Friday, August 13, 2010

IPSC part 2

Senior IP Scholar Session

Paul Goldstein, Stanford Law School

A lot of the industry-specific provisions of copyright have no present justification and should be tossed (e.g., cable licensing); mature industry likes settled rules, but that’s not good enough.

Tom McCarthy, University of San Francisco School of Law

Big changes in scope of field—no such thing really as a TM lawyer when he started; first US TM organization (INTA predecessor) meeting he went to had 200 people.

Rob Merges, UC Berkeley

Big changes over his career: the Supreme Court and you (audience). General visibility of the field is greater; academic scope—patent folks could have fit in a small conference room or back of a cab in 1988.

Pam Samuelson: legislative changes/courts v. legislators?

McCarthy: INTA runs the show in TM. They set the schedule, select witnesses, etc. It’s an association of TM owners. There is no countervailing group; occasionally the ACLU sends someone over dilution. They didn’t want to have to prove dilution, so they got likely dilution, and the courts are giving them even more than that.

Goldstein: when he started, there was a minuet between Congress and courts—courts would defer to Congress; all very civilized with industry associations controlling the agenda, not quite as dramatically as INTA. Evolved into a mosh pit. Courts have moved towards transformative use: much larger concern in judiciary about the scope of copyright, attributable to overreaching by software owners trying to make copyright bear too much weight. Fallout from Eldred was the same: bad odor for copyright in courts, and Congress is now pretty much in stalemate. Academic work on lack of justification for extension has ground Congress to court.

Merges: Similar in patent law. Overall less legislation for various reasons. Rise of West Coast industries bringing to bear their economic clout: patent law used to be an insider game, but the inside was a narrow band geographically and industrially. 20-25 years: rise of computer/software industries and their different uses of patents have made big changes. Same stalemate—veto players—as with Goldstein, except it tends to be industry on industry instead of public interest. This ends up putting a lot of pressure on the courts. Individual cases get freighted with more importance than they intrinsically deserve; we reach a boiling point and wait for the grand solution, but that’s hard to offer in a single case (Bilski). Waiting for Godot, and he doesn’t think it will happen.

Merges: introduction of empirical perspectives into the scholarship is a big change. He sees greater, and productive, use of interdisciplinarity—digesting an outside literature and figuring out how to use it in IP. That’s the way to make our old arguments fresh.

Justin Hughes: third party liability—courts have moved in an opposite direction from their work on damages and transformativeness; they’ve really reached out to impose liability on third parties. What about the impact of internationalization?

McCarthy: patent lawyers were always international lawyers, filing all around the world; that was some extent true for TM too. It’s a problem because it’s dangerous for an American lawyer to opine about the situation in a French court.

Goldstein: An American music publisher has more in common with a German music publisher than with an American restaurant owner who wants to play music.

Second Breakout Session

Derivative Works

Pamela Samuelson, UC Berkeley Law School

The Quest for a Sound Conception of Copyright’s Derivative Work Right

Derivative work right is relatively new. Improvement, abridgement, translations were not within the scope of copyright. Wasn’t the result of literalism but of respect for the contribution of the second comer as well as the thought that there was no competition with the original market. Now, some conceptions of the right put everything as a derivative work. What are the possible limits?

We should look at the examples listed in the statute. Work based on one or more works, then 9 examples, then open-ended language at the end: any other form in which a work is transformed, recast or adapted. Statute ought to say “or other analogous uses,” to force us to think about what the copyright owner is entitled to control. Big contributor to incoherence of the law. We’ve confounded the derivative right with the reproduction right. How would infringement analysis look if we took the two different rights seriously?

Paul Goldstein tried to identify the marker: an original expression contributed by a second party operating in a different market segment; some attention to licensing markets. Castle Rock unmoored the derivative work right into anything the owner wants to license. Seinfeld creators weren’t thinking of trivia books down the line. (Though they are now!)

What to do? Deconstruct the definition and separate out the examples into different clusters. Open-ended language of the last clause should be construed in light of the examples. Focus on proximity and foreseeability of markets; scope of rights should also narrow over time. Clusters: things like translations where the author really wants to be able to choose who’s going to make the translation. Market proximity: a novel is likely to turn into a movie. Variations on the original that operate in the same or similar markets: abridgements, condensations, musical arrangements. If the thing at issue doesn’t fall into one of these clusters, then we should assume that it isn’t a derivative work. Sequels/prequels/continuations may fall within analogous uses. Courts have been suggesting that the owner doesn’t control every possible avenue.

First sale: big advocate of the principle that it’s ok to play with a purchased copy. Annotation does not infringe.

