Cardozo A&ELJ symposium, Trademark

Panel #2, TM, moderated by Vice Dean Felix Wu

Jack Daniels says that use as a trademark is special: like copyright’s bête noire, confusion caused by trademark use is the central concern of trademark law. While I have many questions about and concerns with the Court’s approach, I want to focus today on how the concept of use as a trademark—as an indicator of source for a product or service—interacts with other things that the Court has said about the Lanham Act and how we might reconstruct trademark law to focus on what the Court has told us is important instead of what it’s doing now.

I’m going to start by going back to some earlier cases, Two Pesos v. Taco Cabana, from 1992. As you may or may not recall, §32 of the Lanham Act provides for a cause of action for infringement of registered marks; courts have read 43(a) to provide a similar cause of action for “unregistered marks”—matter that is not registered with the PTO but serves a trademark function of indicating source. In Two Pesos, the Supreme Court tells us that “the general principles qualifying a mark for registration under § 2 of the Lanham Act are for the most part applicable in determining whether an unregistered mark is entitled to protection under § 43(a).”

It doesn’t really tell us what the “for the most part” caveat might mean, but I’m going to focus on this point: “entitled to protection” is not the same thing as “entitled to the same presumptions, scope of rights or remedies.” Indeed, in recent years the Supreme Court has repeatedly emphasized the multiple benefits of registration to a trademark claimant, including in the recent cases of Matal v. Tam and Brunetti, striking down various bars on registration.

Then, in Lexmark v. Static Controls in 2012, a Lanham Act false advertising case, the Court gave us two more principles for interpreting section 43: a statutory cause of action extends only to plaintiffs whose interests “fall within the zone of interests protected by the law invoked.” And here there is a very clear statement in the statute of that zone of interests, which most relevantly includes “to regulate commerce … by making actionable the deceptive and misleading use of marks in … commerce; …; to protect persons engaged in … commerce against unfair competition.” I’ve left out the parts specific to registered trademarks and the reference to treaties.

Second, the Lexmark court says, 43(a) has a proximate cause requirement: a plaintiff suing under § [43](a) ordinarily must show economic or reputational injury flowing directly from the deception wrought by the defendant’s advertising; and that that occurs when deception of consumers causes them to withhold trade from the plaintiff.

So, putting Two Pesos together with Lexmark, that an unregistered mark might be entitled to protection doesn’t mean has the same scope of rights as a registered mark—national scope of registration is just one example; Lexmark’s standing requirement of showing the kind of harm against which 43(a) is directed as part of the plaintiff’s main case, rather than presuming harm, makes sense as another, and some lower courts in trademark cases have noticed that Lexmark by its terms demands this showing from all 43(a) plaintiffs, not just false advertising plaintiffs.

Then, in Jack Daniel’s last term, the Court told us that use as a mark to identify the source of goods is trademark law’s central concern; when there’s use as a mark, the likely confusion test provides enough protection for any First Amendment interests. The Rogers test is for situations in which the use does not indicate source, as in the title of the song Barbie Girl or the use of images of trademarked goods in artwork or movies.

But, by indicating that the Rogers speech-protective test that insulates against claims of infringement could only apply when there wasn’t use as a mark to indicate source, and by saying that it wasn’t ruling on whether Rogers was ever appropriate, the Court necessarily implied that some things that aren’t uses as indicators of source could be actionably confusing.

The obvious candidates are uses that, in the words of 43(a), “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.” That is, Rogers is for uses that might confuse, but not about source.

So I’m going to suggest that Jack Daniel’s, in combination with the last term’s other TM case, points towards a revitalized distinction between the treatment of registered marks used to indicate source—which get the highest level of protection—and everything else; to whit, uses of registered marks that don’t indicate source but might indicate something else, and uses of unregistered marks, where plaintiffs should have to show not only validity but also harm from the use. We already make this kind of distinction in other contexts, for example, someone who has a valid mark can oppose registrations on 2(d) confusion grounds, but if you don’t have a mark but do have fame that you aren’t using as a mark in commerce on goods or services, you can oppose using 2(a), which bars registration of trademarks that "falsely suggest a connection with persons, living or dead” or with institutions. And the practical burden of showing false connection is probably higher, just because the things that make confusion more likely—having competing goods and services in the market, having a commercial reputation and not just a general reputation—are missing if you don’t sell goods or services yourself.

The last piece of the puzzle then is the Court’s rather convoluted decision in Abitron v. Hetronic. The court there was deciding the extent of the Lanham Act’s extraterritorial application, but in the course of their reasoning, several opinions said things about standard trademark use, some of which were embarrassing but at least one majority holding of which has potentially broad implications especially in interaction with Jack Daniel’s.

In Abitron, the opinion of the Court insisted that “Congress has premised liability on a specific action (a particular sort of use in commerce),” and further stated that both 32 and 43(a) “treat confusion as a means to limit liability to only certain ‘bona fide use[s] of a mark in the ordinary course of trade.’” As for that, the Court continued, such use must serve to “identify and distinguish [the mark user’s] goods . . . and to indicate the source of the goods.”

That is, the infringer, to violate 32 or 43(a)’s analogous trademark protection, must be using the mark in a way that is bona fide—that is, not inconsistent or sporadic, but in a way that would—absent the existence of the prior rights claimant—be enough to make it a mark for the defendant. The Court is very clear about this: there is only one definition of use in commerce in the statute, and it applies to both what the plaintiff must do to claim rights and what the defendant must do to violate those rights.

That doesn’t mean that 43(a) couldn’t go beyond classic trademark protection. As Mark McKenna has elaborated, there’s a distinct class of acts that constitute unfair competition that we could easily read as part of 43(a). But the difference is profound, because when a use is not bona fide trademark use, the plaintiff should not get the benefits of the various presumptions we give to trademark claimants—it should have to show harm, even absent some doctrine like Rogers for non-trademark uses. Moreover, when there is non-source confusion, the remedies ought to be different—disclaimers or other redesigns are the appropriate remedy, not prohibitory injunctions.

Indeed, Justice Sotomayor’s concurrence in Abitron for three Justices suggests that the Court is thinking about source confusion, trademark use, and harm as linked concepts; as she points out, along with use in commerce, “Plaintiffs must also generally show, for example, that their “injuries are proximately caused by violations of the statute.” citing Lexmark and concludes that “The Court is thus mistaken that “abstract consumer confusion is sufficient” to recover under the Lanham Act.”

The problem we face, of course, is that lower courts have expanded the concept of confusion so far beyond source confusion that the Court’s thousand-foot vision does not describe the system on the ground. But it could, and in many ways that would be a return to the common law system in which trademarks got special treatment in terms of remedies compared to acts of unfair competition, which could also be enjoined but only on a higher evidentiary showing. And this move—firmly distinguishing source confusion from other kinds of confusion, this time with a statutory basis, would have salutary effects on free speech and competitive freedom to operate, which are topics I believe the other speakers will take up in more detail.

Stacey Dogan & Jessica Silbey: Jack Daniels & the False Promise of Trademark Use.

Kagan/majority is too sanguine about ability of existing doctrines to protect speech that Kagan views as important. Title may seem surprising b/c Dogan advocated for TM use to distinguish b/t secondary and direct liability in TM law. Why not celebrate that Court embraced trademark use? The TM use doctrine is not up to the task given it: policing line b/t expressive speech that deserves special protection and run-of-the-mill commercial uses of marks.

Kagan appears sympathetic to idea that certain expressive uses deserve special protection against infringement suits, whether through varying the confusion test or otherwise. Kagan expresses no explicit view on whether Rogers is the right test for traditional expressive works, though seems to agree some special insulation is required. But no special protection should apply when D has used mark as a designation of source for its own goods/services: in the way the Lanham Act most cares about. Most parodies will be recognized as such and not confusing, but in cases involving TM use, standard likelihood of confusion test applies, regardless of D’s expressive purpose. W/o TM use, can use other doctrines more protective of speech interests.

But what is TM use? Kagan describes it as “designation of source.” Notion of source identification appears throughout as essence of TM function and the type of use that doesn’t perform a source ID function. Completely confident that this will ensure continued protection for the uses Rogers allowed: cases where TM is used not to designate a work’s source but solely to express something: Barbie Girl, Univ. of Ala, LV v. Warner Bros/Hangover II is non-TM/purely expressive and not source indicating v. Hog Farm motorcycle shop, United We Stand v. United We Stand, Timmy Holedigger perfume. But examples are not analytical tools. She thinks it’s intuitively obvious that the former are purely expressive, but what are the characteristics to use going forward.

We might try to distinguish based on the nature of the product: expressive works/other types of products, including mundane products like mugs and other merchandise (Univ. of Ala). But how do we know a mug’s not expressive? What about greeting cards, sneakers, a T-shirt, an NFT, a doll? As in MGA v. Harris, a case about dolls where the court granted Ps new trial after JDI, TM use requires a detailed factual analysis that precludes early resolution before trial.

The lower courts are not drawing the line b/t use and non-use that is consistent w/Kagan’s gut level formulation. Most courts have found Ds to be engaged in use as a mark, for movies, books, and TV show titles, the very type of use that Rogers was designed to protect. HomeVestors: title of TV show. Davis v. Amazon: dismissed claim, but not b/c of Rogers, for movie title. Morlu v. Amazon, Storm Raven title—possibility that it would be perceived as TM use is a problem. JHT v. AMC: Liberty Tax services, granted motion

Didn’t maintain the line Kagan expected it to hold. Implicit line doesn’t survive move to explicitness.

Another possibility: limit “TM use” to situations where D would have plausible claim to TM rights if asserting them, as happened with VIP/Bad Spaniels. Punchbowl News/Hog Farm cases seem like branding uses—establishing independent identity under the mark. This approach could get rid of cases where TM holders’ claim isn’t to source confusion, but sponsorship. But it’s not clear this would achieve Kagan’s goals, b/c courts have found TM uses and source identification in cases involving these protected types of works. [Purely expressive doesn’t work well, but designates something other than source might?]

Real problem w/TM use as device for protecting speech interests: in its original conception, Rogers was devised not b/c TMs in expressive works can’t ever serve TM functions, but b/c there’s something special about speech in this context that requires more than the ordinary LOC test. Titles are hybrid! They have both commercial and artistic elements. Inextricably intertwined: title promotes movie and significant means of expressing what artist wanted to express. Consumers have a dual interest in not being misled, but also in receiving results of author’s free expression.

Another approach: incorporate Rogers-type concerns into LOC test; one might worry whether that test is capable of protecting speech interests, particularly at an early stage of litigation. If it can’t be used that way, lots of Ds will roll over rather than fight.

In the wake of the Barbie movie, does Barbie Girl come out the other way? The argument was in part that people won’t license third parties to make fun of them, but after the Barbie movie, how persuasive is that?

Christine Haight Farley, Parody Perils Post Jack Daniel’s

Sotomayor thought that parody might need a special test at oral argument; the opposite of what the decision said, though it did say that LOC test isn’t blind to confusion. That kind of message matters in assessing confusion b/c consumers aren’t so likely to think that maker of a mocked product is itself doing the mocking. So we don’t need a special test. Maybe generally right; a lot of cases aren’t confusing. But are parodies really safe? If so, how? When the court takes away this speech-protective test, it wants us to believe there’s no sacrifice of speech interests.

3 perils: (1) Survey evidence on authorization; (2) Application of the LOC factors; (3) generally proving parody.

On authorization, Sotomayor concurs to warn that surveys may reflect mistaken beliefs that all parodies require permission from owner. JDI: survey expert found 29% confused. Shari Diamond looked at responses; 29/62 counted as confused were only confused by questions about authorization or affiliation. E.g. “The bottle is mimicked after the Jack Daniel BBQ sauce. So they would hold the patent therefore you would have to ask permission to use the image.” [Like Luke Skywalker says, every word in that sentence was wrong!] Balducci: court accepted survey showing more than half of respondents thought you needed permission to use Michelob name/logo.] What do we do about the statutory language which includes “affiliation” or “approval”? That will take some work (not just assuming that the words in the statute are transparent to consumers).

(2) SG took the position that the dct erred in failing to factor parody into the LOC factors. Chewy Vuiton is a good example of that, but it’s really just strength/similarity that get special treatment in parody. [I think intent does too.] But Balducci said that it was an error of law to do this. In Vans, the court didn’t adjust the factors at all.

(3) Courts say: Not all parodies are nonconfusing, but only successful parodies fare well! What is a successful parody and how do we figure it out? That seems to be a threshold test prior to LOC. That test has to be pretty low-level—not whether you succeed (that would make LOC superfluous) but whether it’s plausibly a parody. But how do you measure that? The LOC test depends on the relevant consumer. The reasonable person? Cliff Notes v. Bantam Doubleday might be an example of that. Pokes fun at the original. The Onion’s amicus brief in a non-TM case notes that parodies are an old literary tradition, and not every one has to slap you across the face to be successful. Demanding a clear message is a tall order! Cliff Notes said that parody has extra 1A protection; but is that true? Tommy Hilfiger says: when a parodist makes TM use of another’s mark, it should be entitled to less indulgence even if this results in residual effect on free speech rights of commercial actors—the opposite of Cliff Notes. Paid out in Vans v. MSCHF: that shoes were conceived of as a parody doesn’t matter—message of ridicule has to be clear; if it leaves confusion, parody hasn’t succeeded and there’s infringement.

[4^th peril: dilution! If anything is TM use, then everything will be excluded from both of the exceptions that were intended to protect noncommercial speech.]

Wu: TM use: suppose you could just decide the role of TM use in the system. How would you design it? What if we took Rogers seriously about titles’ hybridity? If a non-TM use, no infringement; you need Rogers when there’s a hybrid function.

Dogan: the hybrid case is the kind of case that Kagan says need to be bundled into TM use and not treated specially, b/c she says TMs are expressive all the time. As long as it’s a TM use type, then you have to throw it into LOC.

Wu: but you’re saying Kagan’s wrong. So what would you design from scratch?

Dogan: to me, the main value is in distinguishing b/t parties selling their own products or services and third parties that might be enabling them/helping them—informational. Internet intermediaries. Is Google using TMs as marks? No. It should be liable if at all only under contributory infringement standards. TM use can be use in those “easy” cases. What Kagan is trying to do is possible, but TM use is not a very effective boundary-setting doctrine, b/c it’s inherently circular—inevitably brings in questions of confusion. How do we decide whether someone who’s put a symbol on their shirts is engaged in TM use? Ends up a mucky, fact-intensive inquiry. In wake of JDI, we need to do something to firm up the speech protections in LOC analysis and move more in Chewy Vuiton direction and allow courts to make earlier judgments that perceive parody and commentary and throw cases out based on lack of LOC.

Intent should also matter to LOC—intent to comment should be relevant.

Jessica Silbey: How do we know that what Bad Spaniels is doing is use as a mark? If you just saw it, would you think that’s their brand? They had lots of other dog toys with other jokes. None of those were branding uses. They had VIP on the hangtag. We should use other defenses, like nominative and descriptive fair uses to tolerate certain kinds of confusion; we should have that here when there’s hybrid use. SCt is hypocritical in this context; they wave the 1A flag for bakers and florists and website designers; what about this speech? Why are TM/© not within that protected sphere? Property is elevated over speech.

Farley: following JDI, if you’re making TM use, you don’t get nominative fair use (or plaintiffs will argue that for sure). That’s where Gorsuch’s suggestion that anything not in the statute doesn’t count is so worrying. Vans case: TM use was not analyzed; depended on their reaction to the rest of the case. Kagan’s opinion tried to give us insight into what she thought TM use was, but then she pointed to the hangtag with the little dog “logo.” This case was about the trade dress, but the analysis makes it hard to separate TM use from parody products.

RT: Alex Roberts has cataloged TM use at the registration stage; we could start there. The Court was very clear in Abitron that the D has to be making bona fide TM use in order for there to be liability; maybe it didn’t mean that, but we should try holding it to its word.

Wu: what would Farley do?

A: would lean hard on certain things in JDI: parodies should be protected. Along the lines of Ginsburg’s “fair use is the built in 1A protection” take for ©. First decide if it’s a parody at a very low level of inquiry, then filter through LOC factors. If intent is properly applied, we wouldn’t need to flip any analysis: intent to confuse should be the only relevant intent. Really have a theory of the case about why this is a successful parody. If she could change the world, proposes a broader fair use defense that recognizes that LOC is not enough to protect speech.

RT: My view: commercial/noncommercial speech should be the line.

Dogan: Rogers became a victim of its own success—expressive uses did so well [appropriately!]; we’re seeing a swing in the other direction. Trial court hated VIP! Accepted testimony of psychologist who testified that consumers would start thinking of poop every time they drank Jack Daniel’s. If we have a legal standard that gives discretion to these factfinders to reach the equitable result, bad things will happen.

Farley: when the SCt decides a TM case, it’s not paying attention to how it’s working on the ground or how far the law has expanded—it thinks TM law is working pretty well. Hetronic is an example of that. It only applies when you’re making a bona fide use in commerce!

Silbey: big visible wins=people might stop suing as much. But you don’t want to have to take it case by a case to a jury and then to the SCt.