IPSC, day 2 session 2, short presentations

Kevin Hickey, Consent, Refusal, and Fair Use

Consent is irrelevant in black letter law, but consent and disapproval are not binary. Case law’s most extensive treatment: Letterese v. Scientology, 11^th Cir. 2008; use over decades, and then falling out between the parties.  DCt considered ordinary fair use factors + history of parties’ dealings.  11^th circuit rejects, saying consent is irrelevant.

Modeling communications between authors/copyright owners and users. Pre-use and post-use. Author faces choice before use: offer to license, ignore, etc.  After use: adds whether to sue.  Pre-use, user faces choice about using at all, trying to get permission.  Pre-use negotiation, all else being equal, should better trace the social value of the use. Pre-use the user can threaten not to make the use.  Post-use the user is locked in and faces statutory damages. When the communication costs are low (preexisting relationship) or low-cost mechanisms to opt out of a particular use, the model concludes that should matter.

Failure to consider consent allows opportunistic behavior by users and copyright owners. Users can use strategies to avoid detection even in bad faith.  Copyright owner has incentive to engage in wait-and-see strategy to take advantage of improved post-use bargaining situation.  When communication costs are low, we should encourage pre-use negotiation; users won’t be tricked into relying on acquiescence.

If consent isn’t binary, how dealt with? Quantitative dimension is explicit on both sides and silence in the middle. Qualitative is motivations for copyright owners’ and users’ behavior—good and bad faith. Timing of communication also matters.

Implicitly, courts consider consent in a number of ways.  Opt-out in Field v. Google, finding Google’s caching fair; talks a lot about Google’s provision of an opt-out, and that affects the analysis as does the copyright owner’s bad faith.  Mistakes made in iParadigms: 4^th Circuit found it fair based on Perfect 10 and Field v. Google etc.  But iParadigms’ system is not consent-based and individuals can’t opt out.  That should weigh against fair use and possibly swing the balance, and the court doesn’t even address it. 

Another category of implicit consent cases/partial consent: Worldwide Church of God.  Breakaway church wanted to use a text that the WWCG had rejected.  Estate of MLK v. CBS: MLK encouraged CBS to use the footage of his speech, and that should matter when the estate changes its policy and demands a huge licensing fee.

Q: catchall nonstatutory consideration.  What do you do about the fact that courts never really consider additional factors though they can or should?  Are these extras just about whether the judge liked the work or not?

A: true courts are hesitant.  Thinks that’s wrong and is trying to be convincing otherwise.  Moral notions can lead to more efficient outcomes.  Good faith/bad faith is part of his proposal.

Q: we do see the value of tolerated use and not pushing copyright owners to attack each use; doesn’t your proposal work against that?

A: most serious concern.  If consent is properly formulated, the concern over overactive policing isn’t great.  Analogy would be TM (where policing obligation isn’t anywhere near what many lawyers/brand owners claim or think as a way of shifting obligation). His standard would be case by case; tolerating one thing wouldn’t have any bearing on tolerating the next. Change in use would matter—commercializing the Harry Potter Encyclopedia.  Only estoppel would be things allowed for a very long time.

RT: I always worry about what happens when you add a factor; this is one that would ordinarily weigh against users in the cases that get to litigation, and that disturbs me.

I don’t think the description of not considering consent was strictly true of iParadigms: the court talks about the effect of minority and how the students had to use it to graduate and got a benefit from using the system.  Those are consent-related.

A: wants to be neutral for users.  Thinks iParadigms is a close fair use case. They’re unpublished works; the users are minors; there isn’t an optout; they’re using the whole thing; there’s a market harm (!).  Transformativeness is present, but he doesn’t think that’s enough.  (Which is why I’m nervous!)

Eva Subotnik, Intent to Fair Use

The law should formally accommodate those who can document that they tried/intended to make a fair use. Downstream practices by courts—how do they affect upstream creators? References to intent in farflung places, including Georgia State coursepack cases, where the court reviewed 75 claimed infringements and found 5 unfair.  Court concluded that GSU did try to comply with the Copyright Act, but didn’t perfectly predict the court’s results; but intent wasn’t relevant to infringement even though fair use is notoriously unpredictable.  This seems straightforward: either you complied with the law or you didn’t. 

But remaining on the fair use side isn’t like remaining under 65 mph. Fair use can be a difficult call.  Scholars have looked for patterns and groupings.  This paper attempts to close the circle by suggesting that in close cases where a seondary user looked for guidance that should be a thumb on the scale of the fair use defense. That can better guarantee breathing space.

Judge Leval argues for no role of intent.  Tony Reese points in a different direction—early courts did consider good faith in intent to abridge rather than supersede to draw lines in hard cases. Even Folsom makes a passing reference to intent: a reviewer can cite largely, if his design be really and truly use the passage for the purposes of fair and reasonable criticism.

While a bad faith standard might be a bad idea, giving weight to a good faith attempt to comply is utilitarian because it encourages calculated risks that benefit the rest of us.  What is good faith?  Usually discussions of good faith in the case law are actually about what bad faith is—knowing use of purloined work, obtaining copy under false pretenses, removing copyright information, or asking and being denied permission (which last Campbell says is not a problem).  Some of the Court’s language is “whether a parodic character may reasonably be perceived,” but other language talks about the reasons for the defendant’s choice.  Opening for subjective and objective evidence.  Prince v. Cariou doesn’t shut the door on subjective evidence entirely—Prince’s failure to explain himself might have lent strong support to his defense, but that’s not dispositive. The test is not simply what the defendant says.  Court even cites Prince’s deposition to show his drastically different approach.

Aspects on the spectrum: intent to comply, intent to transform (but not necessarily with copyright oriented sensibilities), absence of intent, contrary intent/intent merely to supplant.

A steadfast but wrong claim of fair use, or an attempt to comply with a mistaken understanding of the law, shouldn’t let D win fair use. But attempt to comply should count in D’s favor. One distinction: privilege statements tendered prior to litigation, though SCt has said proper labeling isn’t a prerequisite.  Seeking out legal advice/opinion letters should be favored. With less access to counsel, consulting various sources on law should also count.  Best Practices.  Seeking license to use should count in user’s favor.  Process of informing themselves should lead user to make more acceptable use to copyright owner.  Any distortion in use is natural and ok given that we have copyright.

RT: (1) Lack of knowledge.  People don’t know what they don’t know; don’t know fair use best practices exist.  Level of misinformation is so high.  MPAA distributes curriculum for classes.  People making remixes etc. are often 15—they don’t know they need to seek advice and they don’t know where to go.  TM doctrine: if you’re wealthy enough to conduct a survey and don’t, we weigh the absence of a survey against you.  I could much more easily get behind a standard that was more like that: if you’re wealthy enough to get legal review, then doing so favors you, contra GSU.  (2) I know in TM we have factors that never benefit the defendant. But can you have a factor that never benefits the plaintiff? If there was no good faith attempt, then won’t that weigh against the defendant?  Barton Beebe’s data may have some bearing here; maybe you can have such factors. (3) That’s an interesting spectrum, understandable but it prioritizes law and submission to law over artistic judgment, which may be fundamentally incompatible with the artistic projects of people who most likely be affected by this change, which again means the effects may be different.

Lunney: normative side—what is fair use for? How does intent adjust the balance?

Rosenblatt: implementation is a real barrier.

A: Current situation is asymmetric: the question only comes up when bad faith is addressed.  Even 9^th Circuit has mentioned good faith, though, and she’s trying to give content to that.

Mary Gani-Ikilama, An Analysis of the Effects of Copyright Law on the Creative Autonomy

of the Performing Author in the Nigerian Popular Music

Authors and performers are often the same person in Nigerian popular music.  Legal system inherited from Britain; part of Berne and TRIPS. 170 million people: big market. 15 years ago, American music was more popular, but there’s been a Nigerian renaissance. Growing commercial influence in the African diaspora as well.  At the moment, growing concerns about sameness of sound of Nigerian popular music.  Many complain that you can sing the verse of one song to the chorus of 10 others. 

Research: record label subjects overwhelmingly saw authorship as business measure versus authorship as self-expression. Business managers suggest style and content to artists; many had disputes with artists.  Performing authors overwhelmingly saw authorship as means as self expression and had often experienced pressure on commercial grounds.  Some of their favorites had been rejected.  Label: we define the artist when we choose to sign them; we don’t allow them to express themselves differently. But 90% of artists she interviewed either had or desired record deals because of the financial platforms. Developed world: possibility of independent promotion allows some people to work around the labels.  In Nigeria, internet access/electricity issues make this unviable.  Immediate effect is that music is homogenous.  Adorno: the problem of a technological theory of culture/mechanization.  Economically, reduces cultural output and revenue.  Ratio of songs written to songs released: 4:1.  Culturally: a gift that isn’t used will wither.

