Sunday, April 05, 2009

BC Law Review Symposium, panel 4

The First Amendment & The Construction of Copyright

Alfred Yen, Boston College Law School

Yen accepts Eldred, at least partially, but copyright law still chills speech. In particular, third-party liability poses particular risks because the incentives of the person doing the chilling are different. A speaker will fully weigh the value of her own speech—monetary or otherwise—in deciding whether to resist. But the intermediary has much less incentive to resist.

The expansion of indirect liability over time has made this a bigger deal—e.g., the expansion of vicarious liability from cases of direct financial benefit to indirect and diffuse financial benefit. Vicarious liability is strict—there’s nothing the defendant can do to avoid liability other than to censor another party. That makes it chilling. Contributory: similar issues. What constitutes knowledge for contributory liability? If a copyright owner informs eBay that copyright infringement is occurring on its network, is eBay now liable? That would be strict liability on the cheap. Inducement: another form of fault-based liability, here an intentional tort.

At the time of Sullivan, the common law of libel was also strict liability, including both compensatory and presumed damages (which can be awarded without proof). To the extent that vicarious liability is respondeat superior—responsibility for what your employees do—he has no problem with that. But the more expanded vicarious liability in copyright is constitutionally problematic. Sullivan and Gertz also tell us how contributory infringement should be construed. Notice of generic infringement on a network is constitutionally problematic, converting a fault-based cause of action to a strict liability cause of action. A plaintiff must at least show fault before claiming damages: the defendant needs to have behaved unreasonably under the circumstances, and failing to prevent all infringement isn’t unreasonable.

Following Gertz (which requires greater fault before presumed damages can be awarded than when baseline liability for actual damages in defamation can be awarded), to the extent that copyright allows presumed damages, they should be available in third-party cases only for reckless or intentional behavior.

Even if you don’t believe that this is a constitutional requirement, constitutional sensitivity counsels these reforms.

David Olson, Boston College Law School

Other changes in copyright: elimination of formalities, expansion of term—huge change in our lived experience of copyright law. The First Amendment interests accommodated by copyright law are no longer accommodated so well. The traditional contours analysis can show how our lived experience of copyright has changed.

So what counts as traditional contours? Kahle (certain formalities) and Golan (URAA restoration) (where a new ruling was just issued). The Berne Convention requires us to remove formalities. Golan: Taking material out of the public domain transgresses the traditional contours. The Tenth Circuit therefore sent it back to the district court for a First Amendment analysis; the district court applied intermediate scrutiny (both sides agreed that restoration was content-neutral and that the category of “foreign works” was too general to constitute a content-based category) and found the law insufficiently tailored. People who had used the works had reliance interests, and their First Amendment interests could have been protected while still complying with Berne.

Formalities in general might be a target—even though a generation has now grown up without formalities, it hasn’t been long enough to constitute a tradition. Then the question is whether compliance with the Berne Convention, along with equity and protecting authors from tiny mistakes, is sufficient government interest to justify the abandonment of formalities. The other rationales aren’t sufficient, but the Berne Convention is quite strong; the First Amendment interest might not outweigh it. But maybe we should do narrow tailoring for remedies for authors who haven’t complied with formalities: notice and registration. Also gives us a strong First Amendment interest in orphan works. Could have a requirement that a copyright owner comply with formalities, and if not the only remedy for infringement would be payment of a predetermined fee.

Moderator: Elizabeth Townsend Gard, Tulane University Law School: What would a case challenging formalities look like?

Olson: Ideally, you’d find people, maybe scholars/documentarians who want to make use of orphan works, can’t find the authors, and then could show a real impingement on their speech. File a declaratory judgment.

Yen: Who’s the defendant? Is the claim the statute is unconstitutional as applied?

Olson: The defendant is the US, and you could claim as-applied or facial unconstitutionality.

Gard: Can these two papers be put together?

Yen: He’s particularly concerned with third parties; first parties have incentives to assert their own rights. (Isn’t this a classic question of internalization—as Wendy Gordon noted long ago, fair users routinely don’t actually internalize all the benefits of their uses, for example, so are in similar positions to third parties in terms of incentives?) The First Amendment isn’t a cureall. But there might be room in circumvention and DRM to apply his analysis. He isn’t tying what he’s done here to traditional contours. He does highlight that it’s easy now to become an inadvertent infringer, which it wasn’t in 1909, where notice applied.

Me: (1) I’m now wondering about the role of knowledge/scienter/copyright myths among users—if we care about not imposing liability without fault on intermediaries, why shouldn’t we care about imposing liability without fault on initial speakers/copiers, who may think that copyright requires notice? Or are unconscious copiers? (2) Under Olson’s scheme, presumably the Supreme Court should affirm the Second Circuit case rejecting the settlement of the Tasini follow-on claims because the settlement can’t cover unregistered works.

Yen: The logic of his position probably does extend to individual liability without fault. For a later project!

Liu: Say more about what “traditional contours” means. Implicit in your article is a definition of traditional contours, but what could that possibly mean? The Copyright Act has changed so dramatically since 1789 in scope, works covered, rights granted, term, and lots of other ways (even registration and notice have changed hugely—you used to have to publish notice in a newspaper!). Why single out formalities?

Olson: one reason for traditional contours was a notion of judicial economy. Another: deference to Congress, which he expects the courts would continue to do. Courts will resist full cost-benefit analysis. Sometimes, though, balancing will be necessary.

Liu: Would you treat intermediaries the same regardless of how people are using them? E.g., is there a difference between Napster (which arguably had more limited speech value than the NYT) than YouTube (where there’s more likely to be speech value in what’s up there)?

Yen: thought about the project in terms of an ISP who had full speech rights of its own. Vicarious liability should be rare, and limited to employees; most cases should be contributory liability. (So, I guess, his answer is yes, treat them the same.)

The benefit of using contributory liability is to ask nuanced questions about whether a particular third-party defendant is in a position to take cheap and effective precautions against infringement without also wiping out a lot of valuable noninfringing speech. This is important because the market often doesn’t work that well in letting people held strictly liable for others’ speech pass their costs on and work out for themselves who should be doing the screening.

No comments: