Session 3: Lessons from Boundary Problems
Introduction: Bill McGeveran
What’s on the other side of the boundary? It’s not the case that overlap alone for its own sake is problematic. Why is TM the bad guy here that has to give way? The value judgments are in play because TM’s structure is (1) less explicit about freedom to copy as a value, with the bargain expressed more explicitly with other regimes, and (2) less certain about what TM is for. Tidy boundaries are a means not an end, embodying value judgments.
We spent a lot of time on aesthetic functionality and Dastar. Other overlap issues were mentioned—what about TM extensively overlapping with other regimes when we let it do so? E.g., defamation and dilution. Celebrities in ads: federal and state infringement, federal and state dilution, false endorsement, defamation, and right of publicity are all possibilities. Maybe we don’t have the same concerns about freedom to copy the celebrity, but why not talk about our willingness to get rid of these overlapping rights (no int’l barriers to doing so in comparison)? Maybe we don’t think that the redundancies there are troubling—value clashes/inconsistencies. (That’s unlikely to be true! TM’s boundaries look a lot better than right of publicity’s, though.)
Boundaries also help you think about what the interior is.
Jessica Litman: goal is to improve results in real cases/prevent bad cases from being filed. Legal realism must creep in: if TM owners can tell a free riding story, finders of fact will be sympathetic. (I wonder how inherent this is. There are some people whose efforts seem to be free to ride on, because of ideas about contract and (presumed) consent. I don’t think those have always been the same sets of people even in the past 100 years, so is there room for change?) Some best tools are not doctrinal but stories of overreaching and greed. The flyonthewall case: the court says “no further.”
International harmonization might help sell the idea of limiting §43(a) to false representations and not protecting unregistered marks; it’s not very doctrinally pure, though, and she’s grabbing at an available tool.
Primary Discussants: Michael Grynberg
Distinction between learning from another doctrine and blindly importing from another doctrine. Consider the regulatory nature of a lot of consumer protection law—often administered by an agency. One boundary in IP should be a greater appreciation of the power dispensed by courts/legislatures to private actors to exercise a kind of regulatory power. We have skepticism about this delegation in admin law; have evolved doctrines to ensure that they exercise it legitimately. You see some attempts to limit the scope—Tushnet’s written about prudential standing in false advertising—but the principle that we should be careful who we let muck around with the free market—is a sound one. This may also help explain the appeal of materiality.
Copying elements of trade dress/keywords as a way to communicate category similarity: I am like X. Mostly people in the room think that’s salutory, and if communication of similarity is false we can deal with it through false advertising. If you’re communicating “I’m a generic of drug X,” we might have more discomfort offloading that to false advertising but since we have the FDA behind this it’s not troubling. We are making an implicit judgment of the quality of an underlying regulatory regime in another area. If such cases are common, which they might or might not be, we have to consider the state of play in other areas of law (*cough*antitrust*cough*).
Mark Lemley: Tendency to interpret boundaries as “limits.” Boundary against what? It’s one thing to say TM stops and something else starts, and another to say TM just stops (and then perhaps the right to copy/compete begins). Articulating what’s on the other side of the border helps our theory of why we want to limit TM law.
Group of TM lawyers: no one actually questioned supremacy of patent/copyright in cases of conflict. Why? Could tell a constitutional story, but not clear that works given the Commerce Clause. Could tell a historical story: longer pedigree, though that too isn’t necessarily satisfying. Could tell an earlier vesting story: you don’t get TM until well after whatever actions would have vested a patent/copyright, though after ITU it’s not clear that story works either. Maybe TM just has a less well articulated theory of the public domain—but he thinks patent and copyright don’t have them either, and to the extent articulated it’s been done so in a statutory vacuum.
We do presume TM dominates state and common law doctrines, and to the extent that unfair competition/misappropriation runs afoul of TM’s animating principles, TM wins out. Maybe a federalism story makes sense, but he’s been thinking about maybe a numerus clausus principle for legislation protecting IP. Maybe we need not one but some limited number of some legal regimes and 10 causes of action might not be the right number.
Starting to question whether we need 2 functionality doctrines. You can articulate the rationale/desirability of functionality in a single unified way, which might help a.f. not be treated as the poor stepchild. Part of that articulation might not actually be most desirable as a boundary condition, channeling patent/copyright and TM, but instead as an intrinsic limit. Boundary conditions make the implicit assumption “there’s got to be some law out there that protects this thing,” but that’s corrosive to the idea that market competition is the default unless there’s some significant market failure. Right to copy as a right against which we must bump up is needed if that’s the boundary.
Dogan: patent in particular tends to come in tension w/TM only in product configuration trade dress. Assumption that patent trumps comes from idea that TMs are about information, and there’s some widely shared skepticism about how much info is actually conveyed by design and how hard it is to convey that info by alternate means. Thus the benefit gained from TM feels thin relative to the costs of granting TM protection over things that could threaten patent values. False positive mistakes are costly and true positives are not all that beneficial.
Lemley: he gets that because the protection is at the fringe of TM anyway; but what about copyright?
Dogan: much harder in copyright. Had different instincts about copyrightable works created as logos v. use of Snoopy, created for other purposes, then transferred into TM by licensing.
Functionality convergence: attractive but comes with risks. It does seem like the right thing to do but Lemley thinks it will make a.f. more robust and it might actually water down Traffix functionality if courts have to define it. We think: any time a feature adds utility, it should be functional. That’s appealing, but as a practical matter courts have too much accumulated case law that it’s ok to pick an attractive feature as an identifier and that we should encourage people to choose attractive trade dress.
McKenna: you don’t have to make judgments about other regimes. You should say ‘this doesn’t belong in TM, and if some other system isn’t working then we should fix that system instead.’ Why TM gives way: it’s TM that has grown into patent’s space and not the other way around. It’s not that patentable subject matter has expanded, it’s encroachment into patent’s area. Lemley: copyright is tougher. McKenna: agreed. We have trouble figuring out how this would work in copyright. But one could imagine focusing on boundaries in copyright law—useful articles; limiting protection for logos (which used to be the case!). Staying out of each other’s space need not create a rank ordering. (Feminist me is skeptical of this claim.)
Grynberg: Litman says courts don’t necessarily care about working doctrine pure; Grynberg thinks that they can be attracted to doing so in appropriate circumstances. That’s a consideration in thinking of “this isn’t TM’s job” as an argument.
Kur: in logos/labels, copyright has grown into TM, but that’s not watertight.
McKenna: he’s just making a descriptive argument.
Kur: don’t build a hierarchy on that gut feeling.
TM law doesn’t protect something valuable as such/achievement as such. Everyone is free to copy the thing designated by the TM, but must use another sign to indicate origin. Thus the interference with first-order creation/competition is less. Only true insofar as acquisition of TM remains competition neutral in the sense of not giving competitive advantage to registration. General theory is that this is not the case because TM signs exist in infinite numbers and a different one isn’t competitively disadvantageous. But that’s not true—some signs are limited. Colors, descriptive designations, shapes. Implications: no registration for configuration without secondary meaning, contrary to current EU approach. (Discussion: there is not formally a post-sale confusion doctrine in the EU, but some suggest it is coming soon.)
Dinwoodie: before functionality was added to defenses to incontestability, US courts got around that by calling functionality descriptiveness.
Kur: “used as a mark” can also be deployed this way.
Bently: ECJ has ruled out descriptive use in this way in a case concerning toy replicas of cars (though did use “use”).
Barrett: Rosetta Stone district court was following that instinct, but used the wrong label. More sense to say use is a fair use. It’s the nature of d’s use that matters.
Dinwoodie: benefit is also that it doesn’t knock out p’s mark in its entirety.
McKenna: it doesn’t fit in any of these existing boxes—it’s all the same instinct, d’s using this for something other than source.
Dinwoodie: Kur suggests you can fit it in the box if you understand “descriptive” in a non-literal sense.
Kur: words and drawings can all be descriptive.
McGeveran: so the Louboutin red sole is describing red?
McKenna: if you can use descriptive this way, why not functionality?
Dinwoodie: Vornado uses descriptiveness this way.
McKenna: interesting that we want a harder meaning for functionality.
Dinwoodie: but that’s because of the consequence of invalidating the mark.
McKenna: that need not be the consequence! (This is my Louboutin argument.) Courts are the ones that created the consequence.
Bone: but we run up against the expectation that it would invalidate the mark. So why not shift over to fair use, which might also accomplish the result more effectively. True that we don’t have to make functionality work that way.
Lemley: but to use fair use we have to change the parameters of that doctrine too.
Dinwoodie: true, requires more expansive reading; but if courts are conceiving of product designs as descriptive, it’s not alien. Shakespear/Silstar was a classic example because they couldn’t use functionality as a statutory matter at the time.
Dogan: if there’s actual passing off, the court should be able to require labeling. A cleaner way of getting there than saying the thing is protectable and that people acting in good faith to copy it.
Litman: cleanest way is to say these uses don’t cause likely confusion.
Lemley: but we have to take that to trial.
Heymann: we want to say even if they do cause a little confusion it’s still fair use. Descriptive fair use is like that, use of generic terms with de facto secondary meaning, etc. Unify a bunch of different kinds of activity that are appropriate even if there is some confusion.
Janis: design patent law: there are two types of functionality—a line of cases that uses functionality to restrict scope, as well as to invalidate rights. So it’s conceivable to do that though it’s sub silentio.
Bone: why copyright ought to trump TM in certain respects: one intuition is legislation v. common law. We think of copyright fair use as being a fairly well crafted statute that tries to strike a balance—at least a legislative effort—between protection and copying. While the Lanham Act is so heavily common law; there may be less willingness to intrude in copyright. (But as Litman has pointed out, most of the expansion of copyright came from the courts too, so that is a story but it doesn’t track the cases.)
Boundaries are, in theory, defined in more general terms than other limits, which tend to be more case by case. Our boundary doctrines tend to be case-focused though. Some case-specific things may not be about boundaries.
Sheff: going back to distinction between info with value in itself and info with value that can be realized in other ways because it’s functioning as source indicator and there’s always another source indicator you can use. Does that mean that these distinctions should be embodied in limits on subject matter? Scope? Defenses? Just the remedies? Maybe the last is the most promising.
Grynberg’s point on consumer protection regulation: the interesting Q in that boundary is what conditions are necessary to get the administrative state to move on consumer protection, which may be industry specific. Industries in which TM is alive and well, but there’s some other uncorrected market failure requiring command and control. TM not only serves as info facilitator but can also serve to manipulate consumer decisions in ways that lead to undesirable results, and food and drugs are clear examples of that. Marketing can influence decisions even if not “unfair competition” in ways that lead to very high social costs. That might trigger the regulatory state. The boundary problem is identifying those industries/those conditions. Perhaps we need to say TM is not enough and we need additional measures.
Dinwoodie: note that in Australia Google got hit by the regulators for unfair competition in keyword ads. So we have to ask institutional questions of competence.
Both Dastar and Samara, the Court says, don’t worry there’s another regime out there to take care of abuse. This helps make design patent relevant—but are we comfortable that design patent is effective in the US?
Why prefer patent/copyright to TM? Maybe we need to kick stuff out of copyright where TM does a better job. Two examples of kicking copyright to design instead. Britain has a slightly strange rule that Kur now thinks invalid: restricts term of copyright once there’s an industrial application. If you look at conceptual separability: the idea was to push these claims into the design regime, not destroy them entirely.
Lemley: Omega is an obvious case—copyright is treading on TM’s territory.
RT: Actually, when it comes to tobacco and marketing unhealthy foods to kids, activists and increasingly regulators, though mainly not in the US, say TM should not be allowed. Plain packaging and ad restrictions are about pushing TM out of the realm of certain segments of the market.
My note re: functionality. McCarthy attempts to unify it too! He does it differently than watering down Traffix functionality, but by talking about market demand and pretending there’s no such thing as a.f. while still using the a.f. cases to define the scope of functionality. That’s worth noting as a possible outcome of unification as well.
McGeveran: will courts go for the broad definition of “utility.” We talk about what “courts” generally would be willing to do; the Supreme Court might well be willing to do this if there were proper briefing. If we had an impact litigator willing to try to fix TM, you could design the kind of case to address these problems. The SCt seems to have a very different view of freedom to copy than the circuit courts, especially the Fed. Cir. Doesn’t know why that is, but descriptively true, and it’s a prescription for anyone who actually wanted judicial reform.
Lemley: possible theory: SCt views its job as articulation of principles across cases, at least in theory. Lower cts view their job as resolving cases on facts in front of them. Latter makes you more susceptible to bad facts/free rider feelings. Clearly not always true, but harder to articulate “free riding bad” as a general principle.
McKenna: note that these cases are never close.
Janis: SCt also sees itself as the arbiter of boundaries—taking cases framed as boundary cases.
McKenna: and it sees patent cases and the 7th Circuit doesn’t.
McGeveran: boundary framing as a useful rhetorical tool in amicus briefs and so on?
Austin: we are compartmentalizing what we see as the SCt’s drive to enforce a right to copy from regulatory responses that occur outside the IP appellate decisions. (Also Golan, which can be seen as affirming such a regulatory response.)
McKenna: dangerous to keep things in TM because we think the other schemes aren’t working because of the collateral damage to TM.
Dinwoodie: there has to be some affirmative justification for applying TM—secondary meaning of a product design, for example. But since that also creates problems we can think about whether there are other better regimes.
Sheff: the problem is that allowing TM protection for harmful medicines/devices leads to bad results, which can be dealt with not by limiting TM protection but by setting up a collateral regime to address the bad results in the industry.
RT: I think there’s some ground here to think about the right of publicity/false endorsement, which we haven’t talked much about—courts are hopelessly confused about whether false endorsement is a matter of §43(a)(1)(A) or (a)(1)(B), and that’s no surprise.
Dinwoodie: but it’s picked up because there’s some justification in TM.
Dogan: bleed between false endorsement and the right of publicity. NCAA cases: you see the players saying ‘the NCAA is getting paid, so we should get paid too.’ That’s not without intuitive appeal.
McKenna: that’s the basic RoP theory.
Dogan: but it’s reinforced by the breadth of TM rights—if TM holders are getting paid, that’s juxtaposed to the unpaid use of the players’ images.
Lemley: but that’s contextual. Universities sell T-shirts and tickets, and no one looks at that and says the students have a RoP to share in those benefits.
Dogan: but it’s a self-reinforcing.
Lemley: and it’s about framing. In many other contexts we accept the asymmetry.
Heymann: you see bleed the other way too; in L’Oreal you see the 7th Circuit say the RoP requires false endorsement. (That’s TM trumping RoP, it seems to me, if we want to talk about supremacy.)
Bone: suppose a TM owner licenses a symbol, but it turns out that there’s no protectable mark. Is the contract still enforceable? (E.g., the Facebook ToS.)
McKenna: licensee estoppel?
Lemley: that was thrown out in patent law because of a policy judgment that we want people to challenge overbroad/invalid patents.
McKenna: hasn’t been accepted in TM cases.
Bone: Market solutions can come in to provide consumers with info, not just regulatory solutions.
McGeveran: a lot of RoP cases would be subject to nominative fair use; you’re reporting facts about Lew Alcindor; the court’s just not interested in that as a limit. Interesting that a regime on the other side of TM that doesn’t have baked into it any acceptance of freedom to copy, resulting in no pressure to manage the boundary between TM and RoP.
Mid-point summary: Eric Goldman
Absolutist: TM law is designed to fix only specific defects in a marketplace. Of course TM law does much more than that, and he’s not trying to go back to some halcyon time, but he does think that first principles can help. Institutional competence question is quite helpful. We know that competitor lawsuits are an essential part of an advertising regulatory system—competitors know what’s going on and are willing to sue to protect their interests, but there are also defects in relying on them because they don’t put consumers first. That’s why we have agencies designed to enforce for consumers where industries are more likely to collude/ignore falsity; that’s also why we have consumer lawsuits, foreign TM; then we have certification bodies/private solutions; and the market itself can correct some problems. TM fits into a broader advertising law space. Might be empowering to know TM doesn’t have to be everything to everyone.
He’s a big fan of tidiness for tidiness’ sake. And he thinks tidiness diminishes transaction costs.
Do we think it’s easier to manipulate TM law? Compare ease of modifying copyright’s originality requirement to exclude logos (and ease of dealing with Snoopy as advertising mechanism).
TM as regulatory mechanism to shape the behavior of potential TM owners when they select their marks. He is very keen on this notion. Producing behavior rather than merely reflecting it.
How TRIPS-compatible is US false advertising law, with our without Litman’s suggestion? Although Paris Convention art. 10bis is not very clear, might not be compatible. What would it be like without protection for unregistered marks? Huge explosion in application for registration? Difficulty policing marks, though will force ex ante specificity. But TM office won’t really look for aesthetic functionality ex ante, especially when many marks turn out to be completely valueless.
Litman: there’d be transition problems, as well as First Amendment problems from things like exclusion of disparaging marks from registration.
Bently: Registration has its own fears and costs. Registered product designs have chilling effects too.
Kur: has been threatened with suit for mentioning a TM as an example of a bad case—use for teaching purposes, noncommercial purposes, etc. should be thrown into discussion as explicit exceptions/limitations as a charter of users’ speech rights applicable across IP rights.
Lemley: the problem with regulatory agencies is that they tend to regulate. It’s what they do. Australian Competition Commission ruling on Google: wasn’t that supposed to promote competition? FTC having a false advertising mandate is fine as actually experienced, but there’s a real worry about bringing in a regulatory agency with concurrent/overlapping authority because the end effect is more restriction on freedom of competition.
RT: What we fear depends on where we stand. Don’t think anyone is suggesting roving TM enforcement, but FTC also does antitrust; maybe the enforcers should take harder looks at anticompetitive uses of TMs.
Heymann: thinking more concretely about overlaps/conflicts. I can perform Romeo & Juliet without permission, but not in the middle of 5th Avenue. That’s unobjectionable. Why are some overlaps more objectionable? Is there a higher order consideration about the First Amendment or the public domain?
Think about how we educate our students. The classic issue-spotter: here’s a set of facts; identify as many causes of action as you can. We’re training them to be plaintiff’s lawyers and cautious defense lawyers.
Goldman: You could say “what’s the best theory, what’s the best defense” and still have an issue-spotter. Never underestimate the costs to people who want to do the right thing in the market but can’t figure it out.
Some discussion of fee-shifting in cases where there are some successful claims and some unsuccessful, though Lemley notes that often success is all or nothing at least in patent cases.
McGeveran: multiple theories are a problem when they aren’t doing independent work. That doesn’t necessarily mean that every behavior should have only one legal theory that applies. It’s a criticism of needless duplication. But maybe sometimes there’s more than one troubling thing going on. So kitchen sink complaints are bad (fed infringement, state infringement, fed dilution, state dilution—though they tend to stand or fall together)—but we need some ability to say “several bad things happened.”
TM law is more than a time, place, or manner restriction, as in ‘no performances on 5th Avenue.’ If a TM in Popeye effectively constrains me from making any use I can share with other people, that’s a much bigger problem.
Dinwoodie: TM is affected by the intensity of the interest on the other side of the wall. Lessons for proposed short-term design laws: if the slightly watery conviction we have that there’s limited TM harm here, is our analysis different if we have a design law on the other side of TM instead of the free market?
Dogan: We don’t know! This is an empirical question of whether there’s a market failure in design/whether we’d incentivize more design, and whether the amount incentivized is worth the competitive price. European experiment strikes her as terrifying from a notice perspective—unregistered design rights for any design; you don’t know which feature is protected. (But if you could actually trade product design trade dress as per Litman’s suggestion for short-term design, incentives wouldn’t be the only consideration.)
Goldman: at this point, his energies are going to safe harbors and immunities cutting across doctrines that can protect against cumulation of rights for activities that we want to have happen. §512 fixes only one problem (maybe); §230’s brilliance is that it fixes all doctrines except for those specifically enumerated, so it doesn’t matter if you plead IIED, unjust enrichment, unfair competition, and tortious interference. That works really well.
Dogan: for the intermediary!
Goldman: fixing the transaction cost problem. We need a doctrine that cuts across all advertising law claims, including TM.
Litman: aren’t there costs for the woman who wants Yahoo! to take the bad information down? Many people would like to get intermediaries to remove stuff and might have reasons for that. Transaction costs of finding and going after the original poster go up for her.
Goldman: consider the right to forget. Has received threats based on reporting on the fact that a case was filed. The immunities/safe harbors cut through all this. The problem is not that the threatener is irrational, the problem is that there are so many claims he could at least theoretically allege (e.g., interference with employment opportunities).
McKenna: back to Heymann’s 5th Avenue example: one difference is the meaningfulness with which you’re able to use a work. Vornado says ‘you can use the design however you like, just not to cause confusion,’ but in that case the only reasonable use was as part of a fan. The promise was empty unless you can use it in a meaningful way. Dastar: there’s no meaningful right to use the work if you can’t use the work without attribution, because you’ll be sued if you do use it with attribution.
Grynberg: would we be willing to trade design protection in TM for a short term design right? If that bargain could be struck, we’d have a negative space issue—not knowing what the boundary between patent/copyright and TM is, we’d have the same questions about a design right. So you’d need a very explicit definition/negotiation/preemption provision (Dogan suggests election).