IPSC Breakout Session 2 Copyright Fixation & Subject Matter

Culture Isn’t Transitory: The Disappearance of Music and Film Under the Copyright Regime Amanda M. Whorton & David S. Levine

How could © improve cultural heritage preservation? Modestly change fixation to help archivists. We have only one video recording of jazz great Clifford Brown playing the trumpet b/c of serendipity—Soupy Sales decided to record the broadcast. An estimated 75% of silent era films are lost; 1927-1950 an estimated 50% are lost; many early news broadcasts, the first 10 years of the Tonight Show, first televised presidential address have been lost. Some of the losses are technological challenges/storage media degrading/format incompatibility. But some are attributed to taking fixation as a given. Cartoon Network said that 1.2 seconds wasn’t enough for fixation, but the boundary is unclear.

Proposal: Raise the standard to archival- or preservation-level fixation, closer to permanence, as far as is reasonable. A work must be able to survive past author’s wish to exploit it. Should require author to certify they’ve done so. Would not retroactively restore already-lost works. Better align fixation with Constitution: not just protecting market value, but collective cultural memory and heritage.

Q: why tie to fixation and not just require preservation as separate element?

A: theoretical link.

Q: the tech didn’t used to be valuable; why force people to preserve things that aren’t worth preserving? Why not force archiving of emails in case they’re useful someday?

A: yes, we’re making authors care about something they otherwise wouldn’t care about. There’s nothing new under the sun, though, and if they’re availing themselves of the © system they need to deposit more than a penny’s worth. Bar for protection is low.

Peter Yu: Is this a Berne-noncompliant formality?

Q: is this still a problem v. 70 years ago? Does deposit do enough work?

A: obsolescence will still happen. Certainly deposit can solve some of this but that requires LOC to house all these.

RT: wouldn’t you have to separate the standards for protectability & infringement? Cartoon Network is an infringement case.

A: yes.

Q: example—Nintendo didn’t have to deposit anything, except for the Pokemon movie reels as having cultural significance (they only wanted reels). Valancourt Books case about mandatory deposit as a taking—if yours is tied to seeking © that might escape the problem.

Ambient Copyright Fixation, Brian Downing

Fixation requires the author’s authority over the fixation; occurs more and more w/o author’s knowledge, let alone authority, by ambient recording devices. Creators can’t assert federal © interest over unknown fixation; uncreative device operators reap the rewards from others’ creative works. Operators own the work if they show minimal creativity in fixation.

He proposes notice and adoption as the rule instead of fixation with authority. Authors will use platforms to automatically become aware their work is online: YT and Meta have likeness protection for deepfakes; authors could also manually discover their work is online. Adoption: authors would adopt or reject the fixation. For the fixer, safe harbor, fair use, and news exceptions. Uploader would have to say who is in the video, if they know.

RT: Interesting project. Next problem: What’s the work? You’re assuming that human action creates works. But what about playing with a dog? Under your theory, who is the author of the Zapruder film? What if 2 people are in conversation? What if you are recording a dozen couples on the dancefloor? What is the uploader supposed to do if they believe that the underlying conduct is not copyrightable?

Also: why doesn’t common-law © solve your problem?

Also: Facial recognition mandates are a bad idea regardless of whether they’re supposedly in service of IP rights. The current mechanisms you describe are not used to mandate identification of everyone in a video, nor does Content ID etc perform a fair use or newsworthiness analysis, nor does anyone think that it can do so.

Q: you can make bad © claims on YT today; there’s always ambiguity about the defaults. Most things like a scuffle between people should be left up; a speech is clearer about the underlying work being recorded—notice and counternotice are the right solutions there; right now all the value goes to the wrong person. Failure mode is claim made by person who got in a fight and was recorded. That’s shifting value to a different person who abused the © system, but the value wasn’t created by the recorder and the money is being made. Our fixation rule should address how to reward the participants.

Also common-law copyright is underdeveloped in most states. [That’s what plaintiff’s lawyers are for!]

Maggie Chon: who’s the author? With photos we have doctrine. Operator may be able to claim that their filming meets some kind of test of copyrightability.

A: if there’s no master mind then revenue should be shared.

Peter Yu: 1101?

A: it’s at least ambiguous whether the Writings requirement requires a fixation. UK does allow adoption of unauthorized fixations. But 1101 is good for infringement; it’s not so good where the bootlegger is an automated camera.

The Copyrightability of Living Organisms Cathay Smith

The GloFish: glows under fluorescent light: proteins from jellyfish, sea coral integrated into fish genomes. Living organisms are patentable subject matter, though products of nature aren’t. Patents exist on method of making fish as well as the transgenic ornamental fish themselves. TM also allows for living organisms to be considered goods, so GloFish has a number of registrations like GALACTIC PURPLE and STARFIRE RED.

Copyright Review Board has found lack of copyrightability b/c didn’t owe origin to human authorship or don’t meet fixation requirements. GloFish © was denied despite argument that injecting non-native DNA into GloFish was like a painter using paint on a canvas. CRB found no authorship and no copyrightable subject matter.

Considers doctrinal limits on © protection for living organisms, and policy considerations.

Humans also use living organisms as the medium—plants and flowers; microbial art; Chapman Kelly’s garden. Living organisms can also be used as canvas: tattoos/painting on pigs, hairless cats, cockroaches. Seems like easy PGS separability cases, but are they useful articles? What makes them art is that they’re tattooed on living organisms.

Work itself is the living organism: the GloFish, where the work can’t be separated from the organism. Is there a difference b/t injecting dye into a fish to change its color versus modifying its genes to do so?

