The case has an interesting procedural history. A man saw an ad for Victor’s Secret in a local paper, “was offended by what he perceived to be an attempt to use a reputable company’s trademark to promote the sale of unwholesome, tawdry merchandise,” and sent a copy of the ad to Victoria’s Secret. Litigation ensued, and the Supreme Court in 2006 determined that the lower courts had applied the incorrect standard of review in addressing whether Victoria’s Secret had been harmed. (Specifically, the Court held that only actual dilution by tarnishment was actionable under the Federal Trademark Dilution Act of 1995.)

While the case was still pending, Congress updated the trademark law to overrule the Supreme Court, adding a new claim under which a trademark owner could recover for likelihood of dilution by tarnishment. Under this new cause of action, the owner of a trademark can obtain an injunction or damages if it is likely that a new mark will harm the reputation of a senior mark, and the two marks are associated with each other due to their similarity.

The case was remanded to the district court, which held that it was indeed likely that people would associate Victor’s Secret with Victoria’s Secret due to their similarity, and that Victoria’s Secret’s reputation would be harmed as a result. The grounds for the ruling was that the association with Victor’s Secret was likely to cross that fine line between “sexy and playful” (as Victoria’s Secret described itself) and “sexually explicit” (as Victor’s Secret concededly was).

Yesterday the court of appeals affirmed, holding that “the [updated trademark] Act creates a kind of rebuttable presumption, or at least a very strong inference, that a new mark used to sell sex related products is likely to tarnish a famous mark if there is a clear semantic association between the two,” and that Victor’s Secret failed to rebut that presumption in this case. One judge concurred in the judgment but disagreed that the updated trademark act created a presumption — instead, he would call it simply an “inference.” The third judge dissented, observing that the updated act contains no language at all regarding any presumptions, and would have held that the single incident of association is not sufficient to create a cause of action.

Aside from the obvious problem of the Sixth Circuit creating a presumption where a statute contains none (actually, query whether it is indeed a presumption or an inference; how will district courts interpret this holding?), isn’t this a difficult decision to justify based simply on common sense? If Victoria’s Secret can obtain an injunction because of an association with sex, where will it end? Could a burger joint named McDuffy’s be enjoined if its food was too fatty, or its service too slow? This decision seems to create a slippery slope for courts with respect to the rights of junior mark holders.

On the other hand, perhaps this is a lesson in lawyering. Both the district court and the court of appeals based their decisions in part on the fact that Victor’s Secret failed to present any evidence that the two marks were not associated, or that Victoria’s Secret would not be harmed. Maybe the take-away in this case is simply that even when a claim appears absurd on its face, it still must be defended aggressively.

The Sixth Circuit decision is here [PDF]; the district court’s decision can be found here.

Victor’s Secret, for what it’s worth, plans to seek an en banc rehearing or to petition the Supreme Court for certiorari.

Instead, the prosecution witness — a police officer who was also “trained by several purse and jewelry companies to recognize their trademarks and to recognize counterfeit products” — merely testified that in his opinion the items seized were counterfeit. He also testified that through his “training and experience” he was aware that the trademarks were on the principal register of the USPTO, but he admitted that he had never seen any registration documents. A five-member majority of the court held that this testimony was insufficient to prove beyond a reasonable doubt that the marks were on the principal register. Although the majority doesn’t say so, in effect the holding is that the testimony about registration is hearsay–the companies who owned the marks told the officer that they were registered, and the officer told the court. When viewed in that light, it’s classic hearsay under Evid. Rule 801.

And yet, two justices dissented, arguing instead that the victims are most likely to know which marks were counterfeited, and the fact that the officer was trained by victims qualifies him as an expert with capacity to testify about the registration.

The dissent’s approach strikes me as somewhat similar to the practice rejected by the Supreme Court of the United States in Melendez-Diaz v. Massachusetts. In that case, the Court held that the prosecution may not introduce a forensic report at trial without making the report’s author available to testify (or to be otherwise cross-examined by the defendant). What the dissent in Triosi would allow is something of the complement to what was rejected in Melendez-Diaz — whereas Melendez-Diaz prohibited admitting the documentation without the foundation testimony of the witness, the Triosi dissent would allow the testimony of the witness without the foundation documents.

(For what it’s worth, there was some sentiment in Sixth Amendment circles that Melendez-Diaz, however young it may be, was not long for the world when Justice Sotomayor joined the Court; she was seen as a possible fifth vote to overturn Melendez-Diaz, and the case of Briscoe v. Virginia — which was already on the Court’s docket for OT09 when Justice Sotomayor took the bench — was seen as a possible vehicle to do so. Briscoe was fully briefed, but on January 25 of this year — just two weeks after argument — the Court issued a one-sentence per curiam decision vacating and remanding in light of Melendez-Diaz.)

I’m not familiar with European law on this subject, but in the United States, section 43(a) of the Lanham Act (15 U.S.C. 1125(a)) would control.  That section creates civil liability for

[a]ny person who . . . uses in commerce any word, term, name, symbol, or device . . . which . . . is likely to cause confusion, or to cause mistake, or to deceive as to the affiliation, connection, or association of such person with another person, or as to the origin, sponsorship, or approval of his or her goods, services, or commercial activities by another person . . . .

Under this “likelihood of confusion” test, a person commits trademark infringement whenever a mark is used in a manner which causes confusion as to the source of the goods, services, or commercial activities.  This would seem like a pretty clear-cut case for Portakabin against Primakabin — by purchasing the keyword from Google, Primakabin had users searching for Portakabin’s trademark directed to Primakabin.  That’s likely to cause confusion about the source of goods for the consumer who arrives at the Primakabin web site.  (Note that the test is not whether a user is likely to be confused by the web site.  The test is whether the consumer is likely to think that the goods, services, or commercial activities come from one source when they actually come from another.  Which is to say, there is a likelihood of confusion — indeed actual confusion — in the trademark sense if a consumer never even knows that he has arrived at Primakabin’s web site, and think he’s getting something from Portakabin the entire time.)

Perhaps the more interesting question would be whether Google has any liability under US law.  The Lanham Act does not expressly provide for contributory trademark infringement, but occasionally plaintiffs do succeed on such a theory.  Earlier this year, a California jury hit two web sites with a verdict of $32 million for contributing to the infringement of Louis Vuitton’s trademarks.  (In Europe, several cases of contributory infringement against eBay have been widely publicized.)

Also interesting is what this question does to the AdWords business model.  Google sells advertising space on its search result pages which is customized to the searched-for keywords.  If nobody can purchase a trademarked keyword, Google stands to lose any revenue it otherwise could have obtained from the sale.  (That is, if only the holder of the mark can purchase it, the purchase price will be low or zero.)