Should the right be narrower over time? Works are in a slow dialogue over time, and that’s a good thing. Construe right in light of larger purpose, which shrinks it over time.

Q: to what extent should the amount used matter? Girl Talk’s music, 100 samples. Less problematic than taking the whole work.

Samuelson: it matters, but she’s not that far into the analysis. If you see something in a different way because of its use by a different person, that’s important. Maybe we ought to allow a revenue stream for derivative uses, as the Supreme Court has been saying for a while. That is a payoff for separating the derivative right from reproduction.

Must derive some expression from the original to be derivative. Also fair use will be an important tool, but in some sense we don’t give the second comer’s expression enough weight in the fair use context when we cabin it to reproduction analysis.

Q: is foreseeability/proximity normative or descriptive? Now, trivia books are part of the expectation.

A: to the extent that markets arise over time and copyright owners start acting like they want to exploit them, we might want to take that into account.

Justin Hughes: fixation?

A: Paper will offer the array of different strategies people have had to cabin the right. Fixation is one. Argument for fixation is stronger when you take a look at the statute’s listed examples.

(Is her approach the approach of the RDR Books court? I have the same question as above: why isn’t everything foreseeable and proximate now, the way courts have said the internet makes all products/services share the same channels of communication?)

Chris Newman, George Mason

Identity, Accession, and the Law of Copyright

Diagnose doctrinal incoherence: we should stop talking about “protected expression.” Copyright subsists in “works of authorship.” They may be composed of protected expression, as things are composed of matter; rights should not attach to quanta of expression as such. Property rights are rights in things. Property rights are for discrete objects that have utility in satisfying needs/wants without interference. Right to exclude is a default rule to avoid swamping the system with the costs of coordinating noninterference—does not exclude others from deriving any type of value that can be derived from the owned object. One gathers acorns not to deprive others of the right to use (and if one did that would not be worthy of respect)—property rights exclude uses that might interfere with owners’ ability to assign use.

To calibrate right to exclude, need robust account of what the work of authorship is. It is a planned sensory experience designed by its author to give rise to a specific expressive experience in the mind of an intended audience. Doesn’t mean to include all possible meanings intended by author or socially constructed: means the basic stratum of perception—what anyone equipped to understand the work at all would understand as being present.

Utility of work of authorship lies in this experience. We want to incentivize authors to create to get these benefits. The right to realize exchange value based on demand for the expressive experience designed by the author. The relevant foreseeability is the foreseeable value of the expressive experience, and only secondarily to the specific technical means through which people access it. Author controls experience the work is designed to convey. As long as people are consuming work as she conceived it, doesn’t matter how people are perceiving it—in person, by projector, etc. Use should be regarded as unforeseeable when the benefits derived from the putative copy do not stem from the expressive experience the author designed to convey, but from second-order information or radically discontinuous expressive experiences.

A copy is a material object from which a work can be perceived: must reproduce the sensory experience designed by the author. Has to be complete verbatim or picture perfect copy or virtually identical. Substantial similarity should only be a test for whether there is a derivative work.

Properly speaking a derivative work is designed to convey substantially the same experience as an original work through a modification made to make the expression more desirable to a different audience. Condensations/abridgements (less onerous to process), translations, additions (extra information). Open-ended language does not cover all recasting—a work is recast etc. Has to be a different version of a work—need pervasive nonliteral similarity.

Have I written out fragmented literal similarity and overruled Harper & Row? Fragments of expression can be protected only to the extent they constitute works of authorship in their own right. Don’t just assume any fragment is a work in its own right just because it’s been copied. Does this fragment have any plausible utility on its own as a standalone work, outside the context of the larger work? Could the author reasonably have expected demand for/consumption of the fragment as a standalone work?

2-minute excerpt from the Daily Show: that’s a coherent expressive experience that would have been understood as such. Foreseeable that people would want a 2-minute clip even if they wouldn’t want to watch the entire episode. Thus we’d regard that as a work of authorship in its own right. Contrast that with clips used to make a mashup music video—only a small fraction of a dramatic proceeding. Each clip undoubtedly has expressive value on its own; it’s difficult to imagine anybody actually wishing to watch one of those clips in isolation as a standalone aesthetic experience. This is about the expressive experience the mashup artist is trying to create. Provided that the resulting work isn’t an abridgement, which it only will if it’s selected and arranged to convey substantially the same experience, then the mere fact that it’s composed of material taken from a work doesn’t make it infringing.

My Q: why does authors’ intent/meaning matter so much?

A: Isn’t that author has to intend specific fragment to be a work, but whether they could have so intended reasonably. (I still don’t get it. Now this seems to be more about the social meaning of a work for a reasonable author/audience. I’m just not sure what work intent is doing here or why it’s necessary.)

Q: I run a backup program for my documents, what is that?


A: Computer programs don’t belong in copyright.

Q: are you shifting from copyright to false designation of origin?

A: hasn’t thought about false designation of origin here.

Bruce Boyden: thought he had a nice simple test for what a work was—what the copyright owner chooses to publish. But didn’t you back away from that in your first answer?

A: we could take a hard line and say it’s only a work if you presented it that way. Harper & Row comes out the other way. But it does make sense to say some works of authorship are composed of smaller works. Serialization as chapters can be a successful mode. As long as it’s clear how we decide whether the thing someone copied is a coherent work on its own, or is it something that only has meaning because the second comer wrenched it out of context and used it to make meaning. Would someone think it had meaning if it were examined out of context of the second use?

Rebecca Tushnet, Georgetown

Worth a Thousand Words: The Law Outside the Text

Google Book Search image: the greatest and the best-paid copyright lawyers of our generation can’t figure out what to do about the images. Public discourse about the settlement is that we’ll get access to the books. Not so! We’ll get access to the text—the images disappear

I’ve been struck by the power of images to preclude analysis in many areas of the law—images seem to have a special relationship to reality, except that’s often not true, or not true in the way that would be required for it to have legal significance. My project is mainly critical: identifying this tendency and showing how it affects outcomes.

Images do require interpretation: different cultural conventions for what constitutes realism are the most obvious—the rise of perspectivalism did not mean that the centuries before contained no representations that viewers of the time perceived as realistic. Recent evidence: your view of the accuracy of a picture of a person’s skin color depends significantly on the alignment of your political views with his. But law has consistently struggled with the way that images seem to be pure representations of truth, a short cut from mind to mind as Justice Jackson said about the flag in Barnette, finding that it was unconstitutional to force students to salute the flag.

Worse, the truth of images sometimes enhances their power and sometimes makes them meaningless in legal doctrine. So one of the two poles of legal treatment of images is to treat them as opaque and beyond analysis, while the other is to treat them as transparent and requiring no analysis.

Because of their apparent—and note that we use lots of visual language as dead metaphors for knowledge, so ingrained that we don’t even notice that we’re equating seeing with knowing—anyway, because of this apparent truth value, courts actually don’t see the particular images in front of them. So, in the paper I argue that the law’s difficulty with non-word methods of communication shows up across many fields, from trademark and false advertising law to obscenity, which has essentially given up on the idea that text can ever be obscene.

Specifically as to copyright: Consider McFarlane v. Gaiman, a dispute over ownership of several characters from the comic Spawn. The issue was whether Neil Gaiman co-owned the characters because of descriptions he’d given McFarlane and the illustrators. The argument was that his descriptions had been very general, only a paragraph or so worth, and that his contributions alone were only uncopyrightable stock elements, such that the thing that was copyrightable was contributed by McFarlane and the artists.

Judge Posner rejected this argument, in very strong and general terms, elevating text above image for purposes of copyright’s protections. The writer deserved a share of the copyright because “the description of a character in prose leaves much to the imagination, even when the description is detailed.” Here's an image that Posner didn’t think left much to the imagination:


Among the phenomena Posner’s distinction can’t explain is why generations of supposedly passive fans of audiovisual material, from Star Trek to Lost, have been inspired to write, draw, and otherwise create works extending the initial stories, “completing” the works not just in their minds but on their pages and screens.

Note here also Posner’s dismissal—the literal failure to see—the editing techniques that distinguish film from live performance. A viewer routinely “completes the work in his mind” because film cuts: a character gets in a car, and then is shown elsewhere, a sequence that is sensible only because of the mental operations performed by the audience. But this technique is so naturalized to a modern viewer that it is invisible. Similarly, the scholarly and judicial discourse about why photographs are copyrightable centers on selection: framing, angle, leaving out as well as including.

In Posner’s formulation, words, by contrast to images, have a connection to something more abstract, something that is somehow not quite contained in the words and thus grants an author of a text greater rights than those of someone who works in images. Images here are imagined as so powerful as to not require any interpretation or elaboration (which is of course empirically false), with the surprising consequence that their power makes them less authorial, less qualified for the core of copyright protection.

Contrast Justice Jackson’s characterization of images as a shortcut from mind to mind: a shortcut means the image itself provides less information, with more supplied by the viewer/audience. Formally, one would think copyright would provide less protection when the copyrighted work itself provides less information and the audience responds by bringing to bear all its inferences and associations.

So Posner is wrong both in describing reality (images and audiovisual leave lots out) and in providing a sensible basis for doctrine if we take ordinary copyright principles about thick and thin copyrights into account. The real explanation: Images, by being more real (even when entirely imaginary) lose dignity as creative works of an author when they’re compared to words.

This isn’t just a problem of allocating rights in audiovisual works. It leads to mistakes in thinking about protectability and infringement generally.

Great example from the Third Circuit: Franklin Mint v. National Wildlife Fund.
Plaintiff's work:

Defendant's work:

Painted in court without looking at either:


The court of appeals, upholding verdict for defendant, held that style affects protectability: Monet may have a broader copyright than a painter who attempts photographic realism, so that someone who paints the same subject with similar realism won’t be infringing, whereas someone who painted the cathedral at Rouen in Monet’s style might be. “In the impressionist’s work the lay observer will be able to differentiate more readily between the reality of subject matter and subjective effect of the artist’s work.” This is actually not how the law of photographing the same subject as a previous photographer has worked out, but leave that aside.

This concept of special access to reality structures the court’s reasoning, even though several features of the paintings at issue argue strongly against any such concept. First, the paintings were in a style popularized for the depiction of birds by John J. Audubon, which among other things abstracts the birds and the fragments of plants on which they rest from any background, and structures the arrangement very carefully. It is “realist” in the way that the Hollywood car chase scene is realist, which is to say not at all, even though the representational conventions are so common in Western art that it is easy to perceive this style as realist. Thus, immediately after distinguishing strongly realist from less realist art and suggesting that the litigated paintings fell on the more realist side, the court noted that numerous conventions in ornithological art determined many features of those very paintings.

Second, the initial painting and the accused work were created by the same artist, working from his imagination and synthesis of birds and pictures he’d seen in the past, such that the image he was painting only existed in his head as a sort of Platonic ideal—not a real set of birds and branches. The jury and the court accepted that the second and third paintings were versions of that Platonic ideal, not copies of the first painting. But of course all painting is mentally mediated, not pure representation of reality. And finally, this was not the artist’s unique style, but rather a style copied from Audubon.

None of this can be understood as reflecting reality, even though the opinion takes for granted that the paintings are realist.

Q: courts seem willing to take images on their own v. music.

A: yes, that’s a problem of images’ apparent transparency; whereas they punt too much to experts for music.

Q: visual v. textually delineated characters?

A: yes, there is a problem here, highlighted by Posner privileging the writer over the artist despite making it easier to get copyright in the visual character—turns out what we’re really protecting is the words as instantiated in the images.

Zahr Stauffer, Virginia

Rethinking Intellectual Property Rights in Characters

Courts are getting character cases wrong. First, which characters should be protected; second, which parts or how much; third, which uses to protect against. (TM is a serious problem too.)

Second Circuit’s test is the most widely adopted: characters are protected if sufficiently delineated.

Literary theory: law unintentionally imports aesthetic value judgments. Overprotects first-generation authors. Our understanding of characters is historically contingent, not timeless. Aristotle etc. developed the idea that character shouldn’t matter; plot was all. Modern era: interiority becomes all-important. Whether to accord character protection is a function of that changing estimation of character/importance of psychology—seeped into the received wisdom about literary property.

How much to protect? Narrative techniques: stream of consciousness, blending third and first person narration, other kinds of style that have made it harder to separate the character from the work. (Cf. Salinger?) Different ways to show characters’ interiority have developed—step inside of character and back out. Law rewards modern characters for being individualized.

Popular trope of creator struggling with his characters—Shakespeare with Prospero, Mary Shelley with Frankenstein’s monster and his struggles with his creator Viktor, Six Characters in Search of an Author. (Sherlock Holmes, Falstaff, etc.) Anxiety of influence: authors use others’ characters too—essential part of situating oneself in a lineage. A class of uses we might not want to have infringement reach—historical authorial norms; authors must expect similar reuse.

Law treats characters and our understanding of them as static. But characters and theories are dynamic. Normatively, wants to assert value of interdisciplinarity. And protection’s scope is currently too broad. Why does this tiny corner matter? In this era of remix and mashup and media consolidation/clearance culture, protecting characters has significant chilling effects—should look at millennia of creative reuse.

Q: suggestion to think about divorce of authorship and ownership.

A: yes, though Sam Spade is also earlier than the other cases and the time shift corresponds to a literary theory shift.

Sag: Don Quixote’s second book is about struggling with ripoffs. Authorial norms don’t point in one direction.

My q: Compare the James Bond case—no interiority whatsoever. Infringed only perhaps by a 30-second commercial.

A: Stock character, but also distinctively delineated. Maybe shouldn’t be protected as a matter of public policy.

Pam Samuelson: Talk about Salinger.

A: normatively, shouldn’t be enjoined. It’s exactly the sort of thing literary people have been doing for millennia.

No comments: