Professor of Law, Associate Dean, and Director, Program in Intellectual Property Law, Chicago-Kent College of Law; Professor of Intellectual Property Law, Queen Mary College, University of London.
Professor of Law and H. Blair & Joan V. White Chair in Intellectual Property Law, The University of Iowa College of Law. Copyright 2007, Graeme B. Dinwoodie and Mark D. Janis.
For several decades, the concept of consumer confusion has served as the touchstone for trademark liability.1 The nature and level of actionable confusion, along with the forms of consumer understanding that are properly protected against confusion, have been the principal focus of debate regarding the appropriate compass of trademark law.2
During the last three years, however, a number of scholars have argued that an unauthorized user of a mark is only liable, and should only be liable, when it uses the plaintiff's mark "as a mark."3 According to this argument, sometimes called the trademark use theory, the nature of the defendant's use serves as a threshold filter, requiring courts to engage in a preliminary inquiry regarding the nature of that use, thereby downgrading any analysis of its effects on consumer understanding.4 Indeed, courts following this newPage 1600 theory would not even reach the question of confusion absent the defendant's use being a "trademark use."5 A defendant engaged in non-trademark use would ipso facto be immune from liability.6
Proponents of the trademark use theory claim that requiring trademark use as a prerequisite to infringement has been an implicit (though largely unarticulated) principle of trademark law since before consumer confusion assumed its analytical dominance in the twentieth century.7 The principle, they argue, finds expression (albeit not in haec verbis) in the Lanham Act or has been an underlying principle of trademark law-consistent with standard economic theories of trademark law-that recent developments have brought to the surface. Supporters of this position have been spurred to excavate the theory in hopes of furthering a number of contemporary policy objectives, primarily with regard to online contextual advertising and affiliation merchandising. The trademark use theory threatens, however, to become even more pervasive-an all-purpose device by which to immunize a diverse set of practices from even potential liability for trademark infringement.8
Arguments invoking the theory have been made to courts and legislatures,9 both in the United States10 and elsewhere.11 The trademark usePage 1601 doctrine would function to limit the reach of trademark law, and it would do so in a large number of different contexts.12 Thus, for example, the theory of trademark use would allow Google to sell Athlete's Foot, Inc. the right to have the Athlete's Foot website appear prominently in the search results generated when a user queries for NIKE, even if the presentation of results induced consumer confusion.13 Likewise, it would permit Joe's Yankee HQ,Page 1602 Inc. to sell unauthorized BOSTON RED SOX merchandise, provided that Red Sox fans purchase the merchandise to "show loyalty" to the team without concern for the authorized or unauthorized nature of the merchandise.14 And, it would provide a defense for producers of knock-off MOTOROLA RAZR phones if they were able to show that they copied the distinctive RAZR design for its aesthetic appeal.15 Although courts remain divided,16 the majority of scholars have endorsed some variant of the theory.17
Stacey Dogan and Mark Lemley have offered the most cogent and compelling articulation of the trademark use theory.18 Like other scholars who have recently invoked trademark use, their advocacy of the theory thus far has been limited primarily to the context of use of trademarks on the Internet.19 However, they have sought to ground the theory in the historical and theoretical foundations of trademark law and, illustratively, have suggested its application in a far broader array of trademark settings, both online and offline.20
In this Article, we take on the trademark use theory in its own right. We reject the theory both descriptively and prescriptively. Contrary to emerging indications in recent case law and scholarship,21 we find no foundation for the theory in current U.S. trademark law, and we suggest that its adoption may have consequences unintended by its proponents. Trademark use theory is advocated in order to enhance certainty and thus encourage innovation, but trademark use theorists are pursuing a false and illusoryPage 1603 determinacy. Adoption of the trademark use theory will merely prevent trademark law from policing new information markets. Limiting liability to trademark use, as that term is understood by its proponents,22 will thus result in insufficient marketplace regulation.
The range of contexts (such as product design, music counterfeiting, and brand merchandising) in which a trademark use argument has been advanced of itself makes the theory one of the most important in contemporary trademark law. And its assertion in the worldwide litigation surrounding Google's advertising programs-from which the world's leading search engine generates eighty-five percent of its revenues-gives the theory an immediate and substantial commercial significance.23 With some oversimplification, advertisers pay Google to have their webpages appear on a list of sponsored links in response to a user query consisting of the trademark of a rival producer, prompting trademark infringement suits by the owners of the marks used in this fashion (against both the search engine and the purchasers of the advertising).24
But the debate over trademark use implicates even more profound questions that permeate (and, in some respects, transcend) trademark law. For example, use-based doctrines are among the central elements in the narrative of trademark law as an efficiency mechanism-a narrative expounded by leading Chicago School scholars and embraced by the U.S. Supreme Court.25 Yet, technology is changing the ways that consumers search and shop.26 It is an open question whether the Chicago SchoolPage 1604 analysis of search costs-persuasive in calibrating trademark law generally and invoked by trademark use proponents in support of the theory- adequately accounts for the dynamics of new information markets.27
The broader jurisprudential dilemma of how law adapts to technological change also animates the trademark use debate. In particular, the debate implicates the wisdom of slavishly pursuing analogies between the offline and online world and the pace at which law adjusts to new socio-technological forces. Limiting effective Lanham Act regulation of new market activities (as proponents of trademark use do by opting not to allow trademark law to police certain online activities) may be a premature and unrefined response to what, after the hubbub of today's technological advance has subsided, may be a much more complex social phenomenon.28
Moreover, the trademark use debate serves as a vehicle to consider what one of us has called the difference between proactive and reactive trademark lawmaking.29 Trademark law has become a leading instrument for shaping the forces by which consumer understanding is developed. A proactive view of trademark lawmaking embraces the proposition that trademark law should be used to influence the norms that govern consumers' shopping habits. In contrast, a reactive view relegates trademark law to the role of discerning and protecting extant consumer understanding.30 For example, a reactive scholar might focus on whether substantial confusion exists as a result of the way a search engine sells keywords and presents search results. Such an approach holds open the prospect that search engines would be found liable for the sale of trademarks as keywords, but they would avoid liability if likely confusion were absent.31 Trademark use theorists largely reject a reactive approach.32
For example, in the context of the Google litigation, they adopt an approach that immunizes search engines from trademark liability regardless of the form in which the search engines present search results and, thus, regardless of short-term confusion.33 In contrast, we show that it is possible to adopt a proactive stance without so abruptly discarding confusion-avoidance as a relevant variable. We argue that an appropriate innovation policy should offer immunity, if at all, only on conditions that further the goals of trademark law.34
Finally, the debate over trademark use roughly maps a jurisprudential fault line between formalism and functionalism.35 Proponents of trademark use rely on a single legal concept to do substantial work in limiting a number of perceived excesses in current trademark law. But trademarks increasingly serve a number of roles and, accordingly, our approach relies on a number of legal devices, reflecting a variety of autonomous policy justifications for...