11:51 AM
I have developed what I believe is a novel line of reasoning on copyright law and the Internet. The advantage, as I see it, is that this thinking is consistent with both existing copyright law and how copyright law "should" work on the Internet. However, this line of reasoning contradicts some of my previous statements about copyright law. Unfortunately, as I'm continuing to think about it, my thoughts are getting progressively more entangled, so I am leaving some issues unresolved.
In summary, making copyrighted materials available on the Internet violates the copyright holder's right of display or right of performance, depending on the nature of the material. Both accessing that material and deliberately copying it may or may not be infringing depending on which line of reasoning holds. In the case of file sharing, the provider of the file is violating copyright. The downloader may or may not be, depending on whether the Betamax ruling applies. In the case of publicly sharing iTunes libraries, as provided for example by SpyMac, the owner of the library is committing direct infringement and the sharing service (SpyMac or others) is committing contributory infringement, but listeners are not infringing either if they are shielded by the DMCA's protection of caching or if the Betamax ruling applies.
I will start with the line of thinking that led me to this conclusion. I set out to prove that downloading music from a file sharing network is not a crime. In my original post on the subject, I stated that "copyright infringement is not theft". In response to a comment, I asserted that "downloading is not a crime". In the comments on that post, Derek Slater stated that the Napster case concludes that both sharing a file and downloading are copyright violations. I responded saying that I think the judge got it wrong. Derek then jumped back out to his own blog for further comments. He has since written a followup. After further thought, I've concluded that the Napster ruling is incorrect, but I'm not certain whether downloading is a violation.
Before I get to the meat of the issue, there are a few pieces of business to take care of.
I'm not a lawyer. If you're about to make any important decisions based on this, it's better to assume that I don't know what I'm talking about. It's a much better idea to get actual legal advice now than to find yourself needing legal counsel later.
The most important issue here isn't really whether or not downloading is actually against the law. It's whether or not downloading is "wrong." The question of legality comes up because many people people point to legality to support claims about whether something is right or wrong. Even if it is illegal, that doesn't necessarily mean it's wrong. But its legality can't make it right, either. From this standpoint, whether it's theft or copyright infringement doesn't make much difference, because either way the law is against you.
Copyright law was written to cover copying of physical objects. It was deliberately written to include future technologies, but it didn't consider general distribution through an intangible medium. All copyright rulings involving the Internet represent attempts to fit the existing law to an intangible medium. Some aspects of copyright law adapt to the Internet better than others.
My brother, who is a lawyer, points out that "theft" isn't well defined under the law. What I'm thinking of when I think of stealing CDs is probably the crime of larceny. Let's try that out on file sharing. "When you download music, you're committing larceny!" I hope that sounds as excessive to you as it does to me.
This is really long. Unless you really care, you should probably just make a call on whether or not file sharing is wrong and run with it. Be aware, though, that file sharing is risky. If you run file sharing software, the RIAA may come down on you, hard, and then it won't matter if it's actually legal or not.
The place to start with this is the Napster ruling. The court in the Napster case ruled on the behavior of users of the service on the way to its ruling on Napster itself. This leads to the question of whether the Napster ruling was reasonable and a reconsideration of the general question of making copyrighted works available over the Internet and downloading those works.
To explain why I'm challenging the Napster ruling, I should describe the history of the case. Napster was an Internet service that allowed users to copy music files from other users. The Napster software allowed users to select music files to share with others. Napster maintained a directory of shared files, but not the shared files themselves, which remained on the computers of the individual users. Users could search the directory of files and find files they wanted to download. When they downloaded files, Napster established a direct connection between the downloader's computer and the copier's computer and the file was copied without passing through Napster's computer.
The major music companies sued Napster for copyright infringement on December 6, 1999. Judge Marilyn Hall Patel issued a preliminary injunction against Napster on July 26, 2000. In the preliminary injunction, Judge Patel found that "Plaintiffs have established a prima facie case of direct copyright infringement. . . . [V]irtually all Napster users engage in the unauthorized downloading or uploading of copyrighted music." Napster appealed the preliminary injunction, and the appeals court substantially upheld the injunction on February 12, 2001. The appeals court ruling (PDF) stated that, "Napster users who upload file names to the search index for others to copy violate plaintiffs' distribution rights. Napster users who download files containing copyrighted music violate plaintiffs' reproduction rights." Between the expense of the lawsuits and the inability to keep operating because of the preliminary injunction, Napster declared bankruptcy on June 3, 2002. At that time, there was no final ruling on the court case.
Although the district court ruled that Napster users commit copyright infringement, it was the appeals court that ascribed specific violations to users who upload files and users who download files. (It's worth mentioning in passing that both the district court ruling and the appeals court ruling indicate a certain hostility toward Napster which is reflected in their dismissal of Napster's arguments in its defense.) Despite the ruling coming at the appeals level, I still don't consider the issue settled on the basis that the ruling was a preliminary injunction rather a final judgement and because the case was still ongoing when Napster declared bankruptcy.
Furthermore, Napster did not challenge the copyright infringement claim during the appeal of the preliminary injunction. It may have concluded that its energies were best spent elsewhere at that stage, but it is also true that it would not have served Napster's interests to have argued that sharing files was a violation but downloading them was not. It doesn't matter who was doing the infringing, as long as infringing was occuring.The closely related case of Fonovisa, Inc. v. Cherry Auction, Inc, cited repeatedly in the decisions, concerned the operator of a swap meet where infringing recordings were sold. There, the infringement is clearly only on the part of the sellers, but the argument that the buyers were not committing infringement would not have cleared Cherry Auction of contributory infringement.
In all related cases, it hasn't actually mattered whether the downloader is committing infringement or not. It will only directly matter, and likely will only be directly tested, if copyright holders start pursuing claims of infringement against downloaders. That hasn't yet come to pass, but the RIAA has started pursuing users strictly for sharing files. Until that time, my opinion is that the court is still out.
The appeals court held that the person sharing the file is violating the right of distribution while the person downloading the file is violating the right of reproduction. This ruling is inconsistent with copyright law as applied to physical objects. To show why, I'm going to describe copyright law in general. Copyright law grants five exclusive rights to copyright holders. These are the rights of reproduction, creation of derivative works, distribution, public performance, and public display. Copyright law then places limitations on these rights, including fair use, first sale, compulsory licensing, and other limitations. There is an implied sequence in which the exclusive rights are carried out with physical objects. You can't distribute and sell a CD before you produce the CDs themselves, meaning that the right of distribution can only be exercised after the right of reproduction.
Let's consider a specific example. If I make a copy of "American Life" available for download and Derek downloads it, the court rules that I have distributed the file and Derek has copied it.
Here's where things get weird. Copyright law is written with physical objects in mind. I would have to physically create a CD before I could give it to Derek. If Derek says, "hey, can you give me a copy of that Madonna CD?" no infringement occurs at that moment. I then infringe the law both when I make the copy and then when I give the copy to Derek. I can't give the copy to Derek first and then either produce it myself or have Derek produce it.
The same causality should apply to files on the Internet. Just as it is illogical to give someone a CD and then produce it, it is illogical to distribute a file and then copy it. So then if I'm not distributing the file, what am I doing? Copyright law gives the answer. I'm performing it. Public performance is specifically defined by copyright law to include transmitting the work, whether it is received "in the same place or in separate places and at the same time or at different times." What I'm doing depends on what kind of file it is. Music and movies are performed, while text and images are displayed, but in both cases, making the file available constitutes transmitting it.
This has important implications. Under copyright law, performance of a musical recording is explicitly not protected. When I make a music file available, two copyrights are involved. The author has copyright in the music, and the performer has copyright in the recording. While generally the rights for both are active, protection of the right of performance does not extend to the recording. Radio stations pay licensing fees to authors (the fees are managed by ASCAP and BMI) but pay no fees to the performers. This means that making a recording of a Mozart piano sonata available as a computer file does not violate copyright, because Mozart's copyright has expired and the pianist has no exclusive right of performance.
Generally, however, all other types of files are protected and in the case of "American Life," I would still be violating Madonna's copyright (assuming that she wrote the song. If someone else wrote it, I would be violating that person's copyright.). Most files distributed through file sharing are protected by copyright. The one nonobvious exception is recordings of music where the music itself is no longer under copyright. For other files, the sharer is violating the creator's copyright protections.
The next subject to consider is the copyright implications of downloading files. There are two separate cases to consider. The first is when the file is permanently saved to the user's hard drive. The second is when only temporary copies are made, for example if the file is streamed or if it only resides in the user's browser cache.
When the user makes a permanent copy, as when using file sharing services, the user clearly is violating the right of reproduction. The right of reproduction extends to musical recordings, so the downloader can violate copyright when the sharer is not. In all other cases, copyright law applies to both parties equally.
The DMCA, of all things, may protect temporary copies. Section 512 of the Copyright Act, which was introduced by the DMCA, includes the notorious take-down provision, in which ISPs can be forced to remove their customers' materials from the Internet on the basis of a claim of copyright infringement without due process. But section 512 also protects cached or temporary copies from liability. Section 512(b) reads, "A service provider shall not be liable . . . for infringement of copyright by reason of the intermediate and temporary storage of material . . ." As written, this law applies to service providers rather than users and only covers "intermediate and temporary storage". The end user is obviously not making intermediate storage. Nonetheless, I would argue that this should be applied to end users anyway.
Suppose that I put a link on my website which I label as "Madonna's 'American Life.'" Without further labeling, it is not reasonable for a visitor to know whether the link goes to a copy on my site, in which case the visitor would be committing a copyright infringement by playing the song in the web browser, or Madonna's site, in which case the user would not be. Even more significantly, if I put an image of the cover art of the album on my website or include the lyrics of the song on my website, visitors would have no way of knowing that by visiting my site they would be violating copyright until the violation is already made. The threat that by visiting a webpage, users could automatically violate copyright without any way to know whether the page is "safe" or not without visiting the page could have a severe chilling effect on the useability of the Internet.
The "caching" protection offers a way out. When a user visits a webpage, the page is typically automatically saved to the user's hard drive in the web browser software's cache. In the course of normal web browsing, files in the cache are regularly deleted without the user's involvement or knowledge. Just as service providers are protected from liability for caching data, users should be as well. This leaves users free to surf the web without fear that they will unknowingly become liable for copyright infringement.
This protection does not extend to deliberately making permanent copies or making copies with the intention of deleting them. There is no protection for making copies for 24 or 48 hours with the intention of "sampling" or "trying out" the copy, despite what any number of websites say to the contrary.
This line of reasoning probably would not hold up in court. It is my contention, however, that the public good would be served by adopting this reasoning or modifying the law to reflect this reasoning. Otherwise, fear of inadvertent infringement might make the Internet unuseable.
As I mentioned at the beginning, there is another line of reasoning based on the Betamax ruling that I have not fully explored. The Napster ruling dismissed arguments based on the Betamax ruling. However, I have already demonstrated that the reasoning of the Napster ruling is suspect, and the Napster defense was based on the "non-infringing uses" part of the Betamax ruling. I turn instead to copying for personal use. The Betamax ruling established the principle that despite the exclusive right of reproduction granted by copyright law, copying for personal use qualifies as a "fair use" exception. Thus, you are not violating copyright when you record Buffy, even if you are making a permanent recording, provided that you are not intending to sell the copy or otherwise make non-personal use of the recording.
I would have to study the ruling more carefully to make a definite determination, but I believe this should apply to downloading. Under this reasoning, permanent copies would be noninfringing and the downloader is never liable (provided that the download is for personal use). This would obviously supercede the DMCA argument.
The upshot of all of this is to set reasonable rules of liability related to copyrighted materials on the Internet. Individuals are generally liable for infringement for making copyrighted materials available on the Internet, with the exception of recordings of music when the music itself is not protected by copyright. Under the DMCA reasoning, downloaders are not liable for temporary copies, but are liable for permanent copies. Under the Betamax reasoning, downloaders are never liable.
This reassessment of direct liability also shapes contributory liability. Under the DMCA reasoning, Napster is still liable for contributory infringement, because all downloads are infringing. Under the Betamax reasoning, a significant noninfringing use appears. Sharing and downloading music files where the music itself is not protected by copyright is not direct infringement, and therefore Napster cannot be liable for contributory infringement. Under either the DMCA reasoning or the Betamax reasoning, Spymac can claim substantial noninfringing use.
There are still significant unresolved questions related to this analysis. One topic which has been subject to continuous debate since the passage of the Copyright Act of 1976 is whether music recordings should have an exclusive right of performance. If this thinking were to be adopted, the case for the right of performance would clearly be strengthened. The reasoning behind my DMCA and Betamax arguments also requires and deserves further study. In whole, the argument I have presented should inject rationality in to the application of copyright law to the Internet.
3:40 PM
Today you get two major rants for the price of one. I'm leading off with the story of an RIAA mistake that nearly took a university server offline during finals. If that's not your thing, scroll down to where I challenge a National Review Online article claiming that file sharing will be the death of creative work. I hope my response is level headed, but I'm afraid it might not be. When you read the article, you'll understand why. After that, I have a bunch of smaller news items that didn't get me quite so worked up.
RIAA mistake
CNET reports on a RIAA mistake that nearly caused havok in the Pennsylvania State University astronomy department. The RIAA sent a message to the Penn State computer security office claiming that the astronomy department FTP server was unlawfully distributing recordings by the musician Usher. The RIAA requested that Penn State have the songs removed or that the site be shut down. Under the DMCA, copyright holders can contact ISPs with claims of copyright infringement. If the service provider "acts expeditiously to remove, or disable access to, the material", the service provider will not be subject to infringement claims.
Penn State told Matt Soccio, the astronomy department's system administrator, to remove the infringing files or the server would be disconnected from the network. Soccio searched for infringing files and couldn't find any. He "had been worred that the server would be yanked from the network during the middle of Penn State's final exams last week. 'If our site was shut down as this was being investigated, I wouldn't even be able to have a conversation with you because (there would) be so many people in my office wanting to know when it would be back up,' he said."
He eventually determined what had happened. "The department has on its faculty a professor emeritus named Peter Usher whose work on radio-selected quasars the FTP site hosted. The site also had a copy of an a capella song performed by astronomers about the Swift gamma ray satellite, which Penn State helped to design." The combination of a professor with a similar name to that of a musician and any music file at all earned the letter from the RIAA.
The RIAA states that it has sent out tens of thousands of these letters and this is the first error that's been reported. The RIAA has apologized to the university. The RIAA has no responsibility for the incorrect letter, because the DMCA limits liability to cases of knowing material misrepresentation. Accidents, like this one, can be made without fear of liability.
I have trouble believing that this is the first error the RIAA has made, even if it is the first one reported. Most ISPs do not the relationship with their customers that Penn State University has with its astronomy department. In response to a notice of infringement, most ISPs are likely to remove the material without question, to avoid facing liability themselves. Likewise, most customers are unlikely to be able or willing to contest the removal of noninfringing material.
In review, copyright holders can request that any file that appears to be a copyright infringement be removed by an ISP, with no penalty if they are wrong. ISPs face no liability if they remove the allegedly infringing materials, but may face copyright infringement claims if they do not remove the material and it is infringing. It is up to the customer to prove that the file is not infringing, at which point their only relief is to have the file restored, regardless of any damages they may have suffered. Does anyone else think this is a little unbalanced?
For more on this, see Slashdot.
National Review article
National Review Online has an article about the Grokster ruling titled "Digital Communism". As the title suggests, this piece is overflowing with overblown rhetoric. According to the author, legal file sharing services will "usher in an age of digital communism." He repeats the myths of copyright as property and file sharing as theft, and compares content creation without copyright protection to building a factory in the Congo.
I feel like proponents of weak copyright are at a serious disadvantage in the rhetoric game. Supporters of strong copyright seem to have no trouble mischaracterizing our position and writing pieces like this that are so far off base that the only rational response is to hit the author with a chair.
In spite of the fact that my head is reeling from even trying to comprehend this piece, I'm going to attempt to take it down. If you can look past his invocations of Karl Marx and the Congo, what he is saying is that "the ability to exclude is the essence of property rights. If I 'own' land but anyone can trespass I don't really have any property rights. Similarly, if I own a movie, but anyone can freely watch it, my rights have disappeared."
This is the critical false assertion by advocates of strong copyright. Creation of a movie does not equate to ownership of the movie. The nature of land is that physical borders can be placed on it. There are no such boundaries on a movie. If I watch a movie, that doesn't prevent you from also watching the movie. The idea of a movie as a space occupied by all people who have watched it is bizarre, frankly, but I'll entertain it for the sake of argument. James D. Miller says that, as the creator of a movie, he is entitled to set the terms of all people who enter that space. What happens if I see his movie and then tell a friend about it? Does my friend enter the space? Is someone who watches the movie obsessively somehow more in the space than someone who fell asleep partway through? Does entrance into the space entitle the moviegoer to any other rights? Does a book adaptation of the movie occupy the same space as the movie? How different does my movie have to be to not overlap his movie, and how much overlap is too much? Unlike physical land, the boundaries are amorphous and the rights of the occupant are unclear.
This misses the bigger issue. US law explicitly provides for public use of private land. Easements give a legal right of access to property by people other than the owner. If we allow the conception of copyright as a space controlled by its owner, what sort of easements should it have? Physical easements serve the public interest, by providing access to land that would otherwise be inaccessible or by allowing utility connections to other land that would otherwise be unable to get that service. Easements on copyrighted works should similarly serve the public interest.
Once again, however, I feel like I've been distracted by a detail. The major claim of this article isn't that copyright is property. It's that file sharing will destroy copyright and without copyright protection, creators will have no means of making money. This, too, is a claim that is hysterically overblown. Legalized file sharing may reduce the difficulty of copyright infringement, but copyright holders haven't shown any hesitation in going after the most egregious violators. And the Internet is just one medium. Even if he can't stop copies of his movie from appearing online, he can still arrest me for selling DVD copies of his movie.
The article also presents copyright as either/or. Either he has total control over everything he creates to the level of choosing who can or can't watch his movie on an individual basis or he has no copyright protection at all. Very few people would argue for no copyright protection at all. But the opposite extreme is equally extreme. Creators deserve to be paid. Whether they deserve to control who can experience their creations and what they can do with them until the end of time is another question.
File sharing does nothing to undermine copyright. It may make copyright enforcement more difficult, but whether that is actually the case remains to be seen. It does, however, give some people an excuse to declare that the sky is falling and that the only response is to jack it up as high as possible. (from FurdLog)
Other news for the day:
As the Apple Turns has a "news" item which points out the absurdity of thinking that online music distribution is the same as conventional distribution. Some songs originally available through the Apple Music Store, including songs by Van Morrison and Radiohead, are no longer available. As the Apple Turns has the perfect explanation: the Apple Store is sold out. They tell readers not to worry, because "we're told that the record companies have Van Morrison and Radiohead chained up in a studio somewhere, cranking out copy after copy of the depleted songs, which will soon be freighted to Apple and placed back on the shelves at the iTunes Music Warehouse."
MacSlash has more on the legality of iTunes library sharing. It points to an analysis by a non-lawyer who went a step further than I did by actually referring to the relevant law on webcasting. He concludes (as I did) that iTunes does not qualify as webcasting, and therefore there is no copyright infringement. Users are getting confident of their case. The question is whether the RIAA will agree. The MacSlash comments include a statement from an actual lawyer which reminds non-lawyers of the danger of taking casual legal interpretations seriously while stating that it probably comes down to a fair use argument, which is impossible to resolve outside a courtroom.
Lawrence Lessig picks up on the Bruce Lehman interview. Like the comments I referred to yesterday, he doesn't have good things to say about Lehman.
Lawsuits related to Napster keep coming. Universal Music has sued Bertelsmann because of Bertelsmann's involvement with Napster. Bertelsmann invested $85 million in Napster while Napster was engaged in its legal fight with the music industry. That investment was itself striking because Bertelsmann is the owner of BMG, one of the five dominant music companies. Universal's lawsuit alleges that by investing in Napster, which committed contributory copyright infringement by allowing users to copy copyrighted songs, Bertelsmann was committing vicarious infringement. This follows on Universal's lawsuit with EMI against Hummer Winblad, a venture capital firm which also invested in Napster.
Here's another news item that I dismissed originally. Eminem has stopped "Weird Al" Yankovic from making a video parody of one of his songs. Weird Al regularly gets permission from the original musicians for the parody songs he creates. Eminem was willing to let Weird Al record a song based on "Lose Yourself," but he refused to allow Weird Al to make a video based on the new song, titled "Couch Potato." Weird Al says that he was very disappointed, and the video, a montage of parodied scenes from Eminem's videos, was going to be the best video he had ever created.
So the question is, why is Weird Al getting permission in the first place? Isn't parody considered fair use? There are a few possible answers. First, Weird Al may deliberately try to stay on the good side of the music industry, either out of a need to work within the industry or because he genuinely respects the musicians he parodies. Second, and more ominously, he may not be willing to fight any lawsuits that may arise. The problem with fair use is that there are no clear lines on what qualifies. If he creates a parody without permission, he may spend more time defending the song in court than he does making music, and Weird Al may not want to (and shouldn't have to) follow that course.
Derek Slater beats me to the punch on the legality of file downloads. My thoughts are still coming, just as soon as I can get everything else out of the way. His thoughts are definitely worth reading and thinking about.
![[Valid RSS]](http://matthew-morse.com/blog/valid-rss-lee.gif)
