Friday, February 24, 2017

Bridging the Gap Between the Federal Courts and the U.S. Patent and Trademark Office

Boston University Journal of Science and Technology Law Symposium

Sorry, guys, I did not follow the patent panels.



Panel III: Trademarks
Rebecca Tushnet, Fixing Incontestability: The Next Frontier

Incontestability is a nearly unique feature of American trademark law, with a unique American implementation.    Even if a symbol is merely descriptive and lacking secondary meaning, incontestability means that federal law will treat the symbol as if it were functioning as a mark.  This is a powerful protection, given that descriptive terms lacking secondary meaning are otherwise free for all competitors to use, and that competitors are likely to have good reasons to do so—after all, the symbol describes relevant features of the product or service.  Although competitors retain the defense of descriptive fair use, this defense is difficult to prove and often requires at least extensive discovery .

Given the power of incontestability, one might think it would be carefully granted.  Unfortunately, claims of incontestability are not subject to substantive examination at the PTO.  Moreover, because incontestability has specific legal requirements that may be unfamiliar to lawyers who aren’t trademark specialists, it’s simple for negligence alone to produce invalid incontestability filings, which then appear on the Register and may distort decisionmaking, both by competitors and by courts. 

Worse, and of particular relevance to this Symposium, courts may misunderstand either the prerequisites for or the meaning of incontestability, allowing trademark claimants to assert rights that they don’t actually have.  Courts even appear to be inconsistent about meeting their statutory duty to notify the PTO about the filing and conclusion of all actions concerning registered trademarks, which has implications both for validity in general and for incontestability in particular.  Competitors, many of them likely to have little in the way of sophisticated trademark advice, are in an even worse position to evaluate the validity of incontestability.

In the last two years, several legally significant cases—including one that went to the Supreme Court on apparently mistaken premises—were complicated by unwarranted Section 15 approvals.  I’ll talk about one, but my paper discusses a few others. 

The most striking example of apparently mistaken incontestability comes from B&B v. Hargis.   This case has a complicated procedural history, and it went to the Supreme Court in 2015 on the question of whether preclusion applies to TTAB rulings in subsequent infringement litigation.  The Supreme Court answered in the affirmative, at least for some situations.  That result makes the stakes of registration higher, and reinforces the need to have the register reflect reality.  However, the course of the B&B case itself reveals trouble with the accuracy of incontestability claims.  B&B registered SEALTIGHT for fasteners in 1993 and then sued Hargis for infringement in 1998, relying on its registration.  In 2000, Hargis proved at trial that SEALTIGHT was descriptive without secondary meaning for fasteners.  Yet neither the district court, nor the court of appeals that affirmed the ruling, ordered the PTO to cancel the registration, despite that registration’s invalidity.   Meanwhile, Hargis filed a petition to cancel B&B’s mark, which proceeding was suspended pending the resolution of the infringement lawsuit. 

When the infringement lawsuit was resolved in favor of Hargis, Hargis attempted to amend its cancellation petition to add the ground of invalidity on the basis of mere descriptiveness, but the TTAB denied the motion to amend because by that point B&B’s registration was more than five years old, and thus could no longer be challenged on mere descriptiveness grounds.   B&B subsequently renewed its registration in 2003, and filed a Section 15 affidavit in 2006 claiming that the requirements for incontestability had been satisfied.   (As a reminder, those requirements include that there be no pending challenge to the validity of the mark, or successful and final challenge to validity.) 

Hargis applied for a registration of its own, which was rejected by the TTAB on grounds of likely confusion with B&B’s registered mark.  After filing for incontestability, B&B sued Hargis again in 2006, and the court of appeals found that preclusion from the first trial didn’t apply because of the change in the mark’s circumstances from contestable to incontestable.  Then, after the Supreme Court decision, the court of appeals found that preclusion from the TTAB ruling did apply, meaning that Hargis was an infringer. B&B, that is, filed an affidavit claiming that there had been no final decisions adverse to its ownership of the SEALTIGHT mark, after losing an infringement case on the ground that it did not own a protectable mark.  Then it parlayed that affidavit into a decade of new litigation, including a successful TTAB opposition and a duplicate infringement suit, having precluded its opponent’s previously successful defense. 

Hargis contributed to the debacle by not seeking an order  cancelling B&B’s trademark registration upon its initial victory.   But this too was almost certainly a function of the relative exoticism of registration practice to ordinary trial lawyers : It is not necessarily obvious to non-registration specialists that a final order from an Article III court finding a claimed trademark to be merely descriptive needs to be followed up by another order from the court directing the PTO to cancel the registration in order to have preclusive effect. 

