendant’s trade mark would indicate a connection between the goods or services covered by that trade mark and the plaintiff, and is likely to damage the interests of the plaintiff (see s 8(4)(b)(i) of the current TMA); or

(b)     the plaintiff’s trade mark is well known to the public at large in Singapore (and not merely “well known in Singapore” (per s 8(4)(a)), and use of the defendant’s trade mark would either “cause dilution in an unfair manner of the distinctive character of the [plaintiff’s] trade mark” (see s 8(4)(b)(ii)(A) of the current TMA) or “take unfair advantage of the distinctive character of the [plaintiff’s] trade mark” (see s 8(4)(b)(ii)(B) of the current TMA).

For ease of discussion, we shall hereafter refer to the conditions set out in s 8(4)(b)(i), s 8(4)(b)(ii)(A) and s 8(4)(b)(ii)(B) as, respectively, “the ‘damaging connection’ condition”, “the ‘unfair dilution’ condition” and “the ‘unfair advantage’ condition”.

216    With regard to the international protection of well-known trade marks, the 2004 TMA amendments did not make any changes to the substance of the protection provided by s 55(2) of the 1998 TMA (see s 55(2) of the current TMA). However, a new subsection, viz, s 55(3) of the current TMA, was added. Section 55(3) of the current TMA allows the owner of a well-known trade mark (ie, the plaintiff) to obtain an injunction restraining the defendant from using an identical or a similar trade mark on (inter alia) dissimilar goods or services if the same conditions as those which we have just outlined vis-à-vis s 8(4) of the current TMA (see [215] above) – ie, the “damaging connection” condition (see s 55(3)(a), which corresponds to s 8(4)(b)(i)), the “unfair dilution” condition (see s 55(3)(b)(i), which corresponds to s 8(4)(b)(ii)(A)) and the “unfair advantage” condition (see s 55(3)(b)(ii), which corresponds to s 8(4)(b)(ii)(B)) – are satisfied. These amendments were intended, of course, to bring our trade mark legislation in line with Art 4(1)(b) of the Joint Recommendation.

217    In addition to the above amendments, a definition of “dilution” was added. Section 2(1) of the current TMA now states that:

‘dilution’, in relation to a trade mark, means the lessening of the capacity of the trade mark to identify and distinguish goods or services, regardless of whether there is —

(a)    any competition between the proprietor of the trade mark and any other party; or

(b)    any likelihood of confusion on the part of the public …

This definition, as noted earlier (see [184] above), is identical to that which used to be contained in the 1996 version of s 45 of the Lanham Act.

218    The similarity between the provisions in the current TMA and those in the Joint Recommendation means, naturally, that the arguments as to why the confusion requirement is part and parcel of Art 4(1)(b)(i) of the Joint Recommendation (see [205][211] above) also support the same conclusion in relation to the “damaging connection” condition in s 8(4)(b)(i) and s 55(3)(a) of the current TMA – ie, confusion must likewise be proved before these two provisions can be invoked. We shall now go on to address other arguments peculiar to the current TMA.

219    In Ong’s article ([208] supra), it is argued that the confusion requirement is absent from s 8(4) and s 55(3) of the current TMA (which protect the plaintiff’s trade mark from, respectively, the registration and the use of an identical or a similar trade mark for (inter alia) dissimilar goods or services). The author further suggests that the ambit of the protection provided by s 8(4) and s 55(3) can be delineated in the following manner:

(a)     the “damaging connection” condition in s 8(4)(b)(i) and s 55(3)(a) is directed at protection against tarnishment, including tarnishment caused by confusion and dilution by tarnishment (see Ong’s article at 1242–1244 and 1247–1249);

(b)     the “unfair dilution” condition in s 8(4)(b)(ii)(A) and s 55(3)(b)(i) is directed at protection against dilution by blurring only (id at 1244–1245 and 1251–1252); and

(c)     the scope of the protection provided by the “unfair advantage” condition in s 8(4)(b)(ii)(B) and s 55(3)(b)(ii) is to be derived from UK cases on s 5(3) and s 10(3) of the English Trade Marks Act (id at 1255–1256).

220    Ong argues that the protection covered by the “damaging connection” condition includes dilution by tarnishment because the “connection” referred to in s 8(4)(b)(i) and s 55(3)(a) of the current TMA does not incorporate the likelihood of confusion. We have dealt with this argument earlier (at [205][211] above) in the context of Art 4(1)(b)(i) of the Joint Recommendation, and the views which we expressed there likewise apply here.

221    With regard to the protection covered by the “unfair dilution” condition, Ong argues that it is limited to protection from dilution by blurring for two reasons. First, the definition of “dilution” in s 2(1) of the current TMA, ie, “the lessening of the capacity of the trade mark to identify and distinguish goods or services”, refers only to dilution by blurring and not dilution by tarnishment. From a literal reading of s 2(1), this is true. But, as we have seen (at [185][186] above) from the history of protection against dilution under the Lanham Act, Congress did not intend the term “dilution” to be so narrowly understood. Ong’s view mirrors that of the US Supreme Court in the Victoria’s Secret case ([185] supra), in which the US Supreme Court doubted that the FTDA had made dilution by tarnishment actionable. The TDRA put paid to that view and clearly indicated the true intention of Congress – ie, that dilution under the FTDA covered both dilution by blurring and dilution by tarnishment. It is ironic that Singapore, in its enthusiastic attempts to synchronise its trade mark law with American trade mark law pursuant to its obligations under the FTA, obligingly adopted the 1996 version of s 45 of the Lanham Act wholesale (by incorporating the definition of “dilution” therein into s 2(1) of the current TMA) only to find out that the US itself jettisoned that provision merely two short years later. We think that s 9A of the Interpretation Act (Cap 1, 2002 Rev Ed) requires us to take a purposive rather than literal interpretation of the definition of “dilution” in s 2(1) of the current TMA until such time when Parliament may see fit to effect the necessary amendments.

222    The second reason put forth by Ong as to why the “unfair dilution” condition provides protection from dilution by blurring only and not protection from dilution by tarnishment is that s 8(4)(b)(ii)(A) and s 55(3)(b)(i) of the current TMA mention only “dilution”, whereas their counterpart, Art 4(1)(b)(ii) of the Joint Recommendation, speaks of both dilution and impairment of the distinctive character of a well-known trade mark. Ong argues that the word “impair” in Art 4(1)(b)(ii) must refer to dilution by tarnishment and the omission of this word (ie, “impair”) from our definition of “dilution” in s 2(1) of the current TMA must have significance. Again, this is literally true. However, this interpretation of Art 4(1)(b)(ii) of the Joint Recommendation does not accord with the stated intentions of the drafters of that provision. The Explanatory Notes (at para 4.4) comment on Art 4(1)(b)(ii) thus:

This item would apply, for example, if the use of a conflicting mark is likely to impair or dilute in an unfair manner the unique position of a well-known mark in the market. A further example of dilution is where the conflicting mark is used on goods or services which are of an inferior quality or of an immoral or obscene nature.

223    The first sentence of para 4.4 of the Explanatory Notes treats the terms “dilute” and “impair” as synonymous; both are taken to indicate dilution by blurring. The second sentence of that paragraph uses “dilution” to indicate dilution by tarnishment. Apparently, in the drafters’ minds, there was no material difference between dilution and impairment.

224    Perhaps, it was this lack of any substantial difference between the two concepts (viz, dilution and impairment) which led to the omission of the words “impair” and “impairment” from our trade mark legislation. This view is certainly supported by the Minister for Law’s statement during the second reading of the Trade Marks (Amendment) Bill 2004, where he said (see Singapore Parliamentary Debates, Official Report (15 June 2004) vol 78 at col 111):

[T]here can be situations where the reputation and value of the well-known mark is diluted by blurring or tarnishment, even though there may not be consumer confusion, for example, if someone uses a well-known mark on inferior quality goods or on goods, say, of an obscene nature. [emphasis added]

225    Both the passage quoted at [222] above and that quoted at [224] above weigh against the proposition that any significance should be placed on the omission of the words “impair” and “impairment” from the current TMA. “Dilution” in the context of the current TMA clearly refers to both dilution by blurring and dilution by tarnishment.

226    From this perspective, it is difficult for Ong’s three-point categorisation of the protection afforded under the current TMA (as set out at [219] above) to stand. The “damaging connection” condition in s 8(4)(b)(i) and s 55(3)(a) of the current TMA will be satisfied only where there is a confusing connection between the plaintiff and the defendant’s goods or services (ie, only where there is confusion) because non-confusing connection is covered by the “unfair dilution” condition in s 8(4)(b)(ii)(A) and s 55(3)(b)(i). The “unfair advantage” condition in s 8(4)(b)(ii)(B) and s 55(3)(b)(ii) sets out an entirely separate category of protection against “free-riding” uses. This conclusion is supported by three other considerations.

227    First, it is intuitively strange for dilution in its two forms (viz, blurring and tarnishment) to be split between two different limbs of the current TMA (namely, on the one hand, s 8(4)(b)(i) and s 55(3)(a) vis-à-vis dilution by tarnishment and, on the other hand, s 8(4)(b)(ii)(A) and s 55(3)(b)(i) vis-à-vis dilution by blurring), each of which applies to different kinds of trade marks. If Ong is right, in respect of a trade mark which is merely a “well known trade mark” as defined in s 2(1) of the current TMA, an action for dilution by tarnishment can be brought (in view of s 8(4)(b)(i) and s 55(3)(a)), but not an action for dilution by blurring because the latter is limited to a trade mark which is “well known to the public at large in Singapore” (see s 8(4)(b)(ii) and s 55(3)(b)). This very unconventional bifurcation of actions for dilution has no precedent elsewhere.

228    Second, s 27(3) of the 1998 TMA, which described the circumstances in which the use by the defendant of its trade mark (being a trade mark that was identical or similar to the plaintiff’s well-known trade mark) on goods or services dissimilar from those of the plaintiff would constitute infringement of the plaintiff’s trade mark, was not amended in 2004. The conditions under which infringement is made out, as stated in s 27(3) of the current TMA, remain substantially the same as the conditions which previously applied under s 27(3) of the 1998 TMA, ie, there has to be a “connection” (see s 27(3)(b) of the current TMA) between the plaintiff and the defendant’s goods or services arising from the use of the defendant’s trade mark, “a likelihood of confusion” (see s 27(3)(c) of the current TMA) on the part of the public because of such use and damage to the plaintiff’s interests as a result of such use (see s 27(3)(d) of the current TMA). It might be argued that the express mention of the confusion requirement in s 27(3)(c) of the current TMA makes the lack of such mention in s 8(4)(b)(i) and s 55(3)(a) thereof all the more significant, but, for practical reasons, we think the converse is true. Specifically, if the confusion requirement were not read into s 8(4)(b)(i) and s 55(3)(a) of the current TMA, the following situation would pertain: The plaintiff which relies on s 27(3) of the current TMA to sue for infringement of its well-known trade mark must show a likelihood of confusion on the part of the public due to the use of the defendant’s trade mark on dissimilar goods or services, but, if the plaintiff seeks to stop the registration of the defendant’s trade mark based on s 8(4)(b)(i) or to restrain (by way of injunction) the use of the defendant’s trade mark under s 55(3)(a), it need not show a likelihood of confusion. This situation is anomalous. There is no reason why the confusion requirement should not apply across the board to all these sections.

229    Finally, it will be recalled that it is not too difficult for a trade mark to be regarded as “well known in Singapore” – essentially, the trade mark in question need only be recognised or known by “any relevant sector of the public in Singapore” [emphasis added] (see s 2(7)(a) of the current TMA), which sector could in certain cases be miniscule. In recognition of the fact that many trade marks are potentially “well known in Singapore”, Parliament has granted such trade marks only one advantage over ordinary trade marks, namely, the former are protected from the registration and/or the use of identical or similar trade marks on dissimilar goods or services – such protection takes the form of the protection covered by the “damaging connection” condition in s 8(4)(b)(i) and s 55(3)(a) of the current TMA. The protection covered by the “unfair dilution” condition in s 8(4)(b)(ii)(A) and s 55(3)(b)(i) and that covered by the “unfair advantage” condition in s 8(4)(b)(ii)(B) and s 55(3)(b)(ii) are reserved for trade marks which are “well known to the public at large in Singapore” (per s 8(4)(b)(ii) and s 55(3)(b)). The policy question in issue is the extent of the protection which we wish to give to trade marks which are merely “well known in Singapore” as opposed to “well known to the public at large in Singapore”. In the absence of a clear legislative intent, we do not think that trade marks which are merely “well known in Singapore” should be given protection against the use of a similar or an identical mark on dissimilar goods or services where such use does not give rise to confusion. Such protection (ie, protection despite the absence of confusion) should, for now, properly be the preserve of a rare and privileged few. In this regard, it is interesting to note that s 2 of the TDRA also amended the 1996 version of s 43(c) of the Lanham Act by adding a definition of a “famous” trade mark. Under s 43(c)(2)(A) of the Lanham Act (as amended by s 2 of the TDRA), a “famous” trade mark is one which is “widely recognized by the general consuming public of the United States”. This amendment was made specifically to eliminate the concept of “niche fame”, which was recognised in some earlier cases (see, for instance, Wedgwood Homes, Inc and Wedgwood Homes of Portland v Victor L Lund 294 Or 493; 659 P 2d 377 (1983)). Thus, in the US, both dilution by blurring and dilution by tarnishment are now actionable only vis-à-vis trade marks which are well-known on a national level. This development reflects Prof J Thomas McCarthy’s earlier observation in McCarthy on Trademarks and Unfair Competition (West Pub Co, 4th Ed, 1996) at § 24:108:

To save the dilution doctrine from abuse by plaintiffs whose marks are not famous, a large neon sign should be placed adjacent wherever the doctrine resides, reading: ‘The Dilution Rule: Only Strong Marks Need Apply.’ [emphasis in bold in original]

(6) The position in Australia

230    We wish also, for completeness, to make a passing reference to developments in Australia. Section 120(3) of the Trade Marks Act 1995 (Cth) (“the 1995 Australian TMA”) states:

A person infringes a registered trade mark if:

(a)    the trade mark is well known in Australia; and

(b)    the person uses as a trade mark a sign that is substantially identical with, or deceptively similar to, the trade mark in relation to:

(i)    goods (unrelated goods) that are not of the same description as that of the goods in respect of which the trade mark is registered (registered goods) or are not closely related to services in respect of which the trade mark is registered (registered services); or

(ii)   services (unrelated services) that are not of the same description as that of the registered services or are not closely related to registered goods; and

(c)    because the trade mark is well known, the sign would be likely to be taken as indicating a connection between the unrelated goods or services and the registered owner of the trade mark; and

(d)    for that reason, the interests of the registered owner are likely to be adversely affected.

[emphasis in bold italics in original]

231    Section 120(3) of the 1995 Australian TMA is based on Art 16(3) of the TRIPS Agreement. Like s 55(3)(a) of the current TMA, s 120(3) of the 1995 Australian TMA does not expressly incorporate the confusion requirement. The similarities between these two provisions do not end there. Reminiscent of the different approaches taken by our legislature to the protection of well-known trade marks prior to and after the 2004 TMA amendments, the Australian parliament had earlier passed the Trade Marks Act 1994 (No 156 of 1994) (Cth) (“the 1994 Australian TMA”). (That Act, although passed by the Australian legislature, never eventually came into effect, and was instead repealed by and replaced with the 1995 Australian TMA.) Section 129(3) of the 1994 Australian TMA (the then equivalent of what is now s 120(3) of the 1995 Australian TMA) required a likelihood that the defendant’s dissimilar goods or services might be “mistaken as goods or services dealt with or provided by the [plaintiff]” (see s 129(3)(c)(i) of the 1994 Australian TMA) before the defendant’s “identical … or deceptively similar” (per s 129(3)(b) of the 1994 Australian TMA) trade mark would be treated as having infringed the plaintiff’s registered well-known trade mark, but the 1995 Australian TMA omitted this requirement. This has led to some debate as to whether the confusion requirement is implicit in s 120(3) of the 1995 Australian TMA.

232    There has been support, on many of the grounds relied on by the Respondents in the present case, for the proposition that s 120(3) of the 1995 Australian TMA means exactly what it says, ie, that this provision does not incorporate the confusion requirement (see Maurice Gonsalves & Patrick Flynn, “Dilution Down Under: The Protection of Well-Known Trade Marks in Australia” [2006] EIPR 174). There is also some dicta in the Australian High Court case of Campomar Sociedad, Limitada v Nike International Limited (2000) 202 CLR 45 which suggest that s 120(3) of the 1995 Australian TMA should be considered as an anti-dilution provision (ie, as not requiring proof of confusion). However, it has been argued powerfully in Handler’s article ([175] supra) that the opposite must be true. Apart from highlighting the historical reasons rooted in Art 6bis of the Paris Convention and Art 16(3) of the TRIPS Agreement (along the lines which we discussed earlier), Handler’s article also observes that the wording of s 120(3) of the 1995 Australian TMA and the structure of trade mark protection under Australian law as a whole suggest that confusion is a requirement under s 120(3). These factors are largely idiosyncratic to the 1995 Australian TMA and do not assist us in interpreting s 55(3)(a) of the current TMA. Further, and crucially, unlike Singapore, Australia has not incorporated into its legislation Art 4(1)(b)(ii) and Art 4(1)(b)(iii) of the Joint Recommendation, which address, respectively, the unfair dilution and the taking of unfair advantage of the distinctive character of the plaintiff’s well-known trade mark. Nevertheless, despite these differences between Australia’s s 120(3) and Singapore’s s 55(3)(a), it would be of interest to see how the Australian courts definitively resolve the issue of whether confusion is required under the former.

(7) Our view on whether confusion is required under section 55(3)(a) of the current TMA

233    Parliament has given proprietors of “well known trade mark[s]” (as defined in s 2(1) of the current TMA) the right to have their trade marks protected in Singapore even if those trade marks are not registered here. Furthermore, it is relatively easy for trade marks (whether or not registered in Singapore) to be considered “well known in Singapore” so as to be entitled to the protection set out in the current TMA. In our view, the widespread availability of protection to trade marks which are “well known in Singapore” should be balanced by interpreting s 55(3)(a) of the current TMA as requiring a likelihood of confusion to be shown before an injunction to restrain the use of the defendant’s trade mark on (inter alia) goods or services dissimilar to those of the plaintiff can be obtained. A second, much more extensive level of protection is granted to trade marks which have attained the coveted status of being “well known to the public at large in Singapore”. These trade marks, which form a rare and exclusive class, are entitled to protection from use of the defendant’s trade mark on dissimilar goods or services even in the absence of a likelihood of confusion; that is, such trade marks are entitled to protection against the unfair dilution and the taking of unfair advantage of their distinctive character.

Our decision on the Respondents’ claim for trade mark infringement

234    In the instant case, we agree with the Judge’s finding (at [74] of the Judgment) that the tests to be adopted for the purposes of the “connection” requirement and the “likely to damage the [plaintiff’s] interests” requirement in s 55(3)(a) of the current TMA would yield the same results as those obtained from applying the corresponding tests vis-à-vis the Respondents’ claim for passing off (which tests are, respectively, whether the Appellant has made a misrepresentation to the relevant sector of the public (ie, the persons delineated at [66] above) which causes that section of the public to mistakenly think that the Project has the same source as or is connected with the Respondents’ resorts, and whether such misrepresentation has resulted in or is likely to result in damage to the goodwill attached to the “Aman” names). We thus uphold the Judge’s decision that the claim under s 55(3)(a) of the current TMA has been made out. However, we pause to clarify that while the tests relating to misrepresentation and damage under the law of passing off are substantively the same as the tests relating to, respectively, the “connection” requirement and the “likely to damage the [plaintiff’s] interests” requirement in s 55(3)(a) of the current TMA, the two sets of tests are not identical in one aspect. An important distinction is that the tests relating to misrepresentation and damage in passing off actions concern the plaintiff’s goodwill (see [69] above), whereas the corresponding tests under s 55(3)(a) of the current TMA concern the interests of the plaintiff and not its goodwill (indeed, as the definition of “well known trade mark” in s 2(1) of the current TMA makes clear, the plaintiff’s trade mark may be a “well known trade mark” for the purposes of s 55(3)(a) even if the plaintiff has no goodwill in Singapore).

Conclusion

235    The Respondents succeeded in both their passing off claim and their claim under s 55(3)(a) of the current TMA in the court below, and we see no reason to vary the Judge’s orders, which were properly limited to a more modest scope than what was initially prayed for (see [4] and [75] of the Judgment). This appeal is therefore dismissed with costs and the usual consequential orders.

Reported by Tan Kai Liang.