Sunday, February 10, 2008

Columbia fair use conference: transformative use pt. 1

Panel 1: What is “Transformative” Use?

Moderator: Jane Ginsburg: This concept used to be known as productive use, and it concerned whether there was a second author or merely “a facile use of the scissors.” Productive use was pooh-poohed in Sony, but resurrected as transformative use by Judge Leval and Campbell. It’s now the favored statement of the first factor, but the concept of critical commentary has been stretched.

Laura Heymann: Courts approach transformativeness from the wrong direction. They focus on the user as author, asking how much the defendant contributed and what sort of artistic process she used. Instead, they should ask how the reader engaged with the defendant’s use. The current focus leads to judicial skepticism when the defendant hasn’t added much, but readers recognize change more readily.

When we try to assess artistic meaning – as opposed to merit – we need to keep in mind that meaning is contextual and dependent on readers’ interpretations. Magritte’s ceci n’est pas une pipe is a key work here – Magritte is right that it’s not a pipe, but a representation of a pipe. He’s trying to disrupt the conventions of connection between representation and reality. There is no original in art itself, but always a reference to something else, and it’s up to the reader to decode and interpret. Campbell transformed Leval: from adding value/yielding new insights to adding new material to the work. This has encouraged lower courts to focus on what the second author added, less on the meaning of the work.

Campbell asked whether the character of a parody may reasonably be perceived. This is consistent with Bleistein, but also with the idea that we should not be focused on the second author but on the reader. Everything is transformative to some extent – there’s no originality in art – so we need other concepts to help draw lines. Her proposal: discursive communities form around particular works, engaging with and commenting on them. We should see if separate discursive communities form around two works; if so, then the second is sufficiently transformative for fair use. This is something like looking for secondary meaning/consumer reaction in trademark.

Discursive communities need not be formal scholarly communities – commenters on a blog post can count. We should look at critical reception, context (art world/museum reactions); second authors’ statements about the work, especially if those authors have stature (e.g., Andy Warhol). Accounting for critical reception allows us to include more appropriation art as transformative. Appropriation art provides a different context or frame and generates a separate interpretive community, as with Duchamp’s urinal and Cage’s 4’33” of silence.

Reader focus could explain the difference between Rogers v. Koons and Blanch v. Koons – there was a shift in the latter decision’s analysis to the interpretive response to Koons. Koons and his lawyers perhaps knew that they had to educate the court about the shift in meaning. They brought the court into the new interpretive community. (Comment: in my reading of the cases, Koons said the same thing in all of them; he was just believed in Blanch.)

CleanFlicks: wouldn’t be transformative, because viewers would perceive the edited films pretty much the same as the originals.

Another effect of her test: could help authors feel that they need not distort their works to ramp up the explicitly parodic elements.

Her desire: shift the inquiry away from courts and materials created for litigation to actual reception – don’t ask who’s speaking, but who’s listening.

Comments: Heymann’s paper made me wonder about the Google cases – the possibility of asking new questions about history or literature because of the ability to search through large quantities of texts might indicate that what Google is doing is creating a separate or new interpretive community than those that form around the individual component texts in the database. Dan Cohen’s Digital History blog routinely addresses such issues.

On the other hand, it seems to me that Heymann’s proposal has some difficulty with performance rights. This is not a unique problem – copyright theory in general has a rough time understanding performance. Songs and plays will generate substantially different interpretive communities depending on the interpreter – e.g., Eminem’s Bonnie & Clyde ’97 versus Tori Amos’s cover of the same song. (The right is one to compensation, not control, with respect to records of song covers, perhaps giving some leeway to transforming covers, but the right to control public performance of dramatic works is the standard control right.) As attractive as I find her theory, the existence of a performance right suggests that the presence of a new interpretive community doesn’t, as a descriptive matter, exhaust the scope of a copyright owner’s rights.

Still working on my notes – more soon.

Tony Reese: He examines the relationship between transformativeness and the derivative works right. Whenever he teaches Campbell, some students say it guts the derivative works right. And the CleanFlicks court seemed to hold that if a use isn’t transformative for fair use purposes, then no derivative work is created. So he analyzed how courts have treated transformation in recent cases.

There are two obvious kinds of transformation: (1) change in the content of a work; and (2) change in purpose from what the copyright owner did/intended. Both appear in Campbell.

A lot of uses that transform the work would fall within the derivative works right, so if we weigh any transformation of the work in favor of fair use, that could swallow the derivative works right. Very few cases expressly attend to the question of derivative works even when it’s clear that the defendant has created one, e.g., a translation. It’s usually impossible to tell whether the court thinks the derivative works right has been infringed, but it is possible to tell whether there’s been change within the four corners of the work.

A descriptive chart showing the possibilities and their outcomes:

Content transformed

Content not transformed

Transformative purpose

Favors fair use (e.g., Mattel v. Walking Mountain)

Nuñez, Bill Graham, Kelly, Perfect 10: favors fair use

No transformative purpose

Cat NOT in the Hat: disfavors fair use

Disfavors fair use (e.g., Worldwide Church of God)

The transformative purpose/no content transformed cases tend to be defendant victories, and make high-protectionists unhappy (like Paul Goldstein is unhappy). The transformed content/no transformative purpose cases tend to be plaintiff victories, and make low-protectionists unhappy.

What should we make of this focus on purpose? There’s no evidence in circuit court decisions that fair use is swallowing the derivative works right. The mere fact of preparing a derivative work doesn’t count in favor of fair use. Courts rarely even mention it.

The transformative purpose/no content transformed cases, however, may be a small subset category. They involve still visual images, whose content is often very hard to alter and which may need to be used in their entirety even for a transformative use. Quoting a review of a Grateful Dead concert tells you a lot about the entire review, whereas a few square inches of a Grateful Dead poster doesn’t tell you much about the poster; a Google text search excerpt tells you much more than 10% of an image result.

Courts should avoid the CleanFlicks view of the connection between fair-use transformativeness and the derivative works right. Asking whether the defendant prepared a derivative work would be a bad question. The boundaries are unclear: is a review with a 50-word quote a derivative work? Is a collage a derivative work of the photos it contains? (Comment: this echoes the orphan works problem of defining what works users should be able to continue using even if the owner reappears.) Infringement analysis often doesn’t require answering this question because of the overlap with the reproduction right. Courts shouldn’t buy trouble. Tying the two concepts together wouldn’t clarify our understanding of derivative works, especially if the inquiry is really about purpose. The court might be right in CleanFlicks that there’s no transformative purpose, but making an abridged version of a movie could still be a derivative work.

The circuit courts are doing the right thing on this aspect of transformativeness.


Unknown said...

These are great; they should put the video on the web, but short of that, these are a great substitute.

Anonymous said...

Thanks, Rebecca. I had thought there might be a question about covers in the Q&A, so I appreciate your raising it here. At the moment, I do tend to think that the compulsory license scheme is a recognition of the transformative nature of cover songs, but this (and performance rights re dramatic works) will certainly be an area for further thought in the final paper.