Response

Self-Regulation Through Sui-Genericide: When the Law Stands in the Way

 

In Sui-Genericide,1 Professor Jorge L. Contreras illuminates an area of practice that might seem counterintuitive to those who only know the “standard” story of how trademarks function to create value for consumers and competitors. That standard story posits that, by allowing individual firms to monopolize distinctive terms, consumers benefit from easier searches for the products and services they want and producers benefit from being able to internalize the benefits of a good reputation—knowing that trademarks are reliable, consumers will rely on them, thus rewarding those who invest in creating a good product or service.2 The producers’ incentives in this system are to distinguish themselves, to create differentiation either through physical difference or mental impression. In trademark terms, producers have incentives to avoid genericity, which means interchangeability and commoditization and thus lower prices. Far better to sell Diet Coke than to sell commodity sugar!

To this standard model, Professor Contreras introduces an apparently strange wrinkle: the deliberate genericization of terms by the same bodies that popularize those terms.3 Especially when potentially trademark-eligible symbols are parts of a larger product, and especially when that product class is not necessarily familiar to consumers, genericization may help create a standardized base that makes the entire class of products more attractive to consumers. The examples in the article reflect this intuition: synthetic textile fibers and the components subject to various technical standards are both directed at parts or features of a larger product, which can generally be differentiated by its maker using a trademark that applies to the overall product.4

Pharmaceutical marketing doesn’t quite fit that model, but the policy rationale is essentially the same: At some point, the underlying drug product will become “generic” in the pharmaceutical sense, which is to say off-patent and free for other producers to make. The existence of a generic name simplifies communication about the drug and allows a rising tide to lift all producers’ boats.

In Part I of this Response, I will explore the distinction between sui-genericide, as Professor Contreras defines it, and other methods of promoting a class of products. In Parts II and III, I discuss both the doctrinal developments and the possible effects of those developments, respectively, that may make Professor Contreras’s recommendations difficult to implement, as desirable as they might be. In Part IV, I briefly conclude that preserving genericity, including sui-genericide, requires a normative foundation, not merely a descriptive account. Professor Contreras’s explanations of the rationales for and benefits of sui-genericide reinforce that normative foundation.

Sui Genericide as Distinguished from Other Methods of Organizing a Group’s Definitions or Standards

Professor Contreras identifies sui-genericide as a practice in which an industry, usually guided by a formal industry group, explicitly names a term as generic for a particular thing. There is a clear similarity between sui-genericide, certification marks, and geographic indications (“GIs”). Certification marks are marks used by entities other than the certification mark owner to indicate their compliance with the mark owner’s standards.5 And geographic indications are, as the name indicates, terms that are linked to the products of a particular area, such that only producers from that area—and sometimes only producers from that area who follow additional standards about production—can use the term.6 Europeans have long campaigned to get the United States to expand protection for such indications beyond that allowed by regular trademark law, with notable success only in the area of wines and spirits.7

As Professor Contreras explains, certification marks might accomplish much of what sui-genericide does in terms of allowing markets to form and protecting fair competition among market participants who offer the same features.8 But they also involve extra work for the certifier that might not be worth it if other standard-setting or regulatory structures are already heavily involved in quality assurance, as with pharmaceuticals in particular.9 Unlike Professor Contreras, I’m skeptical that the potential for consumer protection lawsuits exercises significant disciplinary force on most market players—many of whom certainly sell their products with shrink-wrap arbitration agreements that make it uneconomical to attempt to get any consumer redress.10

Nonetheless, it may well be that compliance with many standards is simple enough to ensure, and to evaluate in an online review, that most sellers not engaged in outright fraud will not find misrepresentations about compliance with standards to be worthwhile. However, whether these other pressures deter deception depends very much on the standard and the market segment. For example, when high-definition TVs were starting to take over from older models, several major players found it worthwhile to at least obscure the differences between the superior 1080p and inferior 1080i signal options, claiming to offer 1080p while providing only 1080i.11

Indeed, Professor Contreras argues that entities that adopt the sui-genericide strategy are relying on other enforcers to ensure that individual market participants conform to the standards they purport to adopt. But this reliance can also be described as free riding on the efforts of others to enforce laws or regulations designed to ensure truth in advertising. Unless our regulators are active and well-resourced, this strategy could easily fall apart in standard “tragedy of the commons” fashion.12 This regulatory failure may help explain why, though pharmaceuticals seem to have a successful regime, there’s nothing similarly useful for dietary supplements, an underregulated market full of false claims not just about efficacy but about what ingredients are actually in a given supplement.13

