Jack Daniel's™ Properties, Inc. v. VIP Products LLC, 599 U.S. ___ (2023)

NOTICE: This opinion is subject to formal revision before publication in the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, pio@supremecourt.gov, of any typographical or other formal errors.

SUPREME COURT OF THE UNITED STATES

_________________

No. 22–148

_________________

JACK DANIEL’S PROPERTIES, INC., PETITIONER v. VIP PRODUCTS LLC

on writ of certiorari to the united states court of appeals for the ninth circuit

[June 8, 2023]

Justice Kagan delivered the opinion of the Court.

This case is about dog toys and whiskey, two items seldom appearing in the same sentence. Respondent VIP Products makes a squeaky, chewable dog toy designed to look like a bottle of Jack Daniel’s whiskey. Though not entirely. On the toy, for example, the words “Jack Daniel’s” become “Bad Spaniels.” And the descriptive phrase “Old No. 7 Brand Tennessee Sour Mash Whiskey” turns into “The Old No. 2 On Your Tennessee Carpet.” The jokes did not impress petitioner Jack Daniel’s Properties. It owns trademarks in the distinctive Jack Daniel’s bottle and in many of the words and graphics on the label. And it believed Bad Spaniels had both infringed and diluted those trademarks. Bad Spaniels had infringed the marks, the argument ran, by leading consumers to think that Jack Daniel’s had created, or was otherwise responsible for, the dog toy. And Bad Spaniels had diluted the marks, the argument went on, by associating the famed whiskey with, well, dog excrement.

The Court of Appeals, in the decision we review, saw things differently. Though the federal trademark statute makes infringement turn on the likelihood of consumer confusion, the Court of Appeals never got to that issue. On the court’s view, the First Amendment compels a stringent threshold test when an infringement suit challenges a so-called expressive work—here (so said the court), the Bad Spaniels toy. And that test knocked out Jack Daniel’s claim, whatever the likelihood of confusion. Likewise, Jack’s dilution claim failed—though on that issue the problem was statutory. The trademark law provides that the “noncommercial” use of a mark cannot count as dilution. 15 U. S. C. §1125(c)(3)(C). The Bad Spaniels marks, the court held, fell within that exemption because the toy communicated a message—a kind of parody—about Jack Daniel’s.

Today, we reject both conclusions. The infringement issue is the more substantial. In addressing it, we do not decide whether the threshold inquiry applied in the Court of Appeals is ever warranted. We hold only that it is not appropriate when the accused infringer has used a trademark to designate the source of its own goods—in other words, has used a trademark as a trademark. That kind of use falls within the heartland of trademark law, and does not receive special First Amendment protection. The dilution issue is more simply addressed. The use of a mark does not count as noncommercial just because it parodies, or otherwise comments on, another’s products.

I

A

Start at square 1, with what a trademark is and does. The Lanham Act, the core federal trademark statute, defines a trademark as follows: “[A]ny word, name, symbol, or device, or any combination thereof ” that a person uses “to identify and distinguish his or her goods . . . from those manufactured or sold by others and to indicate the source of the goods.” §1127. The first part of that definition, identifying the kind of things covered, is broad: It encompasses words (think “Google”), graphic designs (Nike’s swoosh), and so-called trade dress, the overall appearance of a product and its packaging (a Hershey’s Kiss, in its silver wrapper). See Wal-Mart Stores, Inc. v. Samara Brothers, Inc., 529 U.S. 205, 209–210 (2000). The second part of the definition describes every trademark’s “primary” function: “to identify the origin or ownership of the article to which it is affixed.” Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 412 (1916). Trademarks can of course do other things: catch a consumer’s eye, appeal to his fancies, and convey every manner of message. But whatever else it may do, a trademark is not a trademark unless it identifies a product’s source (this is a Nike) and distinguishes that source from others (not any other sneaker brand). See generally 1 J. McCarthy, Trademarks and Unfair Competition §3:1 (5th ed. 2023). In other words, a mark tells the public who is responsible for a product.

In serving that function, trademarks benefit consumers and producers alike. A source-identifying mark enables customers to select “the goods and services that they wish to purchase, as well as those they want to avoid.” Matal v. Tam, 582 U.S. 218, 224 (2017). The mark “quickly and easily assures a potential customer that this item—the item with this mark—is made by the same producer as other similarly marked items that he or she liked (or disliked) in the past.” Qualitex Co. v. Jacobson Products Co., 514 U.S. 159, 164 (1995). And because that is so, the producer of a quality product may derive significant value from its marks. They ensure that the producer itself—and not some “imitating competitor”—will reap the financial rewards associated with the product’s good reputation. Ibid.

To help protect marks, the Lanham Act sets up a voluntary registration system. Any mark owner may apply to the Patent and Trademark Office to get its mark placed on a federal register. Consistent with trademark law’s basic purpose, the lead criterion for registration is that the mark “in fact serve as a ‘trademark’ to identify and distinguish goods.” 3 McCarthy §19:10 (listing the principal register’s eligibility standards). If it does, and the statute’s other criteria also are met, the registering trademark owner receives certain benefits, useful in infringement litigation. See, e.g., Iancu v. Brunetti, 588 U. S. ___, ___ (2019) (slip op., at 2) (noting that “registration constitutes ‘prima facie evidence’ of the mark’s validity”). But the owner of even an unregistered trademark can “use [the mark] in commerce and enforce it against infringers.” Ibid.

The Lanham Act also creates a federal cause of action for trademark infringement. In the typical case, the owner of a mark sues someone using a mark that closely resembles its own. The court must decide whether the defendant’s use is “likely to cause confusion, or to cause mistake, or to deceive.” §§1114(1)(A), 1125(a)(1)(A). The “keystone” in that statutory standard is “likelihood of confusion.” See 4 McCarthy §23:1. And the single type of confusion most commonly in trademark law’s sights is confusion “about the source of a product or service.” Moseley v. V Secret Catalogue, Inc., 537 U.S. 418, 428 (2003); see 4 McCarthy §23:5. Confusion as to source is the bête noire of trademark law—the thing that stands directly opposed to the law’s twin goals of facilitating consumers’ choice and protecting producers’ good will.

Finally, the Lanham Act creates a cause of action for the dilution of famous marks, which can succeed without likelihood of confusion. See §1125(c); Moseley, 537 U. S., at 431. A famous mark is one “widely recognized” by the public as “designati[ng the] source” of the mark owner’s goods. §1125(c)(2)(A). Dilution of such a mark can occur “by tarnishment” (as well as by “blurring,” not relevant here). §1125(c)(1). As the statute describes the idea, an “association arising from the similarity between” two marks—one of them famous—may “harm[ ] the reputation of the famous mark,” and thus make the other mark’s owner liable. §1125(c)(2)(C). But there are “[e]xclusions”—categories of activity not “actionable as dilution.” §1125(c)(3). One exclusion protects any “noncommercial use of a mark.” §1125(c)(3)(C). Another protects a “fair use” of a mark “in connection with . . . parodying, criticizing, or commenting upon the famous mark owner or [its] goods.” §1125(c)(3)(A)(ii). The fair-use exclusion, though, comes with a caveat. A defendant cannot get its benefit—even if engaging in parody, criticism, or commentary—when using the similar-looking mark “as a designation of source for the [defendant’s] own goods.” §1125(c)(3)(A). In other words, the exclusion does not apply if the defendant uses the similar mark as a mark.

B

A bottle of Jack Daniel’s—no, Jack Daniel’s Old No. 7 Tennessee Sour Mash Whiskey—boasts a fair number of trademarks. Recall what the bottle looks like (or better yet, retrieve a bottle from wherever you keep liquor; it’s probably there):

“Jack Daniel’s” is a registered trademark, as is “Old No. 7.” So too the arched Jack Daniel’s logo. And the stylized label with filigree (i.e., twirling white lines). Finally, what might be thought of as the platform for all those marks—the whiskey’s distinctive square bottle—is itself registered.

VIP is a dog toy company, making and selling a product line of chewable rubber toys that it calls “Silly Squeakers.” (Yes, they squeak when bitten.) Most of the toys in the line are designed to look like—and to parody—popular beverage brands. There are, to take a sampling, Dos Perros (cf. Dos Equis), Smella Arpaw (cf. Stella Artois), and Doggie Walker (cf. Johnnie Walker). VIP has registered trademarks in all those names, as in the umbrella term “Silly Squeakers.”

In 2014, VIP added the Bad Spaniels toy to the line. VIP did not apply to register the name, or any other feature of, Bad Spaniels. But according to its complaint (further addressed below), VIP both “own[s]” and “use[s]” the “ ‘Bad Spaniels’ trademark and trade dress.” App. 3, 11; see infra, at 8, 17. And Bad Spaniels’ trade dress, like the dress of many Silly Squeakers toys, is designed to evoke a distinctive beverage bottle-with-label. Even if you didn’t already know, you’d probably not have much trouble identifying which one.