This comedy of errors ended in further farce when B&B failed to renew its registration during the pendency of the Supreme Court litigation.  As a result, the registration expired.  The trial court then held that, without the incontestable registration, B&B was once again bound by the earlier trial result finding that it had achieved no secondary meaning and thus no protectable mark.  All of this could have been stopped much earlier if either the PTO had possessed sufficient information to see that the statutory requirements for incontestability had not been met or the courts had understood those statutory requirements and independently examined whether they’d been satisfied.

There are other recent examples with similarly tortured histories.  And one recent Sixth Circuit court case apparently had no understanding at all of incontestability, deeming it applicable to a mark that had been cancelled (and then subsequently registered by the same party, but not for five years).

General unfamiliarity with incontestability probably explains why many parties who would benefit from pointing out erroneous incontestability designations fail to do so.  The cases I have found are examples that turned up in my research fortuitously; they were not produced from any attempt to determine how often problems with Section 15 affidavits occur.  Moreover, each case appears to involve courts that didn’t pay detailed attention to, and were almost certainly unfamiliar with, the legal prerequisites for incontestability.  If a Section 15 affidavit is mistaken (whether or not knowingly false), then the requirements of Section 15 haven’t been met, and thus courts should not accord the affected registration an irrebutable presumption of distinctiveness. But it is a disturbing fact that the error has been repeated multiple times by different courts in case-dispositive ways, including in a case that made it all the way to the Supreme Court and through two trials. 

What might we do?  Changes to the process for Section 15 affidavits similar to those already underway for Section 8 renewal filings could help registrants avoid the kinds of inadvertent errors that may well have underlain these cases.  Fundamentally, while a mark that has become incontestable cannot be challenged on the ground that it lacks secondary meaning, it should be possible for challengers to identify marks that never satisfied the statutory prerequisites for incontestability. Specifically, possessing secondary meaning isn’t itself a prerequisite for incontestability.  But lack of a pending challenge or final successful challenge to validity is.  A mark that was in fact subject to such a successful or pending challenge, therefore, has not satisfied the statute and is not entitled to the cloak of incontestability.

Nor should courts be precluded by the statute from asking the relevant questions about incontestability.  Incontestability is for marks that have satisfied all the requirements thereof, not just for marks that happen to have a Section 15 affidavit on file—should be the same.  As the Second Circuit court commented in a very slightly different context, it would be a “perverse result” if the use of “recording—a ministerial act—[was transformed] into a mechanism for conclusively defeating allegations ... challenging the legality of the assignment.”   So too with the ministerial act of accepting a Section 15 affidavit, or counting the years a registration has been on the record.

The PTO has discovered significant overclaiming through random audits of renewal filings.  Although incontestability is rarer than mere renewal, the cost to the public of a single faulty incontestability designation may be higher than the cost of a single extra goods or services designation, given the ability of a registrant to use incontestability as leverage in disputes and the ongoing confusion in the courts about the effect of incontestability on a mark’s strength.

Small steps could prevent many of these mistakes in the first place.  The PTO has made changes to the standard declaration that an applicant signs under penalty of perjury when filing a trademark application or renewal. Instead of a traditional block of text, the form now presents a separate series of clauses with check boxes next to each statement.  The applicant should now more readily understand that it must claim use, or intent to use, the mark on every good or service it lists in the application.  Similar measures could improve incontestability by making the requirements of Section 15 more salient.  For example, the affiant could be required to check a box, yes or no, about whether there are any pending or resolved cases involving the mark.    If the answer is yes, the form could ask the affiant to provide more information.  Alternatively, the affiant could be required to list all proceedings, pending and terminated, involving any challenge to the validity or ownership of the mark in question, and state the status or outcome of the proceeding with respect to the mark.   Random audits of trademarks for which the PTO has received a “Notice of Suit Incoming” could also improve the situation, assisting the PTO and registrants in determining when the register needs to be updated to reflect reality.

Likewise, the PTO’s proposal to institutionalize random audits of trademark maintanence filings could be expanded to a percentage of Section 15 affidavits, at least to the extent of having examiners run a Westlaw search and check whether any pending cases appear in TSDR and then, if the trademark appears to be the subject of litigation, asking the trademark owner to explain why the cases aren’t relevant.  The PTO could also consider a similar procedure for verifying incontestability if a question arises, making a more formal practice of the correction it made in response to my blog post on one such case.

Most registrants who provide incorrect information are not determined to distort the register.  Negligence and ignorance are bigger problems.  It is therefore worth giving more detailed cues about the statutory requirements for a Section 15 affidavit to affiants at the moment those cues are most needed. 

Trademark attorneys can improve their procedures as well.  Registration counsel is sometimes, but not always, litigation counsel.  When lawyers discuss renewals and Section 15 affidavits with their clients, they should take care to ask about whether there is any pending litigation or PTO proceeding involving the mark, not just whether the client is still using the mark and wants the registration renewed.

The courts also have a role to play.  Under current rules, notices should be sent to the PTO about pending cases involving registered trademarks.   In theory, the plaintiff should file the relevant form with its complaint, and the clerk’s office should pass the form on to the PTO.  These forms are then recorded by the PTO and become part of the formal record for the registration at issue.  The form should be filed again, with appropriate updates indicating the outcome, when the case is terminated.

If a mark’s record is complete, then it will be a simple matter to identify whether further inquiry is required to evaluate whether the requirements for incontestability have been met.  Although the notification process is already in place, experience indicates that compliance is spotty: many litigated cases involving registered marks never show up in the registration record.  For example, the highly significant case Christian Louboutin S.A. v. Yves Saint Laurent America Holding, Inc.,  which involved a partial cancellation of Louboutin’s red sole mark for its shoes, never showed up in the PTO’s records until correspondence between the parties about how to implement the partial cancellation order.  If litigators and clerks’ offices around the country became more aware of the problem, the records could be improved.  I have a research project ongoing to figure out where the notice of suit falls through the cracks and whether there are steps that the PTO or clerks’ offices could take.

Similarly, it would be beneficial to improve judicial understanding of the requirements for incontestability.  Even though courts do not evaluate whether an incontestable mark has secondary meaning, they can and should evaluate whether incontestability was properly secured. 

Fundamentally, no one should be made better off by providing inaccurate information to the PTO.  Unfortunately, it is currently possible to benefit from doing so. Incontestability offers low-hanging fruit for improving the accuracy of the trademark register and the courts’ understanding of what registration means.

Q: petition to cancel incontestability as a possible solution?

A: good idea.  Similar to a random audit; I’m more interested in things that the PTO could do in the absence of congressional help.

Q: patent cases—it’s second entry in the docket in the federal district court; why is it not showing up in the PTO’s records.  Also, why require act of PTO to make it incontestable?

A: The form might be in all the TM dockets in federal district court—the gap may be between the dcts and the PTO, and I want to find out where that gap is.  I neglected to talk about the fact that no act of the PTO is necessary to make incontestability apply in cancellation proceedings, which is what happened to Hargis.  We could do the same with incontestability for litigation purposes, but the incontestability filing may be our only chance to get TM owners to look at this and see whether they actually meet the requirements, so I’ve now talked myself into wanting to keep the affidavit—and maybe to expand it to the cancellation context.

Alex Roberts: Your proposals are conservative; what would you do if you could rewrite the law?

A: I might keep incontestability as part of a more substantive approach to registration. That said, I’d want to have clear requirements for substantive examination that the prerequisites had been satisfied.

Christine Farley, American University Washington College of Law, No Trademark—No Problem: About Belmora, cert petition pending.  Today is the day it was distributed for conference.  Flanax: Bayer had no rights; Belmora didn’t look too good as a copier. 

Person’s: strict territoriality (CAFC); Grupo Gigante, famous marks (9th Cir.); Punchgini, no such exception to territoriality for famous marks even if it would be a good idea (2d Cir.).

4th Circuit discussion was unsophisticated and didn’t engage w/territoriality at all; TTAB discussion was much more nuanced and understanding of the waters in which it was wading.  TTAB knew: territoriality is definitely a hurdle and TM rights arise solely from use in commerce in the US.  Case law didn’t address whether Bayer’s use in Mexico would be enough for a claim, but they decide to extend the law—unlike Person’s, there was a reputation in the US, and Belmora was exploiting that reputation.

Why didn’t Bayer act earlier?  They probably didn’t think they had a successful claim to bring, until there was enough advertising to act on. That’s reflected in the kind of petition/complaint they filed—named 3 different int’l treaties as basis of claims, one of which Mexico wasn’t even a signatory to.  Also, series of TTAB decisions granting Bayer’s motion to amend again and again until everything was gone but misrepresentation of source based on the goodwill that flowed across the border. 

