Monday, December 19, 2011

Snowball's chance: TM claim as false advertising


Southern Snow Manufacturing Co. v. Sno Wizard Holdings, Inc., --- F. Supp. 2d ----, 2011 WL 6287955 (E.D. La.)
Mark McKenna should be interested in this one, which also features Stacey Dogan and Glynn Lunney.
The parties sell ice-shaving machines and flavoring concentrates to vendors. SnoWizard registered ORCHID CREAM VANILLA, then sent a C&D to Southern Snow.  Southern Snow filed suit claiming that SnoWizard had obtained the trademark under false pretenses and adding unfair competition claims for ORCHID CREAM VANILLA and WHITE CHOCOLATE & CHIPS.  SS also filed an administrative cancellation action at the USPTO, and the TTAB cancelled the registration on the ground that the term was merely descriptive.  The court then granted summary judgment to SS for the same reason.
Plaintiff Parasol Flavors, LLC also sued after receiving a C&D from SnoWizard about flavor names SNOSWEET®, HURRICANE, PRALINE, KING CAKE, BUTTERED POPCORN, MUDSLIDE, GEORGIA PEACH, DILL PICKLE, CAKE BATTER, and BUTTERCREAM. SnoWizard counterclaimed, and the original district judge dismissed everything but the SNOSWEET claims, finding all the other flavor names generic.
SS also sued SnoWizard for infringement of SOUTHERN SNOW® and FLAVOR SNOW® and challenging the validity of various SnoWizard trademarks: CAJUN RED HOT, CHAI LATTEA, COOKIE DOUGH, MOUNTAIN MAPLE®, SNOFREE®, SNOSOUR®, SNOBALLS (design), SWISS ALMOND COCO, TIRAMISU, and ZEPHYR. SnoWizard counterclaimed for cancellation of SOUTHERN SNOW® and FLAVOR SNOW® and for infringement of MOUNTAIN MAPLE®, CAJUN RED HOT, and GEORGIA PEACH (the last of which was kicked out on summary judgment).
The court commented that plaintiffs’ challenge to the validity of 22 of SnoWizard’s trademarks was only a first step in its false advertising/Louisiana state law unfair competition claims.  Those were based on the proposition that SnoWizard made false statements of fact in appending ® and TM to various flavors that were in fact generic.  Plaintiffs argued that these claims were literally false.  They contended that the parties’ customers (people who operate snowball machines) weren’t familiar with trademark law, and that their purchasing decisions were likely to be distorted by all those TM claims.  Customers who bought supposedly exclusive flavors from SnoWizard might well buy other flavors and supplies from it, to benefit from shipping discounts and for convenience.
The court was unimpressed by the argument that there was no express authority that false claims to trademark protection are actionable.  Neither state law nor the Lanham Act purports to “enumerate the universe of potential unfair trade practices that are actionable.”
However, the false advertising claims here weren’t just misleadingness claims; they were literal falsity claims.  The court wasn’t willing to go that far.
First, the court went into some trademark theory, which I think it got wrong (though there are ways to reach this result that make sense to me).  Plaintiffs’ literal falsity claim, it determined, depended on the marks at issue being generic rather than merely descriptive marks.  Merely descriptive marks can be protected with sufficient evidence of secondary meaning.  (At which point they aren’t merely descriptive; secondary meaning negates the “merely” part of that.)  Thus, descriptiveness isn’t inherently inconsistent with trademark protection; the proponent just has a more difficult road. 
As a result, even when the USPTO or a court finds that a party failed to establish secondary meaning, “it would not be fair to say that his assertion of the trademark via the TM was a literally false statement. The problem may be one simply of failed proof at the time that the claim is pressed. Therefore, if a trademark symbol is ever to constitute a literally false statement grounded solely on the classification to which the mark belongs, it could only arise with respect to generic marks.” 
Comment: This is what I think is weird.  Where a party invalidates a registration—as occurred with a number of the marks here—the court has misdescribed the burdens; the successful challenger has shown that the word at issue is not a mark.  Maybe in some other world, a world in which it has secondary meaning, it could be, but it’s not in this one.  We ask courts to evaluate the state of the world to judge falsity, not its potential states.  One could argue that we should “channel” various disputes into TM instead of false advertising, but I’m not sure why in this case, though an implicit theory requiring fault rather than error—not usually a requirement in false advertising cases, but implied in Lanham Act false advertising cases brought on the basis of false patent claims—might explain what the court does.  Perhaps the broader class of cases making it difficult to maintain false advertising claims against “false” statements about the law is more helpful: courts usually require intent in those cases because the law can be a tricky thing.
Anyway, the court then goes on to address generic terms: even though they’re never marks, “it is quite another matter to conclude that a competitor's use of a trademark symbol with a generic mark constitutes a literally false statement of fact.”  First, food flavor names can’t be inherently generic because the PTO was willing to register several of the flavors at issue.  But the real problem with plaintiffs’ theory was that classifying marks “is really somewhat of a subjective exercise, particularly at the descriptive/generic end of the category continuum where flavor names are often found…. It is hard to escape the conclusion that in the lower end of the categorization continuum some marks could be classified as either descriptive or generic depending upon who does the classifying. Classification is by no means an exact science.”  Thus, for example, five federal judges have ruled that “chocolate fudge” is descriptive for soda, while three found it generic.  “For Plaintiffs' theory to succeed, a mark would either have to be generic or not but the Canfield decisions demonstrate the fallacy of this otherwise axiomatic statement.”  Genericity can also vary over time, and by context (generic for one product, descriptive for another). “Clearly, genericness is not an intrinsic, immutable characteristic of a mark.”
There is no definitive answer to whether a term is generic or descriptive.  Any such determination “will necessarily come after the fact, meaning after a party has been the subject of legal action challenging his use of the TM symbol, and after he has perhaps exposed himself to civil liability under the Lanham Act for unfair competition.”  The court found the function and purpose of the TM symbol inconsistent with a scheme that would deter competitors from using the TM symbol to claim rights.  “Of course, the Canfield decisions demonstrate why a competitor could not even rely upon a favorable court ruling as to non-genericness when hoping to avoid potential liability for the cause of action that Plaintiffs urge in this case. Even a federal appellate court cannot bind courts located in another circuit.”
Thus, using a TM symbol with a term later found to be generic doesn’t constitute a literally false statement. 
“That said, the Court recognizes that some marks are so incredibly common and descriptive that it would be intellectually dishonest to believe that its categorization as generic vel non could ever be a close call. And it would seem that no party would ever be justified in believing that he could claim that mark as his own to the exclusion of others.”  The court identified peppermint and chocolate as perfect examples, and noted that plaintiffs would surely contend that SnoWizard’s claims to praline and dill pickle fell squarely into this category.  So, what if a party used “Vanilla TM”—literally false?
The court turned to the question of what “Vanilla TM” means: In order to figure out whether it was literally false, “it is necessary to ascertain exactly what information or statement can fairly be attributed to placing a ® or TM next to any product name.”  ® means registered, of course, and conveys to the customer and to other competitors that the trademark is in fact registered with the USPTO, but SnoWizard wasn’t accused of using ® on unregistered marks.
TM isn’t statutorily defined.  The USPTO’s website says “If you claim rights to use a mark, you may use the ‘TM’ (trademark) ... designation to alert the public to your claim of ownership of the mark, regardless of whether you have filed an application with the [ USPTO].”  SnoWizard’s expert Stacey Dogan “explains that businesses commonly use the TM symbol to put the public on notice that they have adopted a term as a trademark, with the hope that the mark will gain traction in the minds of the relevant public.”  Thus, the statement made by the symbol is essentially that the producer claims trademark rights to the term.  “Whether a party actually has a valid and enforceable trademark in the term is a wholly different question and one that a court will likely determine at some juncture.”  In other words, “TM” is not a performative utterance.  (But will they come when you do call for them?)
By contrast, the court found it plausible that it was literally false to use ® on an unregistered mark.  That’s a statement that can be empirically falsified.  Plaintiffs’ expert Glynn Lunney noted that the Copyright and Patent Acts provided penalties for false claims of copyright or patent protection.  Getting another significant detail wrong, the court commented, “[O]f course, a false claim that a work is copyrighted or patented, like a false claim that a mark is federally registered, is a statement that admits of empirical verification. A work, invention, or mark is either federally registered or it is not, and the owner's rights to use the relevant symbols are dictated by statute based on that status.”  Mistaken as to copyright, of course: since 1978, all original works are born copyrighted; registration is unnecessary, and not what’s indicated by the © in any event. 
Regardless, the court found that “TMs do not denote the same unequivocal information that a symbol like the ® does.”  Moreover, plaintiffs’ claims based on improper use of the ® didn’t fit within the scope of literal falsity.  Even if SnoWizard obtained its registration for ORCHID CREAM VANILLA® through fraud on the PTO, 15 U.S.C. § 1120 (which creates liability to any person injured by a fraudulently procured registration, and which plaintiffs were asserting as a separate cause of action) was the appropriate remedy.  Loss of registration on §1120 grounds (or any other grounds, as with some of the other names) didn’t retroactively turn ® into a literally false statement. 
Because “Vanilla TM” doesn’t make a statement that can be judged true or false in a way that admits of empirical verification, it might be an oxymoron but it’s not literally false.  Instead, it makes the statement that the party is claiming rights to the flavor name as a mark, “as preposterous as that assertion might be.”  It doesn’t mean (and therefore, the court took as given, does not inherently communicate to consumers) that the party has rights in the name, “or can ensure that the mark is capable of serving as a trademark or that the party could meet its burden of proof as to secondary meaning if a court were so indulgent so as to allow a party to try.”  Indeed, the court wasn’t even certain that a continued use of TM after rejection by the PTO or the courts would be literally false: some of plaintiffs’ claims were based on continued use by SnoWizard of TM on terms the court had already found generic.  While the court viewed SnoWizard’s chances of success on appeal as slim, there was still some chance that the court of appeals would view the terms as descriptive, allowing SnoWizard the opportunity to establish secondary meaning.
Plaintiffs argued that the parties’ customers weren’t familiar with trademark law and likely to overvalue IP claims, rather than appreciating the nuances teased out by the court.  But this didn’t matter to a literal falsity evaluation.  “The determination as to literal falsity cannot depend on who hears the statement and the meaning that the unsophisticated listener might attach. The intended audience may be relevant to determining the tendency of a statement to deceive but it is not relevant to determining whether a statement is in fact literally false.”  (Comment: But see cases finding literal falsity, or lack thereof, based on the circumstances that a statement was made to people in a particular field to whom it had field-specific meaning, e.g. “catastrophic failure” to medical practitioners or “sunscreen” to retailers.)
(Incidentally, here’s where plaintiffs’ insistence that the state law analysis was the same as the Lanham Act analysis is a potentially dispositive concession.  The state law is, the court says, more broadly written; I am willing to bet a fair amount of money that if there’s any precedent supporting an implicit/explicit falsity division in state law it comes only from other cases in which no party argued that the state law was different.  There is simply no reason that a state law must adopt a requirement that literally true but misleading claims have to be shown to mislead by survey evidence, and many reasons not to do so, including the historical basis of state unfair trade practices laws in (1) unfair competition law that didn’t make this distinction and (2) copying the FTCA, which doesn’t make this distinction.)
Turn now, therefore, to the possibility that the use of TM or even ® could be misleading.  “[E]vidence of some sort demonstrating that consumers were misled is necessary. Plaintiffs cannot simply rely on legal argument that appeals to the fact finder's intuition.”  But they didn’t have any such evidence.  “Plaintiffs have made a strategic business decision to proceed without customer input via surveys, etc., in order to avoid alienating their customers or discomfiting them by serving subpoenaes and other litigation-related materials.”  They therefore lacked evidence of materiality and actual or likely deception.  In the Fifth Circuit, market surveys are required to show these things.  The argument that unfamiliarity with trademark law caused customers to overvalue the TM was wholly speculative.  SnoWizard’s motion for summary judgment on the false advertising claims was granted.
Furthermore, plaintiffs challenged SNOSWEET,® SNOFREE,® and ZEPHYR on the grounds that SnoWizard hadn’t used them in commerce.  The court found that they weren’t entitled to judgement as a matter of law as to these marks, “particularly ZEPHYR which is not federally registered. Based on the explanation of the TM symbol contained in this opinion, the Court can see no basis to apply the use in commerce requirements that apply to federally-registered trademarks to ZEPHYR.” Comment: what? Rights come from use, and indeed only an ITU (federal) or an extension of protection for a foreign mark (federal again) can provide rights in the absence of use, and those only for limited periods. In an opinion so focused on what trademarks “are,” this is a whopper of an error, though if there’s evidence of use it may be harmless.

6 comments:

Mark Edw. Andrews said...

Thanks for this analysis. Southern Snow and Parasol, the plaintiffs here, and my clients, ended up joining with several other snowball business and filing civil action LAED 11-1499, asserting civil-RICO and antitrust claims in addition to Lanham-Act and unfair-competition claims. That action also asserts claims for invalidity and fraudulent obtaining of 2 patents. At least patents are obviously a property right, and their false assertion or misuse is recognized as being wrong. But with trademarks, this idea that there is nothing wrong with claiming 23 bogus trademarks is very frustrating.

Mark Edw. Andrews said...

The first of the consolidated cases was filed in Louisiana state court and asserted only state-law claims. The now-dismissed literal-falsity claims were an artefact, but they had the benefit of being covered by insurance, unlike all the fraud-based claims, which were the bulk of the 3 complaints. Most of the fraud-based claims are still in the cases.
There are cases in the LAED and 5th Circuit holding that federal and Louisiana unfair competition are analysed the same way by federal courts. That was not a concession, it was an honest identification of the caselaw.

Rebecca Tushnet said...

I don't doubt there are such cases; what I doubt is that they are founded on any analysis rather than on ease of disposing of two claims together. If you look at cases brought by consumers, where there's no coordinate Lanham Act claim, state laws are routinely interpreted differently (most obviously, that consumers have standing under them). One can construct an argument that the state laws should be interpreted differently when asserted by competitors than when by consumers so that in the former case the Lanham Act standards apply, but I'm not particularly persuaded by that, and I think courts' unwillingness to do a two-stage analysis has snarled and thwarted state laws as applied in competitor actions.

Mark Edw. Andrews said...

As a practical matter, getting a separate analysis of Louisiana unfair-competition law does not get you very far, because Louisiana requires "ascertainable loss", and actual damages are mighty hard to prove when the damage occurs incrementally every time a private businessperson makes a private business decision to place an order with the company that claims to have a bunch of trademarks on the flavor names. Also, half the circuits in Louisiana apply a 1-year peremption or statute-of-repose, even for undiscovered causes of action. Under the state's interpretation of the state law, you're going to get bounced even faster than in federal court.

Rebecca Tushnet said...

Those are significant barriers, no doubt--I would just prefer to see courts acknowledging the differences between the state and federal laws. I'm primarily concerned with the mission creep of the implicit/explicit falsity divide making a survey a prerequisite for success in any competitor action not founded on literal falsity; that's a bad idea that doesn't get less bad when it's imported from the Lanham Act to a state scheme that never contemplated such a distinction but instead uses FTCA-like language.

Mark Edw. Andrews said...

I agree with you, and I appreciate your analysis of this decision. (The La. law is indeed a mini-FTCA, watered down.) The consolidated cases are set for trial in January 2013, so any appeal of these interlocutory decisions is likely to be in 2013. More I cannot say, but if you've got caselaw or scholarly work that comes down on the proper side of this issue...