Copyright’s position: moral rights theoretically exist.  Producer still has edge over performer—though ownership initially vests with author, that can be changed by contract, and performers lack bargaining power.  Nigerians aren’t keen on litigation and there isn’t much case law.  Also business problems. Inconclusive restrictions on transfer of moral rights, which are negative/waivable rather than property rights; she’s seen it a lot.  360 deals are also used; the performer has to sing the songs that promote the record label.  Also there aren’t distinct payments for assignment of publishing rights and recording contracts.  Reversions aren’t recognized in Nigerian law.

Shyam Balganesh, Copyright and Good Faith Purchasers

Alienability is crucial to property.  Good faith purchaser for value doctrine: if a transfer from A to B is voidable/tainted (fraud, etc.) and before A learns of the taint B alienates the chattel to C.  This sale itself is clean and C lacks knowledge of the earlier taint.  C’s title in common law originally was void: nemo dat—you can’t give what you don’t have.  Problematic, because chattels pass through many hands; if everyone must worry about tracing chain of title back to untainted first sale, value of chattels will drop.  Voidable title doctrine developed: transfer to third party bona fide purchaser (BFP) in good faith and at arm’s length/for value gives perfect title to the BFP.  Codified in UCC and in many recordation statutes for land.  Avoids clogging title.  As between A & C, there is a sale we can reconstruct.  (Void title—stolen property—is different.)

What about with distribution right? Things are very different.  A makes an infringing copy and transfers it to B.  Assume that B had no knowledge that the copy was infringing and that it’s an arms-length transaction.  B gets imperfect title: under current law, B will be infringing if B sells the copy to C. Distribution right’s only mechanism of parsing ownership is first sale, but the very first sale of the copy has to be authorized and lawful.

Adopting a BFP doctrine would move the consequences of infringement from innocent purchasers to copyright owners; incentivizes copyright owners to publicize infringement and provide notice; variable determination from one purchaser to the next; avoids the need for innocent infringer doctrine.

Q: think about other requirements—for land, there’s often a registration requirement before taking advantage of BFP doctrine. Why is copyright so different?

A: doesn’t have a historical answer for why copyright differs.

James Grimmelmann: 3 other things of interest from UCC—the warranty of noninfringement/incentives for purchasers.  UCC’s distinguishing of void from voidable, going to care taken by copyright owner; are there differences that we could look at to see what the copyright owner has done to avoid voidable sales; rules for conflicting transfers from a common owner—could they be borrowed?

Andrew Gilden: how would this actually work?  Would this hypothetical sale be visible to the copyright owner?  Contributory infringement is also relevant.

A: mechanisms where you look whether the person has tried to make an effort to distinguish genuine from nongenuine goods—eBay sales.  Big brands refuse to validate whether a resale is valid or not.

[Adrian Johns has a lot of writing about the historical reasons for making distribution a right; though he doesn’t put it in those terms, the issue is the whack-a-mole quality of having to go after printers.  Also Wendy Gordon’s paper, which I’ll be presenting later, gets into the source of the distribution right: again, it was historically hard to go after manufacturers who could just disappear.]

Sarah Burstein, Costly Designs

We’re not necessarily worried about screening out bad art; we don’t mind giving copyright to really crappy paintings.  But bad design patents are a problem.  This week a patent was issued on an underwire for a bra. Not only is it functional, but also no one sees it in ordinary use, which seems to defy the point.  Partial designs are claimed; Apple uses continuations/divisions to claim little parts.  This allows immense scope for people smart enough to game the system. Also note that we’re not getting anything new: we already have the whole device, so if you care about incentives it’s not clear what extra the partial claim allows.  Patent on the bottom of a thought bubble icon.

Costly screens for patenting: what gets through are things with high private value, whether they have high or low social value.  Things with low private value get screened out; the question is whether there are designs with low private value and high social value that we’d lose, because they took a while to be successful.  Are there differences in industries?  Industrial designs v. fashion/experimental and discursive design?  How IP-sensitive are the areas that we want to encourage v. those that are just incidentally covered?  Costs can be a feature, not a flaw.

Sprigman: if you think design is mostly cheap to make, you might think that the prospect of getting there first and tying it to your TM is enough; a costly screen might be a way of saying that we don’t think design generally needs extra incentive.  Designs that are very expensive also tend to be highly functional—design of a jet airplane. What you’ll have left with costly screens + some functionality screen, that will whittle design patent down a lot. 

A: might be good.  In theory functionality is part of the examination, but not so much in practice.

Collins: hard to manipulate utility patent standard to get rid of the many patents, one object issue.  Is the partial design patent problem something you could address with a costly screen, or does that require some other fix?

A: doesn’t think there’s no role for partial claims, but doesn’t like claiming every tiny bit.  The next step in the project is to figure out what should cost more—so maybe every continuation and every divisional should be more expensive.