Enhanced Distinctiveness and Why ‘‘Strong Marks’’ Are Causing Us All Confusion

A ‘‘strong’’ trade mark, one which has a high level of distinctiveness, is entitled to a greater level of protection in the European Union, the United Kingdom and the United States than other weaker marks. The psychological and marketing evidence suggests that, in fact, stronger marks are less likely to be confused by consumers. Accordingly, courts and tribunals are required to ﬁnd confusion in cases involving strong marks even where the facts do not support it. The antecedence for this legal ﬁction (normative correction or public policy choice) is confused, it makes the law conceptually uncertain and now that marks with a sufﬁcient reputation are entitled to protection from dilution, the ﬁction has become unsupportable from a normative perspective. Indeed, it appears that a neutral approach – that is not taking into account that it is a ‘‘stronger’’ mark during any confusion analysis – does little, if any, harm to the proprietors of strong marks and clearly beneﬁts competitors. It is argued therefore that the rule giving more protection to strong marks (whether it is inherent or acquired strength) should be abolished.


Introduction
A ''strong'' trade mark, one which has a high level of distinctiveness, is entitled to a greater level of protection than a mark which has not reached that threshold.So, put another way, by reason of that distinctiveness courts and tribunals are required to find a likelihood of confusion in cases where otherwise they would not.This finding of confusion is the result of one of the legal fictions,1 public policy choices or normative corrections 2 which exists in the trade mark law of the European Union, the United Kingdom, 3 and the United States.However, recently the Australian High Court in Self Care IP Holdings v. Allergan Australia 4 concluded that reputation should not play any role in the assessment of the likelihood of confusion at all. 5 This ruling means that evidence showing a mark is strong neither improves nor diminishes the chances that consumers will be confused.This case presents an opportunity to re-evaluate the role of strength more widely and to assess whether it should be removed from the assessment of confusion elsewhere. 6The discussion here crosses the Atlantic because the rationales for the rules are sometimes more clearly stated in the United States than Europe and the mischief the rule taints both equally.The hypothesis of this article is that the enhanced protection given to ''highly distinctiveness'' or ''strong'' marks is a legal fiction which at best is a remanent of a time when rules against dilution either did not exist or had narrow scope.But the rule is now but a long and vestigial tail and it is something from which trade mark law should be cut free. 7

The Statement of the Rule
The very first case on the likelihood of confusion test in the trade marks directive, 8 C-251/95 Sabel v. Puma, 9 introduced the rule that more distinctive marks (whether inherently or through acquired distinctiveness) get more protection. 10Oddly, however, enhanced protection was not even claimed or in issue in those proceedings. 11But it was in issue in C-39/97 Canon v. MGM 12 and it was there the now familiar enunciation of the rule originates.After the rule was linked to a recital to the Directive 13 as well as Sabel, it was explained in the following terms: … marks with a highly distinctive character, either per se or because of the reputation they possess on the market, enjoy broader protection than marks with a less distinctive character.It follows that … registration of a trade mark may have to be refused, despite a lesser degree of similarity between the goods or services covered, where the marks are very similar and the earlier mark, in particular its reputation, is highly distinctive. 14 other countries in Europe did not follow this rule in their pre-harmonised law it may have crept into European law for no reason other than the first three preliminary references on confusion to the Court were from German courts.Indeed, the principle was almost accepted without question; 15 and by the time the fourth reference to the Court of Justice was heard, 16 this time from a Dutch court, 17 it was too late. 18The rule had become embedded.It has since been confirmed 19 on many subsequent occasions 20 and what it means in practice is that a low degree of similarity between the trade marks or goods and services can be counterbalanced 21 by an earlier mark which is highly distinctive in the assessment of likelihood of confusion. 22So the more distinctive a mark becomes the wider its scope of protection expands. 23In stark terms this means that where -in fact -there is no confusion between two marks but the earlier mark has sufficient enhanced distinctiveness, the court may have to find confusion on the basis of a legal fiction (that confusion is deemed to exist, contrary to fact, on the grounds of public policy).
It is not a rule confined to Europe, across the water in the United States the Federal Circuit of Appeals stated a similar rule in Kenner Parker Toys Inc. v. Rose Art Indus., Inc: 24 Famous or strong marks enjoy a wide latitude of legal protection … .Thus, a mark with extensive public recognition and renown deserves and receives more legal protection than an obscure or weak mark.Achieving fame for a mark in a marketplace where countless symbols clamor for public attention often requires a very distinct mark, enormous advertising investments, and a product of lasting value.After earning fame, a mark benefits not only its owner, but the consumers who rely on the symbol to identify the source of a desired product.Both the mark's fame and the consumer's trust in that symbol, however, are subject to exploitation by free riders … Thus, the Lanham Act's tolerance for similarity between competing marks varies inversely with the fame of the prior mark.As a mark's fame increases, the Act's tolerance for similarities in competing marks falls.
Accordingly, both sides of the Atlantic, the more distinctive or ''stronger'' 25 a mark is the more likely it is to be confused with other marks, but how did the law get to this intuitively unusual proposition?

Origin of the Rule
In C-425/98 Marca Mode v. Adidas AG, 26 the Advocate General 27 and the Court 28 both found the rule that more distinctive marks get greater protection to originate (or be supported by) what is now the Sixteenth Recital 29 to the Directive which, in part, reads: …whereas the likelihood of confusion, the appreciation of which depends on numerous elements and, in particular, on the recognition of the trade mark on the market … While ''recognition'' has been assumed to mean ''distinctiveness'' from almost the beginning, 30 there is no clear indication in the Travaux pre´paratoires what role ''recognition'' was meant to play and whether ''on the market'' refers to the current, future or notional marketplace.The general tenor of the Travaux, however, would caution against a wide reading being given to this phrase.The original version of the recital was much stricter (and did not mention recognition at all).It was concerned with certain rules in existing national law being contrary to free movement. 31It appears that the Commission had in mind German law, which was considered to give too much protection to trade marks with a reputation. 32Even when an amendment was agreed to broaden the ''likelihood of confusion'' test 33 it was made clear that the Commission's view remained: 34 that too wide an interpretation of the concept of likelihood of confusion would unnecessarily increase the number of conflicts between marks, with damaging effects on the free movement of trade marked goods.
But we know little more than up to this point the Commission did not think the German approach was the correct one and that a narrow conception of ''likelihood of confusion'' was intended.It is also clear when the the recital was amended to add ''recognition'' and by whom.It is mentioned in Reports of the Working Party on Intellectual Property in October and November 1987. 35But there is nothing in the available papers to indicate why recognition was added or what was intended.Or, indeed, whether the Commission maintained its view that the likelihood of 29 Directive (EU) 2015/2436 to approximate the laws of the Member States relating to trade marks (recast), recital ( 16) at the time it was Directive 89/104/EEC, recital (10). 30C-39/97 Canon v. MGM [1998] ECR I-5507, [AG42]. 31The original version of the recital in the proposal was much stricter ''The purpose of protection is to guarantee the trade-mark's function as an indicator of origin.It is essential to give strict interpretation to the above-mentioned concepts so as not to impede the free movement of goods and the freedom to provide services beyond the limits required for the protection of trademarks.The current case-law in several of the Member States affords to trade-marks a degree of protection which is to some extent inconsistent with the specific purpose of trade-mark law.'':New Trade-Mark System for the Community: Proposed Directive and Regulation (Bulletin of the European Communities, Supplement 5/80), p 8. 32 This is what is claimed in Vierheilig (1982), pp.506 and in particular p. 511. 33Removing the restriction that it was limited to cases where the likelihood of confusion was ''serious''.confusion should not be given a wide interpretation.Its early position in submissions to the Court of Justice do not add much clarity. 36evertheless, it was not this recital which was the origin of the rule in Sabel itself.In that case the Court more or less adopted without question the German position that marks with a reputation should get enhanced protection. 37It is therefore worth looking at how the pre-harmonised German law developed this principle and how it was received in its country of origin.
In early cases, German courts took the view that extensive advertising did not mean confusion was more likely, particularly where it was a weak sign. 38But from 192339 the German courts more or less consistently held40 that the more the mark had been used in trade the more protection it received.A typical statement of the rule can be seen in Karo-As: 41 The protection given by a designation depends on its degree of distinctiveness; it is therefore -depending on whether the distinctiveness is weak, normal or strong -narrower or wider.
So by the time the rule was blindly adopted in Sabel it had nearly 75 years of history behind it.But it was not unchequered.In many cases, the German courts42 just stated the rule without justification43 and throughout its history its normative basis was questioned because it did not reflect how marks were perceived in the actual marketplace. 44For instance, Alfons Kraft suggested the words ''likelihood of confusion'' in trade mark law were ''a simplified expression for what the legislature actually means, namely not only protection against confusion, but also the protection of the sign itself and its distinctiveness.''45There were of course those who saw the assessment of ''likelihood of confusion'' as purely a question of fact and not law 46 (meaning that better known marks do in fact cause greater confusion).But it might be said that the German courts essentially extended the likelihood of confusion test to grant protection against dilution; even though this had been possible under the law of unfair competition for almost as long. 47Yet this nuance has never been acknowledged or considered by the Court of Justice.
In the United States, the origin story is little more satisfying. 48In the nineteenth century, the courts had recognised ''trade marks'' (fanciful and arbitrary marks) which were protected like property and did not require any showing of confusion when applied by others; and ''trade names'' (descriptive and other marks requiring secondary meaning) which were essentially protected under the law of unfair competition and required proof of deception. 49While the need for confusion for both types of mark eventually coalesced over the second quarter of the twentieth century there remained a view that trade names warranted less protection than trade marks.The crux of the difference arose from a dispute among the judges of the Second Circuit of Appeals whether there was a need to show ''harm'' for trade names (no such thing was required for trade marks). 50When the first judicial 51 multifactorial test (in Polaroid 52 ) was introduced no distinction was drawn between trade marks and trade names but the ''strength'' of the mark (i.e.whether it was a trade mark or trade name) was made a relevant factor to be considered (albeit no harm requirement subsisted). 53In simple terms, it turned a black and white approach for distinguishing between two types of mark into one which had various shades of grey. 54Thus, the original role of the ''strength'' was to determine whether it was a mark which might be classed as ''low distinctiveness'' (formerly, trade name) or one which is of ''ordinary'' distinctiveness (formerly, trade mark).The suggestion that very distinctive marks benefited from a legal fiction (causing confusion in law when there was none in fact) did not appear to be on the Second Circuit's mind at all and neither does it appear in the pre-Polaroid case law.The fiction is a much more recent invention stretching suggestions that stronger marks get more protection than weaker marks towards it being the case that stronger marks get ''enhanced'' or bonus protection.This rule now exists in some form across all the circuits, 55 but the widest version appears to be that stated in Kenner.
4 Judicial Reasoning Relating to the Rule It had been shown that the antecedence of the rule propagated by the Court of Justice (and the Federal Circuit) can extend to ''fictional'' confusion.Yet this is a fiction which judges in the German courts, in the United States, and elsewhere, try to justify as one based on the factual situation in the actual marketplace.This is not however a facet of the jurisprudence of the General Court or the Court of Justice neither of whom have ever properly evaluated any factual basis for the rule; 56 even when it has been extensively discussed. 57The problem is not all one sided, however, those who criticise the rule often do so without an explanation. 58Nevertheless, the factual assertions made by judges about consumer behaviour to justify the application of the legal fiction will now be briefly explored.Critically, each and every one of these assertions is based on the court taking judicial notice of a consumer behaviour without any evidence being led on the matter.

Theories Why Confusion Is More Likely
There are few clear statements as to why factually distinctiveness or reputation leads to greater risks of confusion. 59A common factual assertion by US courts is that a more distinctive mark is more easily recalled to mind than a less well-known mark.And once the mark has been recalled it is more likely a consumer, when confronted with a similar mark, will assume a connection with the known mark. 60A second assertion develops this so that once confronted with what at first appears to be a better known mark, consumers stop thinking and become careless so they do not notice the differences. 61While an imperfect recollection of the earlier mark is clearly part of the assessment in trade mark law, 62 it is also the case that the consumer is ''reasonably circumspect''. 63Consumers cannot be both circumspect and ''careless''.But even so it is difficult to imagine that many consumers would see a mark that looked similar to a mark they know well and instantly stop thinking, assume it is the mark they know, and go on to purchase; unless of course they truly are a ''moron in a hurry''. 64 further rationale put forward by the court is that where the mark is an arbitrary or fanciful and it has developed a reputation, consumers will infer that its use in relation to other goods or services must be connected 65 because consumers are more likely to associate a little-known mark with one which is familiar to them. 66This 59 In some instances it might be seen as nothing more than giving stronger marks the benefit of the doubt: Planter's Nut and Chocolate Co v. Crown Nut Co Inc, 305 F.2d 916, 924-25 (CCPA 1962). 60Virgin Enters Ltd v. Nawab, 335 F.3d 141, 148 (2d Cir.2003); Brookfield Communications.Inc v.  West Coast Entertainment Corp, 174 F.3d 1036, 1058 (9th Cir.1999); Hormel Foods Corp v. Jim Henson  Prods Inc, 73 F.3d 497, 503 (2d Cir.1996); Restatement (First) of Torts, § 731 (1938). 61 Salamander (1943) GRUR 300, 302 ;Specialty Brands v. Coffee Bean Distribs Inc, 748 F.2d 669, 675 (Fed Cir. 1984); BVD Licensing Corp v. Body Action Design Inc, 846 F.2d 727, 730 (Fed Cir.1988);  TCPIP Holdings v. Haar Communications Inc, 244 F.3d 88, 101 (2d Cir.2001); but the comment in Aveda Corporation v. Dabur India [2013] EWHC 589 (Ch), [48]: ''The human eye has a well-known tendency to see what it expects to see and the human ear to hear what it expects to hear'' relates to marks which are very similar rather than well known. 62Levi Strauss & Co v. Blue Bell Inc, 632 F.2d 817, 822 (9th Cir.1980); Rolex Watch v. Meece, 158 F.3d  816, 830 (5th Cir.2008). 63In the US, in general the reasonably prudent buyer acts with some degree of care: Hiram Walker & Sons v. Penn-Maryland Corporation, 79 F.2d 836, 839 (2nd Cir.1935).But the standard seems less fixed than in the EU: see McCarthy on Trademarks and Unfair Competition (5th edn), § 23:92- §23.94. 64A phrase used in the English law of passing off: See Morning Star Cooperative Society v. Express Newspapers Limited [1979] FSR 113; Newsweek v. BBC [1979] RPC 441; but see a different phrase used by the Canadian Supreme Court Mattel, Inc. v. 3894207 Canada Inc, 2006 SCC 22, [56] (adopting the casual consumer somewhat in a hurry). 65Osram (1937) GRUR 461, 463; TCPIP Holdings v. Haar Communications Inc, 244 F. 3d 88, 94 (2d  Cir.2001); Restatement of Torts (First) §731, Comment (e) (1938); James Burrough Ltd v. Sign of Beefeater Inc, 540 F.2d 266, 276 (7th Cir.1976). 66McDonalds Corp v. McKinley, 13 USPQ 2d 1895, 1900 (TTAB 1989).might be because there is a family of marks,67 but more commonly it arises in the context of what is sometimes called indirect confusion (or in the United States, affiliate confusion 68 ).
In Sabel v. Puma 69 the Belgian, Luxembourg and Netherlands governments submitted there were two types of likelihood of confusion: 70… (1) where the public confuses the sign and the mark in question (likelihood of direct confusion); (2) where the public makes a connection between the proprietors of the sign and those of the mark and confuses them (likelihood of indirect confusion or association) … 71 The mental processes behind direct and indirect confusion 72 were explained in the British case of LA Sugar Ltd v. Back Beat Inc: 73 Direct confusion involves no process of reasoning -it is a simple matter of mistaking one mark for another.Indirect confusion, on the other hand, only arises where the consumer has actually recognized that the later mark is different from the earlier mark.It therefore requires a mental process of some kind on the part of the consumer when he or she sees the later mark, which may be conscious or subconscious but, analysed in formal terms, is something along the following lines: ''The later mark is different from the earlier mark, but also has something in common with it.Taking account of the common element in the context of the later mark as a whole, I conclude that it is another brand of the owner of the earlier mark''.Thus, indirect (or affiliate) confusion, while not limited, 74 usually falls into a handful of categories.First, where the common element of the mark is so strikingly distinctive that a consumer would think no-one but the brand owner would be using the trade mark at all even if other elements of the mark can distinguish (e.g.''26 RED TESCO''). 75Secondly, where the addition adds a non-distinctive element where one might expect a brand extension (e.g., adding EXPRESS, WORLDWIDE or LITE). 76Thirdly, the change of one element appears logical and consistent with brand extension (from ''FAT FACE'' to ''BRAT FACE'' for a younger audience). 77ourthly, where there is a belief there might be co-branding creating a link (for instance, ''BARCLAYS PREMIER LEAGUE''). 78Indeed, these are just some examples because most cases of brand extension -using the same mark on a new product -requires some degree of mental processing to see if a new product fits the brand image. 79While indirect or affiliate confusion might be seen as more likely to be dependent on how well known the mark has become it is suggested in the next section that even this sort of confusion is not really dependent on the strength or acquired distinctiveness of the earlier mark.But first the discussion turns to judicial statements taking the opposite view.

Why Confusion Might Be Less Likely
There are judicial statements suggesting why confusion with better known marks is less likely but there are fewer of them.The main reason suggested is knowledge improves recollection.So as the Australian court stated in C A Henschke & Co v. Rosemount Estates Pty Ltd: 80 It is that, in assessing the nature of a consumer's imperfect recollection of a mark, the fact that the mark, or perhaps an important element of it, is notoriously so ubiquitous and of such long standing that consumers generally must be taken to be familiar with it and with its use in relation to particular goods or services is a relevant consideration … This represents an almost polar opposite view to the ''careless'' reasoning put forward above -a consumer becomes more careful the better they know the mark because they will spot the differences.The idea that consumers spot differences as familiarity grows has been suggested in many jurisdictions over the years. 81If this factual assessment were right, it might be another way of expressing the idea that once a mark is sufficiently well-known something close to a side-by-side comparison becomes appropriate.This would have a profound effect on observing differences between the marks.Indeed, as the EUIPO Guidelines suggest an ''insignificant difference'' is one that a reasonably observant consumer would only perceive by way of a side-by-side comparison 82 and so if the mark is clearly imprinted in the mind of the consumer even insignificant differences would be noticed.

The Psychological and Marketing Evidence
The nature of confusion in the marketplace has received considerable attention in the marketing, psychology and consumer literature. 83In this extensive literature there is nothing directly answering the question of whether a stronger mark is more likely or less likely to lead to confusion in the marketplace 84 and many of the related studies might not withstand sustained scrutiny by a court.Nevertheless, over the decades many experiments have concluded that stronger marks make a difference to recollection.In studies of memory there is a ''familiarity effect'' meaning the more familiar we are with a thing or person the more likely we are to recognise it. 85ikewise as familiarity increases finer distinctions are more likely to be detected. 86hen we move to look at consumer products this means that as Foxman et al state: [i]nvolvement with the product class, memory of the original brand, and certainty of judgment are found to be significantly related to the likelihood of confusion for [products] … 87 [but some] ''consumers appear to be more vulnerable to confusion, regardless their brand/product class experience … 88 .This is in the context of identifying differences between marks for the purpose of determining whether two marks are treated as identical. 83For a recent and extensive literature review see Chauhan and Sagar (2021), p. 445; earlier reviews include Kasabov (2014), p. 1777; Mitchell and Papavassiliou (1999), p. 31; as to legal literature see, for instance, Jacoby (2001), p. 1013. 84Fhima and Gangjee (2019), p. 138. 85McClelland and Chappell (1998), pp.724, 724.

123
P. Johnson better able to discriminate the elements of a Market Leading brand, as compared to two weaker competitors. 89 Alba and Hutchinson put it: The cognitive structures used to differentiate products become more refined, more complete, and more veridical as familiarity increases. 90d there may be attention biases directed towards a person's favourite brand. 91n other words, these all suggest that the better a trade mark is known the easier it is to tell another mark is different.So it appears that -in fact -it would be contrary to the existing empirical evidence to suggest a stronger mark makes confusion more likely.Indeed, it would be incredible if it were true because it would mean that marketing of a brand beyond a certain point would be counterproductive and lead to the loss of sales through trade diversion.Furthermore, the studies show that those less familiar with a brand or product category are more likely to be confused when they see advertising for a product. 92hese studies do not strictly apply to indirect or affiliate confusion where differences are noticed and processed, which makes the question whether the consumer using the information considers a brand expansion likely or not.However, it is questionable that this has much to do with the reputation or distinctiveness of the mark, but rather it relates to the previous or expected future practices of a brand. 93There are some brands which consumers might expect to expand into new and unusual markets and others that they will not.The Virgin brand may be the ultimate example of surprising expansion: 94 consumers would not be surprised to see Virgin on almost any type of product and when they see the mark they would link it back the well-known mark.But this is nothing to do with it being a strong mark, but rather it relates to the behaviour of the brand in the past.Consumers expect the Virgin mark to travel to new products, but the same cannot be said for say McDonalds, another very strong mark which has only ever been used in relation to aspects of its fast-food restaurant business.But in relation to the likelihood of confusion the system does not (and should not) give Virgin more protection than 89 Friedman and Leclercq (2015), pp. 1, 20 and 25. 90  Alba and Hutchinson (1987), pp.411, 412 (this was one of propositions behind the paper).It is supported throughout the paper. 91 Domaradzka and Bielecki (2017), p. 6 92 Brengman, Geuens and De Pelsmacker (2001), pp.231, 241. 93 Gallagher and Goodstein (2004), pp.1229, 1252 (where the brand images are inconsistent confusion may be more likely because it does not fit the ''template'' in the consumer's mind). 94Virgin began as a record company but expanded into many things including retail, package holidays, airlines, ballooning, radio broadcasting, cable television, cola, vodka, cinema, music festivals, banking and then credit cards, book publishing, operating trains, cosmetics, mobile telephone networks, wines, business support, gyms, motor racing, hotels and, of course, space travel: see https://www.virgin.com/about-virgin/timeline.It also diversified into condoms, but the mark Virgin was not used.Not all the companies are still part of the original group or remained connected to Sir Richard Branson.
McDonalds. 95In general, consumer expectations in respect of brand expansion have more to do with product category than they do with the strength of a mark.A clothing brand might launch a perfume whether it is well-known or not, but the same would hardly be expected of an internationally famous oil company.Likewise, an own brand version might be launched for almost any product by a newly established supermarket.
6 So It Is Not About ''Confusion'' At All?
It can be seen that the factual case for ''highly distinctive'' or ''stronger'' marks being easier to confuse is at best weak. 96The counter argument is not only intuitively more likely, but it seems to fit the empirical evidence.It does not mean, of course, that nobody mistakes a mark they know well.We are all familiar with seeing a person across the street and thinking a stranger is a friend. 97In any event, it is not really surprising that the rule is not about confusion in the real world.Even in its German homeland, this factual basis was long ago described as ''untenable'' 98 and in the Kenner case itself, Judge Radar was dismissive of the significance of factual confusion: While scholars might debate as a factual proposition whether fame heightens or dulls the public's awareness of variances in marks, the legal proposition is beyond debate. 99ere any many other commentators and courts who have likewise criticised the factual proposition.100So clearly there is another normative justification for the rule.The old German rule was, as mentioned above, to protect the distinctiveness of the mark -to prevent dilution.In Kenner it was described as a rule to prevent ''free 95 It would also lead to the undesirable outcome of parties filing evidence about their own brand expansion practices (and this could cause brands to use their mark on strange products simply for litigation advantage). 96There are other presumptions applying to confusion.Most notably that confusion is presumed when the marks are identical and the goods and services are identical: see TRIPS Art.16(1); and see the absolute protection granted by Directive (EU) 2015/2436 to approximate the laws of the Member States relating to trade marks (recast), Arts.5(1)(a), 10(2)(a) and recital ( 16). 97This is acknowledged in the context of identification testimony: see R v. Turnbull [1977] QB 224. 98 Vierhelig (1975), pp.534, 540 (''kein Streit mo ¨glich sein''); also see Dietze (1949), pp.963 F.2d 350, 353 (Fed. Cir. 1992).
riding'' and to protect investment in stronger marks. 101This suggestion moves the United States somewhat towards the position in pre-harmonised German law where confusion is side-lined in the analysis of likelihood of confusion. 102So why has the rule survived?It is difficult to disagree with the suggestion of Barton Beebe and Scott Hemphill: it fits with the overall trend ''towards enhanced protection for the strong against the weak''. 103This takes us to a key question.Should confusion analysis be bent into the shape of a law to prevent ''free-riding''?

