AUWCL Booking.com debate

David Bernstein, Debevoise & Plimpton: we don’t give controlling authority to dictionary meaning or previous generic use. We consider one thing only: consumer perceptions of the term now as generic or brand name. In the past, courts and scholars have said that a generic term can never be brought back from the dead, but there’s no support in history, Lanham Act, or sound public policy, and Booking.com rejects that.  A TM need not be born, grow old, and be extinguished. Goodyear case itself proves this point. In 1888, Goodyear was considered generic for a process for making rubber. Today, it’s a wellknown trademark for rubber tires, and registered since 1948. That’s good! If consumers grow to perceive a term primarily as a brand name, it should be protected as a mark.

Protection v. enforcement: even if protected as a TM, the scope may be limited. There are numerous limits that prevent anticompetitive enforcement, which is true for formerly generic terms too. Descriptive fair use defense; crowded field doctrine narrowing scope of protection. Injunctions can’t be issued w/o likely confusion. Ds will have strong argument if they’re using other designs, words, or logos. Law, history, and sound TM policy support using consumers’ current perception of a term as a brand name supports TM protection.

Mark McKenna, Notre Dame Law: Consumer understanding makes sense, but consumer understanding as measured by survey should not always determine this. Genericness has never been solely empirical assessment of consumer understanding. It has always included scope. Because the domain name system already provides exclusivity, the value of a TM registration for a generic.com is primarily in enforcement against nonidentical variations. It’s fair to be skeptical that any party that litigates all the way to the SCt is doing so to acquire razor thin rights. More importantly, the rule applies to all other generic.com and all other arguably generic terms.

Experience teachers that we can expect registrations overenforced, such as Freecycle.org which got a registration for Freecycle.org despite freecycle being generic. Enforced against FB groups using only “freecycle.” The promise of limits does not materialize, and it’s worse in C&D situations where they wave registrations around.

Good reason to worry that the claimed association w/a particular term is from market exclusivity (patent, or de facto exclusivity generated by the domain name system). Courts will not be able to tell the difference b/t TM meaning and de facto secondary meaning. 33% of respondents said that washingmachine.com was a brand name after being taught the difference and passing non-.com comprehension tests.

Given these risks of error and overenforcement, the game is not worth the candle, especially since narrow unfair competition remedies would remain available for true abuse of booking.com, e.g. in phishing emails.

Jake Linford, Florida University College of Law: Genericity depends on meaning to consumers; an unyielding legal rule that disregards consumer perception. A word is the skin of living thought, and can vary a lot according to time in which it’s used. A narrow word can broaden to cover a category. A word that identifies a category can become more specific. Consumers change—firms can’t force and courts shouldn’t ignore changes in TM meaning.  De facto secondary meaning courts are more comfortable with dictionaries; surveys are better than that 17th century technology. Better tools to show how consumers use marks. Courts shouldn’t refuse to consider evidence from consumer surveys even if dictionaries suggest that the term was once generic.

In the same survey, eTrade was recognized by 2/3 of consumers and should have been allowed even accepting all survey criticisms. Anyway, now we can do better surveysh

Rebecca Tushnet: The key here is whether the game is worth the candle: what is the marginal impact of allowing generic words plus generic tlds when the claimed TM meaning comes from consumer perception of the combined generic terms.

Bond-OST: cheese not known in this country.  If you asked consumers what it means, they’d choose brand name over generic. But that’s not what it means and a survey should not be used to call it a TM. That happens a lot in an internationalizing economy.

TM claimants are not victims of happenstance; they are actors, and often the most intentional actors in this space, and their acts can hurt competitors and competition as well as consumers. This means two things: (1)  risks of overenforcement, (2) the keys to the prison of meaning shift are in their hands. They don’t need to narrow generics and should not be given incentives to try to shift language in this way. (Notably, Jake doesn’t claim that “booking” has narrowed in meaning or that .com has narrowed in meaning.)

A rule allowing resurrection of generic terms is risky to competition. Ale House: constantly suing other Ale Houses arguing that it has nongeneric meaning. It loses those cases under current doctrine but, according to David & Jake, shouldn’t without getting an [expensive] opportunity to introduce new evidence with each new lawsuit.

Think about scope when you think about validity, as we do in other areas like functionality: because .com will be ignored in an infringement inquiry, it should also be ignored for validity.

The defenses aren’t adequate. Summary judgment is often disfavored in confusion cases, making them expensive. For example, putting a generic term in one’s domain name is likely to be pretty risky for descriptive fair use, given “otherwise than as a mark.”

Consumer recognition is not itself trademark meaning. De facto secondary meaning is a doctrine with a purpose that should still guide us: The Supreme Court recognized that people correctly understood that for a long time all shredded wheat came from Nabisco. That expectation should not be translatable into TM. Even assuming Booking.com is different because of the absence of a patent monopoly, the available tools for identifying trademark meaning won’t be able to reliably tell us whether we’re dealing with expectations around exclusivity, mere recognition, or the kind of trademark meaning that can justify preventing other uses of a term.

Christine Haight Farley, AUWCL: themes: (1) is this about protection or scope; (2) risks of error; (3) tools available to find consumer perception.

Bernstein: This is all about scope. Just b/c it’s hard or some risk of chilling, that’s not enough: we protect descriptive terms all the time, and the exact same issues arise. As a practical matter, there’s likely to be some chilling, but fair use, crowded field doctrine, and increasing willingness to award attorneys’ fees will be among the different tools that will help prevent anticompetitive conduct. There are real issues w/surveys. If people have never heard of the cheese, we need to know only the opinion of people who have heard of the thing—67% of the washingmachine.com people knew that it was generic. We also disregarded the answers of the 33% of those who got wm.com wrong; even still, the vast majority recognized booking.com. The “voice of the consumer” is another way of assessing consumer perception—how is media using it? How are consumers using it? Finally, unfair competition remedies don’t do the trick. There are times when you need a TM. You can’t bring a UDRP action based on unfair competition, or takedowns, or customs recordation. Perfect example: cars.com suffered a problem where a retail used car company had a big lot with a sign called cars.com and they weren’t affiliated at all with cars.com.

Linford: disclaimer requirement is no remedy at all. Disclaimers may push consumers in precisely the wrong direction—it doesn’t give much relief to mark owners. There is some point to thinking about scope itself. The better place to address that is in the scope stage at litigation and not try to force that at validity.

Mark McKenna: It’s never been true that unfair competition is limited to disclaimers. The Kellogg court looks at size, prominence, font; we can’t say there’s some recent trend on disclaimers b/c there’s no recent trend of using unfair competition as a distinct body of law at all. This was a missed opportunity.

Disagree w/David that we should take solace in wm.com. Those 33% were people who had already been told the difference and passed the screener test—they just can’t do this task. It’s one thing to take them out of the survey but those people don’t leave the marketplace; it’s still an inaccurate view of what’s going on in the world. The surveys aren’t telling us about the world!

Farley: you interpret the 33% as people not understanding, but maybe they’re reacting to the compound, making them more likely to interpret it as a brand.

McKenna: and that turns out to be true even for people who have already been trained, including on .coms: they can’t reliably distinguish between things that are de facto secondary meaning and things that are trademark meaning. Yes, it is about scope. Debevoise gets to represent the big owners, not the small businesses that just have to fold.

RT: This is as David says a marginal effect, but it’s a marginal worsening not a marginal improvement. One possibility: more focus on TM function as such rather than just non generic capacity.

For cars.com, I don’t see why unfair competition would have failed there, it’s neither customs nor UDRP. The UDRP would also accept a registration of URL+design, which Booking.com initially tried.

The real question is what happens in new industries or new forms of advertising. It's not surprising that freecycle, a newly invented generic, is one place that problems have developed.

Linford: functionality ignores secondary meaning; the reason is that a functional feature is fixed in ways that language is not and we can safely draw a line b/t functional product features and genericness. [I wish I had said something about this, because it strikes me as plainly wrong (or alternatively, design is a language!).]

Bernstein: there is a difference b/t a suvey that looks at genericism and one that looks at secondary meaning. A secondary meaning study says “is there one thing called American Airlines or is there more than one thing called American Airlines?”  But he wants to emphasize that surveys aren’t the be-all and end-all. There are problems w/surveys & understanding. “Voice of the consumer” is something we’ll see more of. We did in Booking.com have a linguist who talked about actual use, but the survey got a disproportionate amount of attention in argument. Would also support an anti-SLAPP rule for bad claims.

McKenna: that would be welcome, but for a court to find baselessness, the court has to find the claim baseless. If it’s the case that it’s always a question of consumer understanding, and that’s never fixed in time, it’s much harder to find that without a bunch of evidence which itself will be evidence of not-baselessness.

RT: production v. recognition: really huge difference in some cases, including genericity-relevant situations. In the classic Teflon case, one survey found that most consumers could not come up with a name for the pans other than “Teflon,” which sounds like it’s generic. But when you told them “nonstick,” they understood it immediately. So I’m concerned that “voice of the consumer” evidence may mistake production alone for the full scope of consumer understanding of what terms mean.

Bernstein: after Brunetti & Tam, there was already a lot of concern over failure to function, and Booking.com may accelerate that pattern at the PTO. We’ve seen more failure to function/ornamentality refusals and that makes sense.

McKenna: agree descriptively. Jury is out whether that’s good b/c it’s really undertheorized, I know it when I see it. That’s b/c we’ve subsumed so much of the work of identifying what a TM is into distinctiveness (that’s a long term consequence of folding unfair competition into TM). Having done that we’ve now realized that the Abercrombie spectrum doesn’t quite match what we think TMs are.

Linford: Q of whether they’ll submit surveys to the PTO or wait for an appeal to the district court.