Palo Alto, 2020

In Part 1 of this bloggy thought experiment, I set out the following propositions, building on the cited work of Mark Lemley, Alexandra Roberts, Rebecca Tushnet and Mark McKenna:

  • Under the Perception Axiom, consumer perception occupies the core of the definition of trademark;
  • Valid empirical proof of consumer perception regarding secondary meaning — usually understood as survey evidence — is not only elusive, it is essentially a theoretical construct;
  • Because of this, in infringement litigation we typically use proxies such as the elements of LIKELIHOOD OF CONFUSION as fair guesses of consumer perception;
  • In contrast, in ex parte trademark registration matters before the PTO, we do not use the LOC elements as proxies for “empirical evidence” of consumer perception, i.e., secondary meaning, but rather we rely on them entirely because empirical evidence of perception is impractical;
  • Woven through all this (yes, very fuzzy language, I know; it’s a blog, damn it) is the posited dichotomy which, even if not a true dichotomy (per McKenna, if I understand him correctly), represents a legitimate conceptual divide between “use as a mark” versus “secondary meaning”;
  • Both use as a mark and secondary meaning analyses are subject to attack on grounds of whether the USPTO, under its protocols for examination of trademarks for registration, adequately accounts for “context”; and
  • “Context” is a very messy business.

End of list. Let’s start at the end. Why is “context” a messy business?

As some of the authors named above have observed, we have a “context” problem on one level because word trademarks are registered and protected as “plain text,” notwithstanding that they may never be encountered that way in the market. This same “context” problem exists on the Register itself. This seems to violate the Perception Axiom.

Thus it is often possible to use the registration of a weak mark in stylized form as a method of “backing in” registration-level protection for the plain word mark, and this technnique is used all the time. The distinction is seldom appreciated by judges who have almost, to a man and woman, never prosecuted an application to register a trademark in their lives. This is what happened when, in Jews for Jesus v. Brodsky, I failed repeatedly to convince a judge or judges that the registered stylized trademark

was not infringed by use of the formative “Jews for Jesus” (no funny “o” in “for”) in the URL — and that if the plaintiff indeed believed it were entitled (at the time) to rights in the plain word mark it presumably would have sought to register it. No one wanted to hear it, and in that case the strategy (intended or otherwise) paid off handsomely.

There is a related phenomenon whereby a mark using stylized text is registered in its stylized form, or the text portion of a mark is included in a mark that is registered entirely as a design, in order to “fake out” the public: You publish your logo and your circle-R, and who really knows what’s covered by the registration?

Not really for sale

The sort of formal line-blurring that plagues the registration regime should, presumably, be amenable to solution by both diligent requirement for disclaiming unprotected matter and deploying the appropriate level of respect for what, exactly, a trademark registration says, on its face, is being protected. This respect needs to be imported whole, or perhaps re-introduced, into the judicial perception of the trademark registration process and its outcomes. This is especially true, of course, in a post-B & B Hardware world where USPTO determinations of rights at once exclude empirical information about consumer perception and yet are presumptively valid, and potentially dispositive, in litigation.

But the way to do so is not to to refer to “context,” because this way lies only madness. This I argued, also unsuccessfully (although it turned out better in the end) in preparing the Matal v. Tam briefing in the Supreme Court, that our brief should not ask the court to consider the possibility that, in the “context of the use” by Simon Tam of THE SLANTS — i.e., his being an Asian-American seeking to reappropriate a slur — the USPTO and Federal Circuit both erred in finding that there even was disparagement.

That argument went in over my objection, and was swatted away by the court on the way to ruling our way anyway. Still, it was a mistake. As I have argued above, “context” for a trademark registration is, and must be, limited to the goods and services described in the registration (and to the mark as shown therein). This is how it must be, and the best way to appreciate that fact is by considering that a trademark registration is, along with the trademark and goodwill it represents, freely assignable by the registrant.

In other words, the TMEP required the examiner to evaluate whether the mark was used “disparagingly” by considering the “context” of its use — and one of the grounds for refusal was, in fact, that because it was Simon Tam using the mark, the word “slant” could only be referring to him, an Asian-American activist. Our argument meant to turn this problem around by saying yes, that’s how we know it wasn’t disparaging. Apply the Perception Axiom right and we’ll be just fine!

The problem is that once he got the registration, Simon could have assigned the mark and the goodwill with it to anyone, no matter how un-Asian and how bigoted and disparagey – thereby changing the “context of use” after the registration was issued.

Indeed, as I have said many times, the shortcut to getting this registration, in fact, would have been to assign the mark to Thurston Howell III, filing the Asian-content-free application to register the mark THE SLANTS (i.e., two English words suggesting diagonality) and then having it assigned back to Simon.

But that’s not how any of this works!

The same goes with “context” not based on disparingness or even on the identity of the registrant. Adding another dimension of “reality” to the registration process makes lots of sense in the abstract, considering especially that a district court may well add that dimension in an infringement case or, for that matter, on de novo review of a USPTO cancellation or other inter partes proceeding. But here in this world such an approach blows up the whole idea of a “registry” intended to capture existing trademark “facts”: an applicant’s essentially pleading-level assertion of prima facie entitled to registration of a mark whose rights, pursuant to the common law, have been established and are amenable to proof “off stage,” i.e., beyond the Principal Register.

Does that mean trademark registration and trademark rights will be in conflict? Yes, they will be; they will always be. And to some extent this is merely one of many paradoxes we have to live with in the world of trademarks.

“John Welch and the Trademark Paradoxes”

Is Google generic for googling? Of course it is; but Google is also a brand, maybe the greatest brand of the 20th century. People know that it is a brand and yet they use it as a verb, and if you are asked to Google something and someone Bings it instead, chances are that if you got the right answer no one will complain. It’s not like in the restaurant where you order a Coke and they say, no, we only have Pepsi. You have Googled, or googled (small g), just fine.

And what of Now that this is out of the way, I can tell you my personal opinion about that opinion: The dissent by Judge Wynn got it right:

This case addresses a problem that chose to bring upon itself. Because trademark law does not protect generic terms, an online business, like, has two options in choosing its domain name. On the one hand, it can choose to operate under a generic domain that describes the nature of the services it offers, and thereby attract the wealth of customers who simply search the web for that service. However, in electing that benefit, the entity accepts a trade-off. It must forego the ability to exclude competitors from using close variants of its domain name. On the other hand, the entity can choose to operate under a non-generic domain name—and thereby potentially limit, at least before it has built consumer awareness of its branding, the universe of potential customers who will find its business. Trademark law affords an entity that selects this latter option a special benefit. It can bar competitors from trading on any goodwill and recognition it generates in its domain name. chose the former approach—to operate under a generic domain name and forego the ability to exclude competitors from using close variants of its domain name. But in the face of recognizing that “booking” is a generic term, the district court creatively decided that combining the generic term “booking” with the generic toplevel domain “.com” rendered it non-generic. In doing so, the district court’s judgment— which the majority opinion concedes was grounded in legal error, but nonetheless declines to set aside—allows to have its cake and eat it too. gets to operate under a domain that merely describes the nature of its business and exclude its competitors from doing the same. maintains that such a result is warranted to prevent “unscrupulous competitors [from] prey[ing] on its millions of loyal consumers,” Appellee’s Br. at 38. But to the extent fears that its competitors are using the terms “booking” and “.com” in ways that might confuse its customers, “this is the peril of attempting to build a brand around a generic term.”, Inc. v. AOL Advert., Inc., 616 F.3d 974, 980 n.6 (9th Cir. 2010). Accordingly, although I agree with much of the analysis in the majority opinion, I part ways with my colleagues’ decision to nevertheless affirm the
district court’s judgment that BOOKING.COM is a protectable trademark.

I say that Judge Wynn got it right, but also I say (as I must have on some level since my name is on the NYIPLA brief in favor of’s position) that although brought this on itself… and although shouldn’t be allowed to have its cake and eat it… that, damn it, is a trademark.

And there is no distinction between what Judge Wynn refers to as a “protectable trademark,” of course, and… a trademark.

It’s a trademark! It’s even a famous one! We know who provides the services at, not merely what services they provide! It’s a brand, a great brand, a super-duper brand built on top of a damnable thing it brought upon itself. It is. It just is.

If it walks like a duck … if it functions like a trademark… it’s a trademark. And this goes just swimmingly, of course, with the Perception Axiom.

But. But. Should we say, perhaps, “Yes, you can have your trademark, but not your trademark registration. This is the one thing we will hold out from you because of the choice you made”? That has some appeal, but — no.

A thousand times no. If we learned anything in Tam, it is that the Principal Register should not be an expression of value judgments. It is and should be a register (small s). Of trademarks.

In other words, if you have a trademark, it should be registered. There is no distinction between a protectable trademark and a trademark.

So: Trademark Singularity is … that if it is, it just is — and, that, at the end of the day, that’s how we’re just going to figure it out. Surveys be damned, we just have to try to figure it out.

Maybe that, after all, is why they keep calling trademarks “intellectual property.”

Originally posted 2020-03-03 17:37:04. Republished by Blog Post Promoter

By Ron Coleman

I write this blog.

One thought on “The Trademark Singularity: Part 2”
  1. The Trademark Singularity is a brand new series from the MIT Media Lab, examining issues of intellectual property law. We’ll examine why brands are protecting themselves against competition as companies increasingly adopt digital technology and business models to innovate even more rapidly than before. “Sixty years age.

Comments are closed.