Prof. Justin Hughes, Cardozo School of Law, Yeshiva University
September 22, 2011

Patents, copyrights OR trademarks – that’s easy.  Patents, copyrights, AND trademarks – that’s more difficult.  Vexing issues often arise when a party claims that the same thing or same interest is protected by more than one form of intellectual property.  Our general rule is that more than one form of intellectual property (IP) can apply to the same intangible.  That’s because the different forms of IP serve different purposes and have different eligibility criteria; often when two kinds of IP seem to protect the same thing, they actually protect different interests related to the intangible.

But there are situations in which we do not allow two forms of IP to protect the same intangible interest – and these are situations in which protection from one type of IP would frustrate core policy values of another type of IP.  When it comes to such conflicts it’s fair to say that trademark law is the intellectual property most likely to frustrate the policy objectives of the other two.  This is so for the simplest of reasons:  Copyright and patent are exclusive rights “for limited times,” while – to riff on a trademarked phrase – trademarks are forever.1

Patrolling the Patent/Trademark Frontier

The possibility that patent law’s objectives might be frustrated by trademark protection falls mainly (maybe completely) in the area of trade dress, where aspects of a product’s design can become indicators of source, the benchmark for protectable trademark status.  Preventing problems along this frontier falls to the “functionality” doctrine.  The functionality doctrine forbids trademark protection of any design aspect that makes the product easier to make, distribute, or use.  The rational is simple: If the product design “functions” in any of the ways I just described, exclusive rights for the commercial use of that design would confer a “non-reputational advantage” on the person who had the exclusivity.  Trademark law is about conferring narrow reputational monopolies in commercial communications.2

Patrolling the Copyright/Trademark Frontier

Most of the time the frontier between copyright and trademark is peaceful – and when there is any awkwardness, trademark principles often prevail.  For example, trademark law unquestionably protects slogans – short phrases of the sort to which black-letter copyright denies protection.  Same for titles: absolutely no protection under copyright, but protection is often claimed and sometimes deserved under trademark law.

As for images, there are some copyrighted images that undoubtedly serve as indicators of source – the Disney people use Mickey Mouse, Tinkerbell, and Buzz Lightyear this way.  Characters created for marketing certainly fall in this group – Tony the Tiger, Capt’n Crunch, and those quadrennially forgettable Olympic mascots come to mind (or don’t).  When any of these characters fall out of copyright, they should be able to enjoy all the protection offered by trademark law.  At the other extreme, there are tens of thousands of visual characters that will never merit trademark protection – minor characters and almost all villains in comics, films, and graphic novels.  (With the exception of an extraordinary villain like Darth Vader, The Joker, or Doctor Doom, who wants to use a bad guy as an indicator of source?)

And that goes to the best way to patrol the frontier between trademark and copyright when trademark is claimed to protect something that is expressly not protected by copyright: Courts must be rigorous in demanding that the plaintiff establish that it really has been using the visual element (or any original expression) as an indicator of source.  This is how courts have rightly turned back claims of trademark protection for public domain characters3 and publicly visible buildings.4   A recent example of this sort of problem is Fleischer Studios v. A.V.E.L.A. Inc.,5 a dispute over the character Betty Boop.  The Fleischer litigation initially produced a disastrously reasoned opinion in February 2011, which the U.S. Court of Appeals for the Ninth Circuit had the good sense to withdraw six months later.  But as we will see, the panel’s new decision may create its own problems.

Betty Boop Belongs to Whom?

Fleischer Studios was an entity formed by the family of Betty Boop’s creator, Max Fleischer, to regain and then exploit the rights to his iconic character.  Unfortunately for the Fleischer heirs, their recovery of the rights to Betty Boop may have been incomplete.6   Both the trial court and the court of appeals majority concluded that Fleischer had not established ownership of copyright rights to assert against A.V.E.L.A.  The trial court also concluded that Fleischer had not provided timely evidence that it owned either registered or common-law rights in Betty Boop as a trademark.

The appeal of the trademark ruling could have been decided easily on this single point – lack of evidence at trial of trademark rights, whether registered or through secondary meaning as an indicator of source.  Alternatively, the appellate court could have decided the matter on “whether the fractured ownership of the Betty Boop copyright precludes Fleischer from asserting a trademark claim.”7   As we will see, this may turn out to be the most important issue in the case:  If multiple, unrelated, uncoordinated parties held copyrights in different Betty Boop works – and if these various parties were exploiting these copyrights – then it would very likely follow that no one copyright owner could truly be using Betty Boop as an indicator of source.

But instead of a ruling on one of those grounds, Judge Clifford Wallace’s February opinion embarked on a mistaken “aesthetic functionality” analysis.  Claiming that all the other trademark arguments raised by the parties “were mooted by controlling precedent that neither party cited,” Wallace concluded that there could be no TM claim against the defendant because “A.V.E.L.A. is not using Betty Boop as a trademark, but instead as a functional product” and that the Betty Boop name and image “were functional aesthetic components of the [defendant’s] product, not trademarks.”  In short, A.V.E.L.A. was recognized to have an “aesthetic functionality” defense.

This part of the February decision was based on a 30-year-old (and much criticized) Ninth Circuit decision, International Order of Job’s Daughters v. Lindeburg & Co., 633 F.2d 912 (9th Cir. 1980).  In Job’s Daughters the defendant was selling jewelry with the insignia of many different fraternal organizations.  The court concluded that the defendant’s activities would not be mistaken as meaning the defendant had the plaintiff’s sponsorship or endorsement and, further, that vis-à-vis the defendant’s products “the name and emblem were functional aesthetic components of the product, not trademarks.  There could be, therefore, no infringement.”8

This second part of International Order of Job’s Daughters has been so roundly attacked over the years that it was reasonable to think the doctrinal mistake was, as one commentator recommended, “buried.”9   In the February Fleischer Studios opinion, this doctrinal error crept back from the crypt.  Although the panel withdrew this opinion in August – after a small hailstorm of criticism – it is important to understand how the panel’s misunderstanding of aesthetic functionality in its February opinion threatened to substantially undermine trademark rights.

Functionality Conditions Trademark Rights in Toto

The court’s most obvious mistake was to use the aesthetic functionality doctrine as a defense.  “Aesthetic” functionality arose as an off-shoot of regular functionality: just as a design feature could confer a manufacturing or consumer advantage that we would not want controlled through a trademark, some aesthetic element of a product or its packaging might confer a “non-reputational” advantage that we would not want protected by a trademark.  In a classic example, the color blue is aesthetically functional in relation to sleep products because blue is tranquil and reminds us of night – we would not want to give someone a trademark in that color for nighttime medicine.

A finding of functionality (aesthetic or otherwise) should be a complete bar to trademark protection over that element; functionality analysis goes to whether, given our public policy objectives, we want to protect plaintiff’s claimed interests at all.  Functionality doctrine has never been about whether the defendant is engaged in some privileged use; it’s not about how the defendant uses the features claimed as a trademark.

Just Another “Trademark Use” Mistake

Indeed, another way to see the mistake of the “aesthetic functionality defense,” in both the 1980 and 2011 Ninth Circuit decisions is that it is just a variant of the idea that there is a  “trademark use” requirement for infringement, i.e., that there is only liability if the defendant has used the protected names or symbols “in commerce” as trademarks.  After all, in his February opinion Judge Wallace was quite clear about this when he wrote: “A.V.E.L.A. is not using Betty Boop as a trademark, but instead as a functional product” (emphasis added).10

In recent years, this “trademark use” requirement has been argued strenuously by some defendants; it also has been advocated by some commentators11 and embraced in a few courts.   The problem is that there is nothing in the Lanham Act to support such a requirement: Neither Section 32 nor Section 43(a) mentions “trademark use” or “use as a designation of source” in their descriptions of what constitutes infringement.  As Professors Graeme Dinwoodie and Mark Janis have pointed out, reading a “trademark use” requirement into the statute seems to make a mush out of some of the statutory exceptions from liability.12   The Second Circuit recently – and I think rightly – rejected the “trademark use” requirement, concluding: “The gist of a Lanham Act violation” is simply unauthorized use that is likely to cause consumer confusion, deception, or mistake.13

Among variations on the “trademark use” defense, aesthetic functionality is particularly misguided because it would be so disruptive of modern trademark law.  The interlocking C’s of Chanel’s logo are aesthetically attractive – as are a whole variety of sports logos like the emblems for O’Neill, Quiksilver, and Rip Curl.  Capt’n Crunch and Buzz Lightyear certainly have aesthetic appeal to little boys.  In fact, if an aesthetic functionality defense comes into play whenever a trademarked image may also “be merchandised for its own intrinsic utility to consumers,” then a vast array of trademarks are thrown into jeopardy.

In particular, if that intrinsic utility includes “displaying emblems expressing allegiances” to “the schools we attend, … the sports teams we support, [and] the beverages we imbibe”14 then the defense wipes out much of the value of sports team logos and threatens all licensing of university names.  AsUSPTO Commissioner for Trademarks Deborah Cohn rightly noted: “[T]he logic of the [February] Betty Boop decision effectively undermine[d] the validity of federal registrations for trademarks that are used in an ornamental fashion, but are nonetheless proven source indicators.”15   Just five years ago, a different Ninth Circuit panel clearly understood the ramifications of any “defense” of aesthetic functionality: “Taken to its limits … this doctrine would permit a competitor to trade on any mark simply because there is some ‘aesthetic’ value to the mark that consumers desire.  This approach distorts both basic principles of trademark law and the doctrine of functionality in particular.”16

The New Opinion – and the Problem It May Create

The initial Ninth Circuit decision – filed on Feb. 23, 2011 – was withdrawn on Aug. 19.  The panel’s new opinion is completely silent on aesthetic functionality – meaning that the court of appeals needed a new basis to dispose of the trademark claims.

Earlier I said that it would very difficult to imagine common law trademark rights arising if the copyright ownership was fragmented, uncoordinated, “and if these various parties were exploiting these copyrights.”  The reissued court of appeals decision bore down on this point, concluding that (a) while the trial court had adequate evidence of the split ownership of the Betty Boop copyrights, (b) the trial court had not adequately determined that the copyrights were being exploited, i.e., that there really was fragmented, uncoordinated use of Betty Boop in the market.  Comparing the Fleischer Studios situation to a prior case involving fractured exploitation of a family of copyrighted works, the new opinion concludes:

From a logical standpoint, the mere fact of fractured ownership is not, by itself, conclusive evidence of a lack of secondary meaning.  There must be something more.  In Universal, that something more was the evidence of extensive use and licensing of similar images by other companies, the oddly defined trademark that was legally carved out from other images of the same character, the widespread confusion….

Unlike the facts in Universal, those at issue here merely show that more than one entity owns rights to use Betty Boop-related intellectual property….

Without the evidence of the type present in Universal – confusion in the marketplace regarding who owns the Betty Boop image, “extensive merchandising use of [other Betty Boop] images,” or other images of Betty Boop that have acquired secondary meaning – all we have is the possibility that other copyright owners may be destroying the secondary meaning in the Betty Boop mark.  That possibility is not enough to rule in favor of A.V.E.L.A. on summary judgment.  We thus agree with Fleischer that the district court erred on this issue.17

This reasoning seems prudent and yet raises a new problem for the frontier between copyrights and trademarks.  Judge Wallace’s opinion implies that Fleischer Studios could establish common-law trademark rights in Betty Boop despite the fragmented copyright ownership.  And that could mean that some owners of Betty Boop works who do not exploit their copyrights over a period of time might find themselves unable to commercialize their works because of Fleischer’s trademark rights.  In the rare cases when this occurs, it could pose a real dilemma; at a minimum this sort of reasoning could be particularly unfair for individual artists or heirs who inherit copyrighted materials and do not immediately exploit them.


Both of the Fleischer Studios opinions – February and August – offer us cautionary tales on policing the frontier between trademarks and copyrights.

The February opinion’s reliance on the discredited aesthetic functionality defense was undoubtedly motivated by an overbroad reading of the Supreme Court’s 2003 Dastar opinion.18  In Dastar, the Court concluded that a Lanham Section 43(a) action for reverse passing off could not be used for a work in the public domain in relation to that work’s authorial “source.”  After quoting at length from Dastar,19 the court admitted its underlying concern: “If we ruled that [defendant’s] depictions of Betty Boop infringed Fleischer’s trademarks, the Betty Boop character would essentially never enter the public domain.”20

But, again, the right answer to that issue is that courts must be vigilant in demanding proof of the fictional character’s use as a true designation of source.  When that actually happens – as when Toucan Sam adorns every box of Kellogg’s Froot Loops or Tinkerbell flies around the opening of every Disney film – the character deserves the full array of trademark protection and, yes, it might never fall into the public domain.  As Judge Wallace’s February opinion noted, the real concern for us – and for the court in Dastar – is when a trademark “action is essentially a substitute for a copyright infringement action.”  When the plaintiff is engaged in bona fide use of a character as a trademark, the infringement action is not a “substitute” for a copyright claim and there is no tension with copyright’s policy objectives.

The new Aug. 19, 2011 opinion in Fleischer Studios avoided any discussion of aesthetic functionality – and any limitations on trademark rights driven, expressly or implicitly, by concerns about copyright policy.  But the new opinion offers a dilemma in the opposite direction – a potential (although rare) situation in which new trademark rights developed through establishment of secondary meaning could prevent copyright owners from exploiting pre-existing works.  This possibility would arise when a family of related copyrighted works is owned by different people, some of whom exploit the works and some of whom do not.  The party that exploits its copyrighted works might able to establish secondary meaning in visual images from the works (such as characters), making it later difficult or impossible for the other copyright owners to bring their works back into the market.

Someday, some court may have to sort through this dilemma, although one possible way out – familiar to trademark students – would be to permit the exploitation of the copyrighted works owned by others with clear notice to the consumer as to the “source” of those works and a conscientious effort to ensure that no aspect of the marketing contributes to consumer confusion.21   That might not be ideal from the trademark perspective, but a company that develops secondary meaning in an image knowing that others have different intellectual property rights in that image would certainly be inviting that sort of compromise.