Bad Spaniels is about the same size and shape as an ordinary bottle of Jack Daniel’s. The faux bottle, like the original, has a black label with stylized white text and a white filigreed border. The words “Bad Spaniels” replace “Jack Daniel’s” in a like font and arch. Above the arch is an image of a spaniel. (This is a dog toy, after all.) Below the arch, “The Old No. 2 On Your Tennessee Carpet” replaces “Old No. 7 Tennessee Sour Mash Whiskey” in similar graphic form. The small print at the bottom substitutes “43% poo by vol.” and “100% smelly” for “40% alc. by vol. (80 proof ).”

The toy is packaged for sale with a cardboard hangtag (so it can be hung on store shelves). Here is the back of the hangtag:

At the bottom is a disclaimer: “This product is not affiliated with Jack Daniel Distillery.” In the middle are some warnings and guarantees. And at the top, most relevant here, are two product logos—on the left for the Silly Squeakers line, and on the right for the Bad Spaniels toy.

Soon after Bad Spaniels hit the market, Jack Daniel’s sent VIP a letter demanding that it stop selling the product. VIP responded by bringing this suit, seeking a declaratory judgment that Bad Spaniels neither infringed nor diluted Jack Daniel’s trademarks. The complaint alleged, among other things, that VIP is “the owner of all rights in its ‘Bad Spaniels’ trademark and trade dress for its durable rubber squeaky novelty dog toy.” App. 3; see supra, at 6. Jack Daniel’s counterclaimed under the Lanham Act for both trademark infringement and trademark dilution by tarnishment.

VIP moved for summary judgment on both claims. First, VIP argued that Jack Daniel’s infringement claim failed under a threshold test derived from the First Amendment to protect “expressive works”—like (VIP said) the Bad Spaniels toy. When those works are involved, VIP contended, the so-called Rogers test requires dismissal of an infringement claim at the outset unless the complainant can show one of two things: that the challenged use of a mark “has no artistic relevance to the underlying work” or that it “explicitly misleads as to the source or the content of the work.” Rogers v. Grimaldi, 875 F.2d 994, 999 (CA2 1989) (Newman, J.). Because Jack Daniel’s could make neither showing, VIP argued, the likelihood-of-confusion issue became irrelevant. Second, VIP urged that Jack Daniel’s could not succeed on a dilution claim because Bad Spaniels was a “parody[ ]” of Jack Daniel’s, and therefore made “fair use” of its famous marks. §1125(c)(3)(A)(ii).

The District Court rejected both contentions for a common reason: because VIP had used the cribbed Jack Daniel’s features as trademarks—that is, to identify the source of its own products. In the court’s view, when “another’s trademark is used for source identification”—as the court thought was true here—the threshold Rogers test does not apply. App. to Pet. for Cert. 89a. Instead, the suit must address the “standard” infringement question: whether the use is “likely to cause consumer confusion.” Ibid. And likewise, VIP could not invoke the dilution provision’s fair-use exclusion. Parodies fall within that exclusion, the court explained, only when the uses they make of famous marks do not serve as “a designation of source for the [alleged diluter’s] own goods.” Id., at 104a (quoting §1125(c)(3)(A)).

The case thus proceeded to a bench trial, where Jack Daniel’s prevailed. The District Court found, based largely on survey evidence, that consumers were likely to be confused about the source of the Bad Spaniels toy. See 291 F. Supp. 3d 891, 906–911 (D Ariz. 2018). And the court thought that the toy, by creating “negative associations” with “canine excrement,” would cause Jack Daniel’s “reputational harm.” Id., at 903, 905.

But the Court of Appeals for the Ninth Circuit reversed, ruling that the District Court had gotten the pretrial legal issues wrong. In the Ninth Circuit’s view, the infringement claim was subject to the threshold Rogers test because Bad Spaniels is an “expressive work”: Although just a dog toy, and “surely not the equivalent of the Mona Lisa,” it “communicates a humorous message.” 953 F.3d 1170, 1175 (2020) (internal quotation marks omitted). The Court of Appeals therefore returned the case to the District Court to decide whether Jack Daniel’s could satisfy either of Rogers’ two prongs. And the Ninth Circuit awarded judgment on the dilution claim to VIP. The court did not address the statutory exclusion for parody and other fair use, as the District Court had. Instead, the Court of Appeals held that the exclusion for “noncommercial use” shielded VIP from liability. §1125(c)(3)(C). The “use of a mark may be ‘noncommercial,’ ” the court reasoned, “even if used to sell a product.” 953 F. 3d, at 1176 (internal quotation marks omitted). And here it was so, the court found, because it “parodies” and “comments humorously” on Jack Daniel’s. Id., at 1175; see id., at 1176.

On remand, the District Court found that Jack Daniel’s could not satisfy either prong of Rogers, and so granted summary judgment to VIP on infringement. Jack Daniel’s appealed, and the Ninth Circuit summarily affirmed.

We then granted certiorari to consider the Court of Appeals’ rulings on both infringement and dilution. 598 U. S. ___ (2022).

II

Our first and more substantial question concerns Jack Daniel’s infringement claim: Should the company have had to satisfy the Rogers threshold test before the case could proceed to the Lanham Act’s likelihood-of-confusion inquiry?[1] The parties address that issue in the broadest possible way, either attacking or defending Rogers in all its possible applications. Today, we choose a narrower path. Without deciding whether Rogers has merit in other contexts, we hold that it does not when an alleged infringer uses a trademark in the way the Lanham Act most cares about: as a designation of source for the infringer’s own goods. See §1127; supra, at 2–3. VIP used the marks derived from Jack Daniel’s in that way, so the infringement claim here rises or falls on likelihood of confusion. But that inquiry is not blind to the expressive aspect of the Bad Spaniels toy that the Ninth Circuit highlighted. Beyond source designation, VIP uses the marks at issue in an effort to “parody” or “make fun” of Jack Daniel’s. Tr. of Oral Arg. 58, 66. And that kind of message matters in assessing confusion because consumers are not so likely to think that the maker of a mocked product is itself doing the mocking.

A

To see why the Rogers test does not apply here, first consider the case from which it emerged. The defendants there had produced and distributed a film by Federico Fellini titled “Ginger and Fred” about two fictional Italian cabaret dancers (Pippo and Amelia) who imitated Ginger Rogers and Fred Astaire. When the film was released in the United States, Ginger Rogers objected under the Lanham Act to the use of her name. The Second Circuit rejected the claim. It reasoned that the titles of “artistic works,” like the works themselves, have an “expressive element” implicating “ First Amendment values.” 875 F. 2d, at 998. And at the same time, such names posed only a “slight risk” of confusing consumers about either “the source or the content of the work.” Id., at 999–1000. So, the court concluded, a threshold filter was appropriate. When a title “with at least some artistic relevance” was not “explicitly misleading as to source or content,” the claim could not go forward. Ibid. But the court made clear that it was not announcing a general rule. In the typical case, the court thought, the name of a product was more likely to indicate its source, and to be taken by consumers in just that way. See id., at 1000.

Over the decades, the lower courts adopting Rogers have confined it to similar cases, in which a trademark is used not to designate a work’s source, but solely to perform some other expressive function. So, for example, when the toymaker Mattel sued a band over the song “Barbie Girl”—with lyrics including “Life in plastic, it’s fantastic” and “I’m a blond bimbo girl, in a fantasy world”—the Ninth Circuit applied Rogers. Mattel, Inc. v. MCA Records, Inc., 296 F.3d 894, 901 (2002). That was because, the court reasoned, the band’s use of the Barbie name was “not [as] a source identifier”: The use did not “speak[ ] to [the song’s] origin.” Id., at 900, 902; see id., at 902 (a consumer would no more think that the song was “produced by Mattel” than would, “upon hearing Janis Joplin croon ‘Oh Lord, won’t you buy me a Mercedes Benz?,’ . . . suspect that she and the carmaker had entered into a joint venture”). Similarly, the Eleventh Circuit dismissed a suit under Rogers when a sports artist depicted the Crimson Tide’s trademarked football uniforms solely to “memorialize” a notable event in “football history.” University of Ala. Bd. of Trustees v. New Life Art, Inc., 683 F.3d 1266, 1279 (2012). And when Louis Vuitton sued because a character in the film The Hangover: Part II described his luggage as a “Louis Vuitton” (though pronouncing it Lewis), a district court dismissed the complaint under Rogers. See Louis Vuitton Mallatier S. A. v. Warner Bros. Entertainment Inc., 868 F. Supp. 2d 172 (SDNY 2012). All parties agreed that the film was not using the Louis Vuitton mark as its “own identifying trademark.” Id., at 180 (internal quotation marks omitted). When that is so, the court reasoned, “confusion will usually be unlikely,” and the “interest in free expression” counsels in favor of avoiding the standard Lanham Act test. Ibid.

The same courts, though, routinely conduct likelihood-of-confusion analysis, without mentioning Rogers, when trademarks are used as trademarks—i.e., to designate source. See, e.g., JL Beverage Co., LLC v. Jim Beam Brands Co., 828 F.3d 1098, 1102–1103, 1106 (CA9 2016); PlayNation Play Systems, Inc. v. Velex Corp., 924 F.3d 1159, 1164–1165 (CA11 2019). And the Second Circuit—Rogers’ home court—has made especially clear that Rogers does not apply in that context. For example, that court held that an offshoot political group’s use of the trademark “United We Stand America” got no Rogers help because the use was as a source identifier. See United We Stand Am., Inc. v. United We Stand, Am. New York, Inc., 128 F.3d 86, 93 (1997). True, that slogan had expressive content. But the defendant group, the court reasoned, was using it “as a mark,” to suggest the “same source identification” as the original “political movement.” Ibid. And similarly, the Second Circuit (indeed, the judge who authored Rogers) rejected a motorcycle mechanic’s view that his modified version of Harley Davidson’s bar-and-shield logo was an expressive parody entitled to Rogers’ protection. See Harley-Davidson, Inc. v. Grottanelli, 164 F.3d 806, 812–813 (1999). The court acknowledged that the mechanic’s adapted logo conveyed a “somewhat humorous[ ]” message. Id., at 813. But his use of the logo was a quintessential “trademark use”: to brand his “repair and parts business”—through signage, a newsletter, and T-shirts—with images “similar” to Harley-Davidson’s. Id., at 809, 812–813.

The point is that whatever you make of Rogers—and again, we take no position on that issue—it has always been a cabined doctrine. If we put this case to the side, the Rogers test has applied only to cases involving “non-trademark uses”—or otherwise said, cases in which “the defendant has used the mark” at issue in a “non-source-identifying way.” S. Dogan & M. Lemley, Grounding Trademark Law Through Trademark Use, 92 Iowa L. Rev. 1669, 1684 (2007); see id., at 1683–1684, and n. 58. The test has not insulated from ordinary trademark scrutiny the use of trademarks as trademarks, “to identify or brand [a defendant’s] goods or services.” Id., at 1683.

We offer as one last example of that limitation a case with a striking resemblance to this one. It too involved dog products, though perfumes rather than toys. Yes, the defendant sold “a line of pet perfumes whose names parody elegant brands sold for human consumption.” Tommy Hilfiger Licensing, Inc. v. Nature Labs, LLC, 221 F. Supp. 2d 410, 412 (SDNY 2002) (Mukasey, J.). The product at issue was named Timmy Holedigger—which Tommy Hilfiger didn’t much like. The defendant asked for application of Rogers. The court declined it, relying on Harley-Davidson. See 221 F. Supp. 2d, at 414. Rogers, the court explained, kicks in when a suit involves solely “nontrademark uses of [a] mark—that is, where the trademark is not being used to indicate the source or origin” of a product, but only to convey a different kind of message. 221 F. Supp. 2d, at 414. When, instead, the use is “at least in part” for “source identification”—when the defendant may be “trading on the good will of the trademark owner to market its own goods”—Rogers has no proper role. 221 F. Supp. 2d, at 414–415. And that is so, the court continued, even if the defendant is also “making an expressive comment,” including a parody of a different product. Id., at 415. The defendant is still “mak[ing] trademark use of another’s mark,” and must meet an infringement claim on the usual battleground of “likelihood of confusion.” Id., at 416.

That conclusion fits trademark law, and reflects its primary mission. From its definition of “trademark” onward, the Lanham Act views marks as source identifiers—as things that function to “indicate the source” of goods, and so to “distinguish” them from ones “manufactured or sold by others.” §1127; see supra, at 2–3. The cardinal sin under the law, as described earlier, is to undermine that function. See supra, at 3. It is to confuse consumers about source—to make (some of ) them think that one producer’s products are another’s. And that kind of confusion is most likely to arise when someone uses another’s trademark as a trademark—meaning, again, as a source identifier—rather than for some other expressive function. To adapt one of the cases noted above: Suppose a filmmaker uses a Louis Vuitton suitcase to convey something about a character (he is the kind of person who wants to be seen with the product but doesn’t know how to pronounce its name). See supra, at 12. Now think about a different scenario: A luggage manufacturer uses an ever-so-slightly modified LV logo to make inroads in the suitcase market. The greater likelihood of confusion inheres in the latter use, because it is the one conveying information (or misinformation) about who is responsible for a product. That kind of use “implicate[s] the core concerns of trademark law” and creates “the paradigmatic infringement case.” G. Dinwoodie & M. Janis, Confusion Over Use: Contextualism in Trademark Law, 92 Iowa L. Rev. 1597, 1636 (2007). So the Rogers test—which offers an escape from the likelihood-of-confusion inquiry and a shortcut to dismissal—has no proper application.[2]

Nor does that result change because the use of a mark has other expressive content—i.e., because it conveys some message on top of source. Here is where we most dramatically part ways with the Ninth Circuit, which thought that because Bad Spaniels “communicates a humorous message,” it is automatically entitled to Rogers’ protection. 953 F. 3d, at 1175 (internal quotation marks omitted). On that view, Rogers might take over much of the world. For trademarks are often expressive, in any number of ways. Consider how one liqueur brand’s trade dress (beyond identifying source) tells a story, with a bottle in the shape of a friar’s habit connoting the product’s olden monastic roots:

Or take a band name that “not only identifies the band but expresses a view about social issues.” Tam, 582 U. S., at 245 (opinion of Alito, J.) (discussing “The Slants”). Or note how a mark can both function as a mark and have parodic content—as the court found in the Hilfiger/Holedigger litigation. See supra, at 13–14. The examples could go on and on. As a leading treatise puts the point, the Ninth Circuit’s expansion of Rogers “potentially encompasses just about everything” because names, phrases, symbols, designs, and their varied combinations often “contain some ‘expressive’ message” unrelated to source. 6 McCarthy §31:144.50. That message may well be relevant in assessing the likelihood of confusion between two marks, as we address below. See infra, at 18–19. But few cases would even get to the likelihood-of-confusion inquiry if all expressive content triggered the Rogers filter. In that event, the Rogers exception would become the general rule, in conflict with courts’ longstanding view of trademark law.

The Ninth Circuit was mistaken to believe that the First Amendment demanded such a result. The court thought that trademark law would otherwise “fail[ ] to account for the full weight of the public’s interest in free expression.” 953 F. 3d, at 1174. But as the Mattel (i.e., Barbie) court noted, when a challenged trademark use functions as “source-identifying,” trademark rights “play well with the First Amendment”: “Whatever first amendment rights you may have in calling the brew you make in your bathtub ‘Pepsi’ ” are “outweighed by the buyer’s interest in not being fooled into buying it.” 296 F. 3d, at 900. Or in less colorful terms: “[T]o the extent a trademark is confusing” as to a product’s source “the law can protect consumers and trademark owners.” Tam, 582 U. S., at 252 (Kennedy, J., concurring in part and concurring in judgment); see Friedman v. Rogers, 440 U.S. 1, 15 (1979) (rejecting a First Amendment challenge to a law restricting trade names because of the “substantial” interest in “protecting the public from [their] deceptive and misleading use”). Or yet again, in an especially clear rendering: “[T]he trademark law generally prevails over the First Amendment” when “another’s trademark (or a confusingly similar mark) is used without permission” as a means of “source identification.” Yankee Publishing Inc. v. News Am. Publishing Inc., 809 F. Supp. 267, 276 (SDNY 1992) (Leval, J.) (emphasis deleted). So for those uses, the First Amendment does not demand a threshold inquiry like the Rogers test. When a mark is used as a mark (except, potentially, in rare situations), the likelihood-of-confusion inquiry does enough work to account for the interest in free expression.

B

Here, the District Court correctly held that “VIP uses its Bad Spaniels trademark and trade dress as source identifiers of its dog toy.” See App. to Pet. for Cert. 105a. In fact, VIP conceded that point below. In its complaint, VIP alleged that it both “own[s] and “use[s]” the “ ‘Bad Spaniels’ trademark and trade dress for its durable rubber squeaky novelty dog toy.” App. 3, 11. The company thus represented in this very suit that the mark and dress, although not registered, are used to “identify and distinguish [VIP’s] goods” and to “indicate [their] source.” §1127. (Registration of marks, you’ll recall, is optional. See supra, at 3–4.)

In this Court, VIP says the complaint was a mere “form allegation”—a matter of “rote.” Tr. of Oral Arg. 73. But even if we knew what that meant, VIP has said and done more in the same direction. First, there is the way the product is marketed. On the hangtag, the Bad Spaniels logo sits opposite the concededly trademarked Silly Squeakers logo, with both appearing to serve the same source-identifying function. See supra, at 7. And second, there is VIP’s practice as to other products in the Silly Squeakers line. The company has consistently argued in court that it owns, though has never registered, the trademark and trade dress in dog toys like “Jose Perro” (cf. Jose Cuervo) and “HeinieSniff ’n” (cf. Heineken).[3] And it has chosen to register the names of still other dog toys, including Dos Perros (#6176781), Smella Arpaw (#6262975), and Doggie Walker (#6213816). See supra, at 6. Put all that together, and more than “form” or “rote” emerges: VIP’s conduct is its own admission that it is using the Bad Spaniels (née Jack Daniel’s) trademarks as trademarks, to identify product source.

Because that is so, the only question in this suit going forward is whether the Bad Spaniels marks are likely to cause confusion. There is no threshold test working to kick out all cases involving “expressive works.” But a trademark’s expressive message—particularly a parodic one, as VIP asserts—may properly figure in assessing the likelihood of confusion. See, e.g., Louis Vuitton Malletier S. A. v. Haute Diggity Dog, LLC, 507 F.3d 252, 265 (CA4 2007) (Parody “influences the way in which the [likelihood-of-confusion] factors are applied”); Brief for United States as Amicus Curiae 17–22 (same). A parody must “conjure up” “enough of [an] original to make the object of its critical wit recognizable.” Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 588 (1994) (internal quotation marks omitted). Yet to succeed, the parody must also create contrasts, so that its message of ridicule or pointed humor comes clear. And once that is done (if that is done), a parody is not often likely to create confusion. Self-deprecation is one thing; self-mockery far less ordinary. So although VIP’s effort to ridicule Jack Daniel’s does not justify use of the Rogers test, it may make a difference in the standard trademark analysis. Consistent with our ordinary practice, we remand that issue to the courts below. See Cutter v. Wilkinson, 544 U.S. 709, 718, n. 7 (2005) (noting that this Court is generally “a court of review, not of first view”).

III

Our second question, more easily dispatched, concerns Jack Daniel’s claim of dilution by tarnishment (for the linkage of its whiskey to less savory substances). Recall that the Ninth Circuit dismissed that claim based on one of the Lanham Act’s “[e]xclusions” from dilution liability—for “[a]ny noncommercial use of a mark.” §1125(c)(3)(C); see supra, at 9. On the court’s view, the “use of a mark may be ‘noncommercial’ even if used to sell a product.” 953 F. 3d, at 1176 (internal quotation marks omitted). And VIP’s use is so, the court continued, because it “parodies” and “convey[s] a humorous message” about Jack Daniel’s. Id., at 1175–1176. We need not express a view on the first step of that reasoning because we think the second step wrong. However wide the scope of the “noncommercial use” exclusion, it cannot include, as the Ninth Circuit thought, every parody or humorous commentary.

To begin to see why, consider the scope of another of the Lanham Act’s exclusions—this one for “[a]ny fair use.” As described earlier, the “fair use” exclusion specifically covers uses “parodying, criticizing, or commenting upon” a famous mark owner. §1125(c)(3)(A)(ii); see supra, at 5. But not in every circumstance. Critically, the fair-use exclusion has its own exclusion: It does not apply when the use is “as a designation of source for the person’s own goods or services.” §1125(c)(3)(A). In that event, no parody, criticism, or commentary will rescue the alleged dilutor. It will be subject to liability regardless.

The problem with the Ninth Circuit’s approach is that it reverses that statutorily directed result, as this case illustrates. Given the fair-use provision’s carve-out, parody (and criticism and commentary, humorous or otherwise) is exempt from liability only if not used to designate source. Whereas on the Ninth Circuit’s view, parody (and so forth) is exempt always—regardless whether it designates source. The expansive view of the “noncommercial use” exclusion effectively nullifies Congress’s express limit on the fair-use exclusion for parody, etc. Just consider how the Ninth Circuit’s construction played out here. The District Court had rightly concluded that because VIP used the challenged marks as source identifiers, it could not benefit from the fair-use exclusion for parody. See App. to Pet. for Cert. 105a; supra, at 8–9, 17–18. The Ninth Circuit took no issue with that ruling. But it shielded VIP’s parodic uses anyway. In doing so, the court negated Congress’s judgment about when—and when not—parody (and criticism and commentary) is excluded from dilution liability.

IV

Today’s opinion is narrow. We do not decide whether the Rogers test is ever appropriate, or how far the “noncommercial use” exclusion goes. On infringement, we hold only that Rogers does not apply when the challenged use of a mark is as a mark. On dilution, we hold only that the noncommercial exclusion does not shield parody or other commentary when its use of a mark is similarly source-identifying. It is no coincidence that both our holdings turn on whether the use of a mark is serving a source-designation function. The Lanham Act makes that fact crucial, in its effort to ensure that consumers can tell where goods come from.

For the reasons stated, we vacate the judgment below and remand for further proceedings consistent with this opinion.

It is so ordered.

SUPREME COURT OF THE UNITED STATES

_________________

No. 22–148

_________________

JACK DANIEL’S PROPERTIES, INC., PETITIONER v. VIP PRODUCTS LLC

on writ of certiorari to the united states court of appeals for the ninth circuit

[June 8, 2023]

Justice Sotomayor, with whom Justice Alito joins, concurring.

I join the Court’s opinion in full. I write separately to emphasize that in the context of parodies and potentially other uses implicating First Amendment concerns, courts should treat the results of surveys with particular caution. As petitioner did here, plaintiffs in trademark infringement cases often commission surveys that purport to show that consumers are likely to be confused by an allegedly infringing product. Like any other evidence, surveys should be understood as merely one piece of the multifaceted likelihood of confusion analysis. See, e.g., Uncommon, LLC v. Spigen, Inc., 926 F.3d 409, 425 (CA7 2019). Courts should also carefully assess the methodology and representativeness of surveys, as many lower courts already do. See, e.g., Water Pik, Inc. v. Med-Systems, Inc., 726 F.3d 1136, 1144–1150 (CA10 2013); Starbucks Corp. v. Wolfe’s Borough Coffee, Inc., 588 F.3d 97, 117 (CA2 2009).

When an alleged trademark infringement involves a parody, however, there is particular risk in giving uncritical or undue weight to surveys. Survey answers may reflect a mistaken belief among some survey respondents that all parodies require permission from the owner of the parodied mark. Some of the answers to the survey in this case illustrate this potential. See App. 81–82, n. 25 (“ ‘I’m sure the dog toy company that made this toy had to get [Jack Daniel’s] permission and legal rights to essentially copy the[ir] product in dog toy form’ ”); ibid. (“ ‘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’ ”); see also Anheuser-Busch, Inc. v. Balducci Publications, 28 F.3d 769, 772–773, 775 (CA8 1994) (describing a similar situation). Plaintiffs can point to this misunderstanding of the legal framework as evidence of consumer confusion. Cleverly designed surveys could also prompt such confusion by making consumers think about complex legal questions around permission that would not have arisen organically out in the world.

Allowing such survey results to drive the infringement analysis would risk silencing a great many parodies, even ones that by other metrics are unlikely to result in the confusion about sourcing that is the core concern of the Lanham Act. See ante, at 4, 10, 14. Well-heeled brands with the resources to commission surveys would be handed an effective veto over mockery. After all, “[n]o one likes to be the butt of a joke, not even a trademark.” 6 J. McCarthy, Trademarks and Unfair Competition §31:153 (5th ed. 2023). This would upset the Lanham Act’s careful balancing of “the needs of merchants for identification as the provider of goods with the needs of society for free communication and discussion.” P. Leval, Trademark: Champion of Free Speech, 27 Colum. J. L. & Arts 187, 210 (2004). Courts should thus ensure surveys do not completely displace other likelihood-of-confusion factors, which may more accurately track the experiences of actual consumers in the marketplace. Courts should also be attentive to ways in which surveys may artificially prompt such confusion about the law or fail to sufficiently control for it.

SUPREME COURT OF THE UNITED STATES

_________________

No. 22–148

_________________

JACK DANIEL’S PROPERTIES, INC., PETITIONER v. VIP PRODUCTS LLC

on writ of certiorari to the united states court of appeals for the ninth circuit

[June 8, 2023]

Justice Gorsuch, with whom Justice Thomas and Justice Barrett join, concurring.

I am pleased to join the Court’s opinion. I write separately only to underscore that lower courts should handle Rogers v. Grimaldi, 875 F.2d 994 (CA2 1989), with care. Today, the Court rightly concludes that, even taken on its own terms, Rogers does not apply to cases like the one before us. But in doing so, we necessarily leave much about Rogers unaddressed. For example, it is not entirely clear where the Rogers test comes from—is it commanded by the First Amendment, or is it merely gloss on the Lanham Act, perhaps inspired by constitutional-avoidance doctrine? Id., at 998. For another thing, it is not obvious that Rogers is correct in all its particulars—certainly, the Solicitor General raises serious questions about the decision. See Brief for United States as Amicus Curiae 23–28. All this remains for resolution another day, ante, at 13, and lower courts should be attuned to that fact.