Prof. Jim Gibson, University of Richmond School of Law
November 23, 2010
When an individual makes a music or movie file available for downloading by others, without the permission of the copyright owner, is that an infringing act? Or does infringement take place only when the file is actually downloaded?
This thorny copyright issue is at the heart of much of the controversy over file-sharing. It’s relatively simple for a copyright owner to prove that a file has been made available for download, but it’s much harder to prove that a download has actually occurred. So if liability attaches to the mere act of “making available,” record labels and movie studios will have an easier time combating file-sharing.
Peter Menell recently wrote an essay for this series that examined this issue through the lens of copyright’s “distribution” right. Under Section 106(3) of the Copyright Act, only the copyright owner can authorize the distribution of the copyrighted work (subject to certain exceptions not relevant here). This distribution right has been the focus of the various rulings on the “making available” issue, with some courts holding that making files available equals distribution (and thus infringement) and others requiring actual downloads before distribution is said to occur.
In this essay, I would like to offer up an alternative way to view the issue, based on the way that Section 106 of the Copyright Act divvies up a copyright owner’s entitlement into several discrete exclusive rights. The right to distribute is one such right. Another is the right to reproduce. It may be that what is going on in the file-sharing context is not “distribution” at all – regardless of whether a file is actually downloaded. Instead, “reproduction” might be the only right implicated.
Start with the reproduction language in Section 106(1). It gives the copyright owner the exclusive right “to reproduce the copyrighted work in copies.”1 Elsewhere in the Act we learn that a “copy” is the term the law uses to describe the physical form of a copyrighted work. So if a CD contains a copyrighted song, the CD is a “copy” – a tangible thing in which the song can be found. A book whose printed pages contain a copyrighted story is a “copy” of that story. The celluloid on which a copyrighted film is printed is a “copy” of that film.
Therefore, when Section 106(1) gives copyright owners the right to reproduce the work “in copies,” it means something very specific: the creation of a new tangible instantiation of the work. If I play a CD for you, one might say that I am “reproducing” the music on it, because you can perceive the music’s creative content. Nevertheless, I am not infringing the copyright owner’s reproduction right, because that right is concerned only with the making of tangible, lasting “copies.”
If, on the other hand, I stick the CD in my computer and burn a second CD for you, then I am infringing the reproduction right, because the reproduction has resulted in a new “copy.” Where there had been only one copy of the musical work, now there are two.
Turn now to the distribution right in Section 106(3). It too uses the term “copies”: the copyright owner has the exclusive right “to distribute copies … of the copyrighted work.”2 The “copy” term appears again in the first sale doctrine, an important limitation on the distribution right found in Section 109(a): “the owner of a particular copy … lawfully made under this title … is entitled, without the authority of the copyright owner, to sell or otherwise dispose of the possession of that copy.” So notwithstanding the copyright owner’s distribution right, the owner of a particular physical “copy” of the work is allowed to distribute that copy without permission, as long as it was lawfully made (i.e., as long as the copyright owner authorized the initial creation of the copy).
If one reads these three provisions together, one can see how Congress set out to regulate the market for tangible copies of copyrighted works. Section 106(1) addresses the creation of a new copy in the first place. And once a copy is created, Sections 106(3) and 109(a) define the copyright owner’s control over what happens to it – i.e., what happens when a copy changes hands.
What does this all have to do with file-sharing and the “making available” controversy? Well, think about what is going on in the file-sharing context. Suppose User X has an mp3 of a song – what the Copyright Act would call a “copy.” Her file-sharing software makes that copy available for downloading by others. Across the country, User Y uses the software to find the song in X’s folder and download it. Voilà: illegal file-sharing.
What is happening here, in a technological sense, is that there are now two “copies” where before there was only one. When Y downloads from X, it’s not as if the copy on X’s computer leaves her hard drive, travels through the Internet, and lands in Y’s computer. No copy changes hands. Instead, X’s copy stays right where it is. What happens is that a new copy appears on Y’s hard drive.
So what is happening here in a legal sense is arguably not “distribution” at all. It is “reproduction” – two copies where there was previously only one. Viewed in this way, the controversy over whether “making available” constitutes “distribution” is a red herring. Instead, file-sharing never involves distribution at all. It involves only reproduction.
This may seem like legalistic quibbling over arcane terminology, but it has serious implications for how the law handles file-sharing. If file-sharing is all about reproduction, then it is probably the downloader (User Y in my example) who is the primary infringer, because it is the downloader whose conduct creates the second copy. But identifying the downloader is even more difficult than determining whether a download occurred, which is why copyright owners have focused on suing the uploader.
Yet if we are dealing with reproduction, not distribution, then suing the uploader (User X) is problematic. In many cases, the only way to hold X liable will be as a facilitator of Y’s infringing reproduction – what the law calls secondary liability. If X knows (or should know) of Y’s infringement, and materially contributes to it, then X is also liable. That’s certainly a possibility here, but it’s a more complicated case to prove. And if the copyright owners also want to sue the company that provided the file-sharing software, well, then they may have to invent tertiary liability – suing Z for facilitating X’s facilitation of Y’s infringement. Yikes.
Given how complicated this all becomes, it is no surprise that courts have decided instead to deal with the issue as one of distribution. That approach focuses squarely on the party who is easiest to identify – the uploader – and the only issue is whether “making available” suffices for her liability. But when no copy changes hands, and a new copy is created, it is not at all clear that the distribution approach is correct.
Comments From Our Readers
Tom Sydnor: Professor Gibson:
I have enjoyed many of your writings, but I cannot enjoy this one. May I respectfully suggest that you should reconsider and retract the arguments made here?
They make no sense. Worse yet, they deny the practical implications of the realities that they purport to describe.
Indeed, you seem to rely on mere word-play to evade a simple reality:
Distribution, not reproduction, is the act that got KaZaA users Jammie Thomas-Rassat and Joel Tenenbaum sued for copyright infringement. That is why both Thomas-Rassat and Tenenbaum tend to be ridiculed by sophisticated users of such file-sharing programs. Sophisticated users of such programs despise users who are zealous or stupid enough to reproduce and distribute on a massive scale” rather than just reproduce on a massive scale, like sophisticated “leeching” users.
Nor does your argument seem valid even in a technical sense. Programs like KaZaA and LimeWire are all about distribution – not mere copying. When Grokster/Morpheus/KaZaA/LimeWire users search for file XX, they ask three or five database-servers whether those servers can identify a computer “sharing” file XX. If so, one or more servers respond by providing the IP address and other information needed to enable the searcher to instruct an identified computer to make a copy of file XX, to break it up into packets, and to send those packets to the computer of the searcher seeking a copy of file XX. I am at a loss to imagine why you can perceive the “sharing” computer in this example to have infringed only the reproduction right, not the distribution right. It executes instructions to copy and distribute, not just instructions to copy.
In short, if distribution is the real-world act that actually prompts real copyright owners to file real infringement lawsuits against real people, why does it make any sense for an academic to pretend that the only right infringed was the reproduction right? I can conceive of no sensible answer to that question. Let me know if you can.
Jim Gibson: Tom, thanks for your comments, and I am happy to hear that you have enjoyed some of my earlier writings.
As for the “making available” issue, reasonable people can disagree, but I’m surprised that you think my arguments “make no sense,” because they are rooted in the plain text of the statute. In contrast, your response refers to case law, and not to the statutory language on which those cases are supposed to rely. So unfortunately I am unsure of the exact basis of your objection. But let me take a shot at it, using the example you provide.
As you say, file-sharing software enables “the searcher to instruct an identified computer to make a copy of file XX….” True: It is indeed the searcher (User Y in my essay) who identifies the file and causes a copy to be made. So far, that sounds like an act of reproduction performed “by the searcher,” not by the person from whose computer the file originates (User X). So at that point there is reproduction liability on the part of User Y, but there is no primary liability on the part of User X — just as my essay asserts.
You also say that the file-sharing software also enables “the searcher … to break [the file] up into packets, and to send those packets to the computer of the searcher.” I do not think it is accurate to say that a copy appears in X’s computer and is then sent in packets to Y’s computer, although that may depend on one’s view of whether information passing through RAM constitutes a statutory “copy.” I think the better view is that there is simply one act, of reproduction — i.e., a copy is made by sending packets from X’s computer to Y’s.
But for the sake of argument, I’ll concede the point. Let’s say that your description of the technology is accurate and that there is a distribution. Even so, you make it clear that the “the searcher … instruct[s] an identified computer … to send those packets.” In your view, it is the sending of those packets that constitutes the distribution, no? So once more, the person who causes the infringement to occur is the searcher (a.k.a. the downloader, or User Y).
This leaves me wondering again why you view the other party (User X) as primarily liable. Perhaps the answer is found in your next two sentences: “I am at a loss to imagine why you can perceive the ‘sharing’ computer in this example to have infringed only the reproduction right, not the distribution right. It executes instructions to copy and distribute, not just instructions to copy.”
Two responses here. As a preliminary matter, I did not “perceive the ‘sharing’ computer in this example to have infringed only the reproduction right.” My essay argued that the searcher infringed the reproduction right and that the owner of the sharing computer is secondarily liable.
More important, however, is that the sharing computer does not infringe at all, because computers never infringe; only people do. This may sound like a nitpick, but in fact it is essential to understanding the way that primary and secondary liability interact. It is simply not the case that if X’s computer is used in an infringement, then X is liable. Indeed, twenty-five years of copyright jurisprudence give the lie to this reasoning, starting with Sony and continuing through cases like Netcom, Hotaling (a favorite of “making available” devotees, but a case that also addresses volition), and Cablevision. X may be secondarily liable for facilitating Y’s infringement, because X knowingly provided the facilities for that infringement. But that’s precisely the point that my essay made.
Two final points. First, I think it is important to focus on what I am not saying. I am not saying that file-sharers are free from liability. Quite the opposite; I believe they are. I am saying that the basis for file-sharing liability is a combination of primary reproduction liability and secondary liability. Also, I am not saying that my interpretation of section 106 produces good results; I have made no such claim. Maybe it would make more sense for “distribution” liability to attach to the act of making files available for copying by others. But statutory interpretation looks to what the text says, not what we wish the text to say.
That leads me to my final point: I must object to the notion that my approach is one that only an academic could love. Perhaps I am misinterpreting your comments, but you repeatedly contrast the “real world” with my “pretend” approach. But in what way does my essay ignore reality? Yes, you are correct that distribution is the basis (well, one of the bases) for file-sharing lawsuits. My essay said as much, so I am hardly blind to that reality. What I am saying is that that approach is wrong. Surely you do not meant to suggest that, when an issue has been adjudicated by a handful of courts, any contrary view is divorced from reality? And what “practical implications” am I denying? I clearly state that my interpretation makes it harder to prove infringement in file-sharing cases. That is the practical implication of this approach; how my essay can be read as denying it?