Bookforum: Between The Lines
Editing Slouching Towards Alexandria: A Roundtable On Google's Library Project
To many, the dream of a digital Alexandria seemed a giant step closer to reality when, in December 2004, the search engine Google announced that it had entered into agreements with the New York Public Library and four major universities to digitally scan their collections so that Internet users worldwide could search them. Under the current plan, users will be able to view small portions of text surrounding search terms within literary works under copyright (the entire text of works within the public domain are viewable). Users will also be shown links to websites enabling purchase of those books that are in print. Google says that copyright holders who do not wish to have their books "Googleized" may opt out of the program.
Virtually no one contests the value of such a service. But the project has riled some authors and publishers who object to the advertising behemoth's unpaid exploitation of their works. In September 2005, the Authors Guild, the largest society of published writers in the United States, filed a class action lawsuit alleging that Google's Print Library Project-since renamed the Google Books Library Project-constituted "massive copyright infringement." A separate action was filed a month later by the Association of American Publishers (AAP), the principal trade association of the American book publishing industry. Pointing to Google Print for Publishers, an earlier program in which Google partnered with various publishing companies, the Authors Guild and the AAP say that copyright holders' permissions should also be required for the Google Books Library Project and that compensation should at least be discussed for any use of works that results in copies of the original material. The Authors Guild and the AAP also note that other digitization programs are built on licensing schemes or are confined to works in the public domain, as is the case, for example, in similar initiatives recently announced by Yahoo! and Microsoft.
Although Google halted its digitization program for several months last year, it has now resumed scanning. Meanwhile, the lawsuits are wending their way through the federal court system. Nick Taylor, president of the Authors Guild; Allan R. Adler, AAP vice president for legal and governmental affairs; and Lawrence Lessig, Stanford Law School professor and public domain advocate, joined me to discuss what is at stake.
VIRGINIA RUTLEDGE: Google Book Search wants to provide users with "snippets" of text surrounding a particular search term. How does this project differ from compiling a library card catalog?
LAWRENCE LESSIG: Google Book Search will be a better card catalog, because it will create an index of the entire contents of a book, which more traditional card catalogs can't do. It will be a card catalog for the twenty-first century.
ALLAN R. ADLER: We disagree that Google is simply creating a digital card catalog, or even an index for that matter. Library card catalogs basically provide bibliographic information. Neither a card catalog nor an index has the capability to pull up any part of the full contents of a work.
RUTLEDGE: So although the user is not able to access a complete text using Google Book Search, for you the issue is that Google is copying books in their entirety, without the copyright owners' permission?
ADLER: The fact is, that same search capability could be provided in more traditional ways. It may be more laborious. It may take more time and effort to put together a useful index. But it could be done without having to scan the entirety of these works and take ownership of those copies on your own server. That's really what's at issue here.
Clearly this is an economic and entrepreneurial opportunity for the people who hold the electronic rights to the content of the books. We are beginning to see the unbundling of books as a new marketing opportunity---books being offered by the chapter, or even by the page, say, for cookbooks or travel guides.
RUTLEDGE: Shades of Dickens.
ADLER: So why shouldn't copyright holders profit from exploitation of their works by Google Book Search---which we see, frankly, as a very direct promotion of a commercial enterprise.
RUTLEDGE: Some university presses have worried that there are instances in which Google Book Search could supply enough information that certain sales would drop. Not everyone agrees. But to be clear, your concern is not that Google Book Search is a replacement for books, correct? Though certainly it would help users better select what they did and did not want to look at further, whether at a library or by purchasing a copy of the book for themselves.
NICK TAYLOR: That's right. You know, it's a valuable research tool, and of course authors would find it useful. We would hate to see it not go forward. But we believe that a company with the resources and the technological expertise that Google has should be able to figure out a way to license the works that it is appropriating. It's simply a matter of wanting to participate in a revenue stream that currently will flow exclusively to Google, even though there would be no stream at all without the underlying content developed by authors and publishers.
ADLER: Meanwhile, the entire books are sitting there in Google's database, and the reality is they could be reconstructed quite easily. Google says that it will provide adequate security for this material, but as a sophisticated technology company, Google would be the last to assert that it is absolutely immune from hacking. And if there's no licensing agreement between Google and the owners of the property that Google has collected and is storing on its servers, there's also no particular obligation on Google's part to maintain security at a certain level.
LESSIG: I'm not persuaded hacking is a serious threat here. But in any case, the possibility of hacking seems to be removed from the questions of copyright and fair use.
ADLER: Well, we've been debating for a while now what fair use is in the digital environment. Is the doctrine applied in the same way as in the analog world, or are there some differences?
Not many digital fair use cases have been decided. But I expect that judges would realize that in the digital environment, the capacity for destroying the actual or potential market value of the work to the copyright owner is a far greater risk than in the analog world, so that in performing the fair use analysis it would be appropriate to take into account the security of the copyrighted material in digitized format.
LESSIG: I agree that would be an interesting feature to add to the calculation of fair use. And it's accurate that we haven't had many fair use cases in the context of digital technologies---in part because of the high cost of copyright litigation.
But I would argue that there are a number of other things judges should add to their consideration of digital fair use. Copyright owners are now asserting claims in connection with a whole range of uses that they would never have had the opportunity to control before the Internet. For example, there's no copyright issue when you read a book, because in real space, when you read a book, it doesn't produce a copy. There's no copyright issue when you set up a library, because setting up a library doesn't produce a copy. Before digital technologies, those uses were free of copyright regulation. Now, merely because the architecture of cyberspace entails that every use generates a copy, copyright owners can claim they get to control every use.
So one of the adjustments to fair use must be in response to this extraordinary-and unintended-explosion in the reach of copyright. For it is this that produces many of the issues we're seeing. Merely because an architecture of copyright law designed for the twentieth century interacts with twenty-first-century technologies in a way that produces a right to control all uses of copyrighted works, doesn't mean that policy makers should automatically accept the conclusion that this is the appropriate baseline from which to work. The baseline ought to be what makes sense.
ADLER: I don't disagree with that. But we look at it from the opposite perspective, which is that in the digital environment, suddenly users of copyrighted works are asserting that they have consumer rights. Copyright law doesn't deal with consumer rights; the market does. And consumers establish their expectations and desires in the marketplace by voting with their dollars.
The flip side of the way you've put it, Larry, would be to say that consumers and their advocates look at the functionality that digital technology has added to copyrighted works and say, "Because we can now take advantage of this new functionality, we consider it our right to be able to do so."
LESSIG: Yes, we do look at this from the other side. I think when users see the new technologies of control that digital technologies give copyright owners, they say, "That control is more than the copyright owners ought to have." We can describe the same facts by criticizing the assertion of rights on both sides. But that begs the question, What is the balance of control versus unregulated uses that makes sense in the digital context?
ADLER: Well, through enactment of the Digital Millennium Copyright Act, in 1988, Congress did recognize, at least in part, that in the digital networked environment, copyright owners should be able to control access to their works and certain uses of them in ways that owners as a practical matter did not have to control in analog form. That's what the debate about circumvention of technological protection measures is all about.
RUTLEDGE: Congress recently enacted law stating that it is not an infringement of copyright to provide or use filtering software to alter the home display of movies recorded on DVD---in other words, Congress legalized the service provided by ClearPlay, which the movie industry had opposed. While the home viewer has to purchase or rent a copy of the movie, the use of ClearPlay in itself doesn't generate any revenue for copyright owners. Scanning and displaying portions of books already in the collection of a library could be seen as an analogous use. Is there a role for Congress to play in the debate about Google Book Search and projects using similar technologies?
ADLER: You know, Congress legitimized ClearPlay by a very clever sleight of hand. The law says that model is legitimate, provided that no new copy of the film is created and no fixed copy of the altered version of the movie is made. I don't see a similar work-around here, because Google's technological process involves making copies.
RUTLEDGE: I doubt the ClearPlay subscriber is much concerned with whether a copy is being made, however. She has only to push a button to get the content she wants, on demand.
ADLER: Well, one response to that is that your average American provides a very healthy example in the marketplace of two attributes that may not necessarily be attractive in certain contexts: the desire for instant gratification of whatever wish they have with respect to content and its use, and the preference to get it for free. After all, despite the fact that the courts announced that peer-to-peer file sharing of copyrighted music was illegal, it hasn't stopped the practice.
But licensing allows us to get down to the granular level of what the user really wants. How much material? For what purpose? Within the framework of licensing, you can always resolve those issues.
Larry and others may say, "Well, yes, except that we don't want to give up fair use where we don't have to have a license." But then we're left with deciding which uses that consumers want are fair, when we could leave it to the marketplace to find a way of responding that satisfies consumers, or else the people that are in the business of producing and selling those works aren't going to make any money.
LESSIG: But that can take a lot of time and may deliver a windfall to copyright holders that consumers shouldn't have to pay. Why would any rational society choose that approach?
ADLER: Look, there has always been tension between rights holders and users of their work, long before we had to deal with the additional, nuanced problems that arose with digital technologies. It's just that in the digital environment, from both sides, it seems now that the stakes are higher. From the user's perspective, the added functionality of digital technology means the broader fair use is, the more things can be done with copyrighted works. And from the perspective of the rights holder, there are many more risks, because unauthorized reproduction and distribution taking place on a digital network can destroy a market almost instantaneously.
LESSIG: I think it is fundamentally the case that the difference between the digital and analog world is that the analog world offered three categories of possible uses of creative work: One was the core set of regulated uses that ought to be regulated in order to create incentives for authors to produce works. Another was a set of free uses that were just not regulated by the law of copyright, but which were also extremely important to guaranteeing access and the opportunity for people to develop creative works. And the middle category was fair uses, uses which otherwise would have been regulated by the law, because they triggered one of the exclusive rights, but which the courts held to be free because it was important that they be free.
The digital world doesn't have those three categories. The digital world has only two---regulated uses and fair uses. And as much as I acknowledge your concern that the amount of fair use seems to be going up because the functionality increases, it's also the case that the number of free uses has decreased. Now, presumptively, every use triggers copyright law, and before the Internet that wasn't true.
If the use of works in the Google Book Search project is not fair use, we could also ask why it wouldn't follow that Internet search engines themselves are operating illegally. The architecture in both the Google Book Search project and Google's Internet project is built on the presumption that it's OK to copy everything for the purpose of creating an index---though people can also opt out of Google's Internet index. Why is the analysis different in the two contexts?
ADLER: The publishers see these as separate issues. Websites operate on a model of implied consent. Webmasters who don't want their material to be collected by search engines can prevent that by placing it behind firewalls or by using robot text to inform a search engine to crawl away. And most search engines have accepted this protocol because it works efficiently.
Books aren't websites. We're talking about digitizing material that isn't online to begin with, and then trying to shoehorn it into the model of a website. We think the decision to go digital should be made by whoever holds the electronic rights to the work.
RUTLEDGE: Larry, from the legal perspective, has Google undermined its fair use argument by seeking permissions for the Google Print for Publishers program?
LESSIG: No, not at all. The Supreme Court has been explicit that fair use is not a function of whether you've asked permission and been denied it. Because of the uncertainty around fair use, it's often pragmatic for entities to try to secure permission, even if they believe they don't have to, in order to avoid the exposure of a lawsuit.
For many of the reasons we've talked about, fair use is an insanely complex and uncertain analysis, which complicates the opportunity of people to innovate in all sorts of contexts, not only this one.
RUTLEDGE: Allan and Nick, you are not saying there can be no fair use on the Internet, and we all know that a commercial use can be fair. So what would be a fair use of a copyrighted work in a digitized book search, if it is not what Google is proposing?
ADLER: Well, it's unclear how far it would go. The digital card catalog that was truly fair would probably have the same content as the card catalogs we already have. I'd want to spend more time than we have here thinking that through. But what I'm saying is, there are alternative ways to do what Google is planning to do. We spent a lot of time talking with them about this. But [Google] had an idea and a concept that they wanted to go forward with, despite the objections we raised.
TAYLOR: Why does the question have to be, What would an unpaid fair use be? Why can't we discuss a licensing fee?
Allan and Larry are giving eloquent expression to their competing interpretations of fair use in the digital context. I happen to agree with Allan, but as the only nonlawyer here, and the only one at the table who is not making at least a six-figure salary, I have to say that legal nuances aside, I believe the discussion should be about something else altogether-a writer's ability to make a living. I can speak to that. For the past twenty-five years, I have depended solely on my income as a writer. I have written hundreds of magazine articles, and, more recently, books both on my own and in collaboration. I have invested time and money in research and writing. I have provided what the Internet seems to demand-content. Six of my titles are in the library of the University of Michigan, which has offered up its collection to Google without regard to copyright, and which Google is scanning for its Book Search project without regard to copyright. So Google is using my work to enhance the value of its franchise as the world's leading Internet search engine.
We are all forced to seek revenue streams among the new realities created by the explosion of the Internet.
The forward-looking people in the music industry, for example, like the Future of Music Coalition, are dedicated to looking for revenue streams within the digital world. And they don't want to have to clamp down and encrypt everything they do to prevent people from switching their music files from one format to another.
I'm merely saying that if Google is going to make micropennies many times over from selling ads along the channels its users are required to navigate, I want my micropennies too.
I don't believe that Google faces an insurmountable task here. Just as the United Kingdom has created a structure for paying writers library-lending royalties, and the Scandinavian countries have created structures to pay writers photocopy royalties, we have the ability to find ways to pay authors and publishers for Google's exploitation of their works. It is unfair and absurd to say otherwise.
None of this is to deny the appeal of what Google wants to do, or to suggest that writers don't want a part in it. We all understand that, today, to be invisible on the Internet is to be invisible, period. We all want our work exposed. But exposure for no compensation, or no exposure, is not a trade-off we should be asked to make. The choice that Google has offered authors-the choice to opt out of inclusion in its search engine-is no real choice at all. Not to mention that copyright holders shouldn't have to opt out in the first place. That's not required under anybody's reading of copyright law.
RUTLEDGE: You're saying we should be talking about sustainable business models, and that makes sense to a lot of people. But because fair use is important to creativity itself, I don't see that we can or should separate these discussions.
One reply to your question, Nick, might be that we're not always going to have a deep-pocketed Google that's looking to use the work, so we can't afford to give up fair use. And even if fees weren't at issue, one can imagine scenarios where it would be extremely difficult or impossible to get permissions for all kinds of reasons, from purely administrative to political.
LESSIG: Another problem we're facing now is that we don't have enough diversity in the models for creators out there. There are certainly some creators, such as those in the music industry, whose conception of being a musician is, "My job is making money from my creativity."
But as an academic, I know there are huge numbers of people who produce literary works not because they want to get money from them, but because they want their ideas to be spread. That's their return.
TAYLOR: That's easier when you have tenure.
LESSIG: Of course. I'm not saying that academics don't get paid. They've got to be paid. But I would say that it's better that professors get paid and not try to collect money every time somebody reads their articles than it would be if they weren't paid and people did have to cough up a fee in order to read their articles---primarily because that would impose a pretty severe inequality on who could get access to information.
So far, this debate makes it sound like there are only two business models. One is, Everything ought to be free, and the other is, Everything ought to be controlled. If we can move to a simpler way for authors to make it clear what rights they want to choose to reserve, if any, I think a lot of the frustration that happens within publishing in particular would go away.
ADLER: I agree with you fully. But how do we get to those other models? Somewhat ironically, several years ago, around the time the Digital Millennium Copyright Act was being enacted, one of the worst fears of the library community was that publishers were going to conspire to abandon print and go entirely into digital formats, because that would maximize their ability to control access and use of the works.
And yet now what we're debating with Google is whether or not the authors and publishers want to see their works moved fully into a digital environment, when it's not their decision based on a business model that they've chosen, and it's not their decision based on what they see the marketplace asking of them. Instead it's the decision of a technological innovator who has decided that this serves a business purpose, and who has the advantage of having altruistic cover as well.
You know, the music community has had this wonderful ability over the years. Each new technological format that comes along supplants the previous one. Sure, you can still find people who are playing vinyl records, and perhaps you might even find some using 8-track tapes. But the reality is, cassettes knocked those things out of the market, for the most part. Then CDs knocked cassettes out of the market. Now downloading-according to the RIAA, at least-is having an adverse impact on the sales of CDs. Or take the movie industry, which is able to sell essentially the same product to consumers seven different times, starting with theatrical release and working through to the purchase of a DVD.
Unfortunately, neither of those business models exists with literary works, particularly in book format. Publishers have learned that there is a substantial part of their customer base that is not ready to give up ink on bound paper, for a variety of reasons. So they have tried to step gingerly into the digital environment while recognizing that if they want to be able to hold their market base, they still have to produce hard copy.
Because of this debate about Google Book Search, there are people saying, "The publishers are Luddites. They don't understand and are afraid of technology, so they don't want to go near it." Not at all. We're looking at our market in a clear-eyed manner, and what we're seeing is that our base is not eager to have us move everything over to a digital format.
As long as we're in that situation, the question is, Who decides when business models shift? Publishers and authors don't think Google should decide for them when to digitize their work.
LESSIG: OK, but the first question has to be whether in fact authors and publishers should control the set of uses at issue in the Google Book Search project. Rhetoric around this debate is confusing two questions. Google is not saying, "We have the right to reverse copyright law." What they're saying is, "To the extent we have a fair use right, we don't need your permission." But should the courts determine that there is no fair use right here, neither Google nor anybody serious says that it ought not to be copyright owners who exercise control.
ADLER: But that's the core of the debate, because we look at the factor of the fair use standard that talks about the impact on the actual or potential market or value of the copyrighted work. Case law recognizes that there will be new opportunities for licensing to secondary markets. The value of a book to the rights holder isn't limited to the right to sell a copy of the book. So why shouldn't the people who hold the rights to that content be able to exploit them in these new contexts we are seeing?
LESSIG: If they have the rights in those contexts.
It occurs to me that both sides of this debate have their own "potential" defense. You point to one of the factors in the fair use standard and say that if there is a potential market in which to license a work, there can be no fair use there. We look to the Sony case and say that if a technology has the potential for substantial noninfringing uses, it should be free of secondary liability. For example, P2P [peer-to-peer] technologies have the potential for substantial noninfringing uses, so the law shouldn't regulate them.
We could each learn from criticisms of the other. If all it takes is an imaginable market, there's no fair use in a digital age. On the other hand, if all it takes is imagining a noninfringing use, there's no such thing as secondary liability. So on both sides, the word potential is doing too much or too little. And this suggests we need to fundamentally rethink how our fair use system is working.
ADLER: Well, you can't make an argument to the court about a potential marketplace that is purely speculative. So there are limitations. But what this debate has exposed is that obviously there are going to be growing opportunities for unbundling books and having different partners with whom the rights holders will be able to exploit those opportunities.
RUTLEDGE: And we're back where we started. Let's end by asking about what your greatest hopes and fears are with respect to projects such as Google's Book Search.
TAYLOR: Everyone knows most authors don't make a living as authors. They have some other job or jobs that allow them the luxury of writing. And to be denied a potential source of revenue, whether the purpose is altruistic or not, is both galling and disappointing.
ADLER: There's great opportunity here. But we can arrive at the right solution only if all the affected parties are allowed to bargain for their own best interest.
LESSIG: I fear that Google will settle, de facto establishing an unnecessary, unfortunate, and possibly untenable pay-for-all-uses precedent for what could be one of the most important technology-based developments in the history of culture.