TTAB 2000: found for a TM owner with no registration in the US and allowed them to cancel under a little-known treaty (Belmont); again in 2008, Cuban coffee company used same path; in 2009, the claim was unsuccessful b/c there was a requirement to pay a fee to a Bureau that didn’t exist to take advantage of Inter-American treaty.

Why is the TTAB willing to go out on a limb?  She thinks it’s standing.  TTAB requirement: real interest.  TTAB in Belmora: “petitioner has shown that it has an interest in protecting its Mexican FLANAX mark”—a relatively low standard. But SCt just addressed this under the Lanham Act §43—Lexmark asked claimants to focus on zone of interests protected by the statute.  You’d think there’d be no basis for adjudicating a dispute w/o a use in commerce. Maybe that should inform how the standing requirement is being addressed.  

Given the outcome, you should go from TTAB petition to EDVA to 4th Circuit.  The TTAB stuff is very significant, because this can throw open the doors to a wider set of claims.

Arti Rai: The TTAB standing requirement: why is there a requirement for an administrative agency?

A: Because the Act says any person who would be damaged.  TTAB has said Lexmark standard doesn’t cover its proceedings--§43(a) is different, but the “damaged” language is the same.

RT: zone of interests: people protected by §2 that don’t have commercial interests.

A: treaty obligations—obligated by treaty to protect state emblems and well-known marks, maybe that’s w/in the zone of interests in the Lanham Act.

Stacey Dogan: whether the interests at issue in a cancellation proceeding may be distinct from infringement?  TTAB as protecting integrity of the Register, for example by barring registration of deceptive marks. 

A: it’s still the same priority battle in both cases.  Becomes confusing to have a different standard/answer at the TTAB.

Q: curious about unfair competition aspect.  Seems strong rationale for protecting rights of foreign user w/famous mark even w/o use in the US.

A: goodwill is the reason to protect.  There might be a public policy in favor of this, but there has to be some rationale, and the 4th Cir. simply deleted territoriality instead.

Alexandra Roberts, University of New Hampshire School of Law, Trademark Failure to Function

What do we talk about when we talk about use as a mark? Relative size, color, distinctive font, stylization, prominent placement in TM spot, set off by itself, use of TM or ®, all caps or initial caps.  Lee et al. study: consumers are likely to perceive most non-generic terms as TMs when they are in the right “place” on the package.  Even w/o distinctive font, 70% said “Wonderful” was a brand name when it was large and placed right.  Smaller, lack of prominence—much less likely to say it looks like a brand name.  Small: 33% brand name; small and placed along the bottom edge of the package: only 27% said brand name.

Maybe best way to think about this is ex parte v. adversarial.  Applicant sought to register icon of toilet for call devices used in hospitals.  Refused.  Spectrum, in colors, but used in a sentence—TTAB rejected even though it’s not just for a word mark; it was a design/composite mark using color.  Still not using it as a mark/fails to function.


In re Chemical Dynamics (Fed. Cir. 1988): applied to register dropper and droplet, not full image.  The court (and PTO) said you aren’t using that separately, it’s part of the whole mark—no separate commercial impression.  Often the applicant can file a new specimen to rebut a failure to function refusal.  In re Eli Lilly (TTAB 2015). Statement: “syringe contains a solution of Trulicity™ or placebo.”  Just plain text, part of sentence.  TTAB says: this is inherently distinctive; distinctiveness creeps into use as a mark analysis. 

Ex parte failure to function: administrative expertise—examiners do this every day; but: limited guidance/tools for identifying failure to function v. tons of tests, factors for distinctiveness.  Lists in TMEP of what might fail to function; language is super tautological: does it look like a TM?  Not actually that helpful!  Doctrine creep: mixing use w/use in commerce, inherent distinctiveness, acquired distinctiveness, functionality. Ex parte = closed universe of evidence for applicant’s specimen. Atty can manipulate that by working w/client to generate a specimen that looks trademarky.  Examiners have incentive to work quickly—they have a quota.  There’s also an ITU handicap: a huge # are filed on ITU basis; initial application occurs w/o specimen. We expect the review of the specimen won’t be as thorough—may even be completed by paralegal and not examining atty, so will sail through w/o anyone really examining function as a mark.

Adversarial proceeding: open universe of evidence; context is cancellation, infringement, priority; divergent guidance/tools for identifying failure to function; doctrine creep is similar; greater resources but less expertise.  More nontraditional marks--§43(a) unregistered trade dress.  Rights defined around/against opponent. 


Starting to code cases: a lot of overlap w/other use issues and w/distinctiveness.

No comments: