Rebecca Tushnet is Professor of Law at Georgetown Law. This analysis largely tracks that of the intellectual property professors’ amicus brief she authored with Christine Haight Farley, which in turn draws heavily on Tushnet’s forthcoming article in the Notre Dame Law Review, “The First Amendment Walks into a Bar: Trademark Registration and Free Speech.”
In Lee v. Tam, the court will review an en banc decision by the U.S. Court of Appeals for the Federal Circuit holding that the government’s denial of a trademark registration to a disparaging name violated the First Amendment. The Federal Circuit’s ruling has a surface plausibility, but it is profoundly inconsistent with the regulatory functioning of trademark law, whose entire existence involves allocating speech rights. A trademark, and a trademark registration, both function to strengthen a claimant’s right to suppress the speech of others. There is no neutral, non-speech-suppressing choice when it comes to trademark registration, and the Federal Circuit’s failure to recognize this fact distorted its analysis in fundamental ways.
The Supreme Court has told us that content-based bans on speech are subject to strict scrutiny, a heightened level of review that is fatal to many government regulations – at least outside rules that regulate commercial speech generally. Lee v. Tam offers the court a chance to consider how seriously it takes this principle, especially as applied to a regulation that is not a ban. If content-based regulations must always survive strict scrutiny, most of trademark law is likely unconstitutional: Granting a trademark registration inherently requires content-based determinations about what a word means in the context of the goods and services it is used for. And the trademark registration system has a number of features that take it far from the common law of fraud (which would surely survive strict scrutiny). Among other things, its “intent to use” provisions grant rights in terms that have not been used and thus whose use by others could not possibly cause confusion; its incontestability provisions grant rights in terms that are merely descriptive and lack source-signifying function – again, when use by others couldn’t cause confusion – as long as no one challenges those rights for five years; and many of its other exclusions would fall victim to strict scrutiny.
Trademark law is pervasively content-based
In addition to the bar on disparaging marks, Section 2 of the Lanham Act prohibits the registration of any mark that is scandalous; immoral; deceptive; falsely suggestive of a connection with persons, institutions or religions; likely to cause confusion with an existing mark; descriptive; misdescriptive; functional; a geographic indication for wine or spirits other than the place of origin of the goods; government insignia; a living person’s name, portrait or signature without written consent; or a surname. To determine whether the mark itself, or matter contained within the mark, corresponds to any of the bars in Section 2 (with the exception of the bar on utilitarian functionality), the U.S. Patent and Trademark Office must evaluate the content of the mark. For instance, if an applicant applies for a registration of the word mark BRAMLEY, the PTO must determine if that word would be understood by a substantial number of consumers as the name of a living person, a place or a surname. That is an examination of content.
The bars to registration serve policies that go well beyond protecting consumers from deception in the marketplace. A number of the bars do deal with confusing or deceptive marks. Some bars, and the prohibition on registering generic symbols, relate to another central concern of trademark law: competition. Significantly, however, a number of the bars apply to marks that are neither confusing/deceptive nor anticompetitive on their face. These include the bars on disparaging, scandalous and immoral marks, but also the bars on inaccurate but not necessarily misleading geographical indications of origin for wine or spirits, names of specific living persons absent written consent, and government insignia.
It is true that these bars, including the disparagement bar, target categories of symbols that are less likely than other symbols to serve the trademark function because they carry additional meanings that can overwhelm or detract from their potential source-signifying function, as Michael Grynberg has argued. Nonetheless, they are not centrally about deception. For example, the bar on registering symbols that merely truthfully describe the applicant’s goods and services is obviously content-based, and it is also clearly not about deception. The descriptiveness bar, like the others, is connected to the broader policy goals of trademark law, demonstrating that registration serves government interests beyond protecting consumers from deception.
The disparagement bar is not viewpoint-based
The Federal Circuit treated disparagement as distinguishable from the other Section 2 bars because it found disparagement to be not only content-based but also viewpoint-based. Given the meaning of “disparagement” as a concept that could apply to anyone, however, the ban on disparaging marks is more appropriately characterized as content-based than as viewpoint-based.
The Federal Circuit en banc majority claimed that the “legal significance of viewpoint discrimination is the same whether the government disapproves of the message or claims that some part of the populace will disapprove of the message.” However, the “part of the populace” whose views are considered changes depending on the specifics of the applied-for mark, and therefore no particular group or set of viewpoints receives protection denied to everyone else. Like the law of defamation, which is content-based but not viewpoint-based, the disparagement bar protects everyone.
Specifically, the disparagement exclusion denies registrations equally to people trying to disparage a group and to people trying to reclaim terms on behalf of the disparaged group, as long as a substantial composite of the referenced group would see the term as disparaging. Thus, the disparagement bar is also neutral as to the applicant’s membership in and viewpoint on the particular disparaged group at issue. So, for example, if an advocacy group that sought to bar Asian immigration applied to register CELEBRASIANS for its services, the mark would not be disparaging even if the group’s intention were disparaging.
Denial of registration is not equivalent to punishment for speech
As long as everyone is allowed to use a term in non-trademark ways, trademark law does not regulate or suppress either registrants’ or nonregistrants’ uses in anything other than the trademark sense. Under Section 2(a), any speaker can continue to use any symbol it wants for communicating anything it wants, but it can only get a registration for a mark that does not run afoul of a statutory exclusion. This distinction – which allows the government to regulate the mark’s registration, but not to exercise any leverage over the speaker’s other speech – is consistent with the leading unconstitutional conditions cases. In 1991, in Rust v. Sullivan, the court explained that unconstitutional conditions generally involve “situations in which the Government has placed a condition on the recipient of the subsidy rather than on a particular program or service, thus effectively prohibiting the recipient from engaging in the protected conduct outside the scope of the federally funded program.” More recently, in Agency for International Development v. Alliance for Open Society Internal, Inc., in 2013, the court repeated its distinction between conditions that specify what activities Congress wants to subsidize (legitimate) and those that seek to leverage the subsidy to affect speech outside the government program (illegitimate).
Denying Simon Tam rights in any mark as long as his band kept the name THE SLANTS would be an unconstitutional penalty for the exercise of his First Amendment rights. Denying him the right to register any mark he wants is not. In Walker v. Texas Division, Sons of Confederate Veterans, the court held that a speaker is free to put a Confederate flag sticker on the bumper of his car. He just cannot have it on his license plate. Under Section 2(a), likewise, the band members can keep their band’s name, use it as they like, and receive the benefits of registration by relying on a non-disparaging logo or other symbol. They just cannot get the government’s help by invoking only this particular disparaging symbol. The government does not gain any leverage over the message that appears outside the registered matter, any more than it secures any leverage over the message appearing on a bumper sticker because it has the power to approve or disapprove license plates. The government’s unwillingness to take part in the disparaging or racist aspects of the message communicated by a mark can therefore itself be a sufficient interest to justify the bar.
Ultimately, the government must weigh in on the validity of a speaker’s choice of marks if the mark is to be enforced against others. Because the entire system of trademark registration is a system of speech regulation, this case differs drastically from the unconstitutional conditions cases finding that the government overstepped its bounds. The government’s extensive involvement in the definition of and enforcement of trademark rights means that there is no neutral position. Allowing the government to avoid the risk that it could be seen as endorsing a disparaging mark should be held to be constitutional.