In trademarks, generic is generic is generic. That’s what makes it generic.
But there’s a difference between a would-be trademark that is born generic and one that dies of “genericide.” We call the first type generic ab initio to impress baristas. It’s not only a would-be trademark, it’s a never-was trademark. More precisely, it never was a trademark, and nothing anyone can do could ever make it a trademark.
To use a pretty darned generic example, a chair is a chair is a chair. Forever. All the advertising, sales and even social media engagement in the world won’t infuse your “Chair Brand” line of chairs with secondary meaning.
The second type of would-be generic would-be trademark is one which, at one time, was a trademark, and perhaps a darned good one. But through genericide it became synonymous for the genus of the product it describes. Big difference, of course.
So how big of a difference is it in a cancellation context where a trademark registration has become “incontestable“?
Depends whom you ask, it turns out. I was quite astonished to read this tweet a week or so ago by my friend Roberto Ledesma:
— Proof of Use (@ProofofUse) February 9, 2017
I previously reported on a case that managed to find a non-statutory basis for cancelling an “incontestable” trademark registration, specifically that the application for the registration was void ab initio. The plaintiff and trademark owner was NetJets, with a registered trademark for a software program INTELLIJET. The defendant was a company named Intellijet Group. Both companies provided services related to private jets. . . .
In a non-published decision, the Court of Appeals for the Sixth Circuit has reversed. Its reason was the absence of “void ab initio” from Section 14:
The district court did not address the limits of § 1064 in its opinion, but simply moved from the conclusion that NetJets’s mark was not incontestable to its conclusion that it was void ab initio because it was not used in commerce at the time of registration. This determination is incompatible with § 1064.
The defendant’s various theories why a “void ab initio” challenge was nevertheless available didn’t get any traction with the appeals court.
But the appeals court didn’t explain where the district court erred with it’s two-step theory, saying inscrutibly
Under the statutory requirements of § 1065, the district court determined that NetJets’s INTELLIJET mark was not incontestable because internal use of the software did not constitute use in commerce sufficient to satisfy the Lanham Act, and as a result, the registration was only prima facie evidence of the mark’s validity. The court held that because the mark did not satisfy § 1065’s requirements for inconstestability, IntelliJet Group ws not limited to challenges in § 1115(b) seeking cancellation of the mark. Because we find that § 1064 bars IntelliJet Group from bringing its challenge to the mark as void ab initio, we decline to examine whether the mark is incontestable under § 1065 and whether the limitations of § 1115(b) apply.
Not the best explanation, but I expect what they were trying to say is a drum that John Welch has been beating for awhile. John points out that “incontestability” has no role in a cancellation, which is clear from a careful reading of the statute. Section 15 says only that an incontestable registration is conclusive evidence of the validity of the trademark. . . .
Whether the trademark is “incontestable” has nothing to do with the bases available for a cancellation; it is only the passage of time that limits the grounds for cancellation. Enough time passed and therefore NetJets’ INTELLIJET registration was not susceptible to cancellation on the basis that it was void ab initio.
Very analytical, rational and logical. I respectfully submit, however, that both the Sixth Circuit’s “unpublished” opinion and my friends are wrong about this.
We’ve been here before. I argued in this post about the PING PONG “trademark”, trademark registrations can never be better than the trademarks they register. And it is axiomatic that generic terms can never be trademarks. Period. Unsurprisingly, Pamela disagreed with me on that too.
But I don’t see how we escape the fact that if a “trademark” was generic when it was registered, getting a certificate with a shiny foil seal from an office building in Virginia won’t “fix” the genericness problem, no matter how much time has passed. “[I]f a generic word could ever be infused with trademark significance, the word must have ceased to have current generic meaning,” the Second Circuit explained in Harley-Davidson, Inc. v. Grottanelli, 164 F.3d 806, 811 (2d Cir. 1999). This doesn’t happen much, and it’s hard to see how it could. A chair is a chair is a chair, and five years and an affidavit will never make it anything but a chair.
But what about the statutory analysis? It’s internally consistent, yes, but fails from a lack of context. In this case, that context is the requirement that we establish the existence, vel non, of a trademark. And the reason Congress did not specify that cancellation was available on the ground that a purported mark was generic ab initio is because it didn’t have to: The registration should not have been issued in the first place.
So then why did Congress see fit to include genericide as a ground for cancellation? Because you might well think that an incontestable registration genericness-proofs your very, very valid trademark, encasing it in Lucite for all time and giving you an eternal insurance policy against genericide no matter how sloppy you are about genericization. To that, Congress says, “No.”
This must be the rule because Congress could not have intended to permit some kinds of generic marks to be protected as trademarks. If there’s any doubt, though, we certainly should not infer such unlikely intent from the lack of statutory language.
Yet per the statutory analysis by the Sixth Circuit and endorsed by my friends, trademark registrations for generic terms that were always generic, and which by definition can never become trademarks, would indeed be preserved in amber even if they were improvidently granted. Well, excuse me, but I respect trademark registrations too darned much to treat them as royal grants of franchise or monopoly over a term that a pliant Examining Attorney green lights through the sausage machine!
A registration is no better than the substantive right it represents, and the fount of that right is not the USPTO but secondary meaning, which generic claimed “trademarks” can never have.
Ah, but you’ll say to me — stop talking while you’re reading! — if that were so, why did Congress see fit to make fraud an explicit ground for cancellation? Well, as both John Welch (quoted above) and Pamela seem to see it, § 1065 is a statute of repose, although unlike most statutes of limitation it is not self-executing; it must be perfected by the filing of the requisite paperwork. Throughout the law, even claims for fraud have their limits. That would make sense, but here Congress says, explicitly, “No.” If you can unwind a fraudulent registration, we will undo it. That’s a policy choice, and one properly left to Congress.
There’s another answer to that question too. Maybe your application to register a generic trademark is itself a fraud on the PTO ? I didn’t come up with that — the United States Court of Customs and Patent Appeals, predecessor to the Federal Circuit, did — in a well-known case called Bart Schwartz Int’l Textiles, Ltd. v. Fed. Trade Com., 48 C.C.P.A. 933, 938, 289 F.2d 665, 669 (C.C.P.A. 1961). There the CCPA held that a registration that turned out to be generic was “obtained fraudulently” because appellant misrepresented the rights of others to use the word… and cancelled it.
Oh, yes. See also, Nw. Corp. v. Gabriel Mfg. Co., Case No. 95 C 2004, 1996 U.S. Dist. LEXIS 19275, at *53 (N.D. Ill. Dec. 10, 1996) ( “[a] trademark applicant has a duty to disclose the fact that a [proposed registered mark] is generic at the time of application to the PTO”).
That’s a neat trick, but one I wouldn’t necessarily rely on (though I have, in part, in papers I recently submitted on this issue in a pending case, but which I cannot share right now). The fact is, other courts have agreed with this proposition: That it is inherent in any assertion of trademark rights that one have a trademark — secondary meaning or distinctiveness.
Thus “incontestable” registrations for “generic marks” can never be anything but in terrorem tools to intimidate competitors not equipped to challenge a cease-and-desist demand into avoiding the best possible terms to describe their wares, ones they have every legal right to use.
In fact, other courts considering the issue have held that “[e]ven the presumption of validity arising from federal registration . . . cannot protect a mark that is shown on strong evidence to be generic as to the relevant category of products prior to the proprietor’s trademark use and registration.” Harley-Davidson, Inc., supra, 164 F.3d at 811. See also, Soweco, Inc. v. Shell Oil Co., 617 F.2d 1178, 1183 (5th Cir. 1980) (“A generic term can never become a trademark, [and] if a registered mark at any time becomes generic with respect to a particular article, the Lanham Act provides for the cancellation of that mark’s registration.” (footnote omitted) (citation omitted).
Thus a defendant is “free to argue that a mark should never have become incontestible because the mark is generic for the goods.” Nw. Corp., supra, v. Gabriel Mfg. Co., 1996 U.S. Dist. LEXIS 19275 at *53 . And in a 2016 case, ZW USA, Inc. v. PWD Sys., LLC, No. 4:14-CV-1500-CEJ, 2016 U.S. Dist. LEXIS 129549, at *20-21 (E.D. Mo. Sep. 22, 2016), the Eastern District of Missouri held, “It is a complete defense to a trademark infringement claim to rebut the presumption that a registered trademark is valid, by proving the trademark is generic . . . i.e., that the registration was issued in error and is subject to cancellation.”
Stuff happens, folks. Only God and His Law are Incontestable.