Letter vs. Spirit of the Law Dog Toys by Sam Kilb


As part of its most recent, intellectual property-heavy term, the Supreme Court in June unanimously sided with Jack Daniel’s in a dispute with dog toy manufacturer VIP Products over a product design for a chew toy that mocked the famous Jack Daniel’s whiskey bottle. The holding cast doubt over the future of the expressive or artistic use of third-party trademarks on consumer goods. Big brands viewed the decision as a win for trademark enforcement efforts, while creatives saw the ruling as an imposition on freedom of expression—but the case left several major questions unanswered, and likely represents only an opening volley in what may be an ongoing reshaping of the bounds of trademark law and artistic freedoms in the United States.




This case arose because VIP Products has a line of dog toys featuring designs that riff on popular beverage brands, but with a dog-inspired twist—for example, a “Smella Arpaw” toy that imitates Stella Artois, and a “Doggie Walker” toy that imitates Johnnie Walker. One such toy in the line was “Bad Spaniels,” which imitated the Jack Daniels bottle and trade dress, replacing the familiar “Old No. 7 Tennessee Sour Mash Whiskey” with “Old No. 2 on your Tennessee Carpet.” Jack Daniels was unamused, and sued VIP for trademark infringement.

The district court found VIP Products liable for infringement, but on appeal the Ninth Circuit applied the Rogers test to find that the First Amendment allowed the use of Jack Daniels’ trademarks by VIP Products for the artistic expression contained on the dog toy, low brow though the humor may be.

In June, the Supreme Court unanimously vacated the Ninth Circuit’s ruling and remanded for review consistent with its opinion, which rejected the application of the Rogers test in trademark infringement cases where the allegedly infringing use of the trademark is as the designation of origin for the defendant’s own goods or services.

The Rogers test, which has been a frequent topic of conversation in these pages, has for years provided cover for defendants in trademark infringement cases who can successfully argue that their incorporation of the plaintiff’s trademarks (1) has artistic value to the underlying work, and (2) will not explicitly mislead the public as to the plaintiff’s involvement. The original intent of the Rogers test was to balance the First Amendment freedom of expression rights of artists (such as filmmakers, painters, musicians) against the statutory trademark rights of brands. The test was originally applied to allow the use of a celebrity performer’s name in a movie title for a film that depicted characters emulating that celebrity, and over time has been used to excuse the use of third-party trademarks in other expressive content without the permission of the trademark owner, such as in song lyrics, video games, and movies. In the Supreme Court’s view, the Rogers test has remained cabined to expressive use, rather than trademark use, all along. But the bounds of Rogers have been tested as our understanding of art and expressive works, and what can be art or expressive works, has moved beyond traditional mediums and onto items more commonly understood as consumer products, like clothing, skateboards, and dog toys.

So in addition to featuring the unique joy of having the Supreme Court of the United States discussing the merit of dog poop jokes, this case asked the fundamental question: when, if ever, does the First Amendment excuse the use of someone else’s trademark? Not when the material is being used in a source-identifying manner, according to all of the justices.



In so holding, Justice Kagan, writing for the court, focused on the fundamental purpose of trademarks to identify the source of a product and distinguish that source from others. Naturally, according to Kagan, if VIP Products is using Jack Daniels’ trademarks, or trademarks confusingly similar thereto, as its own designation of source, then it would frustrate trademark law and damage consumers to allow such use. The First Amendment does not demand protection for potentially confusing trademark use, even if the use may be clever, funny, or have some underlying meaning or expression.

You may at this stage be wondering—does the design of a dog toy constitute source-identifying trademark use? Unfortunately for VIP Products and for immediate clarity in the law, VIP conceded that its use was source-identifying trademark use during the course of the proceeding. Accordingly, the court did not address what kind of use would qualify as trademark use, leaving us to wonder whether the design of a dog toy—which would not be considered inherently distinctive and protectable as a trademark if VIP Products attempted to register the trade dress—actually functions as an indicator of source. Consumers of dog toys likely look at product names (is “Bad Spaniels” for dog toys, without the bottle design, really an infringement of “Jack Daniels” for whiskey?), hang tags, and other typical source-identifying material like shelf talkers, packaging, or words affixed to the product to identify the source. But is the product design an indicator of source? Does it depend on the nature (and particularly the fame) of the allegedly infringed trademark?

And in the context of parody—which the court said does not excuse infringement, but acknowledged may play a part in assessing a likelihood of confusion—is the use of another’s trademark actually source-identifying? Parody is designed to poke fun at someone or something, and in order to “get the joke,” enough material has to be taken for the reference to land. And yet for the joke to have worked, it’s also important that the consumer understand that the product is not the product of the parodied. It’s a reasoning similar to the “sucks” cases in domain names—it has long been held there is no likelihood of confusion, because no one would think the brand owner would create a [brand]sucks.com website. If the use is being critical of the brand, it should be clear to consumers that the product in connection with which the mark is being used is in fact not a product of the brand.

Consider for example an article of clothing that makes visual reference to the trade dress of another company, with the intent to demonstrate that the original purpose of the brand and perhaps fashion culture as a whole has become warped and barely recognizable. To send that message, the clothing would need to “borrow” enough of the trade dress to get the message across. If that trade dress is source-identifying—and if the product design trade dress exists, then it is presumably source-identifying—then Rogers no longer applies, under this case. But given the parodic nature of the product, the fact that consumers recognize the trade dress is the point. As long as consumers also recognize the differences, there is no confusion. And if they recognize the differences and understand those differences to mean the shoes are not from the original source, then perhaps there is no longer trademark infringement—confusion is unlikely, because consumers would get the joke or at minimum, realize something wasn’t right. The Supreme Court suggests this analysis can be sufficiently undertaken by applying the standard trademark infringement factors, without resorting to Rogers.

The court left untouched the Rogers test itself, specifically not opining on whether it is the appropriate way to balance the First Amendment against trademark rights in instances where the use of the trademark is not as a designation of source. What remains unclear, though, is when exactly that might be. If the use of the trademark is not as a designation of source, can there be a likelihood of confusion? If not, what do we need the Rogers test for, anyway? The instances may be limited to the sort of incidental usage seen in movies or heard in songs—the kind of expressive use Rogers was designed to protect in the first place. The concurring opinion of Justice Gorsuch, joined by Justices Thomas and Barrett, strongly suggested those three would rather dispose of Rogers altogether, though.

The net result of this decision is that brands will feel emboldened to enforce against unauthorized uses of their trademarks, even in contexts they may have previously hesitated due to the potential arguments available under Rogers. Without Rogers as an off-ramp, it will be harder to dispose of cases at summary judgment. As a result, creators may be more hesitant to create products that include a message about other brands, a chilling of expression the Supreme Court does not seem bothered by, perhaps believing it’s expression that never should have been protected in the first place.