The Yale Law Journal


A Trademark Defense of the Disparagement Bar

26 Oct 2016

The Supreme Court will soon hear argument over whether Congress may forbid registering trademarks that consist of “matter which may disparage or falsely suggest a connection with persons, living or dead, institutions, beliefs, or national symbols, or bring them into contempt, or disrepute.”1 The disparagement bar is the basis for the 2014 ruling by the Trademark Trial and Appeal Board (TTAB) that ordered the cancellation of trademark registrations belonging to the Washington NFL team because the term “redskin” disparages Native Americans.2 Late last year, however, the Federal Circuit ruled en banc that the disparagement bar is unconstitutional on First Amendment grounds in In re Tam.3

In re Tam sets the stage for Supreme Court review by considering a variety of doctrines designed to accommodate First Amendment interests to other values. The Federal Circuit’s treatment of the disparagement bar boiled down in large part to the question of whether the bar serves a viewpoint-neutral purpose beyond disapproval of the message embodied by disparaging marks. The majority saw no such purpose.4 In contrast, one dissent argued that the bar promotes commerce by targeting stigmatizing insults that disrupt the marketplace.5

Fair enough, but commerce promotion is a broad goal. Promoting that general end by specifically targeting stigmatizing insults might support an inference that the disparagement bar actually exists to promote anti-discrimination policies external to trademark law. That view corresponds with the majority’s treatment of the disparagement bar as having little to do with trademark policy.6 In re Tam therefore looks like a clash between fundamental principles of free expression and anti-discrimination in which trademark doctrine sits on the sideline—a framing that makes it all the more difficult for the bar to survive First Amendment scrutiny. The provision looks suspect because it seems an interloper unrelated to its surrounding body of law, potentially undermining its claim of viewpoint neutrality.

This Essay challenges that framing. Its purpose is not to argue the constitutionality of the bar, nor is it to set forth a particular First Amendment framework for approaching the question.7 My more modest aim is to give trademark doctrine its due. The disparagement bar implicates more mundane issues of trademark policy that deserve fuller consideration than offered by In re Tam. Whatever the ability of First Amendment doctrine to accommodate the application of external anti-discrimination policies to questions of trademark eligibility,8 the disparagement bar also reflects principles internal to trademark law. More specifically, the bar enforces the traditional trademark policy that insists that would-be trademarks adequately perform the trademark function. Trademark law does not care about the views embodied by candidate trademarks, but it does require that they do a good job of serving as source identifiers. Disparaging marks are just one of several categories of potential marks that fail this test. Understanding the disparagement bar in the larger context of trademark law may therefore offer a basis for defending its constitutionality against First Amendment challenge.

I. in re tam’s first amendment challenge to the disparagement Bar

Although trademarks do not need registration for protection, the Lanham Act offers a range of benefits to registered marks.9 It also provides various grounds for refusing registration. Section 2(a) contains one such exclusion—a bar on the registration of trademarks consisting of disparaging matter10—which was the basis of the ordered cancellation of trademark registrations belonging to the Washington Redskins.11 The team has challenged the cancellation by arguing, among other things, that the disparagement bar violates the First Amendment.12

The Federal Circuit accepted this argument in In re Tam. Simon Shiao Tam attempted to register THE SLANTS as a mark for an Asian-American dance-rock band. The examining attorney refused, concluding that the term is likely to disparage “persons of Asian descent,” and the TTAB affirmed.13 For his part, Tam argued that, as an Asian-American, his use of the term was a form of appropriation and commentary.14 He challenged the PTO’s refusal to register the mark as a First Amendment violation. By a 9-3 vote, the Federal Circuit, sitting en banc, agreed and declared the disparagement bar facially unconstitutional.

The majority opinion sees the disparagement bar as a form of content and viewpoint discrimination.15 In its account, the viewpoint discrimination is not a byproduct of some other policy; it is the very point.16 This characterization makes the disparagement bar’s fate seem overdetermined. The opinion runs through other First Amendment arguments and doctrines, concluding that the bar cannot be justified as a form of government speech or subsidy17 and that it should not be judged under the commercial speech doctrine.18 But the conclusion that the bar’s purpose is viewpoint discrimination is woven throughout.19

Judge Reyna’s dissent takes issue with the view that the bar is meant to suppress speech, defending it as a content-neutral effort to regulate the secondary effects of harmful speech.20 In his account, the point of the bar lies in the Lanham Act’s goal of promoting “the orderly flow of commerce.”21 Stigmatizing language in the form of disparaging trademarks compromises this goal because an insult based on “race, gender, religion, or other demographic identity, tends to disrupt commercial activity and to undermine the stability of the marketplace in much the same manner as discriminatory conduct.”22

The problem with the stigma rationale is that it risks reinforcing the perception that the disparagement bar is a form of viewpoint discrimination. Simply describing the provision’s purpose as one of promoting “commerce” is too general to shift the focus of judges—like those in the Tam majority—already inclined to see viewpoint discrimination in the operation of the disparagement bar.23 For one thing, expressive harm, though real, might appear to have less of an impact on market participation than the discriminatory conduct—for example, discrimination in places of public accommodation—that has traditionally been the focus of governmental remedial efforts. Judge Reyna’s dissent implicitly raises this issue, as he notes that the disruption caused by “discriminatory conduct” is well recognized and documented by the history supporting passage of the Civil Rights Act.24 But there is not similar legislative history supporting the disparagement bar on this basis. The “conduct” in question seems bound in expression.

Moreover, the danger to commerce posed by insulting language is not a traditional trademark concern. The argument as stated by Judge Reyna does not address the ability of disparaging terms to function as marks. Rather, the problem and means of resolution (via a refusal to promote “low-value” speech)25 both lie outside ordinary trademark discourse. To be sure, not all judges will see a problem. One might connect this conception of the disparagement bar with, say, the prohibition against trademarking generic terms as both “promoting commerce.” But this kinship exists only at a high level of generality. It is therefore unsurprising that the majority did not follow along, either as a matter of pattern-recognition—what “looks like” the ordinary operation of trademark law26—or out of concern that the argument could countenance more overt forms of viewpoint discrimination.27

It may therefore be true that disparaging marks disrupt commerce, but the mechanism identified by Judge Reyna nonetheless reinforces the perception that the stakes are not the ordinary functioning of trademark law, but rather trademark law’s ability to be put to the external goal of combating discrimination.28 Many academic defenses of the disparagement bar sound similar notes.29 The challenge is fitting the effort into one of the many nuanced nooks and crannies of modern First Amendment doctrine.30

Although vindicating equal protection ideals is a far more pressing concern than the operation of the Lanham Act, casting the issue in these terms invites judicial skepticism by seeming to bring free expression and anti-discrimination policies into conflict.31 Analogous perceived collisions between the two interests have favored free expression.32

This Essay takes no position on how best to frame the First Amendment analysis of the disparagement bar’s constitutionality. But the debate as joined in Tam makes relevant the question of whether the bar serves a trademark purpose. The majority’s refusal to accept the conception of trademark law forwarded by Judge Reyna illustrates how treating the concerns of the disparagement bar as exogenous to trademark law may start the debate from behind. A better course might therefore be to focus on neither commerce nor discrimination, but rather on the purposes of trademark law—and the means by which trademark doctrine traditionally implements them—at a lower level of generality.

II. trademark law and the value of “empty vessels”

Given the views of the Tam majority, the disparagement bar would be on firmer constitutional ground if it can be shown to reflect viewpoint-neutral trademark principles, even as it simultaneously promotes anti-discrimination policies. Tam rejects this possibility, arguing that “it is always a mark’s expressive character, not its ability to serve as a source identifier, that is the basis for the disparagement exclusion from registration.”33

Not so—or at least, not entirely so.However much the bar promotes anti-discrimination ideals, it also aligns with trademark law’s other exclusions from protection and registration. These exclusions lack any connection to anti-discrimination goals, and, for the most part, have no overt expressive component (which is not to say that they lack any such component).34 The bar reflects a policy that promotes effective trademarks by favoring, to a large, albeit imperfect, extent, signifiers that are initially irrelevant to the market context in which they will be deployed.35 Such marks are better able to serve as receptacles of source-identifying meaning and, in turn, to be a shorthand for other information about the product or service in question. By contrast, trademark law disfavors and often excludes would-be marks that bring market-relevant information to the table.

This tendency is rooted in the function of trademarks. Trademarks exist to identify and distinguish goods.36 Trademarks generate a range of benefits by doing so. They prevent fraud. They lower search costs. They encourage investments in quality by ensuring that consumers know whom to credit for making superior products. They embody the goodwill trademark owners build up with consumers and permit that value to be captured. They enable sellers to efficiently convey information about their products (by advertising attributes about them), and they enable others to do the same (by word of mouth, consumer reviews, or the like). And as cultural objects, they often become part of everyday language. In other words, once a trademark’s source-identifying function is established, the mark becomes an efficient vehicle for a range of other meanings.37 But these benefits flow from the mark’s ability to be source-identifying.

Trademark law has a lot to say about which marks are up to the task.38 It sees some marks—for example, arbitrary marks like APPLE computer or fanciful marks like KODAK film—as being inherently open to new trademark meanings. Those classes of marks are assumed to perform a source-identifying function. Not so with other classes of marks. Descriptive marks, such as TASTY hamburgers, require secondary meaning for protection. Consumers must actually see them as performing a source-identifying function; trademark meaning cannot be taken on faith. Generic marks that describe product categories, like APPLE apples, cannot be protected under any circumstances. Suggestive marks, which suggest product attributes but do not explicitly claim them, sit in between these two classes. They may be inherently distinctive, but they are sometimes treated by courts as being “weaker” than arbitrary or fanciful marks, limiting their scope in infringement litigation.39

One characteristic of this familiar spectrum of distinctiveness is that the more irrelevant a word is to the product in question, the more favorable trademark treatment it receives. On one end, arbitrary and fanciful marks have no meaning in relation to the referenced product; they are, relatively speaking, empty vessels.40 But as one moves down the spectrum to words that have contextual meaning in connection to the product, trademark protection either weakens, in the case of suggestive marks; becomes harder to obtain, as with descriptive marks; or is refused altogether, as with generic marks.

To be sure, this is partly a judgment about the needs of competitors—ensuring that they are able to accurately describe their own products.41 But it also reflects an evaluation of the ability of certain words to function as marks. Judge Leval has argued, for example, that because food sellers have a competitive interest in using the word “delicious” to describe their goods, “[c]onsumers who see the word delicious used on two or more different food products are less likely to draw the inference that they must all come from the same producer.”42

Similar considerations apply to non-word marks. Product design (as opposed to packaging) cannot be protected absent secondary meaning. This rule is generally seen as supporting the functionality doctrine, which excludes functional matter from protection in part to preserve market competition. Too easy protection for product design raises the risk of meritless claims that threaten the competition that the functionality doctrine promotes.43

But the product design rule also promotes trademark quality. Like generic or descriptive marks, a product’s design brings non-trademark meanings—in this case, information about its attributes—to the table. In announcing the product design rule, the Supreme Court asserted that such attribute information interferes with the ability of the design to serve a trademark function.

[W]e think consumer predisposition to equate the feature with the source does not exist. Consumers are aware of the reality that, almost invariably, even the most unusual of product designs—such as a cocktail shaker shaped like a penguin—is intended not to identify the source, but to render the product itself more useful or more appealing.44

In this, product design is quite unlike an inherently distinctive word mark or eye-catching product packaging. To the extent an arbitrary word mark, for example, embodies market-relevant information beyond source, the information flows from the use as a source identifier, as when a trademarked product earns a good reputation. Because the resulting positive associations depend on the trademark use, and not on inherent qualities of the trademarked word, the source meaning of an inherently distinctive word mark will always predominate over other meanings.45 When non-source meanings loom larger, trademark doctrine is less sure that will be the case and accordingly withholds protection.46

III. another view of the disparagement bar

The preceding discussion signals a way to harmonize the disparagement bar with the rest of trademark law. The bar is one of several manifestations of a viewpoint-neutral principle: to be protected, a mark should be good at identifying and distinguishing a good or service. Disparaging marks comprise just one of several classes of marks that do a bad job of performing this source-identifying function.

A. Pre-Filled Vessels

Although they do not necessarily bring the same kind of market-relevant information as, say, a generic term,47 disparaging marks share the characteristic of being infused with non-source meaning. Unlike arbitrary or fanciful marks, disparaging marks fail the irrelevance test because they embody meanings that interfere with their ability to perform a source-identifying function. In context, they are not the empty vessels that trademark law idealizes because consumers perceive a disparaging message. Just as APPLE lacks trademark significance when applied to the fruit instead of computers, and penguin-shaped cocktail shakers denote a product more than its source, slurs leave less room for trademark meanings than do the comparatively neutral terms that trademark law traditionally favors. To be sure, this is not a necessary choice, but trademark doctrine might rationally determine that their emotional impact overwhelms their ability to identify based on a judgment that consumers “are aware of the reality that, almost invariably,” slurs are “intended not to identify the source,” but to provoke or insult.48 Requiring insulted consumers to then determine whether a disparaging term conveys a source-identifying meaning beyond insult (or parody, commentary, or some combination) may not serve trademark goals.

Seen this way, the bar to registering disparaging marks looks less like viewpoint discrimination and more of a piece with trademark doctrine’s general approach of promoting marks whose content is irrelevant to the market in question.49 Arbitrary and fanciful marks receive their status by having no connection to the denoted product at all. As relatively empty vessels, they are more easily filled with meaning by the trademark holder. This sort of neutrality is impossible for disparaging marks. Their tendency to excite emotion makes their content relevant to at least some consumers and thus suspect. These judgments about consumer reactions may be incorrect, but empirically ungrounded best guesses about consumer impact infuse trademark law. They are hardly restricted to disparaging marks.50

B. Secondary Meaning and Non-Discrimination

The tendency of disparaging marks to interfere with source meanings will, perforce, be strongest among members of the disparaged group. This makes sensible the practice of using the perceptions of the referenced group to define disparagement. Trademark law’s use of a group-specific inquiry is not particular to disparaging marks. Most broadly, of course, the multifactor likelihood-of-confusion test considers the attributes of potential consumers of the goods at issue. More analogously to the disparagement bar, the doctrine of foreign equivalents prevents registration applicants from circumventing registration restrictions by substituting words from foreign languages with equivalent meanings.51 It does so notwithstanding the fact that many Americans are monolingual and would not recognize non-English terms.

The power of disparaging marks for the referenced group also answers the question of why registration with secondary meaning is unavailable. For a significant subset of the consuming public, secondary meaning may be impossible: the disparaging meanings will dominate source identification.

Although this argument arises from trademark law, it is also consistent with the anti-discrimination view of the disparagement bar, as the bar enables members of the affected group to have equal access to source-identifying marks as members of the general consuming public. But because the argument focuses on the operation of trademark law, it does not suffer from the same generality problem discussed above.52 The benefit in question—the availability of source-identifying marks—is the specific benefit trademark law provides to serve the more general goal of promoting commerce.

C. Expressive Failure

Disparaging marks may still seem uniquely expressive, increasing suspicion of the bar. But the irrelevance principle also excludes other marks that double as expressive speech because the expression interferes with source identification. Consider the attempt to register I CAN’T BREATHE. The phrase became famous after the death of Eric Garner during his arrest by New York City police officers on suspicion of selling untaxed cigarettes. Garner was placed in a chokehold and brought to the ground. While down, he repeatedly said, “I can’t breathe,” lost consciousness, and died from head and chest compression. The police officers were not indicted, leading to nationwide protests in which the phrase “I can’t breathe” figured prominently. Many famous athletes, including LeBron James, Kevin Garnett, and Kobe Bryant, wore t-shirts with the phrase, further increasing public attention.53

In December 2014, a trademark registration application was filed for I CAN’T BREATHE for use in “[c]lothing, namely hoodies, t-shirts for men, women, boys, girls and infants.”54 The application was abandoned after the PTO identified various reasons for rejecting it.55

The obvious expressive power of “I can’t breathe” was part of the problem.56 The issue was not its political content (either its use by protestors or the potential appropriation effect of the phrase for commercial purposes) but its ability to function as a source identifier. The phrase “does not function as a trademark or service mark to indicate the source of applicant’s goods and to identify and distinguish them from others” because the informational content of a slogan may overwhelm its ability to identify source. “The more commonly a term or slogan is used in everyday speech, the less likely the public will use it to identify only one source and the less likely the term or slogan will be recognized by purchasers as a trademark or service mark.”57 So, too, with I CAN’T BREATHE, which consumers identify with the message “that the wearer supports the ideas and messages conveyed by rallies and organizations dedicated to protesting violence.”58

The fate of I CAN’T BREATHE reflects standard practice at the PTO under the Trademark Manual of Examining Procedure.59Like the bar for disparaging terms, the refusal to register informational marks ignores secondary meaning.60 And, as with disparaging terms, the exclusion potentially captures marks that may also convey First Amendment-protected expression.61


One concern over In re Tam’s final outcomeis the effect on the rest of trademark law. If exacting First Amendment scrutiny applies and the disparagement bar is deemed unconstitutional, much that is taken for granted in trademark law may suddenly be called into doubt. What would be left?62 Defenders of the disparagement bar should be alert to the inverse issue. If the disparagement bar is constitutional under the First Amendment, what else might be? This question worries judges to whom the bar looks like viewpoint discrimination that is divorced from the ordinary operation of trademark law. Harmonizing the bar with the rest of trademark law might assuage that concern.

Some of the efforts to date are problematic because they justify too much. The argument that the bar fits trademark’s purposes by promoting commerce operates at too high a level of generality to provide comfort to the skeptical judge; conclusory appeals to commerce promotion could be used to justify any number of policies.

Examining the bar at a lower level of trademark generality makes clear that tying its fate to resolving a perceived conflict between interests of free expression and equality is incorrect. The disparagement bar is just one of many manifestations of the viewpoint-neutral principle that would-be trademarks should be effective source identifiers. The disparagement bar embodies this principle in a form that also serves important anti-discrimination aims, but it remains consistent with the larger body of trademark doctrine.

Michael Grynberg is Associate Professor of Law, DePaul University College of Law. He thanks Kelly Baldrate and Lisa Ramsey for their comments on earlier drafts. Comments welcome at

Preferred Citation: Michael Grynberg, A Trademark Defense of the Disparagement Bar, 126 Yale L.J. F.178 (2016),