Cardozo A&ELJ symposium: Amy Adler Keynote and Warhol panel

Keynote Address, Amy Adler, (How) Should Courts Interpret the Meaning of Art?

Warhol was confused/confusing; lower courts are all over the place in response. Ushered in sea change in terms of how it was used in art cases. For years, those cases had determined fair use by focusing largely on whether the secondary work was transformative: new meaning or message. Some of the art stars of our day have been caught up repeatedly—Koons sued five times, Richard Prince recently settled two cases and has been sued five times as well. Cases answered transformativeness question in an undisciplined way; the SCt said the question shouldn’t have been so central to the inquiry, instead elevating commercial concerns and the “purpose” of the use over the questions of artistic contribution of the new work.

Nonetheless, courts are still in the fraught business of determining the meaning of art. First, the Court said so: meaning as reasonably perceived should be considered as necessary to determine the purpose of use. Second, the court revived the satire/parody distinction and placed renewed emphasis on whether a choice of material was “justified.” Both of these inquiries require courts to interpret the meaning of works.

The problem is that courts are catastrophically ill-suited for these kinds of inquiries, and Warhol not only failed to resolve those struggles of thirty years, but exacerbated them. The Court tells us what not to do: don’t evaluate the artistic significance of a particular work of art. Nor does the subjective intent of the user or subjective interpretation of court determine purpose of the use. But, the meaning as reasonably perceived should be considered. So: how?

Lower courts have taken different approaches over the years, and the choice of approach is often outcome determinative, but the Court doesn’t say what to do. Instead, it confidently interprets the meaning of the works and finds them the same—Warhol applied his characteristic “style” to bring out “a meaning that was already available” in the photo; his portrayal is “somewhat different[]” from Goldsmith’s.

But can we reduce visual art to meanings and messages at all? And if we could, what approach should courts use? Intent ruled out. Aesthetics seems to be playing a behind-the-scenes role. Is there a “reasonable perceiver” of art?

Warhol himself refuses the idea of assigning meaning to the work, mocking it by saying to ask his assistant, who “did a lot of my paintings.” Cariou v. Prince: Prince’s deposition testimony says he wasn’t trying to create anything with a new meaning or message. His message was: “He’s playing the guitar now. It looks like he’s playing the guitar. It looks as if he’s always played the guitar. That’s what my message was.” (Narrator: It does not look like he’s always played the guitar.)

Deeper problem: WJT Mitchell, “Whatever images are, ideas are something else.” Images are impenetrable to meaning. This causes trouble in law, particularly First Amendment law, where images revel in their multiplicity of meaning—proved fatal to artists in court, Serra v. US GSA. Serra was unable to identify a message conveyed by the sculpture, so there was no 1A problem in removing it. Incompatible w/language of meaning and messages.

SCt has acknowledged the difficulty of interpreting meaning—it frequently isn’t possible to identify a single message conveyed by a monument, especially when there’s no text, Pleasant Grove v. Summum.

Aesthetics doesn’t work for contemporary art, which is no longer something primarily to be looked at (Arthur Danto)—for better or for worse, probably for worse; Warhol was the hinge of the change. Danto: one feature of contemporary art that distinguishes it from art since 1400, which is that its primary ambitions are not aesthetic. Its concept can’t be accessed by looking at it. On Brillo Boxes: two objects look exactly alike, how is it possible for one of them to be a work of art and the other just an ordinary object? That’s the question Warhol asks.

David Smith dispute: masculine abstract expressionist, Cubi XII, April 7, 1963. 109 5/8 by 49 ¼ by 32 ¼ inches; Lauren Clay, No side to fall into, 2012 paper and acrylic, 18 ½ by 7 by 3 inches—triggered a C&D. There’s no way to explain what the Clay work does through aesthetics, and likewise no way to explain Sherrie Levine’s After Walker Evans. Conceptual work is beneath the surface of the work.

Fair use is not just a game for rich artists like Koons and Warhol, though they’re the reported cases; most artists can’t afford high-stakes litigation.

Court’s assumption that there is an objective meaning of a work. Reasonably perceived: by whom? Ordinary person (‘my kid could do this’ view of contemporary art); judge; art expert; consumer. Ahistoric ways of looking at Warhol (both majority and dissent)—de Kooning supposedly told Warhol: “you’re a killer of art. You’re a killer of beauty.” Warhol was attacking art! Visual differences, which dissent dwells on, are not the key.

Panel #1: Professor Jacob Noti-Victor moderating

Christopher Sprigman: how lawyers should think about fair use going forward: Warhol is only about factor 1. Don’t think of it as a test where you care about what facts go into what factors; SCt and lower courts stick things into different factors. The question is, under an equitable rule of reason, is the use fair? That’s a normative question ultimately.

2 innovations in the fair use analysis: (1) focus on particular uses. Hidden in ordinary case b/c the use complained of is the D’s use to create the D’s work. But ultimately Goldsmith didn’t challenge legality of works as works, a wise decision for litigation. In Cariou, the ordered destruction of the Prince works clearly mattered to the Second Circuit’s reversal. So attacking only the use on the Vanity Fair cover was targeted to get what G wanted.

(2) A transformative use is one that has a different purpose or character, and it didn’t perceive much of a difference, but mere difference is not enough. It has to be a difference that gives the work a different purpose or character. What does that mean? These are related to competition, says the Court. Central question relates to whether there is substitution—copyright’s bête noire. Achieve a purpose that is the same as or highly similar to the original work=more likely to substitute for or supplant the original work. “Magazine photos about Prince” is the market in which they compete.

So: when do “uses” compete? Incredibly difficult question from economic/antitrust perspective. The market is full of highly differentiated products, aesthetically. Not two bushels of wheat. With songs, or motion pictures, or any real IP good, consumer demand is very strange. If you’re a huge Taylor Swift fan, other singer-songwriters aren’t good substitutes. In antitrust, we look for small but substantial and nontransitory increases in price.

If price of coffee goes up 5-10% and it’s nontransitory, will you switch to tea? For many people, no: coffee does not compete in the same market with tea. To understand what competes w/coffee, we have to understand whether increased prices will redirect demand to another product. If no, then coffee is in a market of its own.

Applied to Warhol: if the Warhol license becomes 5-10% more expensive, will people switch to Goldsmith? Probably less likely than you think, so substitution is not often on the premises.

Antitrust lawyers now have something to teach copyright lawyers: an environment that might be promising for many fair use defenses. If we focus on particular uses, judges might be willing to find that a work is fair even if some uses of it might not be. And we may be able to teach judges that competition happens b/t differentiated products less than we might like think. There isn’t a market for “Prince images.”

[Q: how does the antitrust view handle the derivative works right?]

Peter Karol (similar presentation recapped here).

Michael Carroll: Disagree with Adler and Karol about how to read Warhol, because he has a much narrower view. Media industries and parties in Oracle and Warhol tried to repeal Campbell and failed. GvO, which analyzes all four factors, ratifies transformative use. Warhol is to the same effect. In context, these aren’t changes b/c Sotomayor is reacting to a straw person version of transformativeness created in the lower courts; because the Warhol Foundation labeled its argument “meaning or message,” the Court rejected the straw person “a tiny change in meaning or message is enough.” Of course that’s not enough; it never has been.

Disagree with these readings of Warhol as significant change; on the other hand, agree w/everything they’ve said about the work to be done going forward. New meaning or message is still part of the test, and that’s going to continue to be very challenging to courts in contemporary art cases. Similarly, whether a use is new or distinct under Campbell is a nuanced inquiry b/c asking whether it’s competitive or not were always implicit in Campbell.

The basic two questions: What is the purpose? A nuanced inquiry. And then: is the amount appropriate to that purpose? That is the ultimate test but it can be difficult to apply.

In risk analysis, you have always needed to analyze the use you’re making rather than all downstream uses. Why did licensing get elevated? It’s a pivot and also it’s the only act that took place within the statute of limitations.

In Warhol, industry/media wanted a ruling that would say that transformative use had “swallowed” the derivative work right, undermining incentives—prospect theory. The Court didn’t do that. Plenty of space left.

Victor-Noti: The Court did muddle factors one and four. What should we do going forward for predictability?

Sprigman: to keep substitution caged up. Most dct judges see very few © cases, whether or not they’re interested in the subject matter; so too with appellate judges, who see more. SCt is to blame for taking very few cases, period. Maybe we should have a German style SCt that works harder!

It’s on lawyers to fix this, and the Court has given us something to work with. What Campbell is really about is something we only learn many years afterwards b/c Court doesn’t take enough cases! Applying the competition rule to Cariou, btw, gets you the same result! There’s no competition in audiences, aesthetics, purposes. The new methodology is one that good lawyers can work with and how you can discern whether competition exists, b/c competition is at least in part empirical and economists/expert witnesses are now in the game. Antitrust has made the normative decision that 5-10% is the right amount; but © could have a different view depending on what we think about incentives.

Karol: for a lawyer, it’s very hard to deal with the answer “there’s no meaning.” Artists are still pushing back on notions of law. He’s more comfortable w/the inevitability that courts will make artistic decisions, as Christine Farley has written—better to have it overt and include experts.

Carroll: more disturbed by exploitation of subjects—artist Prince gave the value to those images; Cariou’s subjects were exploited by Richard Prince—give more active recognition to third party subjects. Then the fair use arguments get easier, where lots of expressive value is attributable to subjects; scope of photographer’s © is narrower. [Discussion of the ways in which Goldsmith’s contributions weren’t reproduced in the Warhol work]

Sprigman: AI doesn’t compete with any particular artist, they compete with every artist. Handled correctly, the model of competition is “copy a particular work to compete with it” which is not what is going on here. Competing with artists in general is not wrongful; the intuition that drove Warhol is not the same here. But that doesn’t mean that courts won’t consult their intuition. Photographs are competition from machines! Painters hated photography! The camera brought a lot of new people into producing art. Coming from antitrust, he thinks that competition is a good thing.

Karol: AI really shows the difference between ingestion and output. The one to one comparison in Warhol suggests that ingesting a ton of images to train AI is nothing like hanging art on a wall.

Carroll: the existing precedents are about classifier AI, not generative AI, but they seem applicability. The courts are going to feel pressure on the concept of “style” and protecting style in the outputs.

Q from Jessica Silbey: Kat von D won a trial, using arguments about differences between tattoo and photo, that we might notice w/o the artist pointing them out. The lack of substantial similarity offramp worked; that may be important going forward. (Warhol tried! I tried! The Second Circuit had no interest.) Hierarchy: derivative works trump fair use? Fair uses are slices off of derivative works? I don’t like the argument, but it they skipped over the question of whether the paintings were fair uses—seems like the question of whether they’re infringing derivative works is ongoing.

Carroll/Sprigman: there are no such things as transformative works—there’s a transformative use that is creating a new work. But you still have to check whether your use of the work is ok. (But where the use is commercial, I don’t think you should have to reanalyze for the next use!)

RT: how could Sprigman’s market approach handle derivative works rights?

A: has to be worked out. Will derivative works compete w/the underlying works? That depends on the underlying work; can’t say anything as a categorical matter. Look at derivative works along a similar competitive analysis—a derivative work that might compete in some reasonably defined market would be the author’s. But the analysis is not the same for factor four, which does consider derivative markets. We would also have to figure that out in factor four. Will Dorling Kindersley/there’s no right to control transformative markets survive? Thinks it should.

After Warhol, you can just say there’s no cross-elasticity of demand between Cariou and Prince; you don’t have to say anything about Beyonce and Jay-Z showing up at Prince’s shows. The challenge will be to do it right.

Karol: Fair use analysis needs to be very small to be workable for small creators; antitrust analysis is big and expensive.

Carroll: majority is clear that commercial and competitive isn’t dispositive—there could still be a justification. That leaves space for the dynamism of fair use.