Rebecca Tushnet: And now for something different! I’ve been asked to speak about the implications of the Pro-Football case and decisions made by the courts with regards to the trademark process and freedom of speech. Thanks to Fred Schauer and Mark Tushnet for illuminating discussions, and to William & Mary students for the opportunity to talk about these important issues. I’ve been trying to work through them and I caution that my conclusions are necessarily tentative at this point.
All or almost all of the §2 bars are also reasons that the courts will refuse to protect a claimed trademark under §43 of the Lanham Act, which provides trademark protection for unregistered marks. These reasons include: the claimed matter is functional, meaning it affects the cost or quality of the product or service; the claimed matter is merely descriptive, meaning that consumers don’t understand that it indicates source and instead think that it just describes some characteristic of the product; the claimed matter is deceptively misdescriptive, which is like descriptiveness except not true; the claimed matter is deceptive; the claimed matter is confusingly similar to an existing registered mark or mark in use in the US; the claimed matter is the name, photo or signature of a living person and there’s no written consent from that person to register the mark; the claimed matter is the flag or insignia of a nation or state or other US political subdivision; the claimed matter creates a false association with people or institutions; and last but not least, the claimed matter is scandalous, immoral, or disparaging.
The constitutionality of § 2(a) is now in question because of some high-profile cases. From the First Amendment side, we have Reed v. Town of Gilbert: last Term’s sign case in which the Supreme Court struck down a city’s sign regulations on the theory that they were content-based and didn’t survive strict scrutiny. The rule suggested, though it’s not clear whether it will actually apply this broadly, is that all state regulations of speech that depend on an evaluation of the content of the speech to determine whether the regulation has been violated are content-based and must survive strict scrutiny. Section 2(a) is nothing but content-based (with a possible exception to be discussed in a bit).
Meanwhile, we have the cases Professor Farley discussed, so the collision with trademark law seems imminent. It’s possible to argue that trademark is a First Amendment-free zone, as copyright basically is except for the internal restraints of the fact/expression division and fair use, but that argument seems unlikely to prevail as easily with trademark and trademark registration, neither of which are mentioned in the Constitution and the latter of which confers rights unknown at common law.
A brief review of the justifications usually given for the exclusions for marks that are scandalous or disparaging: (1) the harm done by the government endorsement represented by a registration; (2) the desire to withhold government resources from disparaging or scandalous terms; (3) the lack of any effect on a user’s ability or right to use the mark, with (a) possible §43(a) or state common law protection against confusing uses despite unregistrability, though this is not at all certain, or (b) even if there’s no protection against confusing uses by other people, there’s still nothing stopping anyone from using a disparaging term on a T-shirt (at least if they’re not going to a public school) and thus there is no government harm inflicted on them.
Justification (1) very much depends on public understanding. Confederate Veterans license plate case from last Term allowed a state government to decline to approve a Confederate Veterans logo for personalized license plates even though it approved a large number of other specialized license plates. The Court ruled that the plate design constituted governmental speech, and thus the government was not bound to neutrality in its decisions.
If registration is the government’s agreement that a term is appropriate for federal registration, equivalent to the government’s agreement that a term is appropriate for placement on a license plate, then there is no problem with refusing the registration. This is the reasoning recently adopted by the district judge deciding the Washington football team’s case. But do people actually understand the difference between the ® symbol, the TM symbol, and nothing? The Supreme Court didn’t exactly look for empirical evidence in the Confederate Veterans case! The dissent argued that, given how many different license plates are approved, no one could reasonably perceive a government endorsement; that’s even more true for trademark registrations—the government couldn’t possibly be both an Apple Mac and a Windows PC, or a UVa alum and a William & Mary alum. But that’s oversimplifying—the government could express an opinion about what the boundaries of tolerable public discourse are; allowing a particular plate, or a particular mark, expresses only the opinion that it’s tolerable, without saying that it’s correct.
There’s a related body of law of some possible relevance: the establishment clause cases, in which Justice O’Connor developed a line of reasoning that the question of whether government endorsed religion with some action had to be assessed from an objective perspective, assuming a person with reasonable levels of historical knowledge/understanding. Anyone to whom the difference between registration and trademark protection is explained can usually grasp it very easily. Given that fact, by analogy §2(a) is if anything an easier case than Confederate Veterans.
Justification (2) doesn’t work, because it takes a lot of resources to enforce §2(a). Also, if §43(a) remains available, then government resources are still being devoted to protecting such marks, perhaps even more resources, insofar as there is no more presumption of validity under §43 as there is for a registered mark and therefore an inquiry into protectability must be made in each case.
