Sunday, April 01, 2007

What Ifs? Copyright Law III

Susan Corbett, Victoria University of Wellington (New Zealand), What if object code had been excluded from protection as a literary work in copyright law?

Opening question: what to do to preserve New Zealand’s early (1970s) videogames, from companies that are now out of business for systems that are now obsolete? The games are obsolete, but the copyright protection endures.

As of 1994, computer programs are protected in New Zealand as literary works. In 1989, there were two conflicting Commonwealth judgments on protection of object code. Australia (1986): object code was neither a reproduction nor adaptation of source code. Canada (1987): it was an adaptation of the source code.

What if NZ had followed the Australian ruling, but found alternative protection for object code as a separate sui generis category of work? It’s unrealistic, given the economic value of software, to posit no protection.

Problems with categorizing computer programs as literary works: (1) archiving digital culture. UNESCO has encouraged the preservation of digital culture. Problems include physical deterioration, commercial obsolescence, the need for many copies for effective archiving, the length of the copyright term, orphan works.

(2) Fair dealing becomes unfair. It’s impossible to copy part of a program, and you get perfect copies every time. The reverse engineering process (which isn’t a right in NZ, though might be fair dealing) can produce multiple copies, which is excessive under the law.

(3) Cookery analogies. NZ’s first reverse engineering case stated that a recipe might be copyrightable, but no one suggest that following the recipe would infringe the recipe. Still, the court found liability when the defendant made packages based on government specifications that had themselves been derived from a copyrighted design.

Canadian decision in 1987: a computer program is a creation as an instruction book is. Copyright wouldn’t prevent making a recipe, but does prohibit copying the cookbook. There are other examples of recipe analogies. The Australian case denying copyright to source code held that people don’t “translate” instructions by carrying them out, and neither do computers in converting electrical impulses. So we get (idea: recipe: human “translation” into action through cooking: pie) is the same as (idea: source code: computer “translation” into action: result).

Sui generis protection would have been better with its shorter term and substantial public interest protections.

Abraham Drassinower, University of Toronto Faculty of Law, What if copyright were really about authors?

If authorship were central, copyright would be less extensive. There would be no grounds for liability for copying for personal use, and the defense of fair dealing/fair use would not be a mere exception. It would be a user right.

Originality v. novelty – copyright and patent. The distinction teaches us that patent is concerned with substantive contributions to existing knowledge, but copyright is concerned with the form of communication of the author’s thinking. Not what the author has to say, but how she says it. To express a trite idea in one’s own words is to be orginal for copyright’s purposes. The determination of novelty is a comparison to present and prior art – it’s relational with the external world. The determination of originality is about source – the author’s relation to the work. There is no such thing as a second inventor, but there is a second author as long as they both drew the works out of themselves. At the same time, an author is engaged in a discourse with the public – the author’s work itself is a mode of discourse addressed to others. By contrast, the subject matter of patent is an act in the world – doing things in and to nature, rather than speaking to other people.

What does this teach us? Canadian law provides that it’s not an infringement to reproduce for criticism when the reproduction is reasonably necessary. This is generally seen as a guideline for when reproduction is lawful. But we can also see it as an indication of the contours of the authors’ rights – the author owns his work for some but not for all purposes. The point is not that the defendant has an excuse, but that the plaintiff only has certain rights – a purpose-centered definition of authorial copyright that enables us to distinguish authorship from ownership.

A TM owner has an exclusive right to use the TM as a mark, to distinguish goods or services. There are no absolute rights over a TM – it depends on a particular purpose. Holding a copyright is similar to holding a TM: an exclusive right to use a work as a work – to present it to the public as an instance of discourse.

Assume I’m enamored of Hegel’s Philosophy of Right, and I make four copies to put under the legs of my desk for support. As a matter of copyright law, it’s a reproduction, but not a reproduction of the work as a work – as a communication to the public. Therefore, there can be no liability for personal use copying.

Fair dealing: a little more problematic on this account, because the defendant both reproduces and presents to the public. There’s no liability because the defendant doesn’t reproduce the work as a work, but as a reasonably necessary aspect of her own expression as an author. A plaintiff who has claimed authorial rights can’t turn around and deny others the right to be authors.

At the end of Michael Madison’s presentation, he left us with romantic authorship v. economics. But there are other images of authorship that open up other descriptions and explanations of copyright. If copyright were really about authors, it would be more like trademark than patent, and the materialist/economic story would be recognized as a fairytale.

Julie Cromer, Thomas Jefferson School of Law, What if man never walked on the moon?

If the US had never sent anyone to the moon, the US government wouldn’t have taken lots of pictures via LANDSAT, which made them government works. Some private entity might have commercialized satellite data instead, raising the copyright question.

Option 1: Could protect a geospatial image as a photograph. Here, there are still questions of authorship. Is the author the satellite? Is there enough creativity? Is the author the people at the control station, what are they controlling? Is manipulating the satellite to take a picture at a particular time creative, when it’s largely determined by where the satellite is at the time? Is processing the image a creative act?

Option 2: Could protect an image as a map, expressly covered since 1790. But Feist dealt a blow to map protection. Maybe it would only get protection for the raw data, but taking portions (e.g. to make derivative works) that wouldn’t get full copyright protection.

Option 3: Could protect the data as a database of comprehensive geographic information, electronically stored. But has it been selected, arranged, or coordinated when in the database? That’s hard to say. The data might not even be “fixed” until they’re assembled into a picture. Is a fixation being made simultaneously with the transmission from the satellite, as required by the definition in §101? Well, the data here are neither sound nor image until assembled; during the transmission there’s not yet a copyrightable image. So maybe there’s no protection under copyright at all.

Why does it make a difference? It makes a difference in terms of who’s the author. It makes a difference in the level of creativity we attribute to the pictures/data. It makes a difference in the scope of the rights.

There are also international issues – what if the satellite was made in Iran, controlled in France, taking an image in Kuwait, sent to a receiver in Canada, then sent to a processor in China, then distributed to Google in the US, then reproduced on a computer in Nigeria? It may not matter what law the US says will apply if any one of those countries recognizes an author and a work. Maybe we need something like the EU Database Directive to protect the data en route to the receiver.

Deborah Tussey, Oklahoma City University School of Law, What if employees owned the copyrights in their works?

The incentives story for employer ownership seems ahistorical – the industries that benefit most from work for hire (WFH) doctrine seem to have flourished for decades before that doctrine was established, and they were well-represented in the leadup to the 1909 Act. WFH is more a reward for political influence than an essential incentive. Radio and TV appeared and prospered even with an uncertain copyright status.

Without WFH, we could have used assignments instead. Would the transaction costs be too big? Not on the front end, at least, given the imbalance of power between employee and employer. This already occurs with independent contractors, where a writing is required. The drawback is of course the renewal right/termination right.

Even so, the employer could get a minimum of 28 years under the 1909 Act, and could demand preassignment of the renewal term, which would only be avoided if the employee had the bad judgment to die. Under the current Act, employers get 35 years to make back their investment.

Reliance entirely on assignments could potentially cause severe impacts on public access, though. It would make things like playing Raging Bull on TV, or using it in a university course, very difficult given all the rightsholders who’d have to agree.

We’d also have a much more developed law of authorship – who counts as an author? Tussey thinks that courts would be inclined to treat most people who worked on a movie as akin to editors of books – not real authors. (I’d agree that courts would strain to find lack of authorship in cases where exercise of a termination right would seem unfair, and would start holding that assignments didn’t estop the employer from later contesting whether an employee was an “author.” Or possibly the assignments would also include standard language that the employee’s contribution was too minimal to constitute authorship, but assigning rights in the alternative, the way music WFH contracts work now.)

McKenna for Drassinower: So you don’t believe in a reproduction right, just a right of communication to the public.

Drassinower: Yes. Personal use is an example of a broader category – copying wouldn’t be the center of copyright.

Q for Drassinower: Isn’t the author’s right to have a conversation with each member of the public? So if I make a copy for personal use, the author’s been denied the opportunity to have that conversation with me because I’ve taken the author’s place?

Drassinower: Assume that you grant my Hegel example, but still object to reproduction for the purpose of reading. He wants to think of “work” as a verb rather than a noun. This particular act is the communication of a message to another. The other is not present when you make a copy for yourself. You aren’t presenting the author’s authority to another.

Q for Drassinower: Isn’t patent law also about shaping relations of persons with respect to each other? E.g., the experimental use exception.

Drassinower: Sure. He still thinks there’s a distinction between the relational work and the patented thing, but is willing to give an account of patent that also has relational aspects.

Strandburg for Drassinower: The idea of conversation shows what bothers us about patenting in the academic context. There’s a distinction between invention in the commercial world and academic science. You do produce things with relations to the physical world, but scientists are also participating in a conversation.

Lemley for Corbett/Drassinower: How to protect software or functional works under copyright with a relational theory? How could there be fair use of software under a relational theory? What then do we do with software?

Drassinower: If the facts don’t agree with my theory, says Hegel, all the worse for the facts. The discordance means we should change something about the facts – here it is that author’s rights aren’t appropriate for software. We should discuss how to protect software without all the baggage of “theft” etc. from copyright.

Landau for Cromer: Kozinski relied on Feist to suggest that protection for maps is basically dead.

Cromer: She’s no fan of that decision. But we need to know why maps were copyrightable from the beginning. Copyright was an incentive to generate maps. Likewise, maps of the universe could be very valuable. Maybe map copyright is dead, but there’s no definitive holding yet – the incentive theory will keep them protected for a while longer.

No comments: