Spelling Confusion: Implications of the Ninth Circuit's View of the "Explicitly Misleading" Prong of the Rogers Test

Journal of Intellectual Property Journal of Intellectual Property

Law Law

Article 2

Spelling Confusion: Implications of the Ninth Circuit's View of the Spelling Confusion: Implications of the Ninth Circuit's View of the

"Explicitly Misleading" Prong of the Rogers Test "Explicitly Misleading" Prong of the Rogers Test

Thomas M. Byron

Part of the Intellectual Property Law Commons

JOURNAL OF INTELLECTUAL PROPERTY LAW

ARTICLES

SPELLING CONFUSION: IMPLICATIONS OF THE NINTH CIRCUIT'S VIEW OF THE "EXPLICITLY MISLEADING" PRONG OF THE ROGERS TEST

Thomas M. Bymn*

I. INTRODUCTION

including a unique product, from

In its traditional form, a trademark is "any word, name, symbol, or device, or any combination thereof" that enables the consuming public "to identify and distinguish his or her goods, those manufactured or sold by others and to indicate the source of the goods."' By this token, a trademark promotes an association between the label on goods and the nature or quality of the goods themselves.2 In a strictly commercial context, such an association inures to the benefit of both the trademark holder and the consumer. On one hand, a the trademark holder "a convenient means for facilitating the protection of ... good-will in trade."3 To build and retain such goodwill among the public, however, the trademark holder must ensure quality and consistency in the goods associated with the mark, lest the public mistrust the mark and balk at the associated good.4 In this way, trademarks benefit the public by facilitating predictable purchasing decisions.5 If a consumer prefers a brand of detergent, trademark protection ensures that consumer will know which brightly-colored bottle to grab from the supermarket shelf.

trademark grants

To promote the reliability of commercial decisions like this, trademark protection must also insulate consumers from being confused into purchasing the wrong brand due to similarity of marks.6 This end is largely accomplished through a trademark holder's right of action for trademark infringement,7 which

reviews purportedly infringing content in comparison with a purportedly infringed mark to determine if the former content might give rise to a "likelihood of confusion" among consumers. Courts assess the presence or absence of a likelihood of confusion by weighing the following factors (or minor variations thereof):

[T]he strength of [the senior] mark, the degree of similarity between the two marks, the proximity of the products, the likelihood that the prior owner will bridge the gap [between the two marks' markets], actual confusion, and the reciprocal of defendant's good faith in adopting its own mark, the quality of defendant's product, and the sophistication of the buyers.8

This list of factors functions well in a side-by-side analysis of two trademarks. In such a case, likelihood of consumer confusion will increase where the original mark is strong, where two marks look or sound alike, where the marks might be placed next to each other on shelves, where consumer survey evidence suggests actual confusion, and so on.

But the utility of the likelihood of confusion test has its limits. As trademarked goods take their place on store shelves and in browser windows, trademarks do not just exist as a means of differentiating goods. Some trademarks-at least the popular ones-have a way of taking on a life of their own in the public consciousness. No longer are trademarks used only to designate goods or services and facilitate commercial decisionmaking, they are also appropriated as part of common expression. For example, people don't to reference the specific search engine, they speak of just say "Google" "googling"-the act of using any search engine at all. They don't just photocopy on a Xerox copier; they "xerox," no matter what copier they are using.9 Or as Judge Kozinski put it, "Some trademarks enter our public

discourse and become an integral part of our vocabulary.. quick fix, but a Band-Aid?"10

..What else is a

In other cases, a trademark might be used expressively on a one-time basis. In this respect, one could recall the legion songs, books, and movies that reference a trademark-from Judge Smails patronizingly asking Danny Noonan if he'd like a Fresca in the movie Caddysback1' to Cee Lo Green's references to Xbox and Atari (among others) in the recent hit song "Forget You."2 Unlike the Google and Xerox examples, trademark uses like these do not treat a trademark as a generic indication for a whole class of goods or services. Rather, they intentionally refer to the specific trademarked good as a means of adding expressive content to the underlying work. The Smails request highlighted his general smarminess and out-of-touch relations with a younger generation,'3 while Cee Lo's trademark uses enabled him to evoke different levels of quality by contrasting an old, low-tech video gaming system with its more modern bells-and-whistles analog.

The use of a trademark in expressive ways such as those listed above is not a particularly good fit for analysis under the traditional likelihood of confusion test. The public consuming expressive trademark use is just not very likely to be confused into believing that an expressively used trademark indicates any affiliation between the expressive user and the trademark holder.14 Consistent with this, the factors that make up the likelihood of confusion test do not fit well with expressive trademark uses. For one, consumers are not likely to be confused due to bridging of a market gap because few companies branch out into strictly expressive works. For the same reason, the trademark holder's product is not likely to be proximate to the book, song, or movie containing an expressive use of the mark. Further, the presumable market distance between the expressive use and original mark would militate against confusion regardless of the buyer's sophistication. Finally, while the mark used expressively is exactly the same as some typically famous mark, that similarity is precisely the point of the communication-to conjure the original mark to capitalize on

SPELLTNG CONFUSION

knowledge common to speaker and listener. Such common knowledge, if sufficiently widely held, is almost necessarily exclusive of consumer confusion. It is clear, then, that other modes of analysis should, at the very least, supplement the likelihood of confusion test in cases of expressive trademark uses that warrant greater First Amendment protection.'5

At least one such mode of analysis, the so-called Rogers test, has been to address a particular species of expressive use of another's devised trademark-use of a trademark within the title of an expressive work.6 The test finds in favor of the user of another's trademark where the trademark use is both relevant to the underlying work and not "explicitly misleading."'7 As to the latter question, whether a trademark use is "explicitly misleading," the original case setting out the Rogers test was fairly clear in the intended bounds of that inquiry. "Explicitly misleading" trademark uses within a title would only occur where the title contains an overt misrepresentation that the third party trademark is the source of the expressive worki8-think The Devil's Wardrobe Was Sponsored ly Prada.'9

But as the Rogers test has been expanded to apply to uses of third party trademarks both in the titles of expressive works and ndthin such works, significant questions remain as to the meaning of "explicitly misleading" when applied to trademark uses within expressive works, particularly within the Ninth Circuit, which has only recently had its first occasion to consider the Rogers test for such an internal trademark use.20 Owing to this potential confusion, this Article attempts to clarify how the Ninth Circuit has applied the "explicitly misleading" prong in such internal situations, and how it might better clarify the boundaries of this analysis. This question, in turn, will only recur as more circuit courts follow the likely path of adopting the Rogers test for third party trademark uses internal to expressive works.

To the end of providing greater clarity to the explicitly misleading prong in its expanded context, Section II of this Article describes the genesis and substance of the two-pronged Rogers test. The Section then discusses how the Rogers test has expanded in applicability from its original proving ground of trademark uses within titles to the much broader class of trademark use anywhere within the body of an expressive work. But as this expanding field of applicability is relatively recent and untested, this Article then focuses on how the "explicitly misleading" prong might be interpreted when applied to uses of a third party's trademark within an expressive work. Along these lines, Section III of the Article offers three different permutations of the "explicitly misleading" prong as applied to such trademark uses within an expressive work. Each of the three subsections attempts to contextualize its proffered interpretation of the explicitly misleading prong along a spectrum from most to least protective of speech rights, or, in the opposite direction, from most to least solicitous of the public interest in avoiding confusion. the importance of the interpretation of the explicitly misleading prong as applied to internal trademark uses by considering a recent case involving the band No Doubt and the use of their trademarked likenesses within the popular video game Band Hero. While the court in that case was able to avoid application of the Rogers test, the very specter of application of the explicitly misleading prong to the video game's use of No Doubt trademarks underscores how differing interpretations of the prong's contours within an expressive work will lead to divergent results in precisely the same case. In view of this, the Article closes by stressing the need for clarity in the Ninth Circuit's recent expansion of the Rogers test to all uses of a third party's trademark within an expressive work.

IV emphasizes

Section

As noted above, in the context of the use of another's trademark in the title of a work of creative expression, some courts2' analyze the propriety of the

trademark use by resorting to the Rogers test. The test derives its name from the case Rogers v. Grimaldi,22 which involved the Federico Fellini film, "Ginger and Fred."23 The title of that film evokes the trademarked names of the famous dance pairing of Ginger Rogers and Fred Astaire.24 The film itself, however, did not chronicle the lives of the famous Ginger and Fred. Rather, it traced the story of two fictional TV dancers whose work in imitation of Ginger and Fred earned them the nickname of that pair.25 Ginger Rogers took umbrage at the use of her name in Fellini's title, and sued the film's distributors for trademark infringement and violations of her common law rights of privacy and publicity.26

In accord with the previous district court decision in the case, the Second Circuit found no infringement of Rogers's trademark rights. Unlike the district court, however, the Second Circuit did not agree that the Lanham Act did not apply at all to titles due to their expressive content. The Second Circuit did concur with the district court to the extent that "the expressive element of titles requires more protection than the labeling of ordinary commercial products."27 But the appeals court equally noted the partially commercial purpose of titles in that "[t]itles, like the artistic works they identify, are of a hybrid nature, combining artistic expression and commercial promotion."28 Because of this hybrid nature, the court concluded, consumers of artistic works "have an interest in not being misled and they also have an interest in enjoying the results of the author's freedom of expression."29 To balance these interests, the court proposed a novel test to assess whether the use of another's trademark in a literary title violated the trademark holder's Lanham Act rights. Under the twopronged test, a trademark violation would be found in uses where "the title has no artistic relevance to the underlying work whatsoever, or, if it has some artistic relevance, unless the title explicitly misleads as to the source or the content of the work."30

In applying the test to the case before it, the Second Circuit found in favor of the makers of the film Ginger and Fred. First, the court concluded that the title had "genuine relevance" to the underlying story involving characters nicknamed Ginger and Fred.3' The title attained further relevance as it

the challenging permitted Fellini to develop an ironic contrast between conditions of 1930s Italy and the glamour of the original Ginger and Fred.32 As to whether the title was "explicitly misleading," the court acknowledged that some viewers might be misled into believing that the film somehow tracked the lives of the original Ginger and Fred. At the same time, however, the court found the title contained "no explicit indication that Rogers endorsed the film or had a role in producing it."33 The title accordingly did not violate Rogers's trademark rights.

A few other circuits have since adopted the Rogers test as a tool for analyzing trademark uses in expressive titles.34 Perhaps more importantly, though, two courts, the Second and Sixth Circuits, have also expanded the Rogers test's applicability to uses of third party trademarks within expressive works.35 In the 2008 decision E.S.S. Entertainment 2000, Inc. v. Rock Star Videos, Inc.,36 the Ninth Circuit became the third court to follow this trend, almost as a matter of course, as the parties to that case did not dispute the extension of the test, and the court could find "no principled reason" not to expand the applicability of the test.37 the more The court undoubtedly deemed the expansion of the test all reasonable in view of previous judicial adoption of the test for similar uses.

Even in view of the test's broadening adoption, the Ninth Circuit was perhaps somewhat facile in its discussion adopting the test.38 The Rogers test, having been designed for use with what amounts to "artistic labeling," does not necessarily apply mutatis mutandis to cases involving "artistic content." While the in expression and avoiding confusion are overriding public undoubtedly present in each case, the use of a mark throughout an entire work would seem to offer many more permutations of trademark use-some more, and some less, confusing-than the range of uses possible in a title. At the very least, the use of a mark within a lengthy work is susceptible to much greater

interests

SPELLlNG CONFUSION

modulation as to the absolute number and type of uses of a third party mark as compared to the limited possibility for repetitive or varied trademark use in a much shorter title.

Perhaps as a reflection of the less-than-perfect fit of the Rogers test to all uses of third party trademarks within a work, different courts might select, and indeed have selected, different implementations of the Rogers test when applying In particular, the "explicitly misleading" it internally to an expressive work. prong of the test has provided differing permutations in support of the public In Rock Star Videos, the Ninth Circuit only interest in avoiding confusion. provided a sketch of its interpretation of what "explicitly misleading" means when applied to internal trademark uses. In view of this limited articulation, the following subsections discuss a few possible interpretations of the prong with an eye toward how they could inform a more explicit articulation by the Ninth Circuit. The options generally fall at points on a spectrum, from least protective of trademark holders and the public interest in avoiding confusion to least protective of the public interest in expression. And as additional circuits follow the few circuits that have adopted the Rogers test for uses of a third party trademark within an expressive work, these possible interpretations of the explicitly misleading prong could equally inform future judicial opinions.

initially designed

As noted above, the Rogers test was

to determine permissible uses of third party trademarks within an expressive work's title. As titles tend to serve as labels on an expressive work, the Rogers test targeted the protection of consumers of expressive works against overt mislabeling of such works, either through irrelevant or misleading trademark uses. But as the test has been expanded to cover trademark uses throughout an expressive work, it encounters a wider range of possible trademark uses, both in type and repetition. The outcome of this expansion is a test that has yet to find its bearings properly in its new setting. In particular, the "explicitly misleading" prong is susceptible to multiple different interpretations when applied to uses of third party trademarks internal to expressive works. This Section outlines three such approaches: (1) the concept of "explicitly misleading" based on the Rogers court's original understanding of the term, (2) the Second Circuit's approach to the Rogers test where "explicitly misleading" seems swept into a likelihood of confusion analysis weighed against the public interest in being exposed to the trademark-driven expression, and (3) a third approach that defines a trademark use as "explicitly misleading" where an unspecified set of likelihood of confusion factors are violated, but which does not explicitly consider the public interest in expression. Each of these interpretations strikes a different balance

between the public interest in avoiding confusion and the countervailing public interest in receiving expression. At the end of the spectrum least favorable to avoiding confusion but most favorable to the public interest in expression is the original Rogers test, while the variation on the second option, a variation on the likelihood of confusion, seems to fall on a point closer to the opposite end of the spectrum. The third option seems to go even further than the other two in protecting against confusion, and accordingly gives the least weight to the public interest in expression. The Ninth Circuit's application of the Rogers test appears to hew to this third option; but for want of data points applying the test, that approach could be better clarified in light of the available options.

On one end of the spectrum most permissive of expression but least protective of trademark holders, courts could treat "explicitly misleading" in a manner coextensive with the original Rogers case. When formulating the "explicitly misleading" prong of the test, the Second Circuit operated under the premise that the title of an artistic work could be misleading in either an implicit or an explicit manner.39 On the implicit side of the dichotomy, the court offered examples like "the hit song 'Bette Davis Eyes,' and the ... film 'Come Back to the Five and Dime, Jimmy Dean, Jimmy Dean.' "4 In other words, the court deemed "implicitly misleading" the title of an expressive work that "include[s] a well-known name without any overt indication of authorship or endorsement."41 On the other side of the dichotomy, the court focused on trademark uses that contained clear misrepresentations as to a work's source or endorsement-for example, where Nimmer on Copyright or Jane Fonda's Workout Book were used to label books that had no relation to Nimmer or Fonda, respectively.42 Or, as the court hypothetically suggested of the case before it, "if the characters in the film in this case had published their memoirs under the title 'The True Life Story of Ginger and Fred,' and if the film-maker had then used that fictitious book title as the title of the film, the Lanham Act could be applicable to such an explicitly misleading description of content."43 So by the Rogers test's logic, only where a trademark use in an expressive work title falls on the explicitly misleading side of the dichotomy does the Lanham Act regulate Implicitly misleading (and otherwise relevant) trademark the trademark use. uses, by contrast, constitute protectable expression outside the Act's purview.

SPELUNG CONFUSION

11I

This analysis has more or less been applied to internal uses of third party marks by the Sixth Circuit in ETW Corp. v. Jireh Publishing Inc.44 That case involved the use of golfer Tiger Woods's image in a painting depicting several other winners of the prestigious Masters golf tournament.45 Under the Rogers test, the court found the use of the image both artistically relevant and not explicitly misleading, a finding not contradicted by the insufficiency of Woods's survey evidence.46 A lengthy dissent faulted the majority for extending the Rogers test's applicability beyond its original context: creative titles.47 The dissent also questioned whether the Rogers test should incorporate the likelihood of confusion factors when deployed in a broader context.48

These criticisms by the dissent may indirectly express a concern that the ETW Corp. majority's application of the "explicitly misleading" prong in its original form only applies to a limited class of trademark uses. If the Nimmer in Rogers represent the outer bounds of "explicitly and Fonda examples misleading" trademark uses, it is hard to imagine many uses of third party trademarks within a work that would run afoul of the Rogers test as applied by ETW Corp. Infringing uses of third party marks would encompass little more than statements falsely claiming that the expressive work was "sponsored by" (or "brought to you by") a company.

to sponsorship, but this result

Escaping regulation, by contrast, would be a wide range of other uses of a third party's trademark. Of course, one-time mentions of a trademark within the context of a work would only explicitly mislead when accompanied by an is effectively overt misrepresentation as tantamount to the normal operation of the Rogers test in the context of a title. A more interesting case occurs where the use of the same trademark is periodically throughout an expressive work without any misstatement of repeated endorsement. For example, when the sitcom Seinfelds George Costanza chooses the word "Bosco"-the name of a brand of chocolate syrup-as his ATM PIN code, his (and other characters) repeated use of that mark might result in hilarity, but certainly not any explicit mention that Bosco was affiliated with the sitcom. Such use of the Bosco mark would accordingly be acceptable under a traditional understanding of the Rogers explicitly misleading prong. the scope of the explicitly misleading prong would be Similarly beyond trademark uses taken a step farther-by characters who swear by (or have an affinity for) specific products. Take, for example, George Clooney's Ulysses Everett McGill in the film 0 Brother, Where Art Thou?-a character who insisted