Justification (3) is the most interesting to me. (3)(a) is unsatisfying to anyone. There are significant differences between the benefits of registered and unregistered marks—nationwide scope, presumption of validity/incontestability after 5 years, the ability to use a US registration to secure rights under foreign registration systems, the ability to invoke counterfeiting laws. If that harm is acceptable, then it should also be acceptable to impose the further harm of not providing any protection against confusing uses at all. So, to me, this reduces to a question of unconstitutional conditions: does the government violate the First Amendment by refusing to protect certain marks because of their content?
And asked that way, it’s clear this challenge opens up numerous cans of worms. All of the exclusions in §2 are content-based. Let’s assume they only operate on commercial speech (though this is not necessarily true). Full understanding now requires a brief excursion into commercial speech doctrine, the centerpiece of which is the Central Hudson rule. Under Central Hudson, false or misleading commercial speech may be banned. Regulations of truthful, nonmisleading commercial speech, however, must actually serve a substantial government interest and be no more restrictive than necessary to serve that interest, though this is not a requirement to use the least restrictive alternative.
Assuming that §2 only operates on commercial speech, then, we still face a division between false/misleading speech and other commercial speech. Since false and misleading commercial speech can just be banned, there’s no doubt that the government can refuse to support it. Bans on other commercial speech must survive Central Hudson. But wait! There’s also mandatory disclosure law. If the government is not suppressing commercial speech, but rather requiring disclosures, the Zauderer line of cases becomes relevant. Zauderer involved attorney advertising that told laypeople that if the lawyer lost their case, there’d be no fee. The Supreme Court held that the state could require an additional disclosure clarifying that fees and costs were different things and that, if they lost, they’d still be liable for costs. The Court did not require the mandatory disclosure to satisfy Central Hudson, but applied something like rational basis review—as long as the required disclosure was truthful and noncontroversial, and not unduly burdensome, it was constitutional.
A refusal to provide trademark protection is arguably much more like requiring disclosures—it may create an economic burden and thus a possible disincentive for the speech, but it doesn’t suppress that speech directly—than like a direct ban. However, lower courts are not in agreement about whether Zauderer only applies when the government is requiring disclosures to correct misinformation, versus when it is requiring disclosures to provide useful information.
Take the broad view of Zauderer—such burdens on commercial speech are allowed when the government is requiring disclosure of truthful, noncontroversial information that isn’t unduly burdensome, even though it is targeting particular content for the required disclosure. Do the 2(a) bars look anything like that?
(1) Is the government determination about being scandalous/disparaging truthful and noncontroversial? A weak point—this may not be an answerable question. There is probably some reasonably objective truth about whether a substantial proportion of the general public, or the targeted group, finds the mark outrageous or disparaging, respectively. If we are confident in the PTO’s factfinding ability, then at least some scandalousness and disparagement determinations—say, the c-word and the n-word—are probably truthful and noncontroversial. Since failure to protect isn’t itself a disclosure, it might not need to be noncontroversial—the Court’s concern seems to have been for forcing commercial speakers to put out ideological messages they didn’t agree with. That concern would be completely irrelevant for a refusal to register. But what it means for a determination to be noncontroversial is, of course, itself highly contestable. The very fact of a litigated challenge might seem to suggest that a disclosure—or a bar to registration—is controversial; the Supreme Court has not been very helpful in identifying what it meant by that term.
(2) Even if we find a disparagement determination to be truthful and noncontroversial, is it unduly burdensome to deny TM registration or, in its entirety, TM protection? I don’t think this Q can be answered except normatively—this isn’t like a disclosure that takes three minutes to recite when most ads are only 30 seconds long. (3) Suppose we decide that we do need to get rid of scandalousness/disparagement registration bars because they don’t meet a Zauderer-like standard. What of the remaining exclusions? This question can be fruitfully grouped with the next issue:
Take the narrow view of Zauderer—such burdens are only allowed when there’s some relationship to preventing deception. Once we thereby get rid of scandalousness/disparagement, there are other §2 bars whose constitutionality remains to be addressed.
(1) Which are at least arguably related to consumer deception? Well, deceptive marks, pretty obviously—though the fact that refusing registration doesn’t stop the use of the mark to deceive people might be a problem of insufficient tailoring. Marks that create a false association with persons, institutions, etc. Geographically deceptively misdescriptive marks, ok. Name, picture, or signature of a living person (or president whose widow or widower still lives) without written consent, seems ok as a legislative prediction at the wholesale level about the risk of false endorsement. Even the bar on registering flags and other indicia of states, nations, and other political units arguably has some relationship to preventing deception in that this is a subclass of signifiers whose potential for abuse is so great that Congress could legitimately make a blanket rule without requiring case by case evaluation.
(2) What about those §2 bars related to consumer understanding, but not so much to deception? For example, merely descriptive (or deceptively misdescriptive, or geographically misdescriptive) marks are not deceptive when a TM claimant uses them. But there are policy reasons we want other producers to be free to use them as well. Claimants could cause harm to the marketplace by gaining monopoly rights over them. Is that enough of a justification to sustain these bars, given that they are content discriminatory by our excitingly broad definition of content discrimination? To me, the fact that the acid logic of current First Amendment protection makes the bar on registering merely descriptive terms seem even arguably questionable shows that something has gone seriously wrong with the Supreme Court’s condemnation of content-based regulation, which has now become detached from the underlying justifications that one might reasonably give for worrying about content-based regulation.
Sure, we could say that the bar on registering descriptive terms survives strict scrutiny, but I don’t see how that can be true without doing serious damage to the concept of strict scrutiny. For example, we already have a less restrictive alternative than barring registration in the very same statute: we could allow registration of descriptive marks, and give everyone else a descriptive fair use defense, just as we do now when a descriptive term has secondary meaning and is registrable. Of course, having the descriptive fair use defense is not as good for competitors as full freedom to use a descriptive term lacking secondary meaning—but is that marginal difference enough to survive full First Amendment scrutiny? At a minimum, the difficulties with this approach indicate that, if Zauderer-like scrutiny should apply to refusals to register, it should be the broad version of Zauderer, allowing the government to pursue non-deception-related objectives, though which such objectives are independently constitutional can still be debated.
Side note: another constitutional jurisprudence could use the fact that a trademark confers some private right to suppress other people’s speech as a reason that the government can make balancing decisions about whom to favor—that is, whatever rule the government makes will affect the speech environment, so saying that each rule has to survive strict scrutiny doesn’t make much sense. Unfortunately, that’s a tough argument to make in our current constitutional order, which tends to assume the existence of particular baselines, such as that the government will protect trademarks.
(3) What about functionality as a bar to registration/protectability? This is a fascinating issue because it is a pure U.S. v. O’Brien situation: by definition, a symbol that is functioning as a mark, but also possesses utilitarian or aesthetic functionality, is not barred from registration/protection because of anything about its communicative status. It is barred because of the things that it does that aren’t communicative at all, just as the government supposedly punished O’Brien for burning his draft card not because of his anti-war message but because it needed that draft card for other purposes. Unlike O’Brien itself, there’s absolutely no reason to disbelieve the government’s content-indifferent rationale for its regulation. With functionality, the government here is regulating conduct (preserving everyone else’s freedom to use a functional feature) and not speech. Nonetheless the government rule has effects on speech—it discourages the communicative use of functional features, because you can’t get protection for them. Under O’Brien, a law that burdens expressive conduct is unconstitutional, even absent any bad state purposes, unless “it furthers an important or substantial governmental interest; and the incidental restriction on alleged First Amendment freedoms is no greater than is essential to the furtherance of that interest.”
In some sense the exclusion for functional matter is motivated by the government’s desire to shape the universe of communicative devices so that it doesn’t overlap with the universe of functional features. The thing is, there’s just nothing wrong with that desire even if characterized as speech-related—this is why Justice Breyer and the Justices in agreement with him are absolutely right that the term “content-based” as recently used in Reed is ridiculously, preposterously, unsustainably overbroad. The government interest is substantial and the restriction is exactly coterminous with the government interest. So let’s conclude that at least functionality would survive Central Hudson scrutiny both as a bar to registration and as a total bar to protection.
Go back then to the category of descriptive terms lacking secondary meaning. Can government deny protection to those? Denying trademark protection to such terms doesn’t directly protect consumers from deception, though it may indirectly prevent them from being fooled by the absence of information from other producers who also want to use those descriptive terms.
So, may the government condition the benefit of trademark protection on a claimant’s speech-related choices, here what kind of word to use? One answer: yes, if the condition is related to the purpose of the program. Here, one core purpose of trademark is to connect nonfunctional terms to particular producers so consumers can use them as indicators of source. A descriptive term without secondary meaning doesn’t do that, so a producer who uses such a term can be excluded from the government program.
But is the only permissible purpose of the program to improve the quality of truthful or nondeceptive signals that consumers receive? Can our concept of “quality” consider the harm done to significant segments of the population by having disparaging or scandalous terms commercialized?
I have to admit that unconstitutional conditions doctrine is a big hairball. There are huge problems of determining what counts as coercion, or as the appropriate baseline from which to measure your entitlement. Kathleen M. Sullivan, Unconstitutional Conditions, 102 Harv. L. Rev. 1413 (1989): “[D]eciding under what circumstances government offers, like private offers, are coercive is an inevitably normative inquiry.”
Nonetheless I want to argue that the §2(a) exclusions, all of them, do not represent unconstitutional conditions. And the reason is something peculiar to trademarks: trademarks can communicate (at least) two different things—but the only thing the government protects as a trademark is the communication “this symbol stands for a particular producer.” Under §2(a), any speaker can continue to use any symbol it wants, but it can only get a registration for matter that does not run afoul of any exclusion—for the Redskins, the team colors and logo.
Friedman v. Rogers, 440 U.S. 1, 12 (1979), says that trademarks have “no intrinsic meaning” and thus that there’s no First Amendment problem with the government barring certain marks for fear they might be deceptive. From a trademark lawyer’s perspective this might be a puzzling or even contradictory statement, but one way to read what the Court is saying is that the source indication function of trademark is not the communication that is protected in commercial speech. As source indicator, a mark is neither truthful nor untruthful in advance of legally enforceable meaning, and—unlike ordinary communication, where changing the words can change the meaning, as the “Fuck the Draft” case Cohen v. California teaches us—a change of mark does not change what is communicated by the trademark function: this is who I am/this is my source. Google is Alphabet.
Unlike the clinics in Rust v. Sullivan, a case in which the Court found that it wasn’t an unconstitutional condition to require any health clinic that took federal money to separate abortion related information and activities from the part of the clinic funded by the government, a speaker doesn’t even h ave to separate its activities to take advantage of the TM registration system.
[slide] To analogize to the Confederate Veterans case, as with the license plate, the speaker’s free to put a confederate flag bumper sticker on his bumper; he just can’t have it on the license plate. Washington’s football team can keep its name and continue to get the government’s help suppressing counterfeits etc. It just can’t get the government’s help using only this particular disparaging symbol. If the whole car is our unit of analysis, like the whole product or service offered by the producer, then the government’s unwillingness to implicate itself in the disparaging or racist aspects of the message communicated by the car or the product may be itself a sufficient interest to justify the exclusion, given that the government is not gaining any leverage over the message appearing outside the license plate or registered matter.
A number of cases and commentators have suggested germaneness to the program’s aims as a standard for determining whether a condition is unconstitutional: you can be denied TANF because you are making too much money, but not because you protested the government—germaneness here, though, reduces to the question of whether it’s ok for the government to have this preference for nondisparaging marks. I think there’s no question that denying Washington’s football team any trademark rights in any symbol as long as it also used the team name would be a penalty for the exercise of First Amendment rights.
Moreover, viewpoint discrimination can be illegitimate in many more instances than content discrimination—my theory doesn’t license Congress to grant trademarks only to Republicans. But: Would my reasoning license Congress to eliminate any trademarks that were themselves pro-choice? Viewpoint discrimination can violate the First Amendment even if the underlying activity isn’t protected by the First Amendment—RAV v. City of St. Paul teaches that you can’t decide that assaults against Republicans because they are Republicans will be subject to harsher punishment, or that only false advertising that degrades men will be banned.
Conclusion: Once a program is exposed to the modern First Amendment in its rawest form, that program may have difficulty surviving, no matter how sensible it is. This is either a problem with First Amendment doctrine—or it’s a problem with the modern rejection of Lochner. If we are really that suspicious of government regulation picking winners and losers, then we shouldn’t even have registration—and maybe not infringement cases! Or, my preferred response, we should rethink the corrosiveness of First Amendment doctrine applied to ordinary economic regulations. In some ways, whether §2(a)’s ban on disparaging marks survives constitutional scrutiny is only a small part of that broader issue. There remains the possibility of continued ad hoc-ism, where the Court applies its strict scrutiny only to the parts of a system that protect less politically powerful groups, as is its trend with campaign finance and affirmative action. But that would be the worst result of all.
Brooke: Is TM property? © and patent are in the constitution. It’s not really property, but represents goodwill. You can pledge TM registrations to a bank and pledge them as security for a loan; Ford pledged all its TM registrations to financiers to avoid bankruptcy. Ambiguity as to whether it’s property. The gov’t is picking winners and losers. Doesn’t think Court will hold 2(a) unconstitutional, but we need a working system.
Allen: Personally, agrees that registration is an affirmation/presumption of validity that there is some weight behind the mark. Not just something that the applicant takes home but can use abroad.
Heymann: is there any approval of the message in a registration?
Allen: Yes, in the sense that it affirms that it survived §2.
Farley: That’s the question of the day! Going to other gov’ts saying that this gov’t has approved the mark. Most countries have this kind of prohibition, usually phrased as public order or morality. It is the US saying “this message is ok”—whether it’s “using LEATHER on these products is nondeceptive” or something else. Fed. Cir. was interested in whether CO could bar protection of pornography under this rationale.
RT: Depends on your theory of the 1A. I think the answer could be yes, but there is nonetheless a big distinction between © and TM. Because TM is about protecting a source link, the Washington team can have the unregistrable word present on everything that is protected by its rights in the team colors and logo. By contrast, a work refused © because it was pornographic would have to change its content to gain protection. So you could easily invalidate such restrictions on © while still sustaining §2.
Heymann: questions of evidence. If there are a bunch of actually disparaging marks on the register because the targets don’t have the resources to challenge, is that a problem?
Brooke: we’re supposed to care about consumers/protecting consumers, not if Ford is trying to protect its property. Many of these marks will fail/be taken care of by the market. People can choose to be offended or not.
Farley: the market won’t fix everything; if we trust the market on that, why not trust it on everything else the TM act is concerned with? If the issue is connecting the gov’t w/the ® of a disparaging term today, then the rule should be that we assess its meaning today. [That creates some real problems for my constitutional analysis, since that would mean there are untrademarkable goods & services, like the Amos & Andy works, or OJ Simpson: any time the public starts to associate the mark with the underlying offensive goods/services, then that mark would be disparaging/scandalous too.]
Problem with evidence: PTO relied on blog saying Slants were rejected for their name, but Slants say that’s not true; blog was mistaken in its facts. The limited evidence available to the PTO creates some risks.
Brooke: Market pressure often takes care of this: Sambo’s restaurants have been abandoned. If Dan Snyder had been smart, this would have been taken care of a long time ago.
Q: is original justification just to prevent bad marks from being in the marketplace? Scandalousness/disparagement doesn’t seem like consumer protection.
A: But that’s not the only justification for TM: ordering producer relations. Incontestability has nothing to do w/consumer protection, for example. Also, historically, disparagement was first before false association—it may well be unconstitutional the way old defamation law was, but it’s deeply embedded in the existing scheme.
Farley: Flag prohibition could be back-justified as deception based, but legislative history is about diplomacy; likewise presidential name prohibition is dignity-based in stated historical justification.
Brooke: would never say that Lanham Act is other than consumer protection in federal court, as a practicing attorney. If you’re litigating a case, you better be talking about consumers.
Heymann: prediction about In re Tam?
Farley: court was arguing with itself, passionately. The fact they’d have to overturn precedent, plus questions about what exactly they’d be invalidating—disparagement, false association, scandalous, descriptive—is something they’d have a hard time wrestling with. Fed. Cir. doesn’t want to undo all of TM law, which might be enough to convince a majority to affirm.
Brooke: would probably have said SCt would find it unconstitutional before Confederate Veterans, but now thinks they’ll defer to gov’t agency.
 Mitchell N. Berman, Coercion Without Baselines: Unconstitutional Conditions in Three Dimensions, 90 Geo. L.J. 1 (2001) (“The Supreme Court’s failure to provide coherent guidance on the subject is, alas, legendary.”) (footnote omitted); Cass R. Sunstein, Why the Unconstitutional Conditions Doctrine Is an Anachronism (With Particular Reference to Religion, Speech, and Abortion), 70 B.U. L. Rev. 593 (1990).
 Rust v. Sullivan (“[A] refusal to fund protected activity, without more, cannot be equated with the imposition of a ‘penalty’ on that activity.”).
 Cf. Sunstein, supra (“In the spending context, the government may be able to invoke justifications that are tightly connected to, and become legitimate because of, the very fact that it is engaging in those activities. In the public school setting, for example, government can plausibly justify its selectivity by the need to ensure that public funds are not spent on religious activities.”).
 Compare Rust v. Sullivan (discussing hypothetical withdrawal of all benefits from any woman who chose abortion).
 Cf. Sunstein, supra (arguing that government funding choices on aesthetic, qualitative, and content-based grounds are all acceptable and distinguishable from viewpoint-based discrimination).