Geographic indications are also similar to certification marks, but they involve specific claims about the importance of the location of production (or “terroir”) to the characteristics of a product, whether those characteristics are empirically measurable or not. As Professor Contreras points out, the United States has largely resisted special protections for geographic indications outside of the trademark system, and often considers terms generic that Europeans would prefer to have recognized as GIs.14 But why distinguish between generic or GI if they both allow multiple different producers to use the term? Precisely because U.S. producers would often prefer a definition of a putative GI as defining a subtype of product regardless of the location of production—Swiss cheese, or French bread, or tequila.

This comparison helps highlight another factor important to sui-genericide—the subject matter at issue is generally something new to the world: a new drug or technological innovation. The “videocassette recorder” or “digital video recorder,” or even the “escalator” or “elevator,” didn’t come with a natural generic name—those had to be invented, whether initially as trademarks or otherwise, and market participants had to agree on what the generic name was. A new-to-market thing or component needs a name in order to be easy to talk about and therefore easier to sell, but by definition, it doesn’t already have a particular name in wide use for a long time. Where, as with food, there are generally pre-existing names, it’s easier for existing producers to have conflicting interests in whether the definition ought to be stretched. Relatedly, there’s really nothing that a U.S. producer can do to make “champagne” if champagne is a GI, whereas it can (absent patent protection) make any synthetic fiber that is otherwise within its capabilities. If the analogy to GIs is correct, though, it would suggest that over time there might be more conflicts over whether genericity is the appropriate approach. At some point, perhaps more producers will see benefits in shrinking existing definitions (to kick out competitors)15 or expanding existing ones (so they can advertise having the relevant characteristic even if they don’t have it in quite the same way as others do).16

If the “standard” law and economics story has room for geographic indications and certification marks, then it has room for any mechanism that benefits a group of producers as a whole. Professor Contreras persuasively explains why industries, and even individual competitors, might prefer a base of common generic terms. He thus explains why industry organizations in particular may support sui-genericity even though some individual market participants might benefit from defecting. The article sits nicely beside a much larger body of work explaining similar dynamics of sharing that complicate copyright’s standard incentive model.17

But trademark law’s current structure will make it harder to implement Professor Contreras’s insights into how law ought to respond. Trademark law offers trademark claimants lots of chances to expand their rights and rarely binds them to prior representations. In our present intellectual property system, non-ownership is often less stable than ownership: In copyright law, for example, the Supreme Court has approved removal of works from the public domain as “restoration” of rights for foreign copyright owners who forfeited U.S. rights by failure to comply with old U.S. formalities.18 The Court disapproved of any analysis that treated this as a “taking” of rights from the public, because the public’s interests were too inchoate to matter, and specific “reliance parties” who had been using the public domain works got special provisions diminishing, though not eliminating, their losses.19 By contrast, taking copyright away from existing works and situating them in the public domain, even if politically feasible, would clearly be subject to a constitutional takings claim with a nontrivial likelihood of success.

A similar dynamic exists in trademark law: Trademark owners regularly find it easier to maintain and expand rights than the public does to push back by not treating a term as a trademark. For example, a trademark owner’s initial claims about the scope of goods and services covered by a trademark, or the elements of a mark that are protectable, do not prevent the trademark owner from later claiming rights over a broader field or a broader concept of the protected mark.20 While Congress has recently added rules making it easier to expunge trademarks that aren’t in use from the federal register,21 federal law still provides for “incontestability” for registered trademarks, which means that the registration can’t be lost just because the putative trademark isn’t actually distinctive to consumers.22

Likewise, as Professor Contreras explains, various doctrines create uncertainty about whether previous generic uses will bind a trademark claimant that changes its mind about genericity.

Barriers to Adoption: Booking.com and why it is a problem for Professor Contreras’s proposals

In particular, the Supreme Court’s conception of the nature of trademark law revealed in the recent Booking.com decision may make Professor Contreras’s recommendations for how lawmakers should honor, encourage, and enforce industry sui-genericide efforts difficult or impossible to implement in current law. Booking.com held that genericity was entirely a matter of actual consumer perception, with no recognition of the pro-competitive goals served by the genericity doctrine.23 The concluding sentence of Professor Contreras’s article highlights the problem: His proposals, he suggests, will “reduce consumer search costs, enhance competition among producers of standardized products, and bring increased efficiency to markets that depend on the unencumbered availability of common names.”24 But none of these things is about direct consumer perception of the meaning of a term, and they are therefore unimportant in genericity evaluations according to Booking.com.

The case before the Supreme Court arose when Booking.com sought to register BOOKING.COM for its travel services. The U.S. Patent and Trademark Office (“PTO”) refused registration on the grounds that: “booking” was generic for the service of booking travel; .com was generic for websites; and the combination of the two didn’t provide any additional meaning.25 The lower courts disagreed, and the Supreme Court had the final word, holding that a .com domain name used to sell a class of goods or services is only generic for that class of goods or services “if the term has that meaning to consumers.”26 Because the evidence before the Court showed that consumers don’t perceive “booking.com” to signify online hotel-reservation services as a class, booking.com was not generic.27 Indeed, as the Court opined, .com “might also convey to consumers a source-identifying characteristic: an association with a particular website,” because only one entity can have a particular domain name at a time.28 Since consumers know that fact about domain names, a consumer can infer entity-designating function.

Unfortunately, the ability of a term to distinguish this entity—booking.com—from other similar sounding entities or domain names didn’t seem to come into the Court’s reasoning at all, even though it is a significant part of distinctiveness. A consumer may well know that there’s only one booking.com, but have trouble remembering whether their past experience was with booking.com or bookings.com or onlinebooking.com. In that sense, consumers’ understanding that there’s one domain name owner per domain name isn’t at all helpful in promoting source identification; instead, the work has to be done by something that makes the entire domain name memorable, which might sometimes be secondary meaning and might sometimes be inherent distinctiveness.

Noticeably, framing the issue in the way the Court majority did not require a showing of secondary meaning for combinations of generic terms—it merely required that consumers not think of travel websites as “booking.coms.” Thus, the Court held that the case could be resolved merely using “the undisputed principle that consumer perception demarcates a term’s meaning.”29 Only if consumers perceive the full term—before and after the dot—as the name of a class would “generic.com” also be generic.30 “Evidence informing that inquiry can include not only consumer surveys, but also dictionaries, usage by consumers and competitors, and any other source of evidence bearing on how consumers perceive a term’s meaning.”31

The Court did caution that consumer surveys “require care in their design and interpretation,” especially as “survey respondents may conflate the fact that domain names are exclusive with a conclusion that a given ‘generic.com’ term has achieved secondary meaning.”32 But the Court wasn’t concerned with that difficulty in the case before it, because the PTO did “not contest the lower courts’ assessment of consumer perception” of booking.com specifically.33

Relevant to Professor Contreras’s argument, the Court noted that the PTO’s real objection wasn’t to Booking.com’s exclusive use of “booking.com” as a mark, but to the risk of “undue control over similar language . . . that others should remain free to use.”34 However, the Court thought that trademark law’s existing protections for descriptive uses sufficed to protect competition even if booking.com were registered. The Court concluded that the competitive advantage conferred by seizing on a descriptive domain name didn’t justify refusing registration.35

In dissent, Justice Breyer defended genericity in terms that resonate with Professor Contreras’s argument for favoring the creation and maintenance of standard generic terms around which competitive markets can form: Genericity “preserves the linguistic commons by preventing one producer from appropriating to its own exclusive use a term needed by others to describe their goods or services.”36 The majority’s “fact-specific approach” rejected the animating principle of prior law: “Terms that merely convey the nature of the producer’s business should remain free for all to use.”37 Breyer’s dissent focused on the risks of anticompetitive outcomes without a strong, normatively focused genericity doctrine.

The result in Booking.com is consistent with the general evolution of U.S. trademark law since the Lanham Act was adopted: Courts have mostly shifted away from prophylactic rules that protect competition—limiting the registration/protection of trade dress,38 territoriality,39 and numerous other rules—to the idea that all interests other than the protection against likely confusion should be considered only in individualized circumstances, usually as defenses.

Of course, it is impossible to replace the normative completely with the empirical in assessing consumer understanding. The normative creeps back in, at the very least in the factfinder’s prior beliefs about the persuasiveness of specific pieces of empirical evidence such as surveys. But the current legal framing of genericity makes it hard to understand why genericity even matters, since non-genericity (trademarkability) is purely a matter of what consumers happen to think at a given time.

Given this result, it is questionable whether the pro-competitive considerations that are the foundation of Professor Contreras’s argument are the law of the land. Professor Contreras uses non-survey evidence to explain how industries can consciously choose genericity, but, on its face, Booking.com demotes such evidence to irrelevance as long as there is survey evidence that consumers think a term is a trademark. This rule will provide individual producers with opportunities to defect from an industry consensus about genericity, if they are big enough to convince consumers that a term is primarily associated with them. By giving producers with dominant market shares the most opportunity to de-genericize a term, the current rule risks seriously anticompetitive results.

What Comes Next?

After Booking.com, genericity may be limited to cases in which there is no “secondary meaning” in a term—cases that may be unlikely to overlap with the instances of sui-genericide Professor Contreras identifies. Especially where the subject matter is technical, consumers may often be simply guessing about whether a term is a trademark or certification mark or instead a generic term for a class of goods.

Because of the close ties between certification marks and generic terms that Professor Contreras identifies,40 it is hard to understand what questions one could even ask consumers to get reliable answers. Indeed, the usual strategies for genericity surveys involve training respondents about the difference between trademarks (Chevrolet) and generic terms (car) and then asking them to classify multiple examples including the term of interest. This is known as a Teflon survey.41 These surveys would have to be substantially modified to train respondents about the difference between certification marks and generic terms. As the Trademark Trial and Appeal Board (“TTAB”) has pointed out, it’s easy to do a bad survey in the ordinary case, by training consumers that highly descriptive, almost-generic terms are canonical exemplars of valid trademarks.42

Training consumers that, for instance, Underwriters Laboratories’ UL certification mark is a valid certification mark and not generic for compliance with electrical testing standards, but USB for computer interfaces is generic for a kind of computer connector, or that TEQUILA for liquor is a valid certification mark, but LONDON DRY GIN for a different kind of liquor is generic,43 will pose even greater challenges. For example, one of Professor Contreras’s examples of certification marks involves the Bluetooth standard. But when that mark was actually challenged, a Teflon survey that was offered to show lack of trademark function ignored the special characteristics of certification marks and didn’t test respondents’ understanding of what a certification mark was, although the court considered that to go to weight rather than admissibility.44 Other certification mark proceedings have rejected non­­-Teflon surveys as not even probative, so there is basically no guidance for the future of such cases after Booking.com.45

But unless courts rigorously police the boundaries—such as requiring surveys to offer a prominent “I don’t know whether this is generic or not” answer and counting such answers against the trademark claimant—it will often be easy to find a surveyor who can craft a claimant-favorable survey.46

Now that survey evidence may trump any policy concerns about market preclusion, much depends on the quirks of individual surveys. In one recent case, for example, the parties disputed whether “organic protein” was generic or merely descriptive (and thus potentially subject to trademark protection) for nutritional supplements.47 The parties offered dueling surveys, and while the court found genericity,48 it did so in a way that further buried the normative pro-competitive considerations behind the existence of the generic category.

The trademark claimant, Orgain, offered a survey that concluded that 74 percent of respondents believed the term “organic protein” was descriptive. In its survey, respondents were instructed that “[g]eneric terms identify a type or category of product,” while “[d]escriptive terms describe an ingredient, characteristic, quality, feature, or function of a product.”49 This is a reasonable recounting of the legal standard set by decades of precedent distinguishing descriptive terms from generic terms, but even trademark lawyers have trouble drawing the lines. It is not surprising that the court thought that ordinary consumers could also be perplexed. The survey listed “automobile,” “ice cream” and “allergy medicine” as examples of generic terms and “hatchback,” “Rocky Road,” and “non-drowsy” as examples of descriptive terms—and yet, as the court pointed out, the terms in the latter also plausibly were a “type” or “category.”50 To the court, the plaintiff’s survey “side-stepped the key inquiry—namely, whether respondents understood ‘organic protein’ to refer to Orgain’s goods or whether they understood the term to refer to a category of products.”51

Consistent with Booking.com, the court reasoned: “[T]he genericness inquiry is not about where consumers categorize the mark along the distinctiveness spectrum . . . . Rather, the genericness inquiry asks whether consumers perceive the term as identifying a common name for a certain type or class of products.”52 That is, surveys can detect secondary meaning, but they’re not going to be great at distinguishing among conceptual legal categories. However, because Booking.com points to surveys as the best possible evidence, secondary meaning will trump genericness.

Assuming that other courts treat Booking.com in the same way, trademark claimants will be able to override the considerations Professor Contreras proposes to consider whenever they have a favorable survey. True, surveys might well find that terms like USB are indeed generic—but surveyors’ ability to design subtly tilted surveys that are likely to find client-favorable results don’t make me completely confident of that.

Conclusion

The approach to genericity canonized by Booking.com reduces a policy question to a question of historical fact by way of epistemology, that is, by prioritizing consumer survey evidence in answering the question “how do we know when a term is generic?” But the policy question keeps re-emerging, including in cases of innovation in computers, fabrics, and pharmaceuticals described by Professor Contreras.

What can be done? Especially where most consumers are likely to answer “how the heck should I know?” in response to many questions trying to parse legal concepts, survey evidence should not be treated as the gold standard, with other evidence reduced to second-best. Professor Contreras, by illuminating an area in which non-survey evidence of genericity matches up well with the policy justifications for having generic, non-trademark-protectable terms, has provided an example of how this might be done even post-Booking.com. In the long run, genericity needs the normative foundation on which his article relies.

  1. [1].     Jorge L. Contreras, Sui-Genericide, 106 Iowa L. Rev. 1041 (2021).

  2. [2]. See e.g., William M. Landes & Richard A. Posner, Trademark Law: An Economic Perspective, 30 J.L. & Econ. 265, 268–70 (1987).

  3. [3]. Contreras, supra note 1, at 1047.

  4. [4]. Id. at 1068.

  5. [5]. See 15 U.S.C. §1054 (2018) (making collective marks and certification marks registrable); 15 U.S.C. §1127 (defining “certification mark” as a symbol used or intended to be used by a person other than its owner “to certify regional or other origin, material, mode of manufacture, quality, accuracy, or other characteristics of such person’s goods or services or that the work or labor on the goods or services was performed by members of a union or other organization”).

  6. [6]. The United States does not generally provide a separate legal regime for geographic indications, preferring to merge them into collective and certification marks more generally. See 15 U.S.C. §1054 (stating that “collective and certification marks[] includ[e] indications of regional origin”).

  7. [7]. A number of other countries have created specific regimes for managing geographic indications and have pressured the United States to do the same as part of international trade negotiations. See generally Renée Johnson, Cong. Rsch. Serv., R44556, Geographical Indications (GIs) in U.S. Food and Agricultural Trade (2017) (providing an overview of GIs recognized in the United States); K. William Watson, Reign of Terroir: How to Resist Europe’s Efforts to Control Common Food Names as Geographical Indications, CATO Inst. (Feb. 16, 2016), https://www
    .cato.org/policy-analysis/reign-terroir-how-resist-europes-efforts-control-common-food-names-g
    eographical [https://perma.cc/NK4G-KDNG] (offering a highly negative view of such initiatives).

  8. [8]. Contreras, supra note 1, at 1077–78.

  9. [9]. Id. at 1078.

  10. [10]. See Myriam Gilles & Gary Friedman, After Class: Aggregate Litigation in the Wake of AT&T Mobility v Concepcion, 79 U. Chi. L. Rev. 623, 627–31 (2012).

  11. [11]. See, e.g., Johnson v. Mitsubishi Digit. Elecs. Am., Inc., 578 F. Supp. 2d 1229, 1232–34 (C.D. Cal. 2008); Date v. Sony Elecs. Inc., 2009 WL 435289, at *1–3 (E.D. Mich. Feb. 20, 2009).

  12. [12]. See generally Carol M. Rose, Crystals and Mud in Property Law, 40 Stan. L. Rev. 577 (1988) (discussing the phenomenon of free riding on, and potentially overloading, public regimes for managing property rights as if the regime itself was a kind of common property).

  13. [13]. See, e.g., Debra D. Burke & Anderson P. Page, Regulating the Dietary Supplements Industry: Something Still Needs to Change, 1 Hastings Bus. L.J. 119, 121 (2005); Michael A. McCann, Dietary Supplement Labeling: Cognitive Biases, Market Manipulation & Consumer Choice, 31 Am. J.L. & Med. 215, 215–17 (2005); Arthur P. Grollman, Unregulated Herbal Remedies—an Accident Waiting to Happen, Chi. Trib. (Dec. 24, 2002), https://www.chicagotribune.com/news/ct-xpm-2002-12-24-0212240094-story.html [https://perma.cc/EMH8-BKQR] (explaining that most Americans are unaware that dietary supplements do not receive premarket testing); Steven G. Newmaster, Meghan Grguric, Dhivya Shanmughanandhan, Sathishkumar Ramalingam & Subramanyam Ragupathy, DNA Barcoding Detects Contamination and Substitution in North American Herbal Products, 11 BMC Med. 222 (2013) (finding substantial divergence between labeling and actual contents in many supplements); Anahad O’Connor, Herbal Supplements Are Often Not What They Seem, N.Y. Times (Nov. 3, 2013), https://www.nytimes.com/2013/11/05/science/herbal-supplements-are
    -often-not-what-they-seem.html [https://perma.cc/W62X-QDFU].

  14. [14]. Contreras, supra note 1, at 1062.

  15. [15]. See generally Kolleen M. Guy, When Champagne Became French: Wine and the Making of a National Identity (2007) (discussing episodes in which champagne producers both expanded and shrank the geographic boundaries of the Champagne region in order to include or exclude certain producers from being able to use the term).

  16. [16]. See, e.g., LG Electronics v. Whirlpool Corp., 2010 WL 2921633, at *4 (N.D. Ill. Jul. 4, 2010) (involving a dispute over whether the defendant’s “steam dryers” actually used desirable “steam”).

  17. [17]. See, e.g., Creativity Without Law: Challenging the Assumptions Of Intellectual Property 114 (Kate Darling & Aaron Perzanowski eds., 2017); Kal Raustiala & Christopher Sprigman, The Knockoff Economy: How Imitation Sparks Innovation 170 (2012); Emmanuelle Fauchart & Eric von Hippel, Norms-Based Intellectual Property Systems: The Case of French Chefs, 19 Org. Sci. 187, 195–96 (2008); Elizabeth L. Rosenblatt, A Theory of IP’s Negative Space,
    34 Colum. J.L. & Arts 317, 318 (2011); Tim Wu, Tolerated Use, 31 Colum. J.L. & Arts 617,
    627–31 (2008).

  18. [18]. Golan v. Holder, 565 U.S. 302, 315 (2012).

  19. [19]. Id. at 329.

  20. [20]. That is, infringement can exist when the goods/services differ and when the accused marks differ, or both, as long as they are similar enough to cause likely confusion. See, e.g., AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 348–49 (9th Cir. 1979). So, the mark the trademark owner is actually using, and the goods and services on which it is using the mark, create a penumbra of rights—the size of which varies based on the fame of the mark and other marketplace factors.

  21. [21]. Trademark Modernization Act of 2020, H.R. 6169, 116th Cong. §§ 1–8 (2020).

  22. [22]. 15 U.S.C. § 1065; Rebecca Tushnet, Fixing Incontestability: The Next Frontier?, 23 B.U. J. Sci. & Tech. L. 434, 436 (2017).

  23. [23]. U.S. Pat. & Trademark Off. v. Booking.com B. V., 140 S. Ct. 2298, 2307 (2020).

  24. [24]. Contreras, supra note 1, at 1089–90.

  25. [25]. Booking.com B. V., 140 S. Ct. at 2304–05.

  26. [26]. Id. at 2301.

  27. [27]. Id.

  28. [28]. Id. at 2306.

  29. [29]. Id. at 2304 n.3.

  30. [30]. Id. at 2304–05.

  31. [31]. Id. at 2307 n.6.

  32. [32]. Id.

  33. [33]. Id.

  34. [34]. Id. at 2307.

  35. [35]. Id. at 2308–09.

  36. [36]. Id. at 2309 (Breyer, J., dissenting)

  37. [37]. Id. at 2313.

  38. [38]. See, e.g., Mark P. McKenna, The Normative Foundations of Trademark Law, 82 Notre Dame L. Rev. 1839, 1909–10 (2007) (summarizing change in courts’ willingness to protect trade dress under trademark law).

  39. [39]. See generally Graeme B. Dinwoodie, Trademarks and Territory: Detaching Trademark Law from the Nation-State, 41 Hous. L. Rev. 885 (2004) (discussing the evolution of territoriality principles in trademark law).

  40. [40]. Contreras, supra note 1, at 1077–79.

  41. [41]. 2 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition
    § 12:16 (5th ed. 2021) (“The most widely used survey format to resolve a genericness challenge is the ‘Teflon’ format. The name comes from a 1973 telephone survey used as evidence by a court to determine that TEFLON was a valid trademark, not a generic name for non-stick coating. A ‘Teflon Survey’ is essentially a mini-course in the generic versus trademark distinction, followed by a test.”) (footnotes omitted).

  42. [42]. Sheetz of Delaware, Inc. v. Dr.’s Assocs. Inc., 108 U.S.P.Q.2d 1341, 2013 WL 5315963, at *17–18 (T.T.A.B. 2013) (criticizing applicant’s survey for training respondents that highly descriptive terms “Quarter Pounder” and “Original Recipe” were trademarks).

  43. [43]. Luxco, Inc. v. Consejo Regulador del Tequila, A.C., 121 U.S.P.Q.2d 1477, 2017 WL 542344, at *8 (T.T.A.B. 2017).

  44. [44]. Bluetooth SIG, Inc. v. FCA US LLC, 463 F. Supp. 3d 1169, 1181 (W.D. Wash. 2020) (survey included both definitions for trademarks and certification marks, but emphasized the trademark definition, and “did not test the respondents on their understanding of a certification mark”). According to the survey, 82 percent of consumers believed that “Bluetooth” was a generic term; two of the three control terms for brand names were trademarks (ASUS and GARMIN) and only one was a certification mark (UNDERWRITERS LABORATORIES). Id. As noted in text, I’m highly sympathetic to the challenges posed for a Teflon certification mark survey—but I think the lesson is that the project of conducting Teflon surveys for such terms is itself probably heading in the wrong direction.

  45. [45]. I have found two such TTAB proceedings. See Luxco, Inc., 2017 WL 542344, at
    *20 (T.T.A.B. 2017) (survey question was “When an alcoholic beverage is identified as [Tequila], does or doesn’t the fact that it’s identified as [Tequila] communicate or indicate to you the country in which it is made?” and 62.9 percent of respondents said yes; not a Teflon survey); Tea Bd. of India v. The Republic of Tea, Inc., 80 U.S.P.Q.2d 1881, 2006 WL 2460188, at
    *17 (T.T.A.B. 2006) (holding that responses to the open-ended question “What is Darjeeling tea?” that didn’t mention Indian origin weren’t probative of genericity). The result in the TEQUILA proceeding also depended on the special rule that a geographic certification mark is still valid as long as it has geographic significance—that is, it can be merely descriptive of the geographic origin and the PTO will still consider it to be a valid certification mark even if such terms would ordinarily be considered unprotectable, descriptive geographic terms. The complications of the geographic rules make them even harder to translate into genericity generally.

  46. [46]. See, e.g., Indianapolis Colts, Inc. v. Metro. Balt. Football Club Ltd. P’ship, 34 F.3d 410, 416 (7th Cir. 1994) (disparaging the “survey researcher’s black arts”); L&F Prods. v. Proctor & Gamble Co., 845 F. Supp. 984, 995–96 (S.D.N.Y. 1994) (“[Courts are] familiar with the subtle ways surveys are structured. Those who believe they can manipulate the structure of consumer surveys to gain a tactical advantage in the courtroom may actually harm their client’s strategic position before the finder of fact.”).

  47. [47]. Orgain, Inc. v. N. Innovations Holding Corp., No. 8:18-cv-01253, 2021 WL 1321653, at *2 (C.D. Cal. Mar. 22, 2021).

  48. [48]. See id. at *10.

  49. [49]. Id. at *9.

  50. [50]. Id.

  51. [51]. Id.

  52. [52]. Id.

*

Frank Stanton Professor of the First Amendment, Harvard Law School.