Deep impact? October 31, 2008
Posted by Kevin Smith in : Copyright Issues and Legislation, Fair Use, Open Access and Institutional Repositories , 2commentsThat a settlement between publishers, authors and Google over the latter’s Book Search project was in the works was not exactly a well-kept secret over the past few weeks. Nevertheless, the announcement of the complex agreement has set many people buzzing, even before its provisions were fully digested. There is a collection of comments to be found here, on Open Access News, and Siva Vaidhyanathan gives his initial view here. As I read over the agreement, I am not sure its impact will be as deep, nor as overwhelmingly positive, as many of the commentators have suggested. There is a nicely nuanced reaction to the agreement here, from Jack Balkin of Yale Law School.
First, it is important to realize that this is a proposed agreement to settle a pending law suit. It must be approved by the court and may change in its details during that process. The plaintiff classes in this class action suit are very large, so the process of notification will be complex and it is likely that class members will object and want to discuss changes in the agreement. This is not the final word.
I also want to note up front that this settlement would not resolve the fair use argument that is at the heart of the lawsuit; the parties have been very clear that they still have a significant disagreement over whether Google’s activities to date infringe copyright or are authorized as fair use. A decision on that issue would have provided libraries with more guidance as we proceed (or not!) with digitization projects, but both sides in the case, I suspect, wanted to avoid getting to that point. The likely result, unfortunately, is that the next time someone considers pushing the envelope on fair use, there will be even more pressure to just pay the costs of licensing up front and not go down the fair use path at all.
Under this agreement, it seems likely that the availability of in-copyright but out-of-print books would improve in the Google Book Search. Google would be able to show both the “snippet view” for such works that is already available and a “preview” view that would display up to 20% of a work, although no more than 5 adjacent pages and not the last pages of a work of fiction. For out-of-print works this would be the default availability, with the rightsholders able to opt out. For in-print books, the rightsholders would have to opt-in. So while it seems likely that, overall, there will be increased access in the Google Book Search product, some in-print works will also likely disappear, even from the snippet view, as rightsholders elect not to opt in.
The participating libraries are in an interesting “in-between” position here. They have no voice in the settlement agreement, and it appears that, for some of them, the options for using the digital scans of books that they receive from Google will be reduced. That depends on how their original agreements were worded, and that wording seems to have varied among the partner libraries. Under this proposed settlement, the libraries that provide books for scanning can receive digital files for any title they hold in their collections, even if they did not provide the copy of that title that was actually scanned. But there are strict limits on how those files can be used. They cannot be made available for reading even on campus, much less linked into a catalog. They cannot be used for interlibrary loan, e-reserves or in a course management system. They are essentially preservation copies, although there is a provision to allow research based on “text-mining.”
All libraries, of course, will be able to purchase institutional subscriptions which will give them access to the full text of many in-copyright works which publishers decide either not to opt out of this use (for out-of-print books) or which are opted in (for in-print works). We do not know much about the pricing structure yet, but, given the rather small amount of money changing hands at settlement, I think that the publishers are counting on making significant profit here. It will be especially interesting to see if some of the partner libraries choose to subscribe to this more robust version of the database to get the level of access that is denied to them with the scanned files of their own works.
Consumers will also be able to purchase digital copies of individual titles; the pricing structure could allow prices anywhere from $2 to $30 per title, but that structure will undoubtedly undergo further revision.
Finally, there are provisions for free access to this “fuller-text” version of the Google product, via dedicated terminals. One such terminal would be offered to every public library, although it is not clear if public libraries that still lack broadband access would benefit much from this offer. A free terminal would also be available to “colleges and universities,” with one such terminal for each 10,000 FTE (one per 4,000 for community colleges). I am sure that the exact definition of what is a college or university for this purpose will be a matter of some debate. It is also interesting that no allowance is made for free access at the K-12 level.
For all three of these approaches to “access uses,” there are pretty strict limits imposed on cutting and pasting, and on printing.
Overall, I believe this agreement would increase access to a lot of books that are currently hard to find or even to know about. But there are significant strings attached to that access; for most people, it will probably come with a hefty price tag, which was not part of Google’s original, Utopian vision for its project. The strict limits on access, both to the libraries’ own digital copies of books and to the public “access use” versions, seem to be what led Harvard to decide to continue to withhold in-copyright works from the project and remain at its limited level of participation. Most troubling to me, however, is that this agreement would seem to move us one more big step in the direction of per-pay-use, where every library resource would be licensed and metered.
Just ’cause you’re paranoid… October 26, 2008
Posted by Kevin Smith in : Open Access and Institutional Repositories, Technologies , add a commentWhen I wrote a post about a week and a half ago called “Can Copyright kill the Internet?,” I worried that my title might be considered a little bit extreme. After all, the Internet is a big, sprawling “network of networks;” surely the puny efforts of legal enforcement cannot really do that much harm. In some senses this is true, since it is difficult to apply national laws to the persistently international Internet. On the other hand, as I pointed out in the earlier post, a business wanting to engage in commerce on the Internet has to take account of national laws around the world, and is frequently circumscribed by the most stringent law to be found regarding any particular activity.
But what really convinced me that my earlier post was not exaggerating the threat was this news item from Ars Technica called “‘Net filters “required” for all Australians, no opt-out.” Incredibly, to my mind, at least, Australia is moving ahead with a plan to force Internet Service Providers to impose filters on ALL Internet access in the country to filter out “illegal” content. The government would maintain two “blacklists” of material that must be blocked. Australians who requested “unfiltered” access would not have material on the “additional material” blacklist blocked, but there would be no way to get access to Internet sites that the government deemed illegal and so put on its prinicple list of blocked content.
There are many problems with this plan, but I want to focus on two. First, filters never work. It is usually possible to get access to “bad” content in spite of the filter, and filters almost always filter out useful content as well as the bad stuff. In the case of this plan, the task of updating the blacklist will be monumental, as banned material can switch locations and URLs faster than the content police can keep track. And even when content is blocked, the blocking itself will serve as a challenge to many sophisticated computer users to find a way around the filter and gain access to the site. Digital locks are usually broken in a matter of days, and the unfortunate result of filters has always been that law-abiding users find their choices of legitimate content constricted, while those who want to violate the rules find ways to do so.
The other problem, of course, is deciding what consititutes “illegal” material. Few would dispute the need to reduce the amount of child pornography available on the ‘Net, but there are lots of other categories of sites where there is a legitimate debate. What is defamatory in some countries, for example, is protected as political speech in the United States. Will Australian officials be able to keep criticism of government policies (like this) off of Australian computers by declaring it “illegal” because potentially libelous? What about material that potentially infringes copyright? Will all such material be blocked? And how will that determination be made? Many sites — YouTube is the most obvious example — contain material that is authorized by the rights holder as well as videos that are clearly infringing. Is YouTube a legal or an illegal site?
Ars Technica has followed up its original post with this one noting that the government in Australia is trying to suppress criticism of its plan. This strengthens the fear that the filtering plan might be used to silence opposition, even though there ought to be a clear distinction made between what is illegal and what is merely dissent. The article also notes that the point made above — that filters seem seldom to work very effectively — is being borne out in this instance.
So here is a concrete example of terribly bad policy that really does threaten the existence of the Internet as the revolutionary tool for democratic communication that it ought to be.
What does PRO-IP really do? October 22, 2008
Posted by Kevin Smith in : Copyright Issues and Legislation, Technologies , 2commentsPresident Bush signed the “Prioritizing Resources and Organization for Intellectual Property Act of 2008″ — PRO-IP — on October 13, making it Public Law 110-403. Since then a lot of news reports and blog posts have denounced the law, and I have noticed that a number of them claim negative aspects of the bill based on previous proposed versions. One article last week linked to a report about the bill that was a year old and announced an aspect (about which I also wrote way back than) that actually was removed from the bill as it was finally passed and signed. So I spent my weekend reading the actual text of the final, adopted version to see what was and was not still there. The link above, from Washington Watch, includes both the text of the bill as signed and some analysis of it; here is a news report that also reflects the content of the bill correctly..
First, what is not in PRO-IP? The two most objectionable features, from my perspective, were both removed before final passage. First, earlier versions included provisions that would have dramatically increased the statutory damages available in copyright infringement cases. The obvious purpose of this provision was to make more money for the RIAA when it sues file-sharers, since the structure of the change would have increased the potential penalty for infringing a music CD by 10 or 12 times. That provision was not included in the final version. Also dropped was a provision that would have allowed the Justice Department to pursue civil (as opposed to criminal) copyright lawsuits, a provision one commentator called making federal employees essentially pro bono lawyers for the content industries. Because the Justice Department itself objected to the provision, it was omitted as well.
So what is left? Plenty of taxpayer money being spent to help out a few large content industries is the short answer. The Congressional Budget Office estimates that PRO-IP will cost over 420 million over four years.
PRO-IP has five sections. The first, dealing with civil enforcement, lowers the procedural barriers for bringing infringement lawsuits, and it allows for seizure and impounding of allegedly infringing products while the lawsuit is pending. It also raises the statutory damages available for counterfeiting of trademarks. The second section “enhances” criminal enforcement measures in a parallel way. Primarily, it allows for the seizure and ultimate forfeit of infringing goods and any equipment used to infringe. The potential effect here is that computer equipment used for widespread and wilfull infringement could be seized in exactly the same way that cars and boats used for drug crimes are now taken by law enforcement.
With sections III and IV, PRO-IP really starts spending your money; over 55 million dollars a year is explicitly appropriated to increase federal and local enforcement efforts. At the top, a new executive branch official is created — the Intellectual Property Enforcement Coordinator, or IP Czar, as the position has been called — whose job is not to seek balance in our copyright law, as is arguably the role of the Registrar of Copyright, for example, but directly to expand the role of the federal government in protecting these private rights. The section also creates a new enforcement advisory committee, replacing an earlier group with one whose membership is significantly expanded. This group is specifically charged with gathering information about the alleged cost of IP infringement that is used by the industry in its lobbying efforts. Now taxpayers will pay for that research. Indeed, this federal official is essentially a Cabinet-level lobbyist for Big Content.
PRO-IP also requires the addition of over a dozen FBI agents to full-time IP enforcement; it is not clear if these are new agents or ones who will be reassigned from less high priority duties. Twenty-five million dollars are also allocated for grants to local law enforcement to pursue those dangerous file-sharers, and 20 million to hire more investigators for the Department of Justice. The bill closes with a “sense of Congress” section that heaps great praise on the content industries and repeats much of the propaganda that those industries distribute to support their claim that federal intervention to protect their out-dated business models is necessary. It also informs the Attorney General of the United States that IP enforcement should be “among his highest priorities.”
As is probably clear, I think PRO-IP is still bad legislation. The provisions that most threatened to have a further chilling effect on higher education have been removed, but the bill still, in my opinion, is a huge gift of money to the major content industries. The result will be that taxpayers will shoulder even more of the burden of fighting their desperate battle to prop up a business model that both consumers and the technologies they use have passed by. Instead of looking for new ways to enhance and market their products, these industries continue to resort to legal enforcement that is bound to fail (see this report from the Electoronic Frontier Foundation on the fruitless campaign of the past five years), and they have now convinced Congress to invest much more taxpayer money in that effort.
OA @ Duke — why it matters very much! October 17, 2008
Posted by Kevin Smith in : Open Access and Institutional Repositories, Scholarly Publishing , 1 comment so farAs part of our Open Access Day celebration at Duke, we held a keynote and panel event on Tuesday, Oct. 14th featuring Duke faculty and a student talking about why open access is important to them and important to Duke. About 50 staff and faculty members attended, and following is a brief summary of the very exciting talks we heard.
Prof. James Boyle of the Duke Law School and the board of Creative Commons began the afternoon with an entertaining and inspiring talk on why Open Access matters. He pointed out that the Web, which was designed to share scientific information, now works very well for sharing pornography or bargain offers for shoe shoppers, but really is not very effective at sharing science. The message of his talk was “Its the links, stupid” — the ability to build links into scientific work is key to speeding up the progress of science and innovation to the pace promised by this powerful technology. Linking permits all kinds of new discovery, whether through text mining or “waterhole searching” (following the tracks of other). But linking depends on information being freed from the access barriers that currently wall off most scholarship on the web.
Boyle offered a vision for open access based on three stages. At “Open Access 1.0,” scientific research and information will be exposed to many more human eyeballs. At the stage of Open Access 2.0, computers will have access to a depth of scientific information that will permit text mining for new and serendipitous discovery. Finally, with Open Access 3.0 computers and humans will work together to create a map of knowledge within in a given field and amongst fields where relationships were previously not discoverable.
Law School Assistant Dean for Library Services Melanie Dunshee followed Boyle with some interesting information about Duke Law’s ten-year-old experiment with open access to legal scholarship. Her talk gave a nice illustration of the path to open access, which consists in aligning faculty interests with the mission of the university to produce and disseminate knowledge. The services provided by the Law School Library, and the many new ways that faculty scholarship is exposed and promoted, made the point about how to accomplish that alignment very concretely.
Next up was Dr. Ricardo Pietrobon from the Medical School, where he chairs the group that is doing “Research on Research.” His presentation really built on Boyle’s call by suggesting that we need to move beyond text mining and data mining (once we get there) to consider what he called “scientific archeology.” Only at that point, when open access encourages not just access but replicability, accountability and transparency, will the promise of the Internet for scientific learning be fulfilled.
The climax of the afternoon, and what made the need for open access very real to our audience, was the remarks by Josh Sommer, a Duke student who was diagnosed with a rare form of brain tumor during his freshman year. Now three years out from surgery, Josh has refused to accept the “average” seven year life span of chordoma patients that he was given. Instead, Josh has co-founded the Chordoma Foundation and has himself become actively involved in research to understand and treat this disease. His story of how the privileged access he has as a Duke student has helped significantly in his research is only part of the story. He also tells of previously unknown connections between other forms of cancer research and the effort to treat chordoma that have been discovered using open access medical literature. Finally, Josh talked about his young friend Justin who died from chordoma earlier this year; a young man who did not have the advantages that have given Josh the ability to fight his grim prognosis (see the link above for more on Justin’s short life). As Josh puts it, there is no reason that the knowledge that could have saved Justin’s life is walled off behind access barriers. Josh Sommer personified for our event the very message he wanted to deliver to those engaged in the effort to acheive more comprehensive open access to knowledge — perseverance.
Can Copyright kill the Internet? October 15, 2008
Posted by Kevin Smith in : Copyright Issues and Legislation, Fair Use, Technologies , 3commentsThe question seems extreme, and it is certainly rhetorical. But the potential for copyright challenges to significantly limit the range of activities and services available on the Internet is very real, and severe limits on the full potential for digital communications could be imposed.
One of the great strengths of the Internet — its completely international character — is also one of its greatest weaknesses. Since laws change across national boundaries, but the Internet goes merrily along, online services can potentially be made subject to the most restrictive provisions found anywhere in the world.
In the US, for example, there is solid case law holding that thumbnail versions of images used in image search engines are fair use. The cases of Kelly v. Arriba Soft and Perfect 10 v. Google are solid examples of this principle. But fair use is a fairly unique feature of US law; it does not exist in most other countries. So when Google’s image searching was challenged in a German court on copyright infringement grounds, they did not have fair use to rely on for their defense, and they lost the case earlier this week. The German court held that this valuable tool infringes copyright if the thumbnail images are used without authorization, even if the use is to provide an index that helps users actually find the original. There are reports about the decision here and here.
How will Google react to this decision? First, they will almost certainly appeal. It is possible, ultimately, that they would have to employ some kind of technological measures that would prevent users in Germany from seeing the image search results with thumbnails, a result that would ultimately harm business in Germany more than Google. It is very unlikely that Google would have to shut down its image search feature, but multiple decisions might force a reexamination of how Google provides services worldwide. A similar case, involving the sale of Nazi memorabilia in France, led Yahoo to exactly that sort of system-wide change.
The general lesson here is that the current copyright regime throughout the world is in a fundamental conflict with the openness and creativity fostered by the Internet. Most companies today want to do business on the Internet, but few are willing to embrace the fundamentally open nature of the medium. The resulting conflict really does threaten to constrict the role the Internet can play in our lives.
The conflict is the subject of an interesting article from The Wall Street Journal by Professor Larry Lessig of Stanford, a short teaser for his forthcoming book “Remix.” Lessig suggests that the copyright “war” over per-to-peer filesharing risks significant “collateral damage.” That damage would come in the chilling effect that frivolous lawsuits and poorly-researched DMCA “takedown notices” could have on new forms of creativity and art — the products of the remix culture which, Lessig argues, offers a return to an era when amateur artists could thrive. This culture offers “extraordinary” potential for economic growth, according to Lessig, if it is not choked off by aggressive enforcement directed at a very different activity. To prevent that, he offers five changes that could make our copyright law less of a threat to the innovation and creativeity the Internet encourages.
Will copyright kill the Internet? No. But copyright will need to be revised to account for the new opportunities that the Internet creates, lest we find ourselves unable to exploit those opportunities.
PS — This story about the McCain/Palin campaign fighting back against DMCA takedown notices that are being used to force YouTube to remove campaign videos that contain short clips from news programming, is another example (if we needed on) of the potential for abuse of the copyright system to chill important speech on the Internet. Good to see the McCain camp fight back, but I wonder if it is really YouTube’s job to evaluate the merits of the takedown claims. A court recently told content owners that they must consider fair use BEFORE sending a takedown notice; I wonder if the better course isn’t to pursue some kind of sanctions against those who send clearly unwarrented notices.