Free Riding and Dilution
Free riding is closely linked to dilution: it is a type of dilution in the European Union 104 and the justification for dilution protection in the United States. 105ignificantly the legal fiction that strong marks more readily cause confusion was created when there were places which dilution did not yet reach.In Germany, when the fiction was first introduced in 1923, dilution had yet to be recognised under the laws of unfair competition 106 and even when it was discovered the courts largely continued to rely on the broader confusion analysis in trade mark law. 107In the European Union, the rule in Sabel was adopted before the rules preventing dilution were extended to similar goods (rather than covering only non-similar goods). 108In the United States, Kenner was decided before the enactment of the Federal Trademark Dilution Act; albeit there were many state laws preventing dilution. 109nce the interplay between dilution laws and confusion are considered a significant divergence appears between the EU and the United States.
In the European Union, the evidence which is used for establishing enhanced distinctiveness is the same as that for establishing a mark is sufficiently distinctive to function as a trade mark in the first place. 110However, enhanced distinctiveness is more closely aligned to reputation than it is distinctiveness itself.The extent of the reputation necessary for protection against dilution 111 is that the mark must on the relevant date 112 be known by a ''significant part of the public'' concerned with the goods or services in question 113 and territorially the reputation must extend to a substantial part of the relevant territory. 114Like reputation, it is sufficient to establish enhanced distinctiveness throughout a substantial part of the relevant territory. 115In contrast, where a mark needs to acquire distinctiveness to be registered in the first place it must do so ''throughout'' the relevant territory. 116This means that a mark which has acquired distinctiveness or a reputation across a substantial part of the territory, but not across the whole of it, cannot be registered. 117n short, there is a ladder with reputation being on a higher rung than enhanced distinctiveness. 118It is therefore possible to benefit from enhanced distinctiveness without having a reputation (but not the other way around). 119Indeed, while ''reputation'' has a distinct meaning in the context of EU trade mark law it is largely what is being gauged when a court considers whether a mark benefits from enhanced distinctiveness.It is difficult to imagine a case where a mark with a reputation sufficient to engage dilution protection needs the legal fiction attached to likelihood of confusion to make it over the line.If it is that close to causing confusion, then the said same use is almost certainly free riding/taking unfair advantage.This leaves the only marks benefiting from the fiction as those which have sufficient renown to warrant enhanced distinctiveness, but not enough to have a reputation for the purposes of dilution.Do we really need a fiction just to protect such a small class of marks?
In the United States the picture is somewhat different as there are variations in who might be eligible for protection against dilution and who can rely on confusion alone.Dilution laws are restricted to marks ''famous'' among the general public120 there is no niche fame121 -whereas fame among a class of consumers is enough to make a ''strong'' mark for the purposes of confusion.122Indeed, fame is black and white for dilution, but there are shades of grey when it comes to strength for confusion. 123Furthermore, the Trademark Dilution Revision Act protects against blurring and tarnishment but it does not extend to free-riding (or unfair advantage). 124So, in contrast to the EU, reliance of dilution alone would mean those with only niche fame would receive no additional protection and nobody would be able to prevent pure free riding.
On the first point, it might be argued that Congress specifically limited protection to marks with a higher threshold of fame. 125Further, by that time ''unfair advantage'' as a type of dilution was widely known about as it had been adopted by the European Union a decade earlier.So if Congress did not extend dilution protection this far, why should confusion be stretched to do so?In simple terms, the purpose of the confusion test was to do just what it says; and Congress was very strict in relation to the marks to which it gave more protection.So, while the class of marks that are protected by the legal fiction (but not against dilution) is far greater in the US than the EU, it requires the courts to tread where Congress did not want to go.This is nothing, however, to the uncertainty created by giving strong marks enhanced protection

Conceptual Uncertainty
A mark which is more likely to be confused because it is ''stronger'' leaves a void in its wake. 126Tribunals have to work out the effect of the legal fiction.But there is no rationale ever put forward as to what is too close -the usual wording ''free riding'' is a matter of intent and not consumer perception.Indeed, the fundamental problem is that once the legal fiction is engaged the law is not preventing confusion at all, but rather it is giving amorphous protection to brand strength.This occurs despite the fact that brand strength is explicitly protected under the law of dilution with its bounds and fetters distinctly set.By protecting brand strength under the law of confusion, the dilution regime and its own internal balance is upset. 127Even if this is put to one side, practical issues arise.
Once confusion is not the touch stone for being too close (because in fact it does not exist) what becomes the determining factor?Is it just being a little too similar?The scope of dilution has touchstones -being detrimental to the distinctive character of a mark (blurring) or to its repute (tarnishment) or taking unfair advantage of the mark -but no such touchstone exists for confusion analysis when applied to strong marks.And as the mark gets stronger so the scope of protection grows.So slightly less similar marks become infringing.And so on and so on. 128ut at each point there is no guidance how wide the scope of protection becomes or what the tribunal is actually assessing other than pure similarity.Indeed, the expanding protection itself causes problems.There must be a limit on how far the penumbra of the mark can expand as the mark get stronger; but no court has given an indication beyond the rule that the marks and goods (or services) must actually be similar to some degree. 129So we have a legal fiction with no clear markers and no guidance to the parties or tribunals.All we are left with is uncertainty.

Inherent Distinctiveness -A Stronger Case Against Enhanced Protection
In both the EU and US, the rule that more distinctive marks (stronger marks) get enhanced protection applies whether that distinctiveness is inherent in the mark itself or acquired from its use in the marketplace.As we have seen, the justification for the legal fiction is that it prevents free-riding on stronger marks.But where a mark is inherently strong there has been no investment in the trade mark at all; and so there is nothing to ''free-ride'' upon. 130In the case of acquired distinctiveness, the greater and more wide spread the use of a mark the stronger it becomes.But where the strength of the mark is not based on any use at all, but it is entirely inherent, an arbitrary decision is made by a tribunal to give a mark a wider penumbra of protection.It has been argued that the legal fiction cannot be supported even where a trade mark proprietor has invested heavily in a mark.In the absence of that investment, there is simply no basis not for the counterfactual to apply at all.

Changing the Rule
This final section considers both the effect of changing the rule, and what continuing role there should be for the strength or distinctiveness of a mark in the assessment of consumer confusion.It concludes by proposing that the best way forward is to remove enhanced protection for strong or highly distinctive marks (whether inherent or acquired) from the assessment of confusion altogether adopting the outcome, if not the rationale, of the Australian High Court in Self Care IP Holdings: reputation or strength should be a neutral factor in assessing whether two marks are confusingly similar.

Practical Significance
If one is proposing a long-standing rule should no longer be applied, it is important to have a realistic assessment of the effect this might have in practice.Over the last two decades, there have been some empirical studies looking at the basis upon which trade mark confusion decision were made, including whether distinctiveness or strength were considered.When relying on results from empirical studies it is important to take into account the external factors that might have affected the results.The most obvious factor is settlement.Overall, in the period between 2005 and 2022, 29.87%131 of oppositions132 which were started at the EUIPO resulted in a decision being issued and the remainder settled. 133Before the USPTO where opposition and cancellation proceedings are substantially more expensive 134 the annual average rate of decisions being issued was far smaller at 1.9% and 2.1%, respectively. 135The data for trade mark infringement cases is more opaque but it appears that the figure before the USPTO is only slightly lower than that for adjudication of disputes before Anglo-American courts. 136In oppositions, cancellations and infringement actions it is probable that cases involving stronger marks are more likely to settle.This is because richer parties tend to own stronger marks 137 and have greater market power to encourage settlement.In other words, many of the disputes where the strength of the mark is material do not get to an adjudication in the first place.This could lead to under-reporting of the significance of the factor.
There are further caveats when considering empirical data, for instance, a party whose use of the mark has made it stronger may rely on the mark's inherent qualities alone to avoid the time and money needed to collect and file evidence demonstrating the mark's strength. 138Even if there is evidence filed, a final decision is given, and the strength of the mark is considered then some of the judge's thinking is not going to be written down in the judgment, or where it is explained it may only be in summary. 139ith these caveats in mind, the empirical studies of the confusion assessment included examination of the role mark strength played in decisions of the General Court of the European Union, 140 the Dutch courts, 141 and also the US courts. 142In the first, it was found that inherent distinctiveness was not considered in the majority of decisions of the General Court 143 and acquired distinctiveness was only considered in 18% of them.Breaking this down further, 7% of cases overall considered distinctiveness irrelevant and a further 3% did not make it clear how it affected the decision.In the 8% of cases where it was considered and relevant it was 134 This is largely because oppositions and cancellations before the USPTO are heard before the Trademarks Trial and Appeal Board and are subject to rules very similar to those before the Federal Courts (including discovery) and so proceeding to judgment is much more expensive than at the EUIPO. 135This is based on the figures provided in the each annual USPTO, Performance and Accountability Report (FY 2005-2022), Table 23 (or in some cases Table 24) which provides a ''Summary of Contested Trademark Cases''.The table sets out how many cases were disposed of before the oral hearing and after such a hearing. 136Mulcahy and Teeder (2022), pp.326, 337-338 (the rate is usually about 4% of cases issued reach a final determination after trial albeit this did not relate to intellectual property claims, but to business in one division of the English High Court); Kesan and Ball (2006), pp.237, 271 (this had a slightly higher figure with about 15% of cases having a court decision).In relation to the Dutch courts see Haarhuis and  Niemijer (2006), pp.71 at 88 (suggesting quite the reverse about 85% of cases reach a final judgment). 137At least when the strength is acquired and not inherent. 138For instance, in AMAZON (O/873/22) the well-known online retailer led no evidence of its reputation and the case was decided on the basis of what a consumer would consider ''Amazon'' to mean without any knowledge of the brand. 139 Edwards and Livermore (2009), pp.1895, 1927-1930.   1Simon Fhima and Denvir (2015), p. 310. 141Anemaet (2020a), p. 1008.
142 Lim (2022), p. 1285; Beebe (2006), p. 1581 143 Fhima and Denvir (2015), say 68% did not consider it.split five to three in favour of there being a likelihood of confusion. 144The Dutch study, looking at its domestic courts applying the same harmonised rules, did not find any cases where enhanced distinctiveness was expressly considered, but the distinctiveness of the mark was mentioned in 80% of cases. 145It also found that where a mark was said to be highly distinctive (whether inherently or acquired is not specified) then the proprietor of the earlier distinctive mark won in 87.7% of cases (compared to 70% when the mark was distinctive and 43.9% when it was slightly distinctive). 146While this does suggest that strong marks do better before the courts, the study did not disaggregate the data.Were the stronger marks found to be more similar to the contested mark in the first place?Were the goods or services closer? 147In any event, in both the General Court and before Dutch courts there seems to be a negligible role for enhanced protection for highly distinctive marksthat is the application if the legal fiction.
In the US, in a study of confusion cases between 2016 and 2021, the strength of the mark was found to come up in about 70% of cases with 47% of the overall cases suggesting it favoured the claimant (and 15% of overall cases favouring the defendant). 148Once the factor was found to favour a party, the plaintiff won in 46% of cases and the defendant in 71%. 149In an earlier study covering 2000-2004, the issue of mark strength came up in 58% of cases, 150 but this test was based on whether a mark was classified as fanciful, arbitrary, descriptive and so on. 151The acquired strength was found to improve distinctiveness in 6.9% of cases and to be lower than the inherent distinctiveness in 7.2% of cases. 152In other words, in 14.1% of cases the acquired strength of the mark was expressly considered, but its inherent strength was much more important.Finally, it was found that marks with greater inherent distinctiveness were significantly more likely to lead to likelihood of confusion than other marks. 153This does not necessarily suggest towards enhanced protection for inherently highly distinctive marks, rather it may be little more than the courts pointing out that marks with low distinctiveness are more likely to be confused.
153 Beebe (2006), pp.1581, 1637.123 Enhanced Distinctiveness and Why ''Strong Marks''… legal fiction helped a mark over the line.But while there is no data addressing this precise question, the direction of the empirical evidence is clear enough to show that that abolishing the legal fiction would be neutral for trade mark proprietors who believe their mark is strong (as it appears around half benefit and half lose out from filing evidence154 ).

