AM Moolla Group Ltd and Others v Gap Inc and Others (123/2004) [2005] ZASCA 72 (9 September 2005)

THE SUPREME COURT OF APPEAL

OF SOUTH AFRICA

Reportable

Case No 123/2004

In the matter between:

A M MOOLLA GROUP LIMITED First Appellant

A M MOOLLA CLOTHING

(PROPRIETARY) LIMITED Second Appellant

SALT OF THE EARTH CREATIONS

(PROPRIETARY) LIMITED Third Appellant

KINGSGATE CLOTHING MANUFACTURERS

(PROPRIETARY) LIMITED Fourth Appellant

MAJESTIC CLOTHING MANUFACTURERS

(PROPRIETARY) LIMITED Fifth Appellant

STAR SHIRT & CLOTHING

(PROPRIETARY) LIMITED Sixth Appellant

RICKEMP (PROPRIETARY) LIMITED Seventh Appellant

and

THE GAP INC First Respondent

GAP (APPAREL) INC Second Respondent

GAP (ITM) INC Third Respondent

Coram: HARMS, STREICHER, BRAND, LEWIS AND PONNAN JJA

Heard: 15 AUGUST 2005

Delivered: 9 SEPTEMBER 2005

Subject: Trade mark “Gap” ─ well-known trade marks ─ art 6bis Paris Convention ─ principle of territoriality ─ non-use of trade marks.

J U D G M E N T

HARMS JA:

INTRODUCTION

[1] This judgment concerns a number of registered South African trade marks, all having the word GAP as a distinctive feature. The first is a word mark THE GAP; there are two device marks consisting of GAP and THE GAP (in lower case) written in a stylised form; and the word marks GAP STORES and GAP KIDS. Salt of the Earth Creations (Pty) Ltd, the third appellant, to whom I shall refer as ‘the proprietor’, is a local company and the registered proprietor of these marks. The respondents, a holding company called The Gap Inc and two of its subsidiaries, are companies registered in the state of Delaware in the United States of America. (There is no need in this judgment to distinguish between them and for the sake of convenience I intend to refer to each and all of them as ‘the respondent’.)

[2] The respondent claims proprietorship of these marks on the ground that GAP is a well-known trade mark that is entitled, without registration in this country, to protection under s 35(3) of the Trade Marks Act 194 of 1993, which incorporates the provisions of art 6bis of the Paris Convention into our law. Daniels J, in the court below, upheld this claim and granted an interdict against the seven appellants, some of the members of a ‘group’ of companies and enterprises that call themselves the AM Moolla Group (not to be confused with the first appellant, AM Moolla Group Ltd). They were enjoined from using a reproduction or imitation of the respondent’s GAP marks in relation to clothing or retail clothing outlets).

[3] The respondent also applied, successfully, to have the proprietor’s trade mark registrations expunged on the ground of non-use. Further subsidiary grounds for expungement based on s 10(3) and s 9(1) read with s 10(12) were raised without success, and this gave rise to a cross-appeal but only the s 10(3) case was persisted in during the hearing. In response to the application for expungement the proprietor in turn lodged a counter-application for an interdict for trade mark infringement, which was dismissed in the light of the expungement. This dismissal of the counter-application is another issue in the appeal. The present proceedings are with the leave of the court of first instance.

[4] Previously some members of the Group sought to utilise the provisions of the Counterfeit Goods Act 37 of 1997 to prevent the respondent from transhipping clothing bearing a GAP trade mark through South Africa but the attempt failed. The declaratory order issued in favour of the respondent, which declared that the respondent did not transgress the provisions of this Act by transhipment, was upheld in a case with the same name as the instant case.¹

[5] The story of Gap began in 1969 when the first Gap store, under the name THE GAP, was opened in San Francisco. The novel concept was the provision of a speciality retail outlet for clothing to exploit the ‘generation gap’ symbolised by the increased popularity of casual clothing such as jeans. The business expanded and on 29 February 1972 a predecessor of the respondent applied for the registration in the USA of THE GAP as a service mark for ‘retail clothing store services’ in class 42 of the international Nice classification. The year 1974 saw the beginning of the use of GAP in relation to goods and an application for the registration of THE GAP (and design) as a trade mark in class 25 (which relates to clothing). Other GAP trade mark applications followed, both in class 25 and 42, beginning in 1977. As the business expanded, trade mark applications were filed in other jurisdictions, for instance, in the United Kingdom since 1983 and in Canada since 1985. GAP stores were opened in the United Kingdom in 1987, and by 1991 there were some 1000 stores in the USA, Canada and the United Kingdom. By 1992 the GAP trade mark had become one of the largest selling apparel brands in the world measured by unit sales, and it is now the second largest. At present hundreds of GAP trade marks are registered in countries alphabetically and otherwise as far apart as Albania and Zimbabwe.

[6] Prior to the 1990s, the respondent showed little interest in the local market. The explanation given was that because of the political climate in this country, as an American company it was under a number of financial, political and legislative constraints that precluded expansion into South Africa. However, if one considers that in 1991 there were GAP stores in only three countries, and since then the respondent’s attempts to enforce its rights locally have been at less than a pedestrian pace, the explanation rings somewhat hollow.

[7] Turning then to the Gap tale in this country: During 1971, at a time when the respondent had 25 THE GAP retail outlets in six states in the USA, and had not yet used a GAP trade mark in respect of clothing, one Hirsch ‘coined’ the trade mark GAP and began using it locally on clothing, and on 21 March 1973 he applied for the registration of the trade mark GAP in class 25 in relation to articles of clothing, including footwear (TM 73/1378). This registration pre-dates any of the respondent’s trade marks in class 25, even in the USA, and there is nothing on record to suggest that Hirsch had copied or derived his inspiration from the respondent’s use of the GAP name on its stores. During the 1970s, the proprietor, to whom Hirsch in the meantime had assigned this mark, manufactured and sold substantial numbers of jeans under the mark. A further GAP mark (‘the first device mark’ TM 80/5548), which consists of the word mark in a stylised form, was registered during 1980.

[8] During 1983 the Group obtained control over the proprietor. Capitalising on the fact that South Africa was subject to trade sanctions and that art 6bis protection did not then exist, the proprietor changed its 1973 GAP trade mark to THE GAP, the corporate and trading name of the respondent. In 1988, the proprietor applied for registration of a second THE GAP (TM 88/4994) device mark, a blatant copy of respondent’s logo; it also applied for GAP KIDS (TM 88/8783), an appropriation of respondents’ name for its children’s stores, first begun in 1985; and in 1989 it applied for GAP STORES (TM 89/5087) in class 42. These applications were all granted. Another application, in respect of GAP STAR, was filed in 1993 but is pending because of opposition by the respondent.

THE PRINCIPLE OF TERRITORIALITY

[9] At the outset I wish to deal with the principle of territoriality applicable to trade marks. Nicholas AJA explained the principle in Victoria’s Secrets² in these terms:

‘In the Moorgate judgment Mr Trollip³ stated that

“. . . a trade mark is purely a territorial concept; it is legally operative or effective only within the territory in which it is used and for which it is to be registered. Hence, the proprietorship, actual use, or proposed use of a trade mark mentioned in s 20(1) [of the Trade Marks Act 62 of 1963] are all premised by the subsection to be within the RSA.”

It follows that the fact that a trade mark is registered and has been used, even extensively used, by one person in a foreign country does not in itself constitute a bar to its adoption and registration by some other person in South Africa.’

Nicholas AJA continued:

‘In the case of a foreign trade mark, there is no legal bar to its adoption in South Africa unless it is attended by something more. Thus in delivering the unanimous judgment of the Full Court in P Lorillard and Co (supra at 356G-H),⁴ Boshoff J said:

“The basis of the challenge on this ground is that the objector was to the knowledge of the applicant the proprietor of such a trade mark in the United States of America and that the applicant improperly appropriated the mark. In the present state of the law a trade mark is a purely territorial concept and there is, generally speaking, nothing to prevent a person from asserting a proprietary right in a trade mark in relation to which no one else has in the same territory asserted a similar right.”’

Nicholas AJA also cited a number of judgments from foreign jurisdictions to show that the principle is not peculiar to this country but is generally accepted.⁵

[10] More recently, in the Barcelona.com case,⁶ a US Federal District Court of Appeals dealt with the same underlying principle (per Niemeyer, Circuit Judge):

‘The relevant substantive provision in this case is Article 6(3) of the Paris Convention, which implements the doctrine of territoriality by providing that “[a] mark duly registered in a country of the [Paris] Union shall be regarded as independent of marks registered in the other countries of the Union, including the country of origin.” . . . As one distinguished commentary explains,

“the Paris Convention creates nothing that even remotely resembles a ‘world mark’ or an ‘international registration’. Rather, it recognizes the principle of the territoriality of trademarks [in the sense that] a mark exists only under the laws of each sovereign nation.”

J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 29:25 (4th ed. 2002).

It follows from incorporation of the doctrine of territoriality into United States law through Section 44 of the Lanham Act that United States courts do not entertain actions seeking to enforce trademark rights that exist only under foreign law. See Person’s Co., Ltd. v. Christman, 900 F.2d 1565, 1568-69 (Fed. Cir. 1990) (“The concept of territoriality is basic to trademark law; trademark rights exist in each country solely according to that country’s statutory scheme”).’

[11] Accordingly, the legality and propriety of Hirsch’s adoption of the first two trade mark registrations is beyond dispute. Although one may query the business morality of the adoption by slavish imitation of the respondent’s trade marks after the Group’s take-over, it was also legally in order in the absence of ‘something more’, ie, ‘factors that may have vitiated or tainted his right or title to the proprietorship thereof’.⁷ The only factor raised in this regard is the allegation that the respondents’ GAP trade marks are well known in South Africa, which brings me to the next issue, namely the protection of well-known trade marks.

PROTECTION OF WELL-KNOWN TRADE MARKS

[12] Special legislative protection is given to well-known trade marks. There are two possible situations. The one relates to the protection of a well-known registered mark against dilution as contained in s 34(1)(c) and which was the subject of the recent Laugh It Off judgments of this court⁸ and the Constitutional Court.⁹ The other, which is the one applicable in this case, relates to the statutory protection based on art 6bis of the Paris Convention for the Protection of Industrial Property (1883), which provides for the protection of well-known trade marks that have not been registered (and if registered, have not been used and therefore otherwise subject to expungement)¹⁰ that belong to ‘qualified’ enterprises. Article 6bis was first introduced with The Hague revision (1925) and its present form, which was adopted during the Lisbon revision (1958), provides as follows:¹¹

‘(1)The countries of the Union undertake, ex officio if their legislation so permits, or at the request of an interested party, to refuse or to cancel the registration, and to prohibit the use, of a trademark which constitutes a reproduction, an imitation, or a translation, liable to create confusion, of a mark considered by the competent authority of the country of registration or use to be well known in that country as being already the mark of a person entitled to the benefits of this Convention and used for identical or similar goods. These provisions shall also apply when the essential part of the mark constitutes a reproduction of any such well–known mark or an imitation liable to create confusion therewith.

(2)A period of at least five years from the date of registration shall be allowed for requesting the cancellation of such a mark. The countries of the Union may provide for a period within which the prohibition of use must be requested.

(3)No time limit shall be fixed for requesting the cancellation or the prohibition of the use of marks registered or used in bad faith.’

[13] South Africa is not only party to the Paris Convention but also to the GATT ‘Agreement on Trade-Related Aspects of Intellectual Property Rights’ (the so-called TRIPS agreement), which requires compliance with a number of the provisions of the Paris Convention, including art 6bis (art 2.1). In addition, TRIPS ‘supplements’ art 6bis: whereas the art 6bis protection is limited to trade marks relating to goods, TRIPS envisages the protection of service marks (art 16.2); and while art 6bis prohibits the use of a well-known mark on ‘identical or similar’ goods, TRIPS (art 16.3) requires wider safeguards, namely, in relation –

‘to goods or services which are not similar to those in respect of which a trademark is registered, provided that use of that trademark in relation to those goods or services would indicate a connection between those goods or services and the owner of the registered trademark and provided that the interests of the owner of the registered trademark are likely to be damaged by such use.’

[14] In anticipation of the adoption of TRIPS, art 6bis was introduced as part of our statutory law in terms of sections 35 and 36(2) of the current Act:

‘35. Protection of well-known marks under Paris Convention.—

(1) References in this Act to a trade mark which is entitled to protection under the Paris Convention as a well-known trade mark, are to a mark which is well known in the Republic as being the mark of—

(a) a person who is a national of a convention country; or

(b) a person who is domiciled in, or has a real and effective industrial or commercial establishment in, a convention country,

whether or not such person carries on business, or has any goodwill, in the Republic.

(1A) In determining for the purposes of subsection (1) whether a trade mark is well-known in the Republic, due regard shall be given to the knowledge of the trade mark in the relevant sector of the public, including knowledge which has been obtained as a result of the promotion of the trade mark.¹²

(2) A reference in this Act to the proprietor of such a mark shall be construed accordingly.

(3) The proprietor of a trade mark which is entitled to protection under the Paris Convention as a well-known trade mark is entitled to restrain the use in the Republic of a trade mark which constitutes, or the essential part of which constitutes, a reproduction, imitation or translation of the well-known trade mark in relation to goods or services which are identical or similar to the goods or services in respect of which the trade mark is well known and where the use is likely to cause deception or confusion.

36. Saving of vested rights.—(1) . . .

(2) Nothing in this Act shall allow the proprietor of a trade mark entitled to protection of such trade mark under the Paris Convention as a well-known trade mark, to interfere with or restrain the use by any person of a trade mark which constitutes, or the essential parts of which constitute, a reproduction, imitation or translation of the well-known trade mark in relation to goods or services in respect of which that person or a predecessor in title of his has made continuous and bona fide use of the trade mark from a date anterior to 31 August 1991 or the date on which the trade mark of the proprietor has become entitled, in the Republic, to protection under the Paris Convention, whichever is the later, or to object (on such use being proved) to the trade mark of that person being registered in relation to those goods or services under section 14.’

[15] The protection conferred on well-known trade marks by provisions such as these differs from ordinary statutory trade mark protection primarily because they provide an exception to the principle of territoriality as discussed above. In addition, they protect by way of trade mark legislation also unregistered marks¹³ although similar protection is available at common law under the lex Aquilia, more particularly, under the wrong of passing off, which also protects the local reputation of a foreign undertaking, whether or not it trades locally.¹⁴

[16] I now turn to consider whether the respondent has made out a case for this kind of protection. Starting at the beginning, the introductory phrase of s 35(3), namely, ‘the proprietor of a trade mark which is entitled to protection under the Paris Convention’ requires that, in order to be able to obtain the protection provided for in s 35(3), the foreign owner has to be a ‘qualified’ person in terms of s 35(1). It is common cause that since the respondent has a ‘real and effective industrial or commercial establishment in a convention country’ (the USA is a member of the Paris Convention and has been so declared by the President in terms of s 63 of the Act),¹⁵ this requirement has been met. Another requirement is that the claimant has to be the ‘proprietor’ of the relevant well-known trade mark in its home territory, something not disputed.

[17] The appellants submitted that the respondent’s claim based on s 35(3) was flawed because the section grants protection to a well-known mark against unregistered marks and not against marks registered locally. The existing registrations per se, according to the argument, constitute an absolute bar to s 35(3) relief. During argument we were presented with a microscopic analysis of the term ‘trade mark’ as used by the Legislature in different sections of the Act – does it refer to registered or unregistered marks or both when used in s 35 and 36(2)? I do not intend to deal with the argument in any detail because it flounders on a simple point. The object of introducing these provisions onto our statute book was to comply with treaty obligations under the Paris Convention and under TRIPS.¹⁶ Article 6bis explicitly requires of countries to protect well-known marks against registered and unregistered marks. There is no apparent reason why the Legislature would have wished to provide otherwise. Since the term ‘trade mark’ is ambiguous in that it can refer to marks both registered and unregistered, the meaning consistent with the underlying legislative intention has to be adopted. If one has to label this method of interpretation, it can either be an application of the ‘soewereine’ rule of interpretation of Dr L C Steyn,¹⁷ namely a determination of the intention of the Legislature, or the ‘purposive construction’ of Lord Diplock, or even Lord Steyn’s ‘context is everything’.

[18] The question then is whether the respondent’s GAP marks are (or were) well known in South Africa and, if so, since when. This raises a number of interrelated questions which were identified during the course of E M Grosskopf JA’s judgment in McDonald’s.¹⁸

1. The first is the identification of the sector of the population ‘interested in the goods or services to which the mark relates’.¹⁹ The respondent’s case was that this ‘universe’ consists of individuals aged between 16 and 50 and living in an A+ income suburb, the reasoning being that such persons were likely to have travelled overseas and would have encountered the GAP marks abroad.

2. Next is whether the mark is well known within the local jurisdiction as a trade mark belonging to an enterprise with a base in another country (although the knowledge does not have to include the fact that the country is a convention country).²⁰

3. The last issue is the determination of whether those who have the requisite knowledge represent a substantial number of the chosen universe.²¹

[19] Reverting to issue (b), the UK hearing officer in the Swizzels Matlow Limited trade mark application (the Polo trade mark),²² relying on McDonald’s gave this useful exposition:

‘As I have indicated above the basic purpose behind Article 6bis would appear to me to be to protect trade marks which are well-known in a country but are not used there. So in the normal course of events there is a presumption that the party seeking protection under Article 6bis has some form of base in a foreign country and in the normal course of events the consumer would be aware of this owing to the absence of the use of the trade mark in the country in which the claim is being [asserted]. Article 6bis states “well known in that country as being already the mark of a person entitled to the benefits of this Convention”. It seems to me that this must be read as requiring the trade mark to be well-known as the trade mark of a person with a base in a country outside the United Kingdom i.e. that the consumer must identify some overseas base with the trade mark.²³ It would however be silly to expect that the consumer should also know that the country is a signatory to the Convention.’

[20] What the respondent primarily sought to establish was that GAP was a well-known trade mark at the time the present proceedings were initiated in the court below. To this end a market survey was conducted using as a basis the said ‘universe’ of individuals. Of these, 48 per cent recognised the GAP logo and they overwhelmingly associated it with casual clothing. A quarter of the universe thought that the retail outlets were local, 11 per cent thought they were international and the rest did not know. As far as manufacturing origin is concerned, 29 per cent believed GAP goods were manufactured overseas (mostly in the USA), 12 per cent thought they were of a local origin and 11 per cent did not know. (The percentages are approximations.)

[21] I have serious doubts that this evidence establishes any one of the three points mentioned. However, since there is some evidence, as the court below noted, which might suggest that the GAP mark was well known to persons in the trade as that of a foreign concern since about 1988 (although that was not the case the appellants were called to meet) I shall assume in favour of the respondent that its GAP marks were well known in South Africa as at 31 August 1991. That is the date referred to in s 36(2) of the Act and was the date on which the bill that led to the current Act was published. It is also the earliest date on which art 6bis protection became available. Section 36(2) by implication distinguishes between two cases: (i) trade marks that were already well known at that date are entitled to protection subject to the recognition of certain prior rights; (ii) trade marks that become well known after that date are entitled to protection as soon as they become so known. There is little doubt that as at 1991 the proprietor as trade mark owner was not making use of any of its trade marks on a ‘continuous and bona fide’ manner as required by s 36(2) and accordingly was not entitled to the recognition of existing rights by virtue of this provision. Although it may have licensed other parties to use some of them, these licences were not registered user agreements as was required by s 48 of the then existing Trade Marks Act 62 of 1963, all matters to which I shall revert in another context.

[22] That does not dispose of the matter in the respondent’s favour. In order to be entitled to the protection granted by s 35(3), the objectionable trade mark has to constitute ‘a reproduction, imitation or translation of the well-known trade mark’ of a ‘qualified’ party. In other words, unless the ‘foreign’ trade mark was well known at the time when the local enterprise reproduced, imitated or translated it, the foreign trade mark is in the light of the principle of territoriality not entitled to art 6bis protection.²⁴

[23] This conclusion conforms to the judgment in Person’s.²⁵ The facts bear a striking resemblance to the facts in this case. In 1977, X first applied a stylized logo bearing the name PERSON'S to clothing in his native Japan. Two years later he formed a company, P, to market and distribute the clothing items in retail stores located in Japan. In 1981, one C visited one of P’s retail stores while on a trip to Japan and purchased clothing items bearing the PERSON’S logo. He then developed designs for his own PERSON’S brand sportswear line based on P’s products and in 1982 he began producing clothing with the PERSON’S logo. All C’s goods bore either the mark PERSON’S or a copy of P's globe logo and many of the clothing styles were apparently copied directly from P’s designs. In 1983, C filed an application for a US trademark registration of the PERSON’S mark. Some seven months subsequent to C’s first sales P entered the US market. P sought the expungement of C’s registration. The matter was first heard by a board and the judgment next quoted was the decision on appeal.

[24] The court said this:

‘In a well reasoned decision the Board held for Christman on the grounds that Person’s use of the mark in Japan could not be used to establish priority against a “good faith” senior user in U.S. commerce. The Board found no evidence to suggest that the “PERSON’S” mark had acquired any notoriety in this country at the time of its adoption by Christman. Therefore, appellant had no reputation or goodwill upon which Christman could have intended to trade, rendering the unfair competition provisions of the Paris Convention inapplicable.’

‘In the present case, appellant Persons Co. relies on its use of the mark in Japan in an attempt to support its claim for priority in the United States. Such foreign use has no effect on U.S. commerce and cannot form the basis for a holding that appellant has priority here. The concept of territoriality is basic to trademark law; trademark rights exist in each country solely according to that country's statutory scheme. Christman was the first to use the mark in United States commerce and the first to obtain a federal registration thereon. Appellant has no basis upon which to claim priority and is the junior user under these facts.’

‘In the case at bar, appellant Person’s Co., while first to adopt the mark, was not the first user in the United States. Christman is the senior user, and we are aware of no case where a senior user has been charged with bad faith. The concept of bad faith adoption applies to remote junior users seeking concurrent use registrations; in such cases, the likelihood of customer confusion in the remote area may be presumed from proof of the junior user's knowledge. In the present case, when Christman initiated use of the mark, Person’s Co. had not yet entered U.S. commerce. The Person’s Co. had no goodwill in the United States and the “PERSON'S” mark had no reputation here. Appellant’s argument ignores the territorial nature of trademark rights.’

‘As the Board noted below, Christman's prior use in U.S. commerce cannot be discounted solely because he was aware of appellant’s use of the mark in Japan. While adoption of a mark with knowledge of a prior actual user in U.S. commerce may give rise to cognizable equities as between the parties, no such equities may be based upon knowledge of a similar mark's existence or on a problematical intent to use such a similar mark in the future. Knowledge of a foreign use does not preclude good faith adoption and use in the United States. While there is some case law supporting a finding of bad faith where (1) the foreign mark is famous here or (2) the use is a nominal one made solely to block the prior foreign user's planned expansion into the United States, as the Board correctly found, neither of these circumstances is present in this case.’

[25] To conclude: A local mark, validly appropriated, cannot lose its value or protection simply because someone else’s reputation overtakes its business.²⁶ Although the marks in contention may at first blush appear to be imitations or reproductions of the respondents’ marks, the crucial point is that when any reproduction or imitation took place, it was not of a ‘well-known’ mark but of a mark not well known within this country. As we have seen, there is and never has been anything wrong with the reproduction or imitation of a mark that is not well known. Salt of the Earth was the first proprietor of these marks in South Africa and became the proprietor at a stage when the respondents’ marks were not yet well known.

NON-USE

[26] The next issue concerns the question of non-use.

‘A trader registers or acquires a trademark primarily not in order to prevent others from using it but in order to use it himself (although exclusivity of use is of course a necessary corollary). Use by the proprietor is indeed a central and essential element of ownership [since] rights may lapse or be unenforceable in the event of non-use.’²⁷

As mentioned, the court below held that because of non-use of the trade marks by the proprietor, the respondent was entitled to an order expunging these registrations. Section 27(1)(b) of the current Act provides as follows:

‘27 Removal from register on ground of non-use –

(1) Subject to the provisions of section 70 (2), a registered trade mark may, on application to the court, or, at the option of the applicant and subject to the provisions of section 59 and in the prescribed manner, to the registrar by any interested person, be removed from the register in respect of any of the goods or services in respect of which it is registered, on the ground either-

(a) . . .

(b) that up to the date three months before the date of the application, a continuous period of five years or longer has elapsed from the date of issue of the certificate of registration during which the trade mark was registered and during which there was no bona fide use thereof in relation to those goods or services by any proprietor thereof or any person permitted to use the trade mark as contemplated in section 38 during the period concerned; or

(c) . . .’

[27] The application for expungement was launched on 13 August 1999. Use of the marks during the preceding three months would not have saved them but nothing turns on this. According to s 27(3),²⁸ the onus to prove that there was ‘relevant use’ of the trade mark rests upon the proprietor. ‘Relevant use’ in this context refers to bona fide use by the proprietor or bona fide use by a third party ‘with the licence of the proprietor’ (the latter is known as ‘permitted use’: s 38(1)).²⁹ We are accordingly concerned with bona fide use by the proprietor or a licensee during the period 13 May 1994 to13 May 1999. In order to qualify as bona fide use it is not necessary that the use was continuous use (as required by s 36(2))³⁰ although intermittent use may be indicative of the fact that it was not bona fide. Use prior to 13 May 1994 and post 13 May 1999 is for this part of the investigation irrelevant.

[28] The following matters have to be considered in this context: (a) Were the marks used during this period? (b) Were they used by the proprietor? (c) Was any other use permitted use? (d) Was the permitted use controlled? (e) Was the use bona fide?

[29] It is convenient to deal with the first two questions at the same time. The appellants concede that neither the first GAP device mark (TM 80/5548) nor the GAP STORES mark (89/5087) was used during the relevant period by any party. (The concession may have extended to GAP KIDS but that does not matter in the scheme of things.) In addition, they accept in spite of generalised allegations to the contrary by Dr Vahed, that the proprietor was at all relevant stages a dormant company and did not use any of the marks. Yet, it cannot be doubted that members of the Group did use the one or the other of the remaining marks (73/1378 and 88/4994) during this period. There is evidence to this effect not only from Dr Vahed but also, for instance, from a wholesaler (Mr Shapiro) who purchased GAP clothing from the Group and distributed it, all of which was not disputed. Which one of the two marks was used though is unclear. As Daniels J correctly pointed out, Dr Vahed

‘indiscriminately uses the plural and singular when referring to the various marks. . . . No distinction is drawn between the marks. He goes so far as to refer to “the Gap and related” marks, and elsewhere the expression “the various marks” was used.’

For instance the sales figures provided relate to sales ‘under the various GAP trade marks’ although, as mentioned, it is conceded that the statement is incorrect, at least as far as the first device mark is concerned. However, a number of labels, which allegedly had been used, were produced and they show use of the second device mark and of the word GAP.

[30] Even though the appellants’ evidence on this aspect of the case is unsatisfactory, that is not fatal to their case. The general rule that a mark must be used in the form in which it is registered³¹ does not necessarily apply to associated marks. Section 31(1) provides:

‘When under the provisions of this Act use of a registered trade mark is required to be proved for any purpose, the registrar or the court, as the case may be, may, if and so far as he or it deems fit, accept proof of the use of an associated registered trade mark or of the trade mark with additions or alterations not substantially affecting its identity, as equivalent to proof of the use required to be proved.’

The provision applies because the original THE GAP word mark and the two device marks (one consisting of the word GAP and the other of the words THE GAP) are associated marks. In addition, according to the sub-section, use of the word GAP on its own can be equivalent to use of THE GAP because the difference between them cannot affect the registered mark’s identity. Also, use of the name THE GAP in any stylised form amounts to use of THE GAP (73/1378) because that registration is not restricted to those words in any particular colour, style, manner or font.

[31] Having found that at least one of the associated marks was used by members of the Group, and in the absence of any reason not to exercise the discretion contained in s 31(1) in favour of the proprietor, the next question to consider is whether such use was with its ‘licence’. Dr Vahed’s evidence in this regard is not only vague to such an extent that it smacks of evasiveness, but it is also contradictory. The problem lies in the loose structure of the Group, which consists of companies and partnerships with an overlap of shareholders and directors. The Group, it appears, tends to ignore corporate identities and moved enterprises between companies. For instance, at times the fourth appellant itself conducted business and at other times a business under its name appeared to be an operating division of the second appellant.³² In line with this approach, Dr Vahed said that when the Group obtained control of the proprietor it (the Group) ‘acquired’ the GAP trade marks. These were dealt with as Group property, any member using them whenever it suited it. If then, in a 200 page affidavit, Dr Vahed dealt with the issue of licensing in a single phrase, saying that the use by the members of the Group since 1983 was ‘with the licence’ of the proprietor, one has to conclude that this was no more than a mere allegation and did not amount to factual evidence. Affidavits in application proceedings must do more than make bald allegations; they must in addition provide the facts that support the allegations. If the licence had been in terms of a written agreement, one would have expected some particularity and perhaps even the production of a copy; if oral, one would have expected some indication of between whom, when and where; if tacit, an allegation to that effect with the facts that gave rise to the tacit agreement would not have been out of place. In any event, the grant of a licence to the Group is not only improbable but also incompatible with the perception that existed within the Group that the trade marks belonged to the Group.

[32] There is more. User agreements in the names of the first and second appellants were registered on 21 January 1999 and, consequently, use by either of them during the period 21 January to 13 May 1999 would have amounted to permitted use. Although we were told that the Group had sales of GAP clothing to the tune of R8m for the 1999 year, we were not informed whether these sales were by the first and second appellants or by other members of the Group. In fact, there is evidence of the existence of a company in the Group with the name The Gap Clothing Manufacturers (Pty) Ltd, which, according to advertisements, may have been the user of the trade marks.

[33] The mystery deepens if regard is had to the first appellant’s annual report for the year 1998. It starts off with an allegation that the GAP trade marks are its registered brands but then, under the heading ‘Gap, Gapstar and Gapkids’, the statement is made that

‘the Sterling division intends launching a high class range of men’s and ladies casualwear/sportswear during the 1999 summer season and has concluded a User Agreement with the registered owner to market these products under a separate division called Gap Clothing Manufacturers.’

The problems with this statement are manifold. It was not confirmed under oath. There is no evidence that the intention to produce the goods came to fruition. According to the trade mark registry the user agreement was between the first appellant and the proprietor whereas if this statement is to be taken at face value it means that it was between the first appellant and one of its divisions. To complicate matters, during 1999 a company called Sterling Clothing (Pty) Ltd was joined as an applicant in proceedings launched by the Group to interdict the infringement of the GAP trade marks, presumably because it was a registered user (only proprietors and registered users have legal standing to enforce registered trade mark rights), something not explained in the papers.

[34] Similar problems are encountered with the unsubstantiated allegation by Dr Vahed that at a time when the first appellant sought listing on the Johannesburg Stock Exchange, during 1997, licence agreements were entered into with both the first and second appellants. However, the pre-listing statement made no reference to these agreements. On the assumption that the pre-listing statement contained a frank disclosure of the state of the first appellant’s business the omission was left unexplained. But again, there is no evidence that since the conclusion of the licence agreements either of these licensees had used the GAP trade marks. Taking into account all these facts I have to conclude that the appellant failed to discharge its onus.

[35] The respondent raised the issue of lack of quality control as an independent factor under s 27(1). The submission was that unless a licence agreement provides for quality control by or on behalf of the trade mark owner, use by a licensee cannot qualify as permitted use. While the appellants agreed with the legal proposition they submitted that the issue was not properly raised by the respondent in the founding affidavit and that there is in any event sufficient evidence on record to gainsay the allegation. I may interpose to note that both parties have dealt with the matter in bald terms and that if quality control were essential to prove permitted use, it would have been for the trade mark proprietor to have raised the issue in discharge of its overall onus.

[36] Daniels J accepted the correctness of the contention that some form of quality control was required

‘since the licensor must have regard to the provisions of s 10(1) and (2)(a) of the Act which require that whatever use is made of a mark it should still remain “capable of distinguishing”. Use, which results in a loss of distinctiveness, cannot qualify as bona fide use. It is for that reason that cogent evidence of quality control is required.’

[37] To assess the cogency of the contention it is necessary to have regard to some basic principles and to the history of trade marks, much of which was dealt with by the House of Lords in a judgment not brought to the attention of Daniels J, namely, Scandecor.³³ The judgment is relevant not only for its persuasive reasoning but also because the history of our trade mark law mirrors the developments in the United Kingdom in most respects. For instance, the repealed 1963 Act (s 48(2)) provided that use by a licensee could only qualify as permitted use if it were in terms of a registered user agreement.³⁴ The same was the position under the 1938 UK Trade Marks Act. And, as mentioned, the current s 27 (as did the UK Trade Marks Act of 1994) abolished this requirement by providing that use by any licensee is permitted use.

[38] In principle, there does not appear to be any reason at present to require quality control by or on behalf of the proprietor as a requirement for permitted use. As Lord Nicholls explained in Scandecor:³⁵

‘A trade mark is a badge of origin or source. The function of a trade mark is to distinguish goods having one business source from goods having a different business source. It must be “distinctive”. That is to say, it must be recognisable by a buyer of goods to which it has been affixed as indicating that they are of the same origin as other goods which bear the mark and whose quality has engendered goodwill: see GE Trade Mark [1973] RPC 297, 325, per Lord Diplock.’

This exposition accords with the position in our law.³⁶ Further, he said:³⁷

‘Although the use of trade marks is founded on customers’ concern about the quality of goods on offer, a trade mark does not itself amount to a representation of quality. Rather it indicates that the goods are of the standard which the proprietor is content to distribute “under his banner”: see Laddie J, in Glaxo Group v Dowelhurst Ltd [2000] FSR 529, 540-541. . . . Thus, in relying on a trade mark consumers rely, not on any legal guarantee of quality, but on the proprietor of a trade mark having an economic interest in maintaining the value of his mark. It is normally contrary to a proprietor’s self-interest to allow the quality of the goods sold under his banner to decline.’

[39] Put differently, although in the ordinary course of events a trade mark owner would wish to ensure the consistency of the quality of the goods or services marketed under its mark, nothing prevents the owner from providing under one mark goods or services of differing, inconsistent or poor quality. The customer has in the event of the purchase of a substandard product or the provision of substandard services no redress based on trade mark principles. Market forces may eventually exact their toll. All a trade mark does, in the words of Laddie J in Glaxo,³⁸ is to identify the enterprise that is responsible for the quality of the goods or services. Again Lord Nicholls:³⁹

‘This approach accords with business reality and customers' everyday expectations. Customers realise there is always the prospect that, unbeknown to them, the management of a business may change. To confine the use of a trade mark to the original owner of a business would be to give the concept of a business origin or business source an unrealistically narrow and impractical meaning. Of course, the new management, the new owners, may not adhere to the same standards as the original owner. But the risk of an unannounced change of standards is ever present, even when there has been no change in management. An owner may always decide to change his quality standards. As already noted, customers rely on it being in the owner’s self-interest to maintain the value of his mark. The self-interest of the owner of a trade mark in maintaining its value applies as much to a purchaser of the mark as it does to the original owner.’

[40] Although Scandecor was concerned with an exclusive licence, its conclusion that the current UK Act countenances bare licensing, ie, licensing without quality control, applies in my view to all licences under our current Act. As Lord Nicholls explained:

‘Customers are well used to the practice of licensing of trade marks. When they see goods to which a mark has been affixed, they understand that the goods have been produced either by the owner of the mark or by someone else acting with his consent.’⁴⁰

‘For their quality assurance customers rely on the self-interest of the owner. They assume that if a licence has been granted the owner can be expected to have chosen a suitable licensee and imposed suitable terms. They also assume that during the currency of any licence the licensee, as well as the owner, is likely to have an interest in maintaining the value of the brand name. Customers are not to be taken to rely on the protection supposedly afforded by a legal requirement that the proprietor must always retain and exercise an inherently imprecise degree of control over the licensee’s activities.’⁴¹

[41] To revert to Daniels J’s ratio, namely that bare licences may lead to loss of distinctiveness of a trade mark, the answer appears to me to be factual: did the grant or exercise of a bare licence in the circumstances of a particular case cause a loss of distinctiveness? If the answer is in the affirmative, the appropriate remedy would be an application for the expungement of the mark on that ground. In this case there was no evidence of loss of distinctiveness because of the grant of bare licences (although there is reason to believe that there was such a loss) and consequently the objection of non-use could not have been upheld on this ground.

[42] That brings me to the last question relating to this issue: did such use as there was amount to bona fide use? The concept of bona fide use has been the subject of a number of judgments, also of this court,⁴² and the area need not be traversed again. For present purposes it suffices to say that ‘bona fide user’

‘means a user by the proprietor of his registered trade mark in connection with the particular goods in respect of which it is registered with the object or intention primarily of protecting, facilitating, and furthering his trading in such goods, and not for some other, ulterior object.’⁴³

This test is similar to that proposed in an opinion by the Advocate General in the European Court of Justice in the Ansul case:⁴⁴

‘When assessing whether use of the trade mark is genuine, regard must be had to all the facts and circumstances relevant to establishing whether the commercial exploitation of the mark is real, particularly whether such use is viewed as warranted in the economic sector concerned to maintain or create a share in the market for the goods or services protected by the mark, the nature of those goods or services, the characteristics of the market and the scale and frequency of use of the mark.’

[43] The essence of the respondent’s complaint was that the appellants were prepared to assign the marks to the respondent at a price bordering on the ridiculous and that they used implied political threats during the negotiations. This, according to the submission, amounted to ‘trafficking’ and that the only reason for the extensive use by the Group of the marks was to coerce the respondent into paying exorbitant sums for them.

[44] There is no merit in the submission. ‘Trafficking’ is an emotive term which appeared in the repealed Act but is not in the current Act. The issue is simply whether or not the Group’s use amounted to real commercial exploitation of the marks in relation to goods or services in the trade mark sense. Exploiting marks in order to increase their value to be able to sell them at a higher price does not amount to an improper or mala fide commercial exploitation. The point can be illustrated with reference to a short dictum of Jacob J⁴⁵ in a case where the registration of a trade mark was opposed on the ground of bad faith:

‘I need say little about this allegation – it was based on a suggestion that somehow a patent monopoly was being extended [by the registration of a trade mark]. That is miles from bad faith.’

[45] To conclude on the subject of non-use: the objection was correctly upheld on the ground that the proprietor had failed to prove use by itself or permitted use during the period concerned.

THE CROSS-APPEAL

[46] The respondent also sought expungement of the registrations under s 10(3) of the Act. This provides that ‘a mark in relation to which the applicant for registration has no bona fide claim to proprietorship,’ if registered, ‘shall, subject to the provisions of sections 3 and 70, be liable to be removed from the register’. The court below, as mentioned, expunged all the registrations on the ground of non-use but dismissed this claim which sought the same relief, albeit under another section of the Act. A cross-appeal may, consequently, have been unnecessary because the respondent could have defended the expungement by relying on s 10(3),⁴⁶ but since little depends on this procedural aspect I need not consider it any further.⁴⁷

[47] The respondent’s argument under s 10(3) was straightforward. The proprietor, since becoming part of the Group, has been a dormant company. It did not manufacture or sell any clothing. All it did was to hold and register trade marks. For instance, it applied for GAP STORES at the behest of OK Bazaars (1929) Ltd and entered into a registered user agreement with it. The other trade marks were not used or intended to be used by the proprietor. Section 20(1) of the repealed 1963 Act provided that ‘any person claiming to be the proprietor of a trade mark used or proposed to be used by him’ was entitled to apply for registration. There were two exceptions. The first, which did not apply, related to the case where someone intended to constitute a body corporate and assign the mark to that body for its use. The second was where the application for registration was accompanied by an application for the registration of someone as a registered user. This exception could have applied in the OK Bazaars case, but according to the trade mark registry the user agreement was never filed, and the appellants never disclosed the date of the agreement, nor did they allege that it had been filed simultaneously with the application for registration.

[48] The respondent thus had a prima facie case for expungement on this additional ground.⁴⁸ The reason the court below dismissed the claim was based on the introductory proviso to s 10 of the current Act, which made s 10(3) – and all of s 10 – subject to s 70. In so far as it is relevant, s 70 states that

‘the validity of the original entry of a trade mark on the register of trade marks existing at the commencement of this Act [1 May 1995] shall be determined in accordance with the laws in force at the date of such entry.’

The appellants relied on s 42 of the 1963 Act,⁴⁹ which provided that in all proceedings

‘relating to a trade mark registered in part A of the register . . . the original registration of the trade mark in part A of the register shall, after the expiration of seven years from the date of that registration, be taken to be valid in all respects, unless –

1. that registration was obtained by fraud; or

2. the trade mark offends against the provisions of either section sixteen or section forty-one.’

(The exceptions are inapplicable.)

[49] We have to assume that these marks were registered in class A because, in spite of no evidence to that effect in a record of 2500 pages, the respondent did not suggest otherwise. For purposes of s 42, the date of registration is, once a trade mark is registered, deemed to be the date of application.⁵⁰ That disposes of the present attack on the 1973 and 1980 marks. The position in relation to the 1988 and 1989 marks is, however, factually different. When the 1993 Act came into force, they had not been on the register for seven years, although, in one instance, the ‘prescription period’ was missed by less than two months.

[50] The current Act does not retain the prescriptive period in relation to ‘old’ registrations nor does it provide for the completion of prescription after its date. All that it provides is that the original registration of an old mark may only be attacked under the provisions of the repealed Act, in other words, the provisions of s 9 and 10 of the new Act do not apply. After seven years under the old Act a registration became immune to any attack based on its original registration. That could only have applied to marks that had been on the register for seven years when the new Act came into operation. Since this is not the situation the cross-appeal has to be upheld to the extent indicated.

CONCLUSION

[51] The outcome of the foregoing is that the appeal succeeds in relation to the s 35(3) interdict but stands to be dismissed on the expungement claim based on s 27(1)(b). Because of the expungement, the proprietor’s attempt to interdict the respondent from infringing its registered trade marks falls away. In addition, the cross-appeal succeeds. For the sake of clarity I intend to redraft the order of the court below bearing in mind our conclusions. I am conscious of the fact that the result may satisfy neither party because their respective ability to prevent the other from using GAP marks in this country hangs in the air and further litigation may be on the cards.

ORDER

[52] The following order is made:

1. The appeal against para 3 of the order of the court (the s 35(3) interdict) below is upheld with costs, including the costs of two counsel.

2. The appeal is otherwise dismissed.

3. The cross-appeal is allowed with costs, including the costs of two counsel.

4. The order of the court below is replaced with the following order:

   1. The following registered trade marks are expunged from the register and the Registrar is ordered to effect the necessary rectification: TM 73/1378, TM 80/5548, TM 88/4994, TM 88/8783 and TM 89/5087.

   2. The respondents are ordered to pay the costs of the application, including the costs of two counsel.

   3. The counter-application is dismissed with costs.

   4. The costs relating to the application to strike out are costs in the application.

_____________________

L T C HARMS

JUDGE OF APPEAL

AGREE:

STREICHER JA

BRAND JA

LEWIS JA

PONNAN JA