Fuzzy categories: trees trained to grow with specific trunk patterns; topiary sculptures that are trimmed—are they the same? Should the process matter to ©ability? Do we want to look at design of/design on concepts or seperability?

Policy considerations: fish have offspring; if their offspring exhibit the same expression as they do, how do we consider right to reproduce under those circumstances? If the work is self-replicating, what then? Taxidermied mice—if the policy considerations push us to avoiding living organisms as © subject matter, what about when they’re dead?

Jacob Noti-Victor: there are multiple doctrines at play: idea/expression; authorship; functionality—disentangle different doctrinal pieces. AI authorship is also relevant here.

RT: Extreme plastic surgery on humans?

Is Copyright a Noun or a Verb? Jacob Noti-Victor (with Jeanne Fromer)

Allen v. Perlmutter, D. Colo.—cited Star Athletica to argue that CO wrongly considered the process instead of the output. Copyright focuses on the verb in certain contexts—tort or agency contexts—copy/copying in fact, perform/performance/transmission, cause/volition, induce/inducement/secondary liability, employ/WFH. But protectability is all about the noun—things are supposed to be legible in the work itself, rather than the process by which the work is made. This falls apart in different places, but the work is supposed to be a coherent thing (even if scope is in flux).

Why insist on the noun? Property: in rem rights generally need a defined thing. Lower information costs when transacting and suing. Normative policing: focusing on work rather than process allows commodification and marketing, reinforcing market-incentives theory; Feist in particular is about avoiding normative contagion from labor as an independent justification for ©. Relatedly: evidentiary—a work is more easily assessable by courts, juries, and the CO. Process narratives are expensive and easily contested.

This is unstable because © is primarily about creating. Our justifications are about the verb: incentives, labor, personality. But the law focuses on the work as coherent object. The verb side leaks back in to the analysis. A lot of verb but a suspect noun: Meshwerks; some photography cases where work went into producing the thing—that’s when process leaks back in. A coherent noun but a suspect verb—the monkey selfie where a human didn’t do it. Maybe fixation/intent to fix as well.

Doctrinal disciplining as a pattern: courts can’t use only noun language, so verb language starts to seep in, and then courts get uncomfortable and announce a rule expelling process from the doctrinal area. Thin works: Burrow-Giles to Rentmeester.

Useful articles: Brandir talks about artistic judgment; Star Athletica says no, it’s about how the article and feature are perceived, not how or why they were designed. Fair use seems like it’s about verbs—what the fair user did—but there’s a similar pattern in cases like Blanch v. Koons and then Warhol v. Goldsmith. Blanch asks for a “genuine creative rationale,” whereas Warhol says subjective intent doesn’t matter, though meaning as reasonably can be perceived should be considered to the extent necessary to determine purpose.

Can authorship ever be separated from verbs? How can you ID author from the work? Well, you can’t! Naruto, Kelly (goes back and forth b/t authorship and fixation); Urantia (divine authorship, kind of about estoppel); AI authorship as additional destabilizing factor.

Possibilities: hold the line: noun is shaky but necessary. Process is an evidentiary and notice nightmare. (2) embrace the verb; stop pretending process isn’t important even if it means simple photos/random creations are unprotectable; (3) no choice—with AI the work can’t perform evidentiary/notice functions; we need to know how it was made so process inquiry is coming whether we like it or not.

RT: Process can also expand rights: selection of a particularly attractive pepper at the garden; the price cases like CCC/Kapes. Consider also public domain works/talk about the Uncle Sam case.

A: for useful articles process might yield less protectability; for software it might lead to more protectability.

Grimmelmann: a paper about the appeal and limits of formalism—everything you need is included within the thing itself! Literary and artistic theory could offer useful comparator—formalism, reader response, etc.

Maggie Chon: joint works and WFH?

A: there’s no way to avoid process inquiries there—you have to ask who superintended the work?

Samuelson: in Sedlik, the testimony about process was really important to the jury’s verdict (though not to the court of appeals).

Identification! Or, How Do You Litigate Against 3,000 Squishmallows? Ari Lipsitz

What is the work? Squishmallows sued Build-a-Bear claiming trade dress in kawaii squishy characters. BAB response: they were mixing and matching 17 different descriptions and tried to claim Squishmallow Godzilla and Squishmallow Warren Buffett made for Berkshire Hathaway. But dct denied motion to dismiss. Clarifying which of the 3000+ Squishmallows falls within the definition was a permissible aim of discovery. P defines rights in ambiguous way and then plans to slice & dice claims in discovery to target whatever D did. But IP rights are supposed to be defined in the abstract—a problem for trade dress and trade secret as well.

Why identify? (1) notice to D; (2) gating discovery—California requires trade secret to be ID’d before discovery; (3) it straddles the line b/t procedure and substance—in trade secret, P should describe subject matter w/sufficient particularity, to separate it from matters of general knowledge.

Alsup said: it’s easy to allege theft of trade secrets w/vagueness, take discovery, and then specify whatever happens to be there as having been trade secrets stolen from P. Allowing everyone to survive MTD; risk of forcing D to reveal its own trade secrets.

Other forms of IP also lend themselves to strong identification requirements: trade dress can be hard to pin down; © also has identifiability issues if it’s unregistered or in a billion different pieces as in Thomson Reuters. Patents shouldn’t be hard to pin down.

Open question: identification and scope. The more diffuse the right, the stronger the identification should be—with trade secret, claimant may not have concrete idea of secret until there’s litigation, so identification is important; patent: you have it or you don’t, so need to identify. Unregistered trade dress and copyrights seem closer to the diffuse side.