Low Distinctiveness
In the EU, in the absence of a challenge to a mark's validity, an earlier mark must be treated as having the minimum degree of distinctiveness necessary to be registered. 155It cannot, therefore, be argued that a registered mark should be treated as descriptive (and so invalid making it impossible to infringe). 156The Court of Justice has indicated that where the common elements between the signs are weakly distinctive it will usually suggest against there being a likelihood of confusion. 157But it is not the case that just because two marks have shared elements with low distinctiveness the other factors in the global assessment cannot be sufficient to create a likelihood of confusion. 158There is, therefore, no rule that the marks have to be almost identical for weakly distinctive marks to be confusingly similar. 159In other words, there is no legal fiction or public policy rule.In the United States, the basic proposition is that a weak mark 160 does not receive the same scope of protection as a strong mark. 161And as McCarthy puts it ''The weaker a mark, the fewer are the junior uses that will trigger a likelihood of customer confusion.'' 162Of course, it remains the case that a weak mark is entitled to protection and will likely be infringed when it is used on closely competitive goods. 163he factual reason for marks of low distinctiveness causing less confusion are not often discussed, but some examples of where it has been are the Eighth Circuit of Appeals suggesting in General Mills, Inc. v. Kellogg Co: 164 Determining that a mark is weak means that consumer confusion has been found unlikely because the mark's components are so widely used that the public can easily distinguish slight differences in the marks, even if the goods are related.
And two British cases also put forward reasons.In Purity Well Group v. Stockroom 165 it was said that: [t]he point about weak distinctiveness is that consumers will be less likely, depending on context, to jump to the conclusion the term is functioning in a distinctive, origin indicative way.
And in Wan Jou Lin v. General Ecology Inc, when describing the effect of combining a trade mark of low distinctiveness with a more distinctive element, it was suggested: 166 [w]hen the badge of origin is combined with other more distinctive elements in the second mark, the origin function of the first mark may fade into the background in the wider context of the mark. 167ese conclusions largely fit the empirical evidence discussed above.In other words, the rule that marks of low distinctiveness are less likely to lead to confusion is really a rule of commercial reality (of fact) rather than being a counter-weight to the legal (policy) rule that marks with enhanced distinctiveness get more protection. 168In simple terms the rule makes sense 169 and that is why it survives.This view of low distinctiveness also provides a role for recognition of the trade mark on the market 170 as required by the recital to the Directive 171 and thereby frees the Court of Justice, if it choses, to cut off at the base the legal fiction that stronger marks cause more confusion.
The basic proposition here is simple and the same as that found in Self Care IP Holdings.The reputation (acquired distinctiveness) of the mark should not be relevant to confusion at all and so there is no need to file evidence on the matter.It is not being suggested that the rule is reversed and stronger marks are less susceptible to confusion; 172 even though it appears this may be what in fact happens.This is because such an approach would provide difficulties.First, the burden of evidence would be on the party seeking to establish its rival had a reputation.But that party would have incomplete information (particularly if they were of small means) as a trade mark proprietor would not provide evidence of strength to undermine its own case.Secondly, it might force a trade mark proprietor to argue its mark is weaker than it might appear so as to make confusion more likely. 173Thirdly, and more troubling, it would mean that a party who seeks to rely on evidence to establish reputation for the purposes of dilution would be running an unfair risk. 174If they do not make the reputation threshold or fail on the merits on dilution then their own evidence of reputation might actually undermine any case based on confusion.Neutrality avoids all these issues: marks should be assessed based on their inherent distinctiveness alone and, accordingly, no mark should get enhanced protection.
It has been established that the rule that stronger marks are more likely to be confused is a counter-factual.It is a ''normative correction'' 175 to reward marks which have earnt their brand recognition.So why is the rule problematic?When it is almost never determinative of a case, what harm does it do?There are those who argue that trade mark protection is too pervasive and normatively there should be less protection for marks generally.This is usually put in the context of keeping signs free for other traders to use. 176Indeed, rebalancing the norms away from right holders is a common theme across intellectual property more generally. 177A normative argument requires the courts to change their underlying policy of giving trade marks more protection the greater the investment made in the brand; and in turn to re-balance the entire system.Even in the absence of any rebalancing, the legal fiction should be abolished because it is conceptually incoherent, costly and uncertain.
Neutrality, that reputation neither makes confusion more or less likely, has numerous advantages.First, it enables a person to view the register and know the scope of protection of a mark. 178It is true that the register will not indicate whether a mark has a reputation sufficient to engage dilution protection,179 but in many cases the marks that meet this threshold (particularly in the United States) would been known to competitors.Secondly, it avoids the difficulty of tribunals having to work out the effect of the legal fiction.They will no longer have to decide -without any guidance or rational and in the absence of factual confusion -that one mark is too similar to another.Thirdly, it avoids the filing of evidence in a wider range of cases. 180This reduces the cost of litigation; particularly where exorbitantly expensively survey evidence is used to establish the strength of a mark.181Fourthly, it does not punish successful trade marks by finding confusion less likely. 182Finally, it provides clear water between dilution and confusion and enables there to be much more conceptual coherence to the system.

Conclusion
The rule that stronger marks get more protection under EU and US trade mark law is based on a legal fiction.One bedded deeply in the laws of both jurisdictions and, in the EU, it has almost foundational status.It will likely require the US Supreme Court to consider the matter before the rule has any chance of disappearing from the United States.And in Europe it might need an amendment to the legislation itself so deeply embedded is the rule. 183But it is a rule of uncertainty.It is uncertain why it exists in the first place.It is uncertain that trade mark proprietors actually benefit from the rule.It is uncertain how big the legal fiction should be once it is engaged.It is uncertain from the register how close might be too close.And it is uncertain why it is needed now there are dedicated and robust laws relating to dilution.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http:// creativecommons.org/licenses/by/4.0/.