on using only "Dapper Dan" pomade to maintain his hair's oily sheen and shape.49 When confronted with only Dapper Dan competitor "Fop" to restock his diminished pomade supply, McGill indignantly refused the alternative as he was "a Dapper Dan man."50 Now, were "Dapper Dan" a real hair treatment (it's not), this use of the mark would not be "explicitly misleading," as a fictional character's affinity for a product does not rise to an explicit statement of affiliation between the filmmakers and the product. Yet both the Seinfeld and, to a greater extent, 0, Brother examples could mislead the public into perceiving such an affiliation, perhaps without a concomitant uptick in expressive value imparted to the work. In that respect, the bare "explicitly misleading" prong as applied to titles by Rogers might not suffice to protect the public interest in avoiding confusion when applied to trademark uses internal to an expressive work.51

Perhaps sensitive to such limits of the original "explicitly misleading" prong, the Second Circuit has proceeded to develop its own distinct variation on the Rogers test as applied to internal trademark uses. This variation, somewhat less protective of expression and more effective at insulating against confusion, only implicitly employs the two prongs of the original Rogers test. More specifically, the Second Circuit has adopted the theoretical backdrop of the Rogers test, but not its specific articulation, for uses of third party marks internal to an expressive work.

The case initially promoting this more limited adoption, Chffs Notes, Inc. v. Bantam Doubleday Dell Publishing Group, Inc.,52 offers a good example of this "explicitly misleading" analysis. The case was initiated by the publishers of the popular Cliffs Notes line of study aids, which provide easily digestible synopses and analyses of various texts widely deemed "literature" and often found on college English course curricula.53 The Cliffs Notes publishers claimed trademark violations arising out of the defendants' publication of "Spy Notes," satirical summaries of less literary texts, where such summaries were presented under a cover that shared certain similarities with Cliffs Notes' famous

bumblebee-colored trade dreSS.54 This evocation of the Cliffs Notes cover design was part of the defendants' intentional parody, though the defendants attempted to buffer the public against confusion by repeatedly writing the words "A Satire" on the cover of their summaries.55 Unconvinced of such buffering, the district court granted a preliminary injunction on Cliffs Notes' trademark claims under a standard likelihood of confusion factors analysis.56

The Second Circuit faulted

the District Court's simple likelihood of confusion analysis, and instead imported the Rogers test to structure its analysis, finding that "the overall balancing approach of Rogers and its emphasis on construing the Lanham Act 'narrowly' when First Amendment values are involved are both relevant in this case."57 Yet the court's adoption of the test to trademark uses within expressive works was less than complete. The court only endorsed the Rogers test to the extent that it weighed the public interest in expression against the public interest in avoiding confusion. But the latter interest, rather than relying on the original two Rogers prongs, "takes into account the ultimate test in trademark law, namely, the likelihood of confusion 'as to the source of the goods in question.' "58 On the facts before it, the court found consumer confusion caused by Spy Notes unlikely. Without calling out specific factors from the Polaroid test,59 the court considered in passing the similarity of the marks. Critical differences in the books, along with the "Satire" disclaimer, militated against a finding of confusion.60 The court also reviewed the relatively high sophistication of the purchaser, who would normally seek a specific Cliffs Notes offering to aid in coursework.61 Further, to a finding of non-infringement, the court noted that the goods were not destined to be placed on proximate shelves.62 On balance, this limited likelihood of confusion was deemed outweighed by the public's interest in being exposed to the parodic content of Spy Notes.63

Like the ETW Corp. decision, this iteration of the Rogers test raises certain questions. Here, the questions relate to the value to be placed in the interest in public expression, as weighed against the Polaroid factors analysis. Obviously, the test must view the interest in public expression as something greater than

zero. Otherwise, the Rogers test would be reduced to a simple likelihood of confusion analysis, and in that respect, rendered a nullity in granting greater leeway to expression. How that interest in public expression is quantified, however, seems open to potentially arbitrary valuation. In Clffs Notes, the court was dealing with parody, a type of expression with a long pedigree of judicial and societal acceptance as a valuable form of expression.6 Accordingly, the court in Clors Notes would face little scrutiny in affording the trademark use Its result was made all the easier by the significant public expressive value. limited likelihood of confusion engendered by the evident distinction between Cliff Notes and Spy Notes. Tougher questions, however, would undoubtedly follow from a slightly more confusing use involving a more aesthetically subjective form of artistic expression. A highly conceptual form of modern art-say, a Marcel Duchamp urinal65 with the trademark of the urinal maker still affixed-is less susceptible to a consistent, well-defined value of the public interest in receiving the expression. Different courts very easily could assign different weights to the public interests in that case; some dismissing its value entirely and finding the use infringing.

But whatever the value attached to public expression, the more rigorous Clffs Notes view of the explicitly misleading prong ensures a lesser range of expressive trademark uses than ETW Cop. and the original Rogers test. The Duchamp art furnishes just one case where a use not "explicitly misleading" under the original Rogers test could be implicitly deemed "explicitly misleading" due to its increased likelihood of effecting public confusion. Clffs Notes might equally thwart Everett McGill's pomade obsession, as the use would cause a substantial likelihood of confusion of endorsement among the viewing public, for a limited uptick in expressive value added by the pomade brand (if it were real). In other words, the Clffs Notes court's indirect treatment of the "explicitly misleading" prong is less permissive of public expression, and more protective of trademark holders, than the traditional Rogers test regardless of the weight applied to the public expressive interest. In the most extreme incarnation of the Clffs Notes test, if the arbitrary value assigned to the public expressive interest is set low enough, a case applying that text will look like little more than a standard likelihood of confusion analysis.66

SPELLlNG CONFUSION

Ostensibly more confusion-averse and expression-unfriendly than both the ETW Corp. and Ciffs Notes applications of the Rogers test, Rock Star Videos provides a vague sketch of how the Ninth Circuit might interpret the "explicitly misleading" prong as it applies to the use of a third party's trademark within expressive works. The case pitted two parties perhaps not known for catering to mankind's better instincts. On one hand, the plaintiff, ESS, operates a Los Angeles strip club known as the "Playpen."67 On the other, defendant Rock Star Games designs and sells "Grand Theft Auto," a video game in which the player can engage in a variety of criminal activities, from frequenting prostitutes to stealing cars and killing other people-all within the gaming environment.68 The dispute centered around an in-game environment in a Grand Theft Auto installment designed to parody the features of real-life Los Angeles. As part of this parody, the game included the "Pig Pen," a send-up of the plaintiffs club.69 Plaintiff duly filed suit for infringement of trademarks associated with its business.

After opting to apply the Rogers test to the use of the Playpen trademarks within Grand Theft Auto,70 the court addressed the plaintiffs argument that the video game was explicitly misleading. In this respect, the court noted that the critical query "is whether the [g]ame would confuse its players into thinking that the Play Pen is somehow behind the Pig Pen or that it sponsors [Rock Star]'s product."71 The court assessed such confusion through a series of observations that partially tracked a likelihood of confusion factors analysis. For instance, the court noted the lack of similarity between the marks and their settings.72 The court also implicitly cited to a lack of evidence of actual confusion as to an affiliation between the Playpen and Rock Star Games.73 The court expressed additional skepticism that any consumer would suspect a bridging of the market

gap between a lesser known strip club and a sophisticated video game.74 Finding nothing explicitly misleading about Rock Star's modification of the original "Play Pen" marks under this analysis, the court concurred in the prior dismissal of the suit.75

The Ninth Circuit's "explicitly misleading" analysis captures elements of both ETW Corp. and Chffs Notes. Like ETW Corp., the court adopted the full two-pronged Rogers test-not just the high-level weighing of the public interest in expression as measured against the public interest in avoiding confusion. At the same time, Rock Star Videos tracked ChO's Notes to the extent that the Rock Star Games court did not limit its explicitly misleading analysis to examples like the Nirnmer and Fonda books cited in Rogers. The court instead demonstrated a willingness to consider a broader concept of actionable confusion beyond that engendered by overt misrepresentations as to source.

While addressing a broader concept of confusion than ETW Corp., the Rock Star Videos court did not seem to go so far as to adopt the Chffs Notes court's view of confusion. Critically, the Rock Star Videos court did not purport to weigh the public interest in receiving the expression. This omission would seem, afortiori, to make the Ninth Circuit view of the explicitly misleading prong less favorable to public expression than the Chffs Notes articulation of the test.

But this conclusion cannot be reached with complete confidence because the Ninth Circuit stopped short of naming the likelihood of confusion factors test as its analytical framework for determining misleading trademark uses. The court instead sometimes implicitly and somewhat haphazardly applied a subset of the likelihood of confusion factors to the "explicitly misleading" analysis. The remaining question, then, is how future Ninth Circuit decisions should build on the Rock Star Videos initial foray into applying the Rogers test for uses of third party trademarks internal to expressive works. While the court did not seem inclined to adopt the original Rogers test view of the explicitly misleading prong as embodying overt misrepresentations, other interpretations of the explicitly misleading prong do seem legitimately viable. The court could, under a large number of possible iterations, more explicitly designate certain factors in the likelihood of confusion test as critical to the "explicitly misleading" query. Under a narrower but more arbitrary view, the court could equally hew to the entire likelihood of confusion test, like the Chffs Notes court. Under that iteration, the court would be more likely to consider the public interest in receiving expression, as well. But for lack of data points at this time, litigants will be free to rely on any of a range of arguments, even the original Rogers test analysis-as the next case profiled will show.

74 Id at 1100-01. 75 Id. at 1101.

2011]1

As discussed in a few hypotheticals above, cases may turn on a court's particular interpretation of the explicitly misleading prong of the Rogers test as applied to trademark uses internal to expressive works. The reality of this was recently cast in even sharper relief in No Doubt v. Activision Publishing, Inc.,76 yet another case involving a video game's use of a trademark. The video game in question was Band Hero, a game that allows players to simulate musical the on-screen representations of musical performance by manipulating performers.77 In other words, Band Hero, in conjunction with the proper video gaming system, allows some aspects of a player's movements to be reflected in digital performers, also known as "avatars," on-screen.78 While Band Hero does enable participants to play the role of fictional performers, they are equally able to perform as popular real-life musical groups.79

One such group is the plaintiff No Doubt, a band that initially achieved worldwide success with its 1995 album Tragic Kingdom, which included slightly edgy, but mostly radio-friendly, hits such as "Don't Speak," and "Just a Girl."80 The makers of Band Hero, defendant Activision Publishing, duly entered into a license agreement with No Doubt that authorized Activision to incorporate the likenesses of No Doubt's members into the game as avatars.81 To this end, No Doubt participated in a day-long studio session in which the band's physical features and performing motions were captured for subsequent reproduction in the game.82

The agreement with No Doubt did limit Activision's license in certain critical ways.83 All uses of No Doubt's likeness were subject to the band's approval, and the band agreed to license no more than three of its songs for performance within the Band Hero game.8 Ultimately, though, the video game did not respect the limits of Activision's license with No Doubt. Specifically, the band's avatars were susceptible to significant unapproved manipulation within the game through an "unlocking" feature that permitted players to perform other bands' songs as No Doubt, separate the members of No Doubt,

and even change the timbre of the No Doubt avatars' voices from male to female, and vice versa.85 When the band expressed its consternation at these capabilities to Activision, it learned that Activision had even hired actors to impersonate No Doubt in support of the unapproved uses of No Doubt's likeness.86

A suit by No Doubt followed, alleging, among other counts, violation of the band's right to publicity and unfair competition.87 Activision unsuccessfully moved to strike those claims on First Amendment grounds at the superior court level, and appealed to the Second Appellate Division.88

That court ultimately agreed with the superior court's denial of Activision's motion. As part of its analysis, the court felt compelled to acknowledge Activision's argument that, as a right to likeness is analogous to a trademark right, Activision's use of No Doubt's likenesses was permissible under the Rogers test. Specifically, Activision contended that its use was not explicitly misleading, where "explicitly misleading" meant "that Activision overtly represented that No Doubt approved the unlocking feature as well as the songs their avatars can be made to sing."89 After a lengthy description of the Rogers test's development and history, the appellate court concluded that it need not reach Activision's argument because Activision's use of the No Doubt likenesses did not qualify for First Amendment protection, a necessary condition to eligibility for application of the Rogers test.90 Based on cases like Comedy III Productions, Inc v. Gary Saderup, Inc.91 and Kirby v. Sega of America, Inc.,92 Use of a third party's likeness must be "transformative" to qualify for First Amendment protection under Rogers. Activision's use of No Doubt likenesses, in the appellate court's opinion, constituted no more than "precise computer-generated reproductions of the band members that do not meld with the other elements of Band Hero to become ... Activision's own artistic expression."93 The Rogers test accordingly did not apply to the non-transformative use at issue in No Doubt.

Activision's line of argument under Rogers, however, puts a fine point on the importance of the interpretation of the explicitly misleading prong of the Rogers

test. As the court noted in passing, if the original Rogers test had applied to the case, the test would have compelled No Doubt to make the "almost impossible showing that Activision's non-transformative use of the No Doubt avatars was 'explicitly misleading.' "94 There was simply no way that the Band Heo game's unlocking feature was tantamount to an overt misstatement that No Doubt endorsed either the feature or the unauthorized actions that the No Doubt avatars could perform within the game. In this respect, the Band Hero makers can be credited for a worthwhile, if ultimately ineffective, litigation strategy.

What the No Doubt court failed to acknowledge, however, was the possibility that the Rogers test could have been applied to the Band Hero trademark use case in at least the two other forms discussed above. In all likelihood, under either of those alternative articulations of the explicitly misleading prong of the Rogers test, the use of the No Doubt band members' likenesses would have run afoul of the Lanham Act. Under the Second Circuit's view in Chffs Notes, the court would have looked to the likelihood of confusion factors test in assessing the public interest in avoiding confusion. That test would likely weigh against the Band Hero use of the No Doubt likenesses. Its use involved an exact replica of the No Doubt likenesses, a strong trademark, in precisely the setting that the band had partially authorized. The unauthorized extension of the band's permission could almost not fail to confuse the consumer under the likelihood of confusion test. Some question would remain, however, as to the value imparted to the public interest in expression under this iteration of the Rogers test. One could envision a court that deems the game-playing enabled by the unlocking feature to be of some public value. That said, the likeliest scenario would not find the possibly unpredictable interest in expression outweighing the significant confusion worked by Activision's unauthorized trademark use.

A finding in favor of the makers of Band Hem would be even less likely under a version of the "explicitly misleading" prong that weighs certain likelihood of confusion factors. Under the Ninth Circuit's current articulation of the "explicitly misleading" prong, the court would review the similarity of the

a bridging of the gap between markets. Each of these would weigh heavily in favor of a finding of infringement. As with the Chffs Notes example, the trademarks used are not just similar to the originals; they are identical. While survey evidence of confusion is lacking in this hypothetical, the marks equally appear in the same market channel, which eliminates the need to consider the bridging of markets and weighs heavily against Activision. And in the absence of a countervailing public interest in receiving the expression, Activision's liability for trademark infringement would approach certainty.

In view of these likely outcomes, the No Doubt court might have done well to discount Activision's argument on the basis that it misconceived the Rogers test's explicitly misleading prong as applied to internal trademark uses. But at an even more general level, the No Doubt case clarifies how the three different interpretations of the explicitly misleading Rogers prong lead to three different probabilities as to a given outcome. Where the original Rogers interpretation of the prong would lead to a certain outcome favorable to the trademark use, the Chffs Notes interpretation would likely lead to the opposite finding, with the Ninth Circuit interpretation taking the Chffs Notes result one step farther-to a point of near certainty.

Future cases should afford greater opportunity for courts, and the Ninth Circuit in particular, to mold an approach to the explicitly misleading prong, whether that approach assumes the form of an endorsement of an amorphous confusion standard, certain likelihood of confusion factors, or the entire likelihood of confusion framework coupled with a countervailing public interest in receiving expression. In so doing, the court will, by extension, establish a fulcrum balancing confusion against expression where expressive use of third party marks is concerned.

As the Ninth Circuit addresses these tensions native to a broader application of the explicitly misleading prong, it will, in turn, inform other circuits considering adoption of the Rogers test. As savvy litigants are hauled into court over the expressive use of third party trademarks within their works, such litigants will undoubtedly argue in favor of the application of the Rogers test, and more specifically, the expression-friendly brand endorsed by the ETW Corp. court. The many courts that have not taken a stance on the applicability of the Rogers test will first need to decide whether the Rogers test forms an appropriate basis for analyzing expressive uses of third party trademarks. But once the court decides this question in the affirmative-as some undoubtedly will-the court will then be compelled to address the subsidiary question of how to handle the explicitly misleading prong of the Rogers test. As the Ninth Circuit gravitates towards either ETV Corp. or Chffs Notes, or alternatively forges its own path, future courts will enjoy a more significant quantum of data points as reference when crafting their analysis. Considering the sheer volume of third party trademark uses within expressive works-like the Caddyshack, Seinfeld, and Cee Lo examples, and untold more just like them-such future decisions in this area will be of no small importance in shaping the type of artistic expression that the public might continue to enjoy.

--- FOOTNOTES ---

FALL 2011

TABLE OF CONTENTS

I.

INTRODUCTION

................................

.................

II.

THE DEVELOPMENT AND SUBSTANCE OF THE ROGERS TEST .......

III.

THREE WAYS OF INTERPRETING "EXPLICITLY MISLEADING".............9 A. "EXPLICITLY MISLEADING" AS ENVISIONED BY THE

ROGERS COURT..................................................................................

B. THE SECOND CIRCUIT VIEW OF ROGERS AS APPLIED TO

INTERNAL TRADEMARK USES...........................................................

C. A THIRD POSSIBLE INTERPRETATION OF THE "EXPLICITLY

MISLEADING" PRONG .............................................

IV. WHY THE INTERPRETATION OF THE "EXPLICITLY MISLEADING"

PRONG MATTERS

....................................

.......17

V.

CONCLUSION

..............................................

* J.D., Emory University School of Law. The author currently serves as in-house counsel for The MathWorks, Inc., the maker of MATLAB and Simulink software packages. The opinions expressed in this Article are solely those of the author.

1 15 U.S.C. § 1127 (2006). 2 New Kids on the Block v. News Am. Publ'g, Inc., 971 F.2d 302, 305 n.2 (9th Cir. 1992) ("In economic terms, trademarks reduce consumer search costs by informing people that trademarked products come from the same source.").

3 United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90, 98 (1918). 4 Elmer William Hanak, III, The Quaty Assurance Function of Trademarks, 43 FORDHAM L. REV. 363, 363 (1974) ("In more recent times trademarks have been recognized as serving [the] function... of assuring the purchaser a certain degree of uniformity or quality in the products to which they are attached.").

5 Id. 6 Mattel, Inc. v. MCA Records, Inc., 296 F.3d 894, 900 (9th Cir. 2002) (referring to "avoiding

confusion in the marketplace" as the "core purpose" of trademark protection).

7 The cause of action is set forth statutorily as follows:

Any person who, on or in connection with any goods or services, or any container for goods, uses in commerce any word, term, name, symbol, or device, or any combination thereof, or any false designation of origin, false or misleading description of fact, or false or misleading representation of fact, which -A) is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person.

15 U.S.C. § 1125(a)(1) (2006).

8 Polaroid Corp. v. Polarad Elecs. Corp., 287 F.2d 492, 495 (2d Cir. 1961). Different circuits employ different variations of this test, but such differences tend to be minimal. See, e.g., AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 348-49 (9th Cir. 1979); Pignons S.A. de Mecanique de Precision v. Polaroid Corp., 657 F.2d 482, 487 (1st Cit. 1981).

9 These are but two examples of trademarks beginning to lose their significance as they refer to a whole class of goods rather than a specific good from a specific source. This gradual process is called "genericism" or "genericide." Words that once functioned as trademarks but have since become generic include Aspirin, Thermos, and Shredded Wheat, among many others. For support for these instances of genericide, see Bayer Co. v. United Drug Co., 272 F. 505 (S.D.N.Y. 1921), King-Seeley Thermos Co. v. Aladdin Indus., Inc., 321 F.2d 577 (2d Cir. 1963), and Kellogg Co. v. Nat'l Biscuit Co., 305 U.S. 111 (1938).

10 Mattel, Inc. v. MCA Records, Inc., 296 F.3d 894, 900 (9th Cir. 2002). 11 CADDYSHACK (Warner Bros. Entertainment Inc. 1980). 12 CEE Lo GREEN, Forget You, on TI-HE LADY KILLER (Elektra Entertainment Group, Inc. 2010). The full line in question is: "I'm sorry/I can't afford a Ferrari/But that don't mean I can't get you there/I guess he's an Xbox, and I'm more an Atari/But the way you play your game ain't fair."

1 Certain members of which he had sent to "the gas chamber" because he felt he "owed it to

them."

14 Cf Mattel, Inc. v. MCA Records, Inc., 296 F.3d at 902:

If we see a painting titled "Campbell's Chicken Noodle Soup," we're unlikely to believe that Campbell's has branched into the art business. Nor, upon hearing Janis Joplin croon "Oh Lord, won't you buy me a Mercedes-Benz?," would we suspect that she and the carmaker had entered into a joint venture.

15 Judge Kozinski again said it well when he commented,

trademark owner asserts a right

to control how we express [WIhen a ourselves-when we'd find it difficult to describe the product any other way (as in the case of aspirin), or when the mark (like Rolls Royce) has taken on an expressive meaning apart from its source-identifying function-applying the traditional test fails to account for the full weight of the public's interest in free expression.

Id. at 900.

16 Rogers v. Grimaldi, 875 F.2d 994 (2d Cir. 1989). 1 Id. at 999. 18 Id. 19 As an alternative hypothetical title to the movie and book The Devil Wears Prada, which is not

explicitly misleading. Id.

20 E.S.S. Entertainment 2000, Inc. v. Rock Star Videos, Inc., 547 F.3d 1095 (9th Cir. 2008).

II. THE DEVELOPMENT AND SUBSTANCE OF THE ROGERS TEST

21 Some, but perhaps not all. Certain courts might still rely on the so-called "alternative avenues of communication" test. That test reviews use of a third party trademark from the perspective of whether the trademark user could have communicated the expressive message without reliance on the trademark. See Dallas Cowboys Cheerleaders, Inc. v. Pussycat Cinema Ltd., 604 F.2d 200 (2d Cir. 1979) (where a theater screening the film Debbie Does Dallas, whose protagonist wore a cheerleader outfit similar in style to the Dallas Cowboys Cheerleaders', was held liable for violating the trademark rights of the Cowboys cheerleaders on the grounds that the film could have delivered its message without use of the mark). The test was later explicitly rejected in Rogers as insufficiently protective of expression, but has seen more recent application. See also Am. Dairy Queen Corp. v. New Line Prods., 35 F. Supp. 2d 727 (D. Minn. 1998) (holding the First Amendment did not preclude entry of preliminary injunction prohibiting use of "Dairy Queen" in a movie title as alternative avenues for expression existed).

2 875 F.2d 994 (2d Cir. 1989). 23 Id. at 996. 24 Id 25 Id. at 996-97. 26 Id 27 Id. at 998. 28 Id 29 Id 30 Id. at 999. 31 Id. at 1001.

32 Id. 33 Id 34 See, e.g., Mattel, Inc. v. MCA Records, Inc., 296 F.3d 894 (9th Cir. 2002); Parks v. LaFace Records, 329 F.3d 437 (6th Cit. 2003); Sugar Busters LLC v. Brennan, 177 F.3d 258 (5th Cit. 1999).

35 ETW Corp. v. Jireh Publ'g, Inc., 332 F.3d 915 (6th Cir. 2003); Cliffs Notes, Inc. v. Bantam Doubleday Dell Publ'g Grp., Inc., 886 F.2d 490 (2d Cir. 1989). See also Facenda v. N.F.L. Films, Inc., 542 F.3d 1007 (3d Cir. 2008) (noting that the Sixth and Second Circuits were the only courts to have adopted the Rogers test for use of a third party mark anywhere within an expressive work. The court ultimately did not reach the issue of whether to adopt the test because the trademark use at issue only occurred in the context of less-protectable commercial speech.).

36 547 F.3d 1095 (9th Cir. 2008) [hereinafter Rock Star Videos]. 37 Id. at 1099. 38 In this respect, the dissent in ETW Corp., supra note 35, at 938 expressed concern as to how the Rogers test should function when stretched beyond its original context as a means of analyzing trademark uses in creative tides.

III. THREE WAYS OF INTERPRETING "EXPLICITLY MISLEADING"

A. "EXPLICITLY MISLEADING" AS ENVISIONED BY THE ROGERS COURT

39 Rogers, 875 F.2d 999. 40 Id. 41 Id. 42 Id. at 1007. 43 Id. at 1000.

44 ETW Corp. v. Jireh Publ'g, Inc., 332 F.3d 915 (6th Cir. 2003). 45 Id. at 918. 46 Id. at 937. 47 Id. at 944. 48 Id. at 945.

B. THE SECOND CIRCUIT VIEW OF ROGERS AS APPLIED TO INTERNAL TRADEMARK USES

49 0 BROTHER, WHERE ART THOU? (Touchstone Pictures 2000). 50 Id. 51 The possibility that the standard Rogers test would not apply to the use of a real pomade may

have driven the makers of 0, Brother to rely on a fictitious trademark.

52 886 F.2d 490 (2d Cit. 1989). 53 Id. at 491.

54 Id. at 492. s Id 56 Id. at 493. 57 Id. at 494. 58 Id. at 490 (quoting Mushroom Makers, Inc. v. R.G. Barry Corp., 580 F.2d 44, 47 (2d Cit.

1978 .

s9 See supra note 8 and accompanying text. 60 ChfNotes, Inc., 886 F.2d at 496. 61 Id. 62 Id 63 Id at 497.

6 See, e.g., Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 580 (1994) (parody "can provide social benefit, by shedding light on an earlier work, and, in the process, creating a new one"); Fisher v. Dees, 794 F.2d 432 (9th Cir. 1986); Elsmere Music, Inc. v. Nat'1 Broad. Co., 482 F. Supp. 741 (S.D.N.Y. 1980).

65 Fountain, Marcel Duchamp (1917). For an image of this readymade chefd'oeuvre, see http://

www.tate.org.uk/servlet/ViewWorkworkid=26850.

6 Such analyses in artistic works have not always favored the artist. Take, for example, Mutual of Omaha Insurance Co. P. Novak, 836 F.2d 397 (8th Cir. 1987), where the court found infringement

C. A THIRD POSSIBLE INTERPRETATION OF THE "EXPLICITLY MISLEADING" PRONG

arising out of a deformed t-shirt version of the Mutual of Omaha Indian logo.

67 E.S.S. Entm't 2000, Inc. v. Rock Star Videos, Inc., 547 F.3d 1095, 1097 (9th Cir. 2008). 68 Id. at 1096-97. 69 Id. at 1097. 70 And, as part of that test, finding the use of the marks relevant to the overall theme of the

Grand Theft Auto game.

71 E.S.S. Entm't 2000, Inc. v. Rock Star Videos, Inc., 547 F.3d at 1100. 72 Id. ("Both [the Grand Theft Auto video game] and the Play Pen offer a form of low-brow entertainment; besides this general similarity, they have nothing in common. The San Andreas Game is not complementary to the Play Pen; video games and strip clubs do not go together like a horse and carriage or, perish the thought, love and marriage.").

73 Id. ("Nothing indicates that the buying public would reasonably have believed that ESS

produced the video game or, for that matter, that Rockstar operated a strip club.").

IV. WH-Y THE INTERPRETATION OF THE "EXPLICITLY MISLEADING" PRONG MATTERS

76 No Doubt v. Activision Publ'g, Inc., 122 Cal. Rptr. 3d 397 (Cal. Ct. App. 2011). 77 Id. at 401. 78 Id. 79 Id. 80 This success, in turn, allowed the band's main vocalist and front-person, Gwen Stefani, to

launch a solo career including notable collaborations with acts like Moby and Akon.

81 No Doubt v. Activision Publ'g Inc., 122 Cal. Rptr. 3d 397, 402 (Cal. Ct. App. 2011). 82 Id. 83 Id. at 402-03. 84 Id.

85 Id. at 402. 86 Id. 87 Id. at 402-03. 88 Id. at 403. 89 Id. at 412. Under the Rogers test, Activision would also have had to argue that the use of the No Doubt band members' likenesses was relevant to the overall work. That burden would presumably be satisfied as use of the likenesses of a notable band would be entirely relevant to a game centered on allowing users to perform as notable bands.

90 Id. 91 21 P.3d 797 (Cal. 2001). 92 50 Cal. Rptr. 3d 607 (Cal. Ct. App. 2006). 93 No Doubt, 122 Cal. Rptr. 3d at 415.

marks

and

market

channels,

survey

evidence

of

confusion,

and

the

possibility

of

94Id

V. CONCLUSION