Justices Agree on Constitutionality of Trademark Restriction, Disagree on How to Approach Such Questions

The underlying methodological debate might also bear on free speech disputes more broadly.

Today's decision in Vidal v. Elster unanimously upheld a federal trademark law provision that bars federal registration of any trademark that "[c]onsists of or comprises a name … identifying a particular living individual except by his written consent." The law doesn't bar people from using others' names as part of a trademark, but bars federal registration of such marks, which would provide trademark owners special protect against infringement. (The case involved plaintiff's attempt to register the mark "Trump Too Small," but the legal issue went far beyond this mark.)

The Court unanimously agreed that this was a content-based, though viewpoint-neutral, classification. But while content-based speech restrictions are presumptively unconstitutional, the Justices agreed that this law is fine.

But the Justices sharply disagreed on the rationale for this conclusion.

1. Justice Thomas, joined by Justices Alito and Gorsuch, concluded that the law is constitutional chiefly because of what they saw as a longstanding historical tradition of allowing such exclusions from trademark registration.
2. Justice Barrett, joined by Justice Kagan and in considerable part by Justices Sotomayor and Jackson, concluded that the law is constitutional chiefly by analogy to the First Amendment doctrine having to do with government-provided benefits for private speakers, under which reasonable, viewpoint-neutral restrictions on the benefits are constitutional; they were quite skeptical of Justice Thomas's historical focus. (Justice Sotomayor also wrote a similar opinion, joined by Justices Kagan and Jackson.)
3. And Justice Kavanaugh, joined by Chief Justice Roberts, tried as much as possible to abstain from this debate.

First, a bit of history that drew a majority, and that was joined by all the conservative Justices (including Justice Barrett):

Most importantly, trademark rights have always coexisted with the First Amendment, despite the fact that trademark protection necessarily requires content-based distinctions. Trademark rights "ha[ve] been long recognized by the common law and the chancery courts of England and of this country, and by the statutes of some of the States," and that protection continues today. As we all agree, this "[h]istory informs the understanding that content-based distinctions are an intrinsic feature of trademarks" [citing both the Barrett and Sotomayor concurrences]. And, for the duration of that history, the inherently content-based nature of trademark law has never been a cause for constitutional concern.

Our country has recognized trademark rights since the founding. At the outset, there were few recorded decisions, and the law developed slowly. Much of early American trademark law "was lifted essentially from that of England." The protection of trademarks under English law was an inherently content-based endeavor. For example, an early English law made it "lawful to and for every Trader, Dealer and Weaver of Linen Manufacture, to weave his Name, or fix some known Mark in any Piece of Linen Manufacture by him made." And, a person could be liable for fraud if he sold a product under another person's mark. So, the content of the mark (whether it was the same as another person's) triggered the restriction.

Although there was an early push for federal legislation to protect trademarks, no such law was enacted during our country's infancy. Instead, trademark law fell largely within "the province of the States" for the 18th and most of the 19th century. For example, Massachusetts passed a private bill incorporating a cotton corporation on the condition that it affix a label to its goods "with the seal of the said Corporation." The law also prevented others from "us[ing] a like seal or label" by subjecting them to treble damages. To be sure, for most of our first century, most commerce was local and most consumers therefore knew the source of the goods they purchased. …

The "first reported American decision that may be described as a trademark case" involved a dispute over the content of a mark—and in particular, the use of a person's name. In Thomson v. Winchester (1837), Samuel Thomson—who sold a medicine under the name "Thomsonian Medicines"—brought suit against another Massachusetts druggist who sold an allegedly inferior product under the same name. The court held that the druggist could be liable for fraud if he passed the medicine off as that of Thomson. Ibid.

In a similar vein, the first reported trademark case in federal court revolved around a trademark's content. Justice Story, sitting as Circuit Justice, granted an injunction to prohibit a seller of spools from infringing on the plaintiff's trademark of "Taylor's Persian Thread." Taylor v. Carpenter (D. Mass. 1844). Justice Story explained that, by using the trademark, the seller "imitated … both descriptions of spools and labels, red and black, of the plaintiffs," and that the principles prohibiting such infringement were at that time "very familiar to the profession" and not "susceptible of any judicial doubt."

Recorded trademark law began to take off in the last decades of the 19th century—after the ratification of the Fourteenth Amendment in 1868—and its established content-based nature continued. American commerce became more national in character, and, perhaps because of this shift, Congress enacted the first federal trademark law in 1870. Although States retained their important role, "Congress stepped in to provide a degree of national uniformity" for trademark protection.

This first law contained prohibitions on what could be protected as a trademark. For example, the law would not protect a trademark that contained "merely the name of a person … only, unaccompanied by a mark sufficient to distinguish it from the same name when used by other persons." It thus restricted a trademark based upon its content (i.e., whether it contained more than a name). As trademark disputes increased, courts continued to assess trademarks based on their content. [Details omitted. -EV]

Then came a historical analysis that was also endorsed by a majority, but this part not including Justice Barrett:

We have acknowledged that trademark rights and restrictions can "play well with the First Amendment." In this case, we do not delineate an exhaustive framework for when a content-based trademark restriction passes muster under the First Amendment. But, in evaluating a solely content-based trademark restriction, we can consider its history and tradition, as we have done before when considering the scope of the First Amendment.

The Lanham Act's names clause has deep roots in our legal tradition. Our courts have long recognized that trademarks containing names may be restricted. And, these name restrictions served established principles. This history and tradition is sufficient to conclude that the names clause—a content-based, but viewpoint-neutral, trademark restriction—is compatible with the First Amendment. We need look no further in this case….

Restrictions on trademarking names have a long history.Such restrictions have historically been grounded in the notion that a person has ownership over his own name, and that he may not be excluded from using that name by another's trademark. As the Court has explained, "[a] man's name is his own property, and he has the same right to its use and enjoyment as he has to that of any other species of property." Brown Chemical Co. v. Meyer (1891). It is therefore "an elementary principle that every man is entitled to the use of his own name in his own business." F. Treadway, Personal Trade-Names, 6 Yale L. J. 141 (1897) (Treadway); see also A. Greeley, Foreign Patent and Trademark Laws (1899) ("The right of any one to place his own name on goods sold by him is recognized as a natural right and cannot be interfered with"). [More historical details omitted. -EV] …

We conclude that a tradition of restricting the trademarking of names has coexisted with the First Amendment, and the names clause fits within that tradition. Though the particulars of the doctrine have shifted over time, the consistent through line is that a person generally had a claim only to his own name. The names clause reflects this common-law tradition by prohibiting a person from obtaining a trademark of another living person's name without consent, thereby protecting the other's reputation and goodwill.

None of this is to say that the Government cannot innovate when it comes to trademark law. A firm grounding in traditional trademark law is sufficient to justify the content-based trademark restriction before us, but we do not opine on what may be required or sufficient in other cases….

And then Justice Thomas, joined by Justice Alito and Gorsuch, had a separate nonmajority section that rejected Justices Barrett's and Sotomayor's analyses (analyses that, as noted above, were generally joined by Justice Kagan and Jackson):

Our colleagues would address the names clause with two analogies. Neither is compelling in this case. Under both analogies, the test would boil down to what a judge believes is "reasonable in light of the purpose" of trademark law. But, no matter the approach taken, we all agree that the names clause does not violate the First Amendment.

Justice Sotomayor would pull "strands of precedent" together to conclude that heightened scrutiny does not apply to trademark registration because it is a Government initiative or benefit. This conclusion rests primarily upon cases in which the Government provides a cash subsidy or conditions the use of a public payroll to collect union dues. But, those cases "occupy a special area of First Amendment case law, and they are far removed from the registration of trademarks." The Government benefit cases are an ill fit for the names clause, and we would not graft this precedent, which Justice Sotomayor acknowledges is not controlling, onto this trademark dispute.

Justice Barrett, echoed by Justice Sotomayor, would import the test that we have used for a "limited public forum." Our precedents hold that the Government "may create a forum that is limited to use by certain groups or dedicated solely to the discussion of certain subjects." Justice Barrett provides little explanation for why that approach makes sense in the trademark context—she simply declares that the limited public forum framework "is apt" due to the content-based nature of trademark law. Although she attempts to cabin the analogy to the content-based nature, the limited public forum test is quite obviously about creating a forum. And, there is reason to doubt that the federal trademark register is analogous to a limited public forum. To start, unlike a speaker in a limited public forum, a markholder does not communicate with customers on the register. Rather, as the Government acknowledges, the register "is a way of warning potential infringers that they risk liability if they use the same or confusingly similar marks." The Government has also previously asserted that it did not create a forum for speech by providing for the federal registration of trademarks. Without an analogous forum, it is hard to see why the test for a limited public forum should apply. We see no need to adopt a potentially fraught analogy to resolve the names clause's constitutionality.

Despite the differences in methodology, both Justice Sotomayor and Justice Barrett reach the same conclusion that the names clause does not violate the First Amendment. On the bottom line, there is no dispute. Rather than adopt a reasonableness test premised upon loose analogies, however, we conclude that the names clause is grounded in a historical tradition sufficient to demonstrate that it does not run afoul of the First Amendment.

Justice Barrett's concurrence began:

While I agree with the Court that the names clause does not violate the First Amendment, I disagree with some of its reasoning. The Court claims that "history and tradition" settle the constitutionality of the names clause, rendering it unnecessary to adopt a standard for gauging whether a content-based trademark registration restriction abridges the right to free speech. That is wrong twice over. First, the Court's evidence, consisting of loosely related cases from the late-19th and early-20th centuries, does not establish a historical analogue for the names clause. Second, the Court never explains why hunting for historical forebears on a restriction-by-restriction basis is the right way to analyze the constitutional question.

I would adopt a standard, grounded in both trademark law and First Amendment precedent, that reflects the relationship between content-based trademark registration restrictions and free speech. In my view, such restrictions, whether new or old, are permissible so long as they are reasonable in light of the trademark system's purpose of facilitating source identification.

She went on to lay out the often content-based history of trademark law, and concluded:

These content-based trademark rules have long coexisted with the Free Speech Clause, and their function is generally compatible with it. Courts have applied content-based rules not to "suppres[s] … ideas," but simply to serve trademark law's purposes. Indeed, these trademark restrictions can actually help prevent "interfere[nce] with the marketplace of ideas," insofar as they ensure that a single producer cannot exclusively appropriate words or phrases in the general domain. This is not to say that the Government could not abuse content-based trademark registration restrictions—as I explain below, such restrictions are not insulated from scrutiny. But they do not set off alarm bells signaling the likelihood that "'official suppression of ideas is afoot.'" Therefore, I agree with the Court that we need not treat content-based trademark registration restrictions as presumptively unconstitutional.

And she drew the analogy to the "limited public forum" cases (in part of an opinion that, like the preceding parts, were joined by Justices Kagan, Sotomayor, and Jackson):

The Solicitor General suggests that we draw an analogy to another area that is inherently content based: the limited public forum. When the government opens its property to speech for a particular purpose, creating a limited public forum, it necessarily must "reserv[e the property] for certain groups or for the discussion of certain topics." Content-based restrictions are "inherent and inescapable" in maintaining a forum for speech "compatible with the intended purpose of the property." It is therefore inappropriate to view these restrictions as presumptively unconstitutional and apply strict scrutiny. Still, not every restriction is permissible. To evaluate these rules, we ask "whether they are reasonable in light of the purpose which the forum at issue serves." This ensures that the government "respect[s] the lawful boundaries it has itself set."

Though I would not shoehorn the trademark registration system into the definition of a "limited public forum," the Solicitor General's analogy is apt Content discrimination is "[i]mplicit in the concept of" the trademark registration system much like it is in a limited public forum. {Justice Thomas mistakenly suggests that I present the federal trademark register as a limited public forum. That is not my position. Rather, I view the content-based nature of the limited public forum as analogous to the trademark registration system.} …

Justice Barrett then went on, in part of her opinion joined just by Justices Kagan and Sotomayor:

I cannot agree with the Court that the existence of a "common-law tradition" and a "historical analogue" is sufficient to resolve this case. Even if the Court's evidence were rock solid [and Justice Barrett argued that it was actually quite weak -EV], I still would not adopt this approach. To be sure, tradition has a legitimate role to play in constitutional adjudication. For instance, the longstanding practice of the political branches can reinforce our understanding of the Constitution's original meaning. The views of preceding generations can persuade, and, in the realm of stare decisis, even bind. But tradition is not an end in itself—and I fear that the Court uses it that way here.

The Court does not (and could not) argue that the late-19th and early-20th century names-restriction tradition serves as evidence of the original meaning of the Free Speech Clause. Nor does it treat the history it recites as a persuasive data point. Instead, it presents tradition itself as the constitutional argument; the late-19th and early-20th century evidence is dispositive of the First Amendment issue. Yet what is the theoretical justification for using tradition that way?

Relying exclusively on history and tradition may seem like a way of avoiding judge-made tests. But a rule rendering tradition dispositive is itself a judge-made test. And I do not see a good reason to resolve this case using that approach rather than by adopting a generally applicable principle. (After all, there is a tradition of the latter approach too. See, e.g., McCulloch v. Maryland (1819) (adopting standard for application of the Necessary and Proper Clause).) In the course of applying broadly worded text like the Free Speech Clause, courts must inevitably articulate principles to resolve individual cases. I do not think we can or should avoid doing so here. As I explained [above], the takeaway from history is that content-based trademark restrictions have long been central to trademark's purpose of facilitating source identification, and they have not posed a serious risk of censorship. This principle offers a generally applicable way to think about whether registration restrictions "'play well with the First Amendment.'"We should bring clarity to the law by adopting it.

In my view, the Court's laser-like focus on the history of this single restriction misses the forest for the trees. It gives secondary billing to what I think is the central point: that the names clause "reflects trademark law's historical rationale of identifying the source of goods." I see no reason to proceed based on pedigree rather than principle. Besides, as the Court admits, its approach merely delays the inevitable: Eventually, the Court will encounter a restriction without a historical analogue and be forced to articulate a test for analyzing it.

Justice Sotomayor offered a similar argument, joined by Justices Kagan and Jackson; a brief excerpt:

This case involves a free-speech challenge to a viewpoint-neutral, content-based condition on trademark registration. In deciding how to evaluate this kind of challenge, the Court faces two options: Either look only to the history and tradition of the condition, or look to trademark law and settled First Amendment precedent. The first option, which asks whether the history of a particular trademark registration bar plays well with the First Amendment, leads this Court into uncharted territory that neither party requests. The other guides it through well-trodden terrain. I would follow the well-trodden path.

In assessing the constitutionality of the names clause and other trademark registration provisions, I would rely on this Court's tried-and-tested First Amendment precedent. This Court has held in a variety of contexts that withholding benefits for content-based, viewpoint-neutral reasons does not violate the Free Speech Clause when the applied criteria are reasonable and the scheme is necessarily content based. That is the situation here. Content discrimination is an inescapable feature of the trademark system, and federal trademark registration only confers additional benefits on trademark holders. The denial of trademark registration is therefore consistent with the First Amendment if it turns on "reasonable, viewpoint-neutral content regulations." Because the names clause satisfies that test, I would uphold the constitutionality of the provision on that ground alone.

These are just excerpts; there's a lot more to the opinions, and people who are interested in First Amendment doctrine and theory should find them very interesting. But I hope these excerpts give a flavor of the debate.