I have blogged twice before on Google Book Search (Nov. 18: “Hypos for My Copyright Class” and Nov. 11: “Fair Use and Inