After the news broke on January 12 that the Supreme Court had asked the Solicitor General for its views as to the CNN v. Cablevision case, which is now under review for certiorari, I overheard two attorneys saying that the case could have a lot to say about the application of copyright law to user-generated content (“UGC”). That struck me as wrong, until I started thinking about it and realized that I had really never bothered to define that term very precisely. Roger Faxon, CEO of EMI Music Publishing, has been quoted as saying that UGC is “taking copyrighted content, whether that is a song or a recording, and the user is, without authorization, turning it into another product.” That is precisely what’s going on in the Cablevision case: users are turning a time-bound television broadcast into a view-on-demand product. Does the lack of “authorship” by the user change anything? Should it?
If somebody uploading a copy of a sound recording onto YouTube gives rise to UGC, why not somebody telling Cablevision to record a movie for them on a digital video recorder (DVR)? Copyright consequences aside, does it matter that the YouTube content is visible to the world and the Cablevision DVR content is accessible only by a single subscriber? What follows is an attempt to think about these fundamental UGC questions without translating them into the language of copyright. To paraphrase Benjamin Kaplan in his 1967 book, An Unhurried View of Copyright, “I shall take you on a stroll over some of the [UGC] terrain and examine various knolls and gullies in a rather desultory, even naïve way; finally we shall climb to the top of a hill and see whether anything can be usefully said about the whole landscape.”
Most literally, user-generated content is any (a) content that is (b) generated by users of a digital network of some kind. Neither term is self-defining. First, by describing the subject matter as “content” we are immediately stepping into a rhetorical frame in which the nature of the creative work doesn’t matter – we don’t call them “user-generated works,” after all, but appropriately opt for the more generic “user-generated content.” What does matter, what turns information or a work of authorship into “content,” is the fact that it is contained. Merriam-Webster’s first definition of the noun “content” is simply “something contained – usually in plural, ‘the jar’s contents’ ‘the drawer’s contents’”.
Thus content is “what’s inside,” as when Dr. Martin Luther King, Jr. spoke of judging people not by the color of their skin but by the content of their character. More prosaically, phone numbers are the content of a phone book; they are what we look inside the book to find. But as the Webster’s definition recognizes, information per se is not content. Take a string of numbers from a phone book, where they comprise the content of the book, and put them on a car’s license plate and they cease to be content, in the sense under discussion here. A police officer entering that license plate number on a traffic ticket may subsequently make the number back into content, i.e., the content of the ticket, and subsequently perhaps the content of a database aggregating ticket information, but the car itself, sitting there being a car, has no content, only a license plate. Content and its container have a dialectical relationship: content is always the content of something.
The same holds for the telephone number assigned to a particular telephone. The phone doesn’t contain its number, despite the extremely strong association between the two, and despite the fact that the number seems, at least in some physical sense, to be inside the phone. The point is, no one would open up the phone to look for the number. (What is actually inside the phone, I assume, is a mechanism that responds when a particular sequence of numbers is entered into a network by another phone on the network. That makes the number, in conjunction with the mechanism and the network, extremely useful, but it does not make it the content of the phone.)
When the web was in its very early stages, circa 1993, large media companies often took comfort in the thought that “content is king,” by which they meant that their content – films, recordings, novels – was the main attraction that would draw people to use the nascent technology in large and lucrative numbers. The tech sector could provide equipment and connectivity, the thinking went, but until there was content worth seeking out, i.e. professionally-produced and promoted entertainment filling the virtual shelves/pipes/airwaves, no one was going to make any real money. That turned out to be wrong, but only because the old, pre-web definition of content was too narrow. Content really is king on the web, but it does not have to be ready for prime time in order to draw a crowd. It just has to be in the right container.
Consider eBay. If I have a spare short-scale fretless electric bass guitar lying around (which I do), it isn’t content. Even my snapshot of it isn’t content. When I post that snapshot to eBay, however, it becomes eBay’s content. I might be tempted to go further and say that the bass itself becomes eBay’s content, because it is the bass that an eBay shopper is looking for, not merely my unattractive snapshot. But even if we stop short of considering the bass itself, and limit ourselves to the snapshot, it is remarkable that eBay, by bringing together large amounts of such extremely boring, unprofessional content, contributed voluntarily and without recompense by the people who created it, somehow found a way to make a lot of money. Make no mistake, content is king on eBay, and on every other website, whether that content is user-generated or not.
Which brings us to “generated.” Just as the noun “content” is vastly broader than the sphere of professionally-produced entertainment, or even the copyright concept of a “work of authorship,” the term “generated” is far broader than analogous copyright-related terms like “created,” “prepared” and “authored.” Although a great deal of UGC does in fact consist of copyrightable works of authorship, not all does.
To cite just one example, there are large numbers of YouTube “video” postings that are no more than complete audio-only copies of commercial sound recordings, streamed on-demand to the visual accompaniment of a slowly changing, unsynchronized, vaguely psychedelic digital wallpaper pattern. The user here hasn’t “generated” anything in the sense of authorship. Surely the audio content was “generated” in the first instance by the label that issued the recording and/or the musicians and producers who made it.
But who “generated” the content in the first instance doesn’t matter very much, unless we move out of UGC territory and start thinking in terms of copyright law. As long as we stay in UGC terrain, the user who posts such a recording to YouTube has indeed generated the content. This is because he or she has put it into the YouTube container to be retrieved later, just as I might do with my electric bass snapshot on eBay, or the hypothetical police officer above might do by putting a car’s license number into the proper space on a traffic ticket. Before the recording is posted to YouTube it may very well be content somewhere else, on a CD perhaps, or even on iTunes, but it is not content on YouTube and for patrons of YouTube that makes a world of difference.
In copyright terms, it probably makes a world of difference to the copyright owner as well. If a recording is posted to YouTube, what matters copyright-wise is not whether the user happened to combine it with original video footage (of a dancing baby, say), but simply that the recording is on YouTube, generating ad revenue in which the copyright owner does not share. The DMCA, to the extent that it applies to YouTube at all, does not distinguish between infringing copies and infringing derivative works, and neither does any other relevant portion of the Copyright Act. The category of user-generated content is thus much broader than the universe of unauthorized derivative works and encompasses content having no authorship or originality whatsoever on the part of the user.
In effect, the “generation” of the content need be no more than sweat of the brow, “industrious collection,” and may in fact be a good deal less. A YouTube contributor who throws a single recording into the virtual hopper hasn’t been particularly industrious, after all, let alone original in the Feistian sense, but nonetheless he or she has been the but-for cause of that recording joining thousands of others in the sea of YouTube content. So it is not a value judgment to say that a user has generated content. It is merely a statement of fact about what people mean by the words. Without being perjorative in either direction it might be fair to say that a user generates content just as much as I generate the bad snapshot of my fretless bass that I want to sell on eBay. In the world of UGC, generating content is not (necessarily) authoring a work.
This is not to take a position on whether the CNN v. Cablevision case does or does not have any bearing on UGC. It all depends on what happens next. But at the very least we need to recognize that the vocabulary of copyright is not the only way in which UGC can be understood.