Bad Faith, Public Policy and Morality: How Open Concepts Shape Trade Mark Protection

Abstract

The grounds for refusal or invalidity relating to trade marks filed in bad faith or being contrary to public policy or accepted principles of morality are based entirely on undefined concepts and require a value judgement to be made by the adjudicating authorities. This openness has led to the exploration of new lanes for these grounds in case law beyond the traditional scope of their application. This paper examines how the flexibility of the concepts at hand has led to considerable uncertainty in assessing whether a particular mark is registered in bad faith or offends against morality and public policy – and in some cases both. In particular, the objective, consistent viewpoint of the average consumer, which is normally applied in other areas of trade mark law, appears to be exiled from the analysis in these cases. In the case of bad faith, the assessment involves a subjective element, as there is a mandatory requirement that there should be bad faith intent on the part of the applicant. The paper argues that much of the uncertainty as to when bad faith applies could be overcome by adopting an objective viewpoint in its assessment. By contrast, the assessment of the offence against morality remains objective. It is based on the perception of a model member of society with average thresholds of sensitivity and tolerance who does not always overlap with the average consumer of the goods covered by a trade mark application. Infringement of public policy is also objective and is linked with offence against the “top-down” rules laid down by the public authority.

1 Introduction

The two refusal (invalidation) grounds for trade marks filed in bad faith or trade marks which are contrary to public policy or to accepted principles of morality are both based entirely on what might be termed open concepts. Indeed, from the standpoint of civil law, bad faith, public policy and morality could be described as “general clauses”, meaning clauses used in legislative text whose application requires a value assessment.¹ Other refusal grounds in trade mark law (from the perspective of the EU and UK systems) may also include flexible concepts that necessitate a value assessment, but only as part of a larger set of requirements, for example, “taking unfair advantage” or “due cause” as elements of a relative refusal ground concerning trade marks with a reputation (Art. 5(3)(a) of the Trade Mark Directive 2015/2436 (TMD)²).³ This value-oriented flexibility of the refusal and invalidation grounds based on bad faith or violating public policy and morality might also explain the trend in recent years for parties to trade mark actions or adjudicating bodies to seek new lanes for applying these grounds beyond their traditional scope of application.

One reason for the flexibility in identifying these concepts may lie in the absence of a single, objective viewpoint, which is provided in many other areas of trade mark law by the average consumer. For example, it is based on the perceptions of the average consumer that registering authorities and the courts will decide if a mark is sufficiently distinctive for registration and whether a third party’s mark overlaps with or infringes a registered mark causing a likelihood of confusion. And, because the average consumer is a legal construct, courts may themselves take on their mantle when making rulings in these areas. The decisive role of this supposedly rational individual allows for the claim that in EU trade mark law, findings of distinctiveness or conflict between marks are based on consistent and objective criteria. Refusal and invalidation grounds based on either bad faith or violation of public policy or accepted principles of morality are, however, two key areas of trade mark law where the utility of the average consumer is uncertain. The resultant vagueness of these concepts might be beneficial precisely because of their flexibility and hence adaptability to novel situations but equally might be detrimental to legal certainty when these concepts are applied to situations that should rather be assessed under other refusal or invalidation grounds.

In fact, both may be true, depending upon whether one looks at public policy and morality grounds or bad faith. In relation to public policy and morality, this paper will argue that the law has developed an objective test for their application, albeit in the context of a wide variety of circumstances in which they may be applied. Conversely, assessments of bad faith have come increasingly to rely on the subjective value judgements of the court or registering authorities, with the objective circumstances a secondary consideration. In both cases, however, the fact that the refusal (invalidation) grounds at hand are entirely based on open concepts has led to considerable uncertainty in making assessments of whether a particular trade mark is registered in bad faith or offends against morality and public policy – and in some instances both. It will be maintained that the lack of consistency concerning the application of both grounds has also allowed parties in trade mark actions, as well as the courts, to widen the factual scope of situations which might be held to fall within their scope.

This article will look first at trade marks filed in bad faith. It will then consider marks contrary to public policy and accepted principles of morality. It will conclude by emphasising the commonalities between the problems raised by each ground for refusal or invalidity and offer suggestions on how they might be overcome. But it will also emphasise the key differences that have emerged through judicial decisions in relation to these grounds, most especially the development of an objective viewpoint for assessing the grounds of public policy and morality and how that contrasts with the present approach to bad faith. Indeed, it is perhaps a measure of the continuing uncertainty in the application of these grounds, as argued here, that recently the European Union Intellectual Property Network has published drafts of Common Practices concerning trade marks filed in bad faith,⁴ as well as marks contrary to public policy and accepted principles of morality.⁵ Finally, for the sake of clarity, it is worth adding, that while the title of the paper suggests that it examines three concepts – bad faith, public policy and morality – these concepts are, however, reflected in the structure of provisions of trade mark law as two refusal or invalidation grounds. The first concerns marks filed in bad faith; the second, marks being contrary to public policy and accepted principles of morality. While discussing the latter refusal ground, it will nevertheless be necessary to examine the relationship and differences between morality and public policy which form the two components of this refusal ground.

2 Bad Faith

According to Art. 4(2) EU Trade Mark Directive (TMD): A trade mark shall be liable to be declared invalid where the application for registration of the trade mark was made in bad faith by the applicant. Any Member State may also provide that such a trade mark shall not be registered. Under the EU Trade Mark Regulation (EUTMR⁶) bad faith can only be invoked as an invalidation ground (Art. 59(1)(b) EUTMR). While in the EU, the UK adopted bad faith as a refusal ground which remains in force. Thus, according to Sec. 3(6) UK Trade Marks Act 1994 (TMA):⁷ A trade mark shall not be registered if or to the extent that the application is made in bad faith. Furthermore, to date, despite no longer being a member of the EU, in interpreting this article, the UK courts have followed the precedents set by the EU, in particular the General Court and the CJEU. Hence, case law from both the EU, Member States and the UK relating to bad faith will be referenced here.

It was in 2009 in Lindt v. Hauswirth,⁸ that the Court of Justice of the European Union (CJEU) first addressed in a substantive way the meaning of bad faith. Later, in 2013 the CJEU explained that “the concept of ‘bad faith’ is an autonomous concept of European Union law which must be given a uniform interpretation in the European Union”.⁹ Nonetheless, it was still possible for the UK High Court to observe in 2021:¹⁰ “The ‘bad faith’ objection to the registration of trade marks has defied any straightforward definition, either by the UK Courts or by the CJEU”. There are indeed a number of unresolved problems with identifying bad faith conduct, and central to these is a failure by the courts to pinpoint a consistent and identifiable viewpoint from which it is possible to assess whether bad faith is present: in other words, whether there is dishonest intention or other sinister motives on the part of the actor accused of bad faith.¹¹ This was not always the case. For example, in the early UK trade mark case Gromax v. Don & Low (1998),¹² which involved an accusation of bad faith, Lindsay J, in the High Court, identified bad faith as “dishonesty” or standards which fall short “of acceptable commercial behaviour observed by reasonable and experienced men in the area being examined”. The attributes ascribed to such “men” (and some 20 years later we can add “women”) approximates those of the average consumer for the goods and services in question not least because both are assumed to be rational, with knowledge of the relevant field and as having no stake in the outcome of the dispute. Crucially, such reasoning also identified an objective and external viewpoint for identifying bad faith, which, if adopted, would presumably avoid having to scrutinise and make judgements about the motivations of the applicant for or the owner of a registered mark.

The notional existence of an external, neutral arbiter of what might constitute bad faith conduct was an early victim of EU trade mark law. Its demise was ensured in the still authoritative case of Lindt. In Lindt, both AG Sharpston and the CJEU in its subsequent judgment made clear that it is necessary to demonstrate a “subjective motivation” on the part of the applicant for a finding of bad faith.¹³ In her opinion, AG Sharpston explicitly rejected the suggestion of the Czech government “that the notion of bad faith in the legislation has a significant moral or ethical element; it implies a contravention of accepted norms of conduct”:¹⁴ a description similar to that offered in Gromax. She also rejected the Swedish approach, shared by a number of other EU Members, that took the view that the applicant’s intention is not relevant, rather the question is whether the applicant was, or should have been, aware that another person was already using the same or a similar mark.¹⁵ Instead, according to the AG the concept of bad faith “in normal usage” implies a “subjective mental state of a general nature”.¹⁶ Acknowledging that bad faith based on a subjective intention might be difficult to establish, she accepted that objective circumstances might help in identifying it.¹⁷ The CJEU agreed with the AG that the applicant’s intention when registering was a “subjective factor” which must be determined “by the objective circumstances of the case”.¹⁸ And, in this case, “the intention to prevent a third party from marketing a product may, in certain circumstances, be an element of bad faith on the part of the applicant”.¹⁹ The CJEU also pointed out that it is possible for the same behaviour to be undertaken in either good faith or bad faith. Although this ambiguity was not an issue in assessing bad faith in Lindt, it would give rise to difficulties in later cases.²⁰

In Lindt, Hauswirth had been selling chocolate bunnies in Austria and Germany since the 1930s. Lindt had sold chocolate bunnies in Austria since the 1950s. In 2000, Lindt registered the shape of a chocolate bunny as a EU trade mark. Hauswirth alleged it had done so with bad faith intent. The CJEU looked at the relevant objective factors which it identified as: whether Lindt was aware or should have been aware of third-party use of the sign when applying for the mark; whether its intention was to prevent a third party from using the mark; and finally what degree of legal protection was enjoyed by the sign.²¹ The Court made clear that this was a non-exhaustive list and that in any event these same factors were not necessarily probative of sinister intent. It follows, that the Lindt test is by no means equivalent to the tests put forward by other EU countries, including, at the time, the UK. In the latter, the test of bad faith was entirely objective and external to the proprietor of the mark. It used accepted standards of behaviour to deduce whether the actions of the proprietor, in the circumstances, would amount to bad faith. Conversely, in Lindt, it was held that objective circumstances merely point to the possibility of a subjective intention. The same is true of later EU cases in which courts have identified other conduct which departs from accepted standards of ethical behaviour or honest commercial and business practices, as evidence of the bad faith intent of the accused, but held such evidence is not probative. The nature of the latter must still be ascertained by the court.²² The result, it will be argued here, is that the boundaries of what constitutes bad faith conduct have remained uncertain. This is particularly, but not exclusively, the case when the alleged bad faith conduct is not directed at any particular third party, as it was in Lindt. It is in these latter cases that the problem of assessing the bad faith conduct remains most acute, not least because the subjective intent of the actor is far more difficult to ascertain.

There have certainly been a number of successful bad faith actions where the allegation is that a trade mark was either registered to prevent a specific third party from using the same sign or with the intention of taking advantage of a particular third party’s goodwill, which often involve well-known personalities. In these cases, it can be argued, it is relatively easy to identify a dishonest or sinister motivation. A prominent example of a case where the bad faith conduct was to take advantage of goodwill, is Carlos Moreira v. EUIPO.²³ Here Moreira obtained an EU trade mark in 2013 for the mark “Neymar”, inter alia for clothing. In 2016, representatives of the Brazilian footballer alleged that the mark had been registered in bad faith. Moreira responded that at the time of registration Neymar was not yet widely known in Europe. However, the General Court accepted, based on objective evidence including contemporaneous press reports, that this was not the case. Neymar had considerable fame even before he moved to Barcelona in 2013. The General Court held that there was ample objective evidence that the proprietor’s motivation had been to “free-ride” on the “attractive force” of Neymar’s name.²⁴ In addition, the applicant filed on the same day a further trade mark with the name of another famous footballer, which pointed out to a pattern of dishonest conduct on the applicant’s part.²⁵Lindt was, of course, another case where the target of the alleged bad faith was clear. Although, ironically, when the case returned to the Austrian court, it was held that the EU trade mark was valid as Lindt had acquired its own right to sell the golden bunnies.²⁶

Arguably, in actions where the targets of the alleged bad faith action are known, it is simpler to identify objective factors which provide strong evidence of a subjective bad faith intent. By contrast, bad faith might also reside in conduct which is not directed at a particular target. The fact that Art. 4(2) TMD 2015 (or Art. 59(1)(b) EUTMR) encompasses such conduct has been frequently acknowledged in the CJEU case law. Thus, in Koton,²⁷ the CJEU stated that bad faith invalidity applies:

… where it is apparent from relevant and consistent indicia that the proprietor of an EU trade mark has filed the application for registration of that mark not with the aim of engaging fairly in competition but with the intention of undermining, in a manner inconsistent with honest practices, the interests of third parties, or with the intention of obtaining, without even targeting a specific third party, an exclusive right for purposes other than those falling within the functions of a trade mark, in particular the essential function of indicating origin … .

In other words, as was put by the UK Court of Appeal in Sky v. Skykick,²⁸ there are two possible ways bad faith conduct might manifest itself, which the CJEU identified in Koton and reiterated in Sky v. Skykick.²⁹ The first is where the bad faith is targeted at undermining the interests of a particular party; the second is where no particular party is targeted but rather what is sought is a right to use a trade mark for purposes other than indicating origin.

In seeking to ascertain whether the actor had a bad faith intent, it is inevitably more difficult in these latter situations. One of the key reasons for this, which was identified in Lindt and in many later judgments, resides in the general principle that bad faith is defined by an absence of good faith. For example, in Walton v. Verweij,³⁰ the UK High Court held:

(i) A person is presumed to have acted in good faith unless the contrary is proved. An allegation of bad faith is a serious allegation which must be distinctly proved. The standard of proof is the balance of probabilities, but cogent evidence is required due to the seriousness of the allegation. It is not enough to prove facts which are also consistent with good faith.

(…)

(vi) Consideration must be given to the applicant’s intention. This is a subjective factor which must be determined by reference to the objective circumstances of the particular case.

What the Walton extract suggests, is that in many cases the court’s approach must be the opposite of Lindt. In Lindt, the approach was held to be that, whatever their possible ambiguity, objective factors should be adduced to decide whether the applicant acted with bad faith intent. In Walton, the approach is the reverse. Since the same conduct might be undertaken in bad faith or good faith, a possibility which was raised by the CJEU in Lindt, it is the subjective intention of the actor which will indicate which is the case.

A recent case which underlines the possible ambiguity inherent in determining whether a party is guilty of engaging in good faith or bad faith conduct is Sky v. Skykick. Sky, an internationally known media company, had a number of EU trade marks incorporating the word “Sky”, registered for a wide variety of goods. Sky brought an infringement action in the UK courts against Skykick, a company which supplies software services relating to cloud migration, back up, and management.³¹ Sky claimed that Skykick had infringed its marks registered for goods including computer software and computer services for accessing and retrieving information. Sky had made extensive use of its trade marks for a wide range of goods relating to its core business of television broadcasting, broad band provision and telephony. However, Skykick alleged that Sky had not used its marks for email migration or cloud backup goods and services and there was no evidence it would do so in the future. It followed that these marks should be declared invalid as they were registered in bad faith. The CJEU made a number of general findings regarding whether in circumstances such as in the main proceedings a subjective intention to act in bad faith can be established. Most importantly, the CJEU explained that there must be objective criteria showing that the applicant registered the mark in order to undermine the interests of third parties rather than to use the marks as a badge of origin.³² It followed, as the Court pointed out,³³ that the bad faith of the applicant “cannot, therefore, be presumed on the basis of the mere finding that, at the time of filing his or her application, that applicant had no economic activity corresponding to the goods and services referred to in that application”. That fact must be proven, especially since under the EU and UK law a lack of genuine use becomes detrimental only upon the lapse of a five-year period of non-use following the registration of the mark.³⁴ In other words, the objective factors which are to be called in aid to identify whether an action was in good faith or bad faith can, themselves, be uncertain and hence not indicate intent.

A further finding in Sky was that it is possible, when making a broad application for goods and services, that the ground for invalidity exists only in respect of some of the goods or services for which the mark is sought to be registered. And, in those circumstances, the trade mark is to be declared invalid as regards only those goods and services. Or as was put by the UK Court of Appeal in considering the CJEU judgement,³⁵ “the facts in the present case … meant that there can be good and bad reasons for applying for the trade marks”. The Court of Appeal also warned against presuming that the overall width or size of the specification can be simple evidence of bad faith. According to Floyd LJ,³⁶ there will only be bad faith where the absence of intention to use is coupled with objective, relevant and consistent indicia of the additional positive intention of obtaining an exclusive right for purposes other than those falling within the functions of a trade mark, which was also the position expressed earlier by the CJEU.³⁷ For example, it could not be assumed, as the lower court had done, that Sky did not have an intention to use all the goods within the broad category of computer software, given Sky’s “reputation and brand recognition”.³⁸ According to the Court of Appeal, justification for a wide claim would depend upon whether “one is discussing bleaching preparations or computer software”. It followed that there was no objective evidence that the broad specification made by Sky demonstrated bad faith rather than a good faith intention to use the registered mark on specified goods in the future. In Sky v. Skykick, as it has moved through the courts,³⁹ there have been conflicting findings on whether or to what extent Sky had acted in bad faith. In other words, the same objective factors could be interpreted and reinterpreted to be either indicative of good faith or bad faith. A decisive factor would of course be Sky’s subjective intention, but how is that to be uncovered without reliable “objective” evidence?

Separating good faith from bad faith is particularly difficult not just in cases where the applicant registers the mark across a broad range of goods but also in cases of what is called “evergreening”, which involves the re-registration of marks which, not having been used during the obligatory five-year period, are in danger of being revoked. The fact that in such cases the objective element of re-filing an identical trade mark is not regarded of itself as indicative of bad faith was made clear by the General Court in pelicantravel.⁴⁰ In pelicantravel, the applicant sought to register as an EU trade mark a mark almost identical to an earlier registered mark for a wide range of goods and services. The General Court made clear that a repetitive application of the same mark can be made with the intention of avoiding the revocation of the earlier mark due to non-use and this fact could be taken into account when assessing bad faith.⁴¹ However, in this case, it was held the refiling coincided with modernising the mark on the anniversary of the existence of the brand “Pelikan” and was also made for an updated, i.e. more specific, list of goods and services in contested classes. In other words, “All this led to the conclusion that there was honest commercial logic behind the filing, which at the same time excluded the possibility of bad faith on the part of the applicant”.⁴²

A different outcome was reached by the General Court in Hasbro,⁴³ which concerned the EU trade mark “Monopoly” re-filed for the fourth time by its proprietor. As with pelicantravel, the General Court noted that since no provision prohibited the re-filing of an EU trade mark application, such application could not, of itself, establish bad faith unless it could be proved by additional evidence. Here, the Court acknowledged the existence of such evidence, including the fact that the applicant had “admitted, and even submitted that one of the advantages justifying the filing of the contested mark was based on the fact that it would not have to furnish proof of genuine use of that mark”.⁴⁴ The Court chose to interpret this statement as indicative of bad faith resting on an intention to evade the requirement of genuine use.⁴⁵ Again, as with pelicantravel, Hasbro demonstrates the flexibility that adjudicating bodies have in interpreting so-called “objective” circumstances (in this case the content of a testimony) in order to establish the “subjective” element of intention, which underlies a finding of bad faith. Hasbro also illustrates that a given conduct is not per se undertaken in good faith just because it is in accordance with trade mark law. Therefore, a sole argument that making repetitive filings of the same mark for the same goods is not prohibited will not prevent the adjudicating body from establishing bad faith when there is “objective” evidence which can be interpreted as pointing to a flawed intention of the applicant.

A recent UK decision, Lidl v. Tesco,⁴⁶ presents an excellent illustration of the fundamental problem of differentiating good faith and bad faith conduct using the same objective facts in the context of “evergreening”. Tesco and Lidl are competing UK supermarkets. The dispute arose from Tesco’s alleged infringement of Lidl’s “wordless” mark: a yellow circle, ringed in red on a blue background which was registered in 1995. Tesco uses yellow circles ringed in red on blue background with words Clubcard Prices. Lidl accused Tesco of trade mark infringement. Tesco responded that the Lidl mark was registered in bad faith because Lidl had never used the wordless mark without text. In particular, Tesco alleged that Lidl had “evergreened” the mark by applying for re-registration in 2002, 2005 and 2007, with the new applications covering the same as well as additional goods. The result, according to Tesco, was that the application for the wordless mark was made solely for the purposes of deployment as a weapon in legal proceedings, not in accordance with its function of being used on goods or services to indicate their origin. Following the Court of Appeal decision in Sky v. Skykick, Smith J stated, citing Hasbro, that “evergreening is not indicative of bad faith on its own”.⁴⁷ Nor did the pattern of Lidl’s consequent registrations give rise to a prima facie finding of bad faith. Smith J did however concede that on the evidence there may have been an error at the time of registration but again that did not provide grounds for arguing bad faith. The court concluded Tesco did not have any prospect of success in claiming bad faith as a defence to infringement.

It is a measure of the difficulty separating good faith from bad based on the “objective” evidence, that when Tesco appealed to the Court of Appeal it was successful. Two of Tesco’s grounds for appeal were that, first, Smith J did not apply the correct test for bad faith and second, that she failed properly to consider the facts and inferences relating to the claim of bad faith. Arnold LJ in the Court of Appeal took the view that Smith J had indeed applied the right test for bad faith. However, she had not properly appreciated the evidence Tesco had submitted as evidence of bad faith. According to Arnold LJ, the fact that Lidl acted in bad faith was a “permissible inference from the facts pleaded …”. Arnold LJ also took as an admission of bad faith intent from Lidl that, as was argued by Lidl’s counsel: “My learned friend says that you must have registered it in order to obtain a wider scope of protection. I did. That is exactly what I did and I am entitled to do so”.⁴⁸ Arnold LJ then looked at the evidence presented by Tesco and took the view that, for an applicant to seek unjustifiably broad protection may amount to an abuse of the trade mark system which constitutes bad faith. He also held that the evidence, not least the successive re-registrations of the unused marks was clearly evidence of “evergreening”. However, perhaps the most interesting statement in his judgment is when Arnold said of the Lidl’s conduct: “Whether it does constitute bad faith is at least to some extent a fact-sensitive question which depends in large part on the applicant’s intentions”.⁴⁹ In other words, in a finding of bad faith, it may only be possible to decide if evidence points to a bad faith intent, if the applicant has “bad faith intentions”.⁵⁰

It is submitted that in Sky, Pelikan, Hasbro and Lidl, the decision of the court as to whether there was good faith or bad faith was not made on the basis of clear, objective evidence but rather depended upon how the court chose to interpret the so-called objective evidence. In other words, the courts were left to base their decision on their own “common sense” judgement much in the same way they might, in other circumstances, have employed the legal construct of the average consumer. Except in the latter case, the average consumer test is confined to consumers for the goods and services in question and is understood as an objective test with no need to find a subjective intention on the part of a defendant. The difficulty in deciding whether there is a subjective intention to indulge in bad faith conduct when the objective facts are elusive is amply illustrated by another recent UK High Court decision, Swatch v. Apple.⁵¹ Swatch applied to register the phrase “One more thing” for watches: a phrase frequently used by the late Steve Jobs, of Apple, at the end of a presentation. The UKIPO Hearing Officer refused the application on the grounds that Swatch was using the mark not for trade mark purposes but rather to harm Apple using parody and was, therefore, an example of bad faith conduct. Swatch appealed. In the High Court, Apple argued that Swatch’s intent was not merely to parody as such but went beyond that to malicious behaviour known as “trolling”. The judge noted that bad faith based on parodic use had not been used previously to found a bad faith objection, but also that a bad faith objection was not limited to any particular bad faith or dishonest conduct. He did conclude that Swatch may have registered the trade mark to “annoy” Apple but, in his view, that that did not amount to bad faith conduct. Looking specifically at parodic use, he held that “The point at which parodic or humorous activity of that kind would transgress the boundaries of honest business practices must depend on the nature of the humour, the intensity of its use and its consequent impact on the business interests of the recipient”.⁵² Finally, he noted that even if the use was parodic this might well be compatible with using the mark in a trade mark sense. Swatch’s appeal succeeded.

In Swatch, the court recognised that bad faith conduct needs intent. But in order to identify intent, which is a subjective factor, it had to parse the difference between parodic use of a mark which might have been designed to amuse and thus arose from good faith intent or alternatively might have been designed to offend and hence arose from bad faith intent. The problems highlighted by Swatch, it is submitted, characterise the uncertainties which arise when assessing bad faith conduct. One consequence of this uncertainty is that a variety of actions have been brought based on an allegation of bad faith conduct where there is an absence of alternative grounds for challenging the registration of a mark. Lindt is a case in point, as is Swatch. Thus, in Swatch, both the Hearing Officer and the Court agreed that the phrase “One more thing” lacked the distinctiveness to be registered as a trade mark. Nor had it acquired a secondary meaning which would give it a prior right in passing off. However, Apple was able to bring an (albeit unsuccessful) action alleging bad faith.

Another area where the application of bad faith allegation is uncertain concerns cases where it is deployed against the registration of a culturally significant sign, which results in preventing other traders from using that sign. In Bundesverband Souvenir v. EUIPO,⁵³ the applicant Freistaat Bayern, owner of a famous German castle and museum, Neuschwanstein, successfully registered the name of this site of cultural heritage for goods in a wide variety of classes. The registration was challenged by Bundesverband who alleged that the application had been made in bad faith, in order to prevent third parties from continuing to use the mark. The CJEU conceded that Freistaat Bayern had registered the sign in order to prevent its use by the appellant but was justified in doing so to prevent the appellant from taking advantage of the name. Referring to the Lindt judgment, the CJEU noted that “it is not apparent from that judgment that the assessment of bad faith must necessarily take the means used to achieve such an objective into account”.⁵⁴ Rather, the CJEU took the view, shared by the General Court, that the applicant was pursuing the legitimate objective of conserving and preserving the museum location. Interestingly, registration of “Neuschwanstein” had earlier been refused in Germany due to a lack of distinctiveness:⁵⁵ the opposite of the view subsequently taken by the CJEU.⁵⁶ The “Neuschwanstein” decision well illustrates the problem with invoking bad faith in cases concerning the registration of culturally significant symbols. In these cases, it might prove difficult to establish, on the basis of objective evidence, a lack of intention on the part of the applicant to use a given sign in the function of a trade mark with the view to merely prevent other traders from using the sign: in other words, with bad faith intent. In “Neuschwanstein” the applicant being a public body which owned and was responsible for a site of cultural heritage was interested in actual commercial exploitation of the contested mark through licensing.⁵⁷ The presence of such an intention might therefore limit the utility of the bad faith allegation in the cases of registrations of symbols of cultural heritage.

Registrations, such as that challenged in “Neuschwanstein” fall into the category of bad faith applications that do not target any specific third party, but which might have effect on competition in general. Consequently, in the absence of objective evidence of a lack of such intention, the utility of bad faith as a tool for preventing monopolisation of culturally significant signs is limited.⁵⁸ The flexibility of the concept of bad faith might, however, support a different assertion: that even when a public body attempts to register a cultural sign as a trade mark with the view to its commercialisation through licensing the mark, the prevailing motive could still be preserving cultural heritage, e.g. by preventing kitschy souvenirs from non-authorised sellers appearing on the market. Such an attempt, however, even if justified by concerns about proper commercial exploitation of a cultural sign, aims to obtain an objective lying outside of trade mark function, by means of trade mark law. It could therefore encourage allegations of bad faith.⁵⁹ This leads to a wider issue whether trade mark law is at all designed to serve as a tool for controlling the use of signs embodying cultural heritage: an aim which might be better served by the introduction of special laws governing this issue.⁶⁰

The uncertainty as to when the criteria for a finding of bad faith have been met is well demonstrated in “Neuschwanstein”. Indeed, in its judgement, the CJEU made an observation that underlined this uncertainty. According to the Court, “even in a situation where that applicant files an application for registration of a sign with the sole aim of competing unfairly with a competitor who is using a similar sign, it cannot be excluded that the applicant’s registration of the sign may be in pursuit of a legitimate objective” claiming support for this approach in the Lindt judgment.⁶¹ Regardless of how one would assess the objective pursued by the applicant in this particular case, a general statement of the CJEU that filing a mark with the sole aim on the part of the applicant of competing unfairly with a competitor might nevertheless be justified by a legitimate objective is puzzling. It contrasts with its judgement in Sky where it ruled that bad faith existed where it was apparent from relevant and consistent indicia that the mark had been filed “not with the aim of engaging fairly in competition” referring further to both categories of bad faith applications, as distinguished in Koton, namely those targeted and not targeted at a specific third party.⁶²

A related issue is the attempt to use a bad faith allegation to prevent the registration as trade marks of works of art as illustrated by two among a number of cases involving the graffiti artist, Banksy which came before the European Union Intellectual Property Office (EUIPO) Cancellation Division.⁶³ Both cases involved challenges to EU trade marks registered on behalf of Banksy by his representatives, Pest Control. The trade marks represented two of his iconic works, the “Flower Thrower” and the “Monkey”. Banksy made it clear that he had registered the marks only to prevent other parties from making commercial use of the images and to protect his works in the US and elsewhere, where, in order for them to have copyright protection, he would have to reveal his true identity. Further, he only began selling any goods once the marks were registered, stating: “Sometimes you go to work and it’s hard to know what to paint, but for the past few months I’ve been making stuff for the sole purpose of fulfilling trademark categories under EU law”.⁶⁴ The Cancellation Division ruled that the proprietor’s (or more specifically Banksy’s) actions were inconsistent with honest practices, in that the sign had been registered as an EU trade mark with no intention to use it as a trade mark according to its function, and so it had been filed in bad faith. While the decision of the Cancellation Division concerning the “Flower Thrower” became final, the decision concerning the “Monkey” was appealed. The Board of Appeal took a notably different view, annulling the contested decision and rejecting the bad faith claim.⁶⁵ Notably, the Board found no support for the allegation of circumventing provisions of copyright, indicating that an artwork can be simultaneously protected by copyright and trade mark law⁶⁶ and stressing that the five-year grace period of non-use which had not lapsed yet, weighed strongly in favour of the applicant.⁶⁷ We might take the Banksy case as an example of the admirable flexibility of the definition of bad faith conduct; or a prime example of how bad faith might insert itself into actions in ways not envisaged when the bad faith clause was included in the first EU Trade Mark Directive 89/104. After all, with his transparent honesty, it would be difficult to claim Banksy had a “sinister or dishonest motive”. On the other hand, it is hard to avoid the impression that the sole reason for filing trade marks containing Banksy’s artworks was the artist’s wish to remain anonymous. The immediate question arises as to whether this is a goal that should be legitimately pursued under trade mark law. These cases also illustrate a more fundamental uncertainty, which arises when judicial authorities seek to interpret the significance of “objective circumstances” when seeking to establish bad faith intent.

The CJEU in Lindt rejected other Member States’ approaches to identifying bad faith conduct which were based on external objective standards such as conduct which falls below acceptable standards of commercial behaviour as judged by “reasonable and experienced men in the area being examined”.⁶⁸ Instead it was necessary to identify the subjective intent of the party accused of bad faith. The CJEU in Lindt also held that this subjective intent could be gleaned from objective circumstances. However, later case law shows that such objective circumstances may themselves be in the eye of the beholder, not least because bad faith is defined as an absence of good faith and the same objective circumstances might potentially be evidence of either. By failing to establish an objective, external standard for bad faith, it is suggested that bad faith is an outlier in European and UK trade mark law. The fact that courts themselves may regret its absence is suggested by the recent General Court judgement in Neratax.⁶⁹ The Court held that the appellant had acted in bad faith when it registered as EU trade marks, national trade marks previously owned by a related company, and licensed them back to that company and others, to avoid their loss in future bankruptcy proceedings. The Court justified its finding, at least in part, on the fact that the EU trade marks were not registered with “the intention of engaging fairly in competition”:⁷⁰ a standard which is certainly compatible with the earlier notion of conduct which is “a contravention of accepted norms of conduct”, rejected by AG Sharpston in Lindt.⁷¹ As we shall see, the approach to public policy and morality, although at the outset was similarly uncertain, is now taking a different course, which does involve the application of objective assessments.

3 Marks Contrary to Public Policy and Morality

Article 4(1)(f) TMD reads: “The following shall not be registered or, if registered, shall be liable to be declared invalid (…) trade marks which are contrary to public policy or to accepted principles of morality”. The same refusal ground is regulated in Art. 7(1)(f) EUTMR. In UK law it is contained in Art. 3(3)(a) TMA. The refusal ground in question has a potentially very wide scope, due to the use of these open concepts. While the main role of this refusal ground is to prevent the registration of offensive marks, in recent years new issues have emerged in case law,⁷² which prompt questions about the possibility of applying this ground in new situations, which were almost certainly not envisaged when it was first included in the trade mark legislation.

From the wording of the provisions of the Trade Mark Directive, the UK Trade Mark Act and the EUTMR, it follows that public order and morality constitute two separate evaluation criteria. Therefore, it is necessary to distinguish between them properly when applying the refusal ground at hand. In practice this does not always happen and adjudicating bodies seem to qualify a certain mark under the heading of violation of both. For example, the General Court carried out its assessment of the mark containing the Soviet coat of arms as offending public policy and accepted principles of morality using the same arguments and without differentiating between those two concepts.⁷³ In the European Free Trade Association (EFTA) Court judgment in the case of the Municipality of Oslo (Vigeland’s works of art),⁷⁴ the EFTA Court noted that, while the two criteria may, in some cases, overlap, each of them requires separate consideration.⁷⁵ The difference, according to the EFTA Court, lies in the fact that, while the conflict with public policy must be determined according to objective criteria, the conflict with morality “concerns subjective values, but these must be applied as objectively as possible by the examiner”. The latter means that offence of morality must be determined by reference to the perception of the sign by the relevant public, unlike the offence of public policy which does not require establishing the perspective of the public.⁷⁶

An interesting approach to distinguishing the two grounds was proposed by Advocate General Bobek in his Opinion in Constantin (trade mark “Fack Ju Göhte”)⁷⁷ recognising, first, the “top down” nature of public policy as expressing the public regulator’s wishes as to the norms to be respected in society. According to the AG the content of the public policy clause should be ascertainable from official sources of law and/or policy documents.⁷⁸ Second, the AG described the clause of “accepted principles of morality” as having a “bottom up” meaning that concerns values and convictions currently adhered to by a given society, set and enforced by the prevailing social consensus within that society at a given time.⁷⁹ This approach appears to correspond with the concept of accepted principles of morality as “subjective values” as expressed by the EFTA Court in the “Vigeland” judgment.

An example of a decision based on establishing the offence of public policy, in the manner described by AG Bobek in Constantin, is the judgment of the General Court concerning the mark “Cannabis Store Amsterdam”.⁸⁰ The Court found that something being against the law is not necessarily the equivalent of it being contrary to public policy as it is also necessary to establish that the unlawfulness in a given case affects an interest considered as essential by the Member State or States concerned, according to their value systems. Using this criterion, the General Court considered the mark composed of the phrase in question surrounded by stylised green lettering as being contrary to public policy, stating that the consumption of the narcotic substance derived from cannabis remained prohibited in many Member States for reasons of public health and that combating the dissemination of this substance is characterised by particular vigilance. Thus, the legal regime regulating the consumption of this substance falls within the concept of “public policy” for the purposes of Art. 4(1)(f) TMD and Art. 7(1)(f) EUTMR.

Turning to the second ground, that is the assessment of a trade mark as being contrary to accepted principles of morality, this was the subject of the landmark judgment of the Court of Justice in Constantin, concerning the trade mark “Fack Ju Göhte”.⁸¹Constantin concerned the traditional area of application for this refusal ground, that is trade marks which are offensive as such. Nevertheless, the CJEU’s judgment provides wider guidance on the concept of accepted principles of morality in trade mark law. In the Court’s view, the notion of accepted principles of morality ought to be interpreted “in the light of its usual meaning and the context in which it is generally used”; that assessment must take account of “fundamental moral values and standards to which a society adheres at a given time” and “due account is to be taken of the social context, including, where appropriate, the cultural, religious or philosophical diversities that characterise it, in order to assess objectively what that society considers to be morally acceptable at that time”.⁸² This might at first glance appear to contradict the position of the EFTA Court in “Vigeland” which saw subjective values as the source of accepted principles of morality. Also EUIPO Guidelines state that while the assessment of violation of public policy is based on objective criteria,⁸³ violation of accepted principles of morality “concerns subjective values, but these must be applied as objectively as possible by the examiner”.⁸⁴ It seems, however, that it is not just the objective assessment on the part of the examiners, who should not base their decisions on their personal views, that counts. Indeed, if accepted principles of morality are to be understood as rules to which a society, and not particular individuals, adheres, it is hard to describe such values as subjective. However, as observed earlier, the word “subjective” appears to have been used by the EFTA Court in a specific sense to juxtapose accepted principles of morality as depending on the public’s perception, with public policy where this is not the case. As a result, there seems to be no contradiction between the EFTA Court and the CJEU in Constantin on this point.

The CJEU also stressed in Constantin that the social consensus on the understanding of morals might change over time and might be territorially diversified. Further, the Court provided more specific guidance by naming two elements that are essential for a proper assessment. The first element is the necessity to take into account the perception of a reasonable person with average thresholds of sensitivity and tolerance. The second element relates to the need to take account of the context in which a mark applied for might be encountered and thus not to limit the analysis to the mark itself (or its components).⁸⁵ These elements are considered below.

In the light of this framework the Court considered the mark “Fack Ju Göhte” filed as an EU trade mark as not being contrary to accepted principles of morality. The mark, containing the misspelled variation of the vulgar English phrase “fuck you”, as well as the misspelled name of the preeminent German literary figure, Goethe, was the title of a very successful German comedy film which was followed by two sequels. The EUIPO rejected the application on the ground of the mark’s being contrary to accepted principles of morality and the Board of Appeal of the EUIPO and the General Court concurred. The Court of Justice took a different view. It set aside the earlier judgment of the General Court and simultaneously annulled the decision of the Board of Appeal. In taking this position, the CJEU pointed out that both the General Court and the Board of Appeal had erroneously relied solely on the vulgar character of the English phrase “fuck you”, having failed to make a contextual assessment and take account of factors that were a priori indicative of the social acceptance of the mark among the German-speaking general public. These factors included the great success of the film, the fact that its title had not appeared to have caused controversy, and the fact that the film had been used for educational purposes, notably by the reputable Goethe Institute which promotes the German language worldwide.⁸⁶ Additionally, the CJEU made a short statement that freedom of expression, as guaranteed by Art. 11 of the Charter of Fundamental Rights of the EU, must be taken into account when applying Art. 7(1)(f) EUTMR.⁸⁷

While after Constantin more marks are likely to pass the “immorality” check, they will often still be subject to a refusal on other grounds, in particular lack of inherent distinctiveness, descriptiveness or the customary character of a mark. A good example could be the EU trade mark “Brexit”, which was not considered by the Grand Board of Appeal of the EUIPO as offending morality just because a part of society in the UK that was against the UK leaving the European Union could be upset when encountering this mark.⁸⁸ It was, however, refused for lack of distinctiveness.⁸⁹ Similarly, several EU trade mark applications for “Black lives matter” were refused as offensive for some goods covered by the applications, but eventually were all refused in entirety because of a lack of distinctiveness.⁹⁰ Another example is the mark “Fucking Awesome”, which was the subject of a recent ruling of the General Court,⁹¹ in which the Court sided with the Board of Appeal in finding that the mark lacked distinctiveness in light of Art. 7(1)(b) EUTMR, as it merely contained a promotional expression that could not serve as an indication of the commercial origin of the goods at hand. This mark, however, had initially been rejected by the examiner as being both offensive and indistinctive. In contrast, the Board of Appeal considered this mark a slightly vulgar promotional expression and did not find it inadmissible in the light of Art. 7(1)(f) EUTMR, citing Constantin,⁹² whereas it sided with the examiner insofar as the latter had refused the mark as lacking distinctiveness.

The framework set out in Constantin leaves a number of open questions. The first concerns the point of reference for assessing the violation of accepted principles of morality. The CJEU stated that “the examination is to be based on the perception of a reasonable person with average thresholds of sensitivity and tolerance”.⁹³ This was the model applied in earlier case law that also stressed that the assessment of an offence against morality cannot be based on the perception of the sign in question by that part of the public who is easily offended or conversely is non-sensitive.⁹⁴ Consequently, the mere fact that a mark is in bad taste, is not a sufficient basis for refusing the mark as being contrary to accepted principles of morality; it is necessary to take into account the average sensitivity threshold.

In Constantin, the CJEU is referring to the model of a reasonable individual with an average threshold of sensitivity and tolerance and not expressly to the average consumer. The average consumer is of course a legal construct who has specific features of being reasonably well-informed, circumspect and observant.⁹⁵ This is different to the criteria relied on by the CJEU in Constantin. Defining the model person relevant for assessments of offensive marks is not necessarily clear in the light of the earlier case law. In some earlier cases courts linked this assessment with the perspective of relevant public composed of reasonable consumers with average sensitivity and tolerance thresholds⁹⁶ or expressly referred to the average consumer of the goods and services in respect of which registration is sought.⁹⁷ The perception of targeted recipients of good or services covered by the mark was not, however, considered as decisive in the judgment concerning the mark “PAKI”,⁹⁸ which is a word perceived in the English-speaking countries of the EU as a racist phrase, offensive to minorities of Pakistani origin. In this judgment the General Court found that even if the mark “PAKI” was filed for specialised goods and services intended exclusively for logistics companies and freight forwarders and this group of recipients did not attach any discriminatory significance to the term in question, persons other than direct buyers of the goods and services in question might still encounter this sign. Consequently, the registration of “PAKI” as an EU trade mark was inadmissible. One should stress that in PAKI the General Court neither referred to the average consumer, nor even mentioned consumers, but instead used the model of the average reasonable person with average thresholds of sensitivity and tolerance,⁹⁹ reiterated in Constantin. It is also worth noting that in the early seminal decision concerning the EU trade mark “SCREW YOU” the Grand Board of Appeal of the EUIPO stressed the need to refer to standards and values of “ordinary citizens” when assessing the offensive character of a mark, not mentioning the average consumer.¹⁰⁰ In UK decisions one can find references to “the concept of right-thinking members of the public” when assessing immorality of marks.¹⁰¹

Criteria linked with sensitivity and tolerance imply a value-oriented perspective of a model person who is neither overly sensitive nor, at the opposite extreme, not sensitive at all. Their perspective does not appear to coincide with that of a reasonably well-informed, circumspect and observant average consumer that is used elsewhere in trade mark law.¹⁰² It also seems difficult to accept that if accepted principles of morality are those to which society adheres at a given time, the violation of those principles needs to be assessed from the perspective of the average consumer of goods and services covered by the mark in question. While the model of the person whose assessment is relevant for establishing offence of morality should be normative,¹⁰³ just as the model of the average consumer, this assessment should rather focus on the perception of a model member of society, who has average threshold of sensitivity and tolerance, but is not necessarily the average consumer of goods covered by the mark. Moreover, normally the relevant public that might encounter the mark is not just the public that is made of actual or potential recipients of goods or services covered by the mark, as illustrated by the PAKI judgment.¹⁰⁴ It will usually be the general public. There might be circumstances when the public which might encounter the mark placed on certain goods will be limited, e.g. because of distribution of these goods in points of sale that are entered by the public with a specific interest. An example is the BoA’s “Screw You” decision, where the BoA allowed for a more relaxed assessment of the mark in question in the case of goods that were sold in sex shops, as opposed to assessing the mark for goods to which the general public, including children, is exposed.¹⁰⁵ This does not, however, preclude assessing the permissibility of such marks using the criterion of the presumed sensitivity and tolerance of an “ordinary citizen”. It reaffirms the view taken here that the perception of the average consumer is exiled from the assessment of whether marks offend against accepted principles of morality. Even if the result of this assessment from the perspective, first, of a member of the society and, second, of the average consumer – or in a wider sense – recipient of goods covered by a given mark, might in certain circumstances factually overlap, the position argued here means, however, that average thresholds of sensitivity and tolerance possess their own significance as the relevant criteria for the assessment of immorality of marks.

Exiling the average consumer from the assessment made under the “morality” refusal ground can, it is argued, be viewed as a desirable development. The model of a person with average thresholds of sensitivity and tolerance whose perception is relevant for assessing violation of accepted principles of morality has particular importance for marks that offend a specific social group, for instance, an ethnic minority, a religious group or the LBGTQ+ community. The traditional UK approach has been that such marks should not be registered.¹⁰⁶ For example, the mark “Jesus” was refused,¹⁰⁷ whereas the mark “Pooja Sweets and Savouries”, containing the term “pooja” describing a form of Hindu worship was registered only because the UKIPO decided it was not objectionable to religious Hindus.¹⁰⁸ Had it been highly objectionable to that group, it would not have been registered. Similar sensitivities were taken into account when refusing the trademark “Budha” in Germany¹⁰⁹ and “Buddha” in Poland.¹¹⁰ The issue of a mark offending minorities is particularly sensitive. As far as the Constantin judgment is concerned, the fact that the title of the German film did not reference a particular social group, meant the CJEU did not address the issue of marks offensive to particular groups, including minorities. It would be undesirable, if the position expressed in Constantin taking the model person of average thresholds of sensitivity and tolerance served as a vehicle to enable the registrability of such marks under EU law.¹¹¹ Arguably, the position of the General Court in PAKI that a racist term that pertains to a specific ethnic minority can be offensive to the general public, in this case the English public, will still be relevant after Constantin, especially as the General Court had employed the same model of a reasonable person with average threshold of sensitivity, which the CJEU used later in Constantin. One could therefore argue that this model person does not have to be directly offended by a mark as a member of a minority group, nor does he or she have to be a consumer of the goods covered by the mark. Rather, they can still find a mark that offends the sensitivity of a member of a minority group unacceptable, because from a wider societal perspective it is not acceptable for such marks to be used in the public sphere as trade marks. In other words, the model person with an average threshold of sensitivity could be viewed as also sensitive to others being directly offended.¹¹² This would also accord with the approach of the draft of Common Practice, which classifies marks that are perceived by a particular group as offensive against members of this group as marks against accepted principles of morality.¹¹³

Turning to the contextual approach, the CJEU stated in Constantin that:

The examination to be carried out cannot be confined to an abstract assessment of the mark applied for, or even of certain components of it, but it must be established, in particular where an applicant has relied on factors that are liable to cast doubt on the fact that that mark is perceived by the relevant public as contrary to accepted principles of morality, that the use of that mark in the concrete and current social context would indeed be perceived by that public as being contrary to the fundamental moral values and standards of society.¹¹⁴

Such an approach means that when assessing whether a sign is contrary to morality, one cannot focus solely on the form and content of the sign, for example, only on the linguistic (dictionary) vulgar meaning of a verbal element contained in a trade mark. Instead, the sign applied for should be assessed in the social context in which it – perhaps – already exists. The analysis of previous courts’ decisions regarding the application of this refusal ground generally shows that the approach to assessing the offensive or vulgar character of signs has been inconsistent, in the sense that the courts either adopted a purely linguistic perspective or took into account the social context as well.

For example, in the case-law of the EUIPO, the trade mark “Die Wanderhure” (German for “the wandering whore”) was considered admissible in the light of the context of this mark being the title of a successful German film.¹¹⁵ Conversely, a dictionary analysis formed the basis for the decision of the EUIPO, Board of Appeal and the General Court in Constantin case that the film title “Fack Ju Göhte” could not be registered. In the CJEU’s view, the Board of Appeal and the General Court had erroneously relied solely on the vulgar character of the English phrase “fuck you” instead of taking into account that the misspelled mark “Fack Ju Göhte” would not necessarily be perceived as vulgar by the German-speaking public. From the perspective of the contextual approach adopted in Constantin, the position of the Polish Administrative Supreme Court, in finding the inadmissibility of the mark “Pożar w burdelu” (Polish for “Fire in a brothel”), which is the name of a well-known Warsaw cabaret¹¹⁶ would arguably be incorrect. The Supreme Administrative Court justified its position by referring to the opinion of Polish Language Council which confirmed the vulgar nature of the word “brothel”, based solely on the analysis of the dictionary meaning of this word. This was the very approach which had been overruled by the CJEU in Constantin.

In Constantin, the CJEU assessed the mark only under the “morality” part of the refusal ground, noting that, since the sign in question had been refused solely on the ground of it being contrary to accepted principles of morality (as it had been established by the General Court), it was not necessary to analyse the other part of Art. 7(1)(f) EUTMR, that is the public policy clause. The contextual approach raises several further issues. First, the need to establish the context in which the trade mark is used might, in practice, benefit only those trade marks containing scandalous elements that were already known to the public prior to the filing and therefore provide a context justifying their use as a trade mark. Examples might be the aforementioned titles of films such as “Fack Ju Göhte” or “Wanderhure” (the “Wandering Whore”), as well as other expressions used in the artistic sphere such as the name of the Polish cabaret “Pożar w burdelu” (“Fire in a Brothel”). In the absence of a specific offensive expression in public speech, which would allow an assessment of whether the context of its use renders registration as a trade mark acceptable, the registering offices would have to rely on language analysis and hypothetical societal perceptions of the mark. Another area where the context might be relevant relates to words and phrases referring to current tragic events, such as “Black lives matter”,¹¹⁷ and “RUSSIAN WARSHIP, GO F**K YOURSELF”,¹¹⁸ but also to neologisms which reference current issues such as “Covidiot”.¹¹⁹ In addition, in such cases one might pose the question of whether even if a given trade mark seems to lack offensive or upsetting content under the contextual aspect, its admissibility could still be viewed as offending against public policy from the perspective of restraining the availability of such a mark in the public domain, should it become the subject matter of a trade mark monopoly.¹²⁰

Second, applying a contextual approach would mean that there should be some evidence for establishing the societal perception of the mark. This raises the question of how the registrar can obtain such knowledge if sufficient contextual evidence is not provided by the applicant. The contextual assessment must, nonetheless, be made by the registering office ex officio since it concerns an absolute refusal ground, as was forcefully emphasised in Constantin.¹²¹ Without such evidence the offices might instead rely on a purely linguistic analysis of the mark (or other analysis of the structure of the mark – e.g. a visual analysis in the case of obscene figurative or shape marks). Of course, from a procedural perspective, the applicant might have available tools in the course of the grant procedure to submit relevant evidence, for instance, when invited by the office to do so, prior to a final decision on refusal. The applicant might, however, not take this opportunity or such evidence might be insufficient or not convincing. Further, following Constantin, it is unclear how detailed such evidence should be. While undertaking a contextual assessment of “Fack Ju Göhte”, the CJEU took into account the popularity of the film, but simultaneously stressed that the contextual assessment of the mark in question allowing it to be considered acceptable was not contingent on establishing that public recognised the mark as the title of the film or as a joke. The Court noted that the absence of such evidence did not serve to establish per se a violation of accepted principles of morality.¹²² The latter statement implies that the contextual assessment of immoral marks does not necessarily require extensive empirical evidence and can be made by the adjudicating body by assessing the general social context.

Third, in the case of EU trade marks, the contextual assessment may sometimes only allow acceptance of the trade mark in certain parts of the EU, for example where a particular film or book was successful. In such cases an EU trade mark may still be refused registration if it can be considered as contrary to accepted principles of morality and public policy in the remaining parts of the EU (Art. 7(2) EUTMR). This could be the case with a trade mark that contains English swearwords, such as “Fack Ju Göhte” that might be recognised as vulgar in non-English-speaking Member States. The question then arises, which the CJEU did not address in Constantin, of whether this trade mark as an EU trade mark should also be considered in the context of countries other than Germany or Austria, where the movie with this title was not necessarily popular or was shown under a different title.¹²³

While scandalous or offensive trade marks are the most common ground for refusal of registration as contrary to public policy or accepted principles of morality, recent developments in case law have added new areas where the application of this refusal ground has been considered, albeit with varied outcomes. One example is its application in cases regarding marks which may have cultural value. The novelty of the “cultural heritage” aspect in the context of the refusal ground in question lies in the fact that in this case the source of concern is not the offensive character of the trade mark “as such”, but rather the attempt to register a culturally significant symbol as a trade mark.¹²⁴ Thus, in the above-mentioned case Municipality of Oslo (Vigeland’s artworks),¹²⁵ the EFTA Court considered the possibility of applying this refusal ground to reject a trade mark representing works of particular cultural value, the copyright protection of which had expired. The questions referred for a preliminary ruling concerned an attempt by the Oslo authorities to register images of artworks by the Norwegian artist Gustav Vigeland as trademarks after the expiry of their copyright protection, which was held by the city of Oslo as a successor of the artist. The position of the EFTA Court certainly offers points for discussion on cultural heritage in the context of the refusal ground at hand, both from the UK perspective, as well as the EU perspective, although the CJEU has not provided guidance on this issue.

The EFTA Court allowed the refusal of registration of marks containing elements of cultural heritage as an offence against public policy, on the basis that the mark consisted exclusively of an artwork in the public domain, provided that the registration of that mark would constitute a genuine and sufficiently serious threat to the fundamental interest of society.¹²⁶ By taking this position the EFTA Court accepted that the need to secure a public domain might be of a fundamental interest for a society.¹²⁷ It should be added that in the main proceedings, in which the EFTA Court had been asked for a preliminary ruling, this was eventually the justification for the rejection of the trade marks containing Vigeland’s artworks.¹²⁸ In addition, as regards violation of accepted principles of morality in relation to works of art, the Court ruled that the assessment must take account of the status or perception of the work in a given EEA state, as some works may enjoy a special status “as prominent parts of a nation’s cultural heritage, an emblem of sovereignty or of the nation’s foundations and values”. Registration of a trade mark containing an artwork enjoying such a status might lead to its “misappropriation” or “desecration”. It is also necessary to take account of the goods or services for which registration as a trade mark has been applied. The EFTA Court does not view this, however, as an absolute requirement, indicating that “desecration” and “misappropriation” might occur “in particular if it is granted for goods or services that contradict the values of the artist or the message communicated through the artwork in question”¹²⁹ and that account should be taken of the goods or services “where relevant”.¹³⁰ Consequently, it seems that misappropriation or desecration of an artwork might be offensive because of a mere monopolisation of the artwork by trade mark registration without necessarily establishing a degrading effect resulting from its attachment to specific goods and services covered by the application.¹³¹ Finally, the Court held that it may also be necessary to consider the extent to which the sign contains elements, which may lead to a desecration of the work.¹³² The Court found that the assessment should be made from the perspective of the average consumer in the EEA state.¹³³ As already argued in this paper, the posited relevant model should be that of a member of the society with average thresholds of sensitivity and tolerance, which does not always overlap with that of the average consumer of goods covered by a trade mark application.

In fact, as far as signs containing elements of cultural heritage are concerned, the adjudicating bodies have more commonly refused their registration on the grounds of lack of distinctiveness, as in the failed attempts to register the paintings “Mona Lisa”¹³⁴ or “Nightwatch”.¹³⁵ Relying on the “gate-keeping” role of lack of distinctiveness might , however, pose problems if a sign containing an artwork for which copyright expired, had acquired distinctiveness, and hence could be registerable.¹³⁶ By confirming the exceptional possibility of refusing such marks as offending public policy or morality, which cannot be overcome by acquired distinctiveness, the judgment of the EFTA Court signalled a significant shift in identifying those tools suitable for preventing the monopolisation of cultural heritage through trade mark registration.

The application by the courts of the refusal grounds in question to marks containing artworks is an important example of how the open concepts of public policy and morality are capable of shaping trade mark law in new areas. While the position of the EFTA Court should be perceived as a positive development, it is a perfect example of how the flexibility of these concepts will inevitably lead to uncertainty. In relation to cultural heritage it is further exacerbated by the contrast with a different approach taken by the CJEU regarding the application of the bad faith refusal ground to a trade mark containing a cultural heritage element. As we have seen, in the “Neuschwanstein” case,¹³⁷ registration of the castle’s name prevented other traders from commercially exploiting it and reserved the monopoly to one party – the German state of Bavaria. Yet for reasons of conserving and preserving the museum, the CJEU could establish a legitimate objective pursued by a public body and denied the existence of bad faith. This discrepancy between the outcomes in the “Vigeland” case and the “Neuschwanstein” case under – respectively – public policy and morality versus bad faith is significant given that in both cases the applicants were public bodies that held the custody and responsibility for the cultural objects in question.

A further question might be raised as to whether the application of the morality refusal ground, as defined by the EFTA Court, could also be important for counteracting attempts to register a trade mark composed of elements other than artworks enjoying the status of cultural heritage, but still embodying publicly valuable signs. This question should be answered in the affirmative. Early cases rejecting signs with a “positive connotation” filed as EU trade marks, such as “ATATURK”¹³⁸ or the image of the Latin cross¹³⁹ can be viewed as forerunners of this trend subsequently confirmed in the judgment of the EFTA Court with regard to cultural heritage.¹⁴⁰ In these cases, not the sign, itself, but the mere attempt to register a valuable sign as a trade mark was considered offensive. It is worth referring in this context to the optional refusal ground Art. 4(3)(b) TMD concerning trade marks that include a sign of high symbolic value. This provision contains an open-ended concept of a sign of symbolic value with religious symbols only mentioned as an example. It has been implemented in few Member States. One of these Member States is Poland, where implementation of this refusal ground mentions cultural and patriotic symbols, next to religious symbols, while still leaving it open-ended.¹⁴¹ Moreover, where this provision has not been implemented into national laws, marks including valuable symbols may be still be refused on the general basis of being contrary to public policy and morality. This is in particular illustrated by the above cited cases concerning rejected EU trade mark applications for marks with a “positive connotation”, in the absence in the EUTMR of a specific refusal ground corresponding to Art. 4(3)(b) TMD.

Furthermore, other valuable signs, not just cultural signs, can be assessed not only under the “morality” aspect but also under the “public policy” aspect. Criteria formulated by the EFTA Court concerning public policy, which might arise with attempts to monopolise signs through trade mark registration, which reference a fundamental interest of the society, could prove particularly useful in that regard. For example, as noted above, EU trade marks “Black lives matter” were considered by the EUIPO several times as offensive and possibly upsetting for people of colour encountering this expression on certain goods.¹⁴² It is also possible to argue that these marks not only offend against morality, but also against their monopolisation from the perspective of public policy because of their importance for the societal discourse. This assertion may be supported by the need to secure freedom of expression. Signs which are valuable, e.g. for political, societal or other reasons, might be employed to express a viewpoint, for example, when used on “expressive” goods, such as T-shirts or coffee mugs. Therefore, granting a trade mark monopoly over such signs for such goods might pose a threat to free speech.¹⁴³ In this scenario, unlike in Constantin, freedom of expression would not be invoked to support a trade mark registration, but to refuse or invalidate a trade mark. Arguably the registration as collective or certifications marks, rather than individual marks, could be – if trade mark law was to offer any possibility in this area – a better method of protecting signs that convey certain values important for the society, such as signs expressing social and environmental concerns,¹⁴⁴ just as the same could be true for signs containing culturally significant elements.¹⁴⁵

Finally, another area where the application of the public policy and morality refusal ground was considered was the CJEU’s judgment in the Sky case.¹⁴⁶ The request for a preliminary ruling placed the issue of too wide specifications of goods and services with lack of intent to use not only in the context of bad faith (discussed in Section 2), but also in the context of marks being contrary to public policy and morality. The High Court, when asking the CJEU whether the term “computer software” qualified as too general, expressed its view that such registration would be against the public interest. AG Tanchev agreed that the refusal ground based of violating public policy would be applicable in such a case.¹⁴⁷ The CJEU did not follow this approach. It held that the refusal ground of a mark being contrary to public policy relates to features of the sign itself rather than to the features of the application, such as a lack of clarity and precision of the terms used in the list of goods or services in the trade mark application.¹⁴⁸ This position might be viewed as contrary to the position of the EFTA Court in the “Vigeland” case, which as we have seen, takes the view that the violation of public policy does not have to rest in the mark itself, but in mere attempt to monopolise a sign. In Sky the CJEU seems to support the traditional understanding of the refusal ground at hand as limited only to the assessment of the mark. The view taken by the CJEU might be read as preventing the refusal grounds of public policy and morality from being too widely applied, in particular by introducing a refusal and invalidation ground which is missing from the text of the TMD and EUTMR. As regards the issue of too wide specifications of goods and services with lack of intent to use is concerned, a different approach was taken by the CJEU with regard to bad faith as discussed earlier, in which the need to assess the subjective conduct of the applicant is a decisive factor.

4 Conclusion

The fact that both refusal and invalidation grounds analysed in this paper are based entirely on open concepts means that the situations in which they may be applied remains uncertain. As we have seen, this uncertainty has allowed parties in registered trade mark actions to seek their application in novel situations, albeit with varying success: a fact which the draft Common Practice directions, which cover large and open-ended areas of application, merely emphasise. For example, attempts have been made to plead both grounds in circumstances which were not envisaged when the trade mark legislation was drafted. A case in point is Sky where, as we have seen, the defendants sought unsuccessfully to apply the public policy grounds not to the nature of the mark itself but rather the content of the application. However, we have not argued here that such openness does not also have important advantages. In the “Vigeland” case, for example, application of the public policy refusal ground allowed preventing the monopolisation of the name of an important cultural artefact and ensured that it remained in the public domain. It is of course of interest that the same result was not reached in the “Neuschwanstein” case, where the CJEU took a contrary view that it was not bad faith to seek to protect the name of a site of cultural heritage from wider commercial use.

We have also seen how the difference in application of these grounds is of fundamental significance. The crux of this difference lies in the contrasting ways that actions alleged to fall within these refusal grounds are assessed. Looking first at bad faith, this ground has been interpreted to include a subjective element in the form of the applicant’s intention. This subjective element inevitably requires an individual approach to bad faith cases. There might of course be difficulties in establishing applicant’s intention, especially in cases of alleged bad faith which is not directed at any particular third party; intention, nevertheless, remains the key element of the bad faith framework. It should be stressed that the manner in which the adjudicating bodies establish bad faith does not imply that there must first be an objective violation of the law or generally applicable rules of conduct before finding a subjective element, since the same behaviour may be undertaken in good faith or bad faith. The assessment depends on how offices and courts evaluate the objective evidence before them and is well demonstrated by the decisions concerning marks containing the artworks of Banksy and those concerning “evergreening” of non-used trade marks.

By contrast, offenses against public policy and morality are not dependent on establishing a reprehensible intention on the part of the applicant. They are to be assessed only objectively. Moreover, it remains unclear whether this objective assessment is confined, as traditionally accepted, to the mark as such, or it could also extend to instances when it is the filing for registration that raises concerns, as in the “Vigeland” case. The latter possibility seems, however, to have been rejected in Sky by the CJEU. In the case of offense against morality, the CJEU in Constantin stated that “the examination is to be based on the perception of a reasonable person with average thresholds of sensitivity and tolerance”, taking into account the objective circumstances in which the allegedly offensive mark would be used. This is a normative model that does not, however, appear to be the same as that of the average consumer of goods and services covered by application or registration concerning a given trade mark. We have argued here that this approach supplies a sound foundation for denying registration to signs which might offend only a minority of the population since the reasonable person does not have to be a member of that minority group or a consumer of the goods at hand. This person can still share the reasonable view that a mark which offends a member of a minority group is unacceptable. In the case of public policy, again the courts need to apply an objective test in the sense that an offense occurs when a trade mark undermines the fundamental interests of society, assessed in a “top-down” manner, that is in the light of the public regulator’s wishes.

In the case of finding of bad faith and a finding that a mark offends against public policy and morality the courts have rejected the utility of making judgements based on the perception of the putative average consumer. Rather, as we have argued here, the vagueness of the open concepts inherent in these clauses is in part a measure of the fact that these registering authorities and the courts have had to go beyond the perspective of the average consumer. In the absence of this average consumer, the courts in relation to public policy and morality have identified two normative perspectives. The first involves the “top down” rules laid down by the public regulator (public policy). The second looks to the “bottom up” values to which a society adheres and assesses the offensiveness of the relevant sign from the standpoint of a reasonable person with average thresholds of sensitivity and tolerance. It is ironic that in early cases concerning bad faith, after the passage of the Directive, a normative viewpoint was also identified in order to measure bad faith: that is the reasonable and experienced man in the area being examined. It is this perspective which was banished by the Lindt decision. It is difficult not to think that by returning to an objective viewpoint in assessing bad faith a considerable amount of the uncertainty regarding its application would be overcome. Of course, the objective versus subjective assessment is not the only difference between bad faith and offence against public policy or morality. In the former case it is the applicant’s conduct that is assessed whereas in the latter case it is the mark as such, with the door slightly open in the “Vigeland” judgement of the EFTA Court to extend the assessment to the very filing of the mark by the applicant. Nevertheless, if objective criteria were used to assess the applicant’s conduct under bad faith then we might expect to see a fairer balance between proprietor and other traders as is now possible in cases concerning public policy and morality.