Copyright Society panel on fair use (Warhol, Seuss, Oracle)

CSUSA Fair Use panel

Jennifer Pariser, MPA

Joseph Salvo, Sesame Workshop

Rebecca Tushnet

II.  Discussion of Three Notable Fair Use Cases Currently on Appeal

A.    Dr. Seuss Enterprises v. ComicMix, 372 F. Supp. 3d 1101 (S.D. Cal. 2019)Q&A:            The MPA and Sesame Workshop filed amicus briefs in support of Dr. Seuss Enterprises, and Professor Tushnet filed an amicus brief in support of Comic Mix.

Salvo: Sesame has done parodies in the past, but came in on DSE’s side here. Fair use is unpredictable, but there are a number of factors. This case in our view departed substantially from preexisting law. Court seemed to give a brand new rule for mashups. Obviated discussion of amount/substantiality of taking and market harm. Other thing was court looked past blatantly commercial nature of D’s work.  Also there was a lot of slavish copying: illustrator tried to go as close as possible, as did rhyming scheme. Finally, harm: fair use is an affirmative defense. District court also conflated derivative work transformativeness with fair use transformativeness, negating the derivative work right.

RT: 9th Circuit has already held that fair use is a defense, but not an affirmative defense. It negates infringement rather than excusing it according to the statute’s own terms. And since we are all textualists now we have to expect that the words of the statute will have a different effect than in the past when it gets to the Supreme Court.

Transformativeness is and should be a broad concept. Sesame is the best example of this. My own memory includes a Sesame book that just tosses off “isn’t it good, Norwegian wood” as monsters are eating furniture. That was just funny; it didn’t constitute criticism of the Beatles. Sesame Street’s own mashups also usually do a separate thing, like teach counting, rather than expressly comment on an adult show. That’s not different from Boldly Go, which contrasts Star Trek’s communality to The Places You’ll Go’s individualism (something that was recognized in the literature long before Boldly Go; David Brooks even has a column about it, believe it or not).

Commerciality: © plaintiffs are hoist on their own petard here. Having convinced courts that everything that gets litigated is commercial use, they have generated the doctrine that says that commerciality isn’t super important to most fair use cases.

Copying of each illustration v. copying a substantial amount of a work. This also comes up in Google v. Oracle: if © owner can slice its works into a large number of microworks, it can manipulate and distort the analysis, not just for fair use but for substantial similarity.

Why factor three rarely matters in transformative purpose cases: it doesn’t matter to meaning and it doesn’t matter to markets. If there’s transformation, you can take what is artistically important to you: pride in your own work as an artist is a good enough reason to do a good job, if you are creating something with new meaning and purpose. And that’s why the court’s comparison to the Liebovitz case was such a good one: in that case there was artistically exact copying of Annie Liebovitz’s photo in order to make the Naked Gun 33 1/3 poster a better poster. But because of the transformation of meaning and lack of impact on the market, the exactness of the copying didn’t change the outcome.

DSE had not licensed mashups; that’s a misdescription of the record, which showed only that the Grinch had appeared as a playable character in a game called Panda Pop; he didn’t interact with or have distinct powers in that game; he was basically a “skin” for the player character.

B.    Andy Warhol Foundation for Visual Arts v. Goldsmith (S.D.N.Y 2019)

Q: Professor Tushnet, can you speak to this idea as it relates to the Warhol v. Goldsmith case and your amicus brief specifically, and also as it relates to any other recent cases in which the focus may or may not have been, incorrectly, on fair use? 

The core argument of our brief: No matter what its purpose or effect on the market, a work that is not substantially similar in its use of copyrightable elements from the original simply does not infringe. Holding such instances to be fair use may mitigate the damage, but it makes the fair use doctrine more complicated and implicitly suggests that the defendants in such cases did take enough to infringe, which encourages future plaintiffs to try their luck with aggressive copyright claims.

Here, the overlapping visual elements between the two works stemmed from the appearance of Prince’s face and the nature of a close up image of a face; those are unprotectable, so similarity in only those things isn’t substantial similarity.

Fair use is intellectually more fun and it’s easier to have a discussion about bigger principles, but it’s not great to have it as the only tool in the basket. Similar cases include a case about a William Faulkner quote, when the Faulkner estate sued over a Woody Allen movie that used part of the quote, as well as the Seventh Circuit case of Kienitz v. Sconnie Nation. If we paid more attention to substantial similarity, we wouldn’t have as many hard fair use cases.

Pariser: agree that where that’s true, it’s bad to assess the issue on fair use grounds, but disagrees that’s the case in Warhol. Substantial similarity is usually discussed when there hasn’t been straight duplication of the senior work: instead, reproduced elements of the senior user, not made a straight copy on the photocopy machine. Tremendous burden to overcome the idea you didn’t take very much when you made a literal copy. Decision relies too much on Warhol’s popularity/fame as transforming the meaning of the picture.

On whom is the burden of showing market harm (also comes up in Seuss): a challenge in all fair use cases. The facts were too weak for Goldsmith on the fourth factor here.

Salvo: Fair use is fact sensitive. The fact that it was Warhol seems to have been magic/talismanic protection. © owners have the right to say no to licensing. As a brand owner, when you choose to associate w/another brand is important.

[RT: Things I didn’t have time to say: The switch to brand language is super important as a signal of an underlying issue that needs resolution: how much will ©/TM be separated, and will TM considerations count in © cases? Dastar is an important case here. Completely agree that relying on Warhol’s fame is problematic. As to literal copying, the doctrine is very clear that copying in fact is ok if you don’t create a substantially similar work. Even if Warhol did make a photocopy, which is unclear from the record, the things w/in the statute of limitations are not substantially similar. A full face shot is not itself creative and the way his face looks is a fact; the record unsurprisingly contains a number of photos of Prince staring at the camera in a hard to distinguish way.]

C.    Google v. Oracle, (N.D. Cal. and Fed. Cir. 2012-present)

Q&A:  For this discussion, we will leave aside the important question of the protectability of the APIs and focus on the question of fair use, but specifically, whether it was correct for the Federal Circuit to overturn the jury’s fair use finding.  

Role of appellate court in assessing jury verdict in © case: Feltner: right to jury for all issues relevant to making an award of statutory damages. But that proves too much. Everything goes to the jury in making an award of statutory damages. Harper & Row: courts of appeal can review district court findings of fair use as a matter of law. A lot of courts have decided they can review all fair use findings de novo, but this has only rarely happened b/c many fair use cases are decided on sj. Rarity of overturning jury verdict doesn’t mean doing so is incorrect. If courts can assess fair use as a matter of law, they can review it de novo. Critics of the outcome point to the issue that the jury didn’t make any specific findings of fact; how can we review that de novo? But the obverse doesn’t work either: if there are no findings of fact by the jury, that doesn’t mean the jury gets more deference. Lack of findings should worry us more. Weird system in which juries make broad rulings entitled to some amount of deference. Legal/practical perspective: review the legal aspects de novo. Can disagree about Fed Cir’s legal analysis, but if we’re just zeroing in on de novo review, thinks their ability to do so did exist.

RT: We have a a lot of jurisprudence about this; could a reasonable jury applying the correct law have reached this verdict? Prediction, which I rarely make: Supreme Court is going to decide this on non-© principles; it will try to figure out whether fair use was equitable in the 18th century and make up a presentist answer, given that fair use is not like the 18th century analogues/ancestors. My unsatisfying answer: it’s a mixed Q of law & fact. Market harm: pretty factual, and the Fed Cir intervened in an unjustified way there. Transformativeness: pretty legal. Commerciality as it is presently understood: pretty legal.

Q: if fair use is highly fact dependent, then that seems relevant.

Salvo: Not a tech person. Fed Cir seemed bothered by commerciality and market harm. Of note: what passes as permissible in visuals might not for music; for documentaries v. print publishing. In music it may be 3-4 notes. Different standards for different industries make comparisons difficult.

Q: contrast Cat Not in the Hat: Commentary on something other than Seuss lost. Is it a different fact pattern or have litigants learned to argue differently?

Pariser: ComicMix doesn’t even cite Not in the Hat; shows weakness of analysis.

RT: Different fact pattern and litigants have learned to argue differently. Subsequent Green Day case is more relevant; earlier Cat in the Hat case argued only parody.

Q: how does the Warhol case reflect back on the earlier Shepherd Fairey/Hope case?

RT: same issues!

Salvo: Koons also lost some cases until he/his lawyers convinced the courts that his view of art was correct.

Q: Mashups of music?

Pariser: same issues; which is why mashups shouldn’t get special deference. If there’s no lyrics in a sound track, it’s harder to show transformativeness. Marrying two disparate things together isn’t necessarily fair.

Salvo: first sampling cases; courts very strict in finding use of preexisting musical works/recordings were infringement. Robust licensing regime evolved. For brands, licensing is the solution here.

[RT: (1) racial bias in music case outcome v. visual work outcome, not understanding the language in which Black musicians were speaking (Bleistein). (2) TM concepts/Dastar issues.]

Q: why should DSE’s failure to license mashups matter if other entities do license mashups?

Salvo: they shouldn’t. Right to control subsequent use in derivative markets is important. We curate our brands.

Q: does the codification of fair use preclude equitable fair use that requires witholding injunction/profits but requiring payment of reasonable royalty?

Salvo: Congress said what to do.

RT: Twice: Tasini & the Rear Window case—has been suggested; interesting Q why more lower courts haven’t taken the Court up on the invitation.

Q: how do you reverse jury findings w/o examination of jury instructions?

Pariser: you do need to do both. Fed Cir started w/the Q of whether the district court’s understanding of the law of fair use was correct; didn’t say it explicitly but that must mean whether the court gave the jury the right factors to consider. Could remand to do it again if they weren’t right to take another whack. [As the Q suggests, of course, the Fed Cir didn’t actually do that.]  Fed Cir must have at least implicitly considered whether a remand would be a cleaner way of skinning this particular cat.

Q: has transformative use gone too far?

Pariser: of course!

RT: of course not!

Salvo: depends, leans towards Pariser. Courts have seized on transformativeness as alpha and omega, and that’s not right.

Pariser: that was our amicus in Oracle: whatever else you do, Court, don’t reverse on transformativeness. Google’s articulation of transformativeness was that they had a new product that was great. Our brief argues that you have to look at whether the © work was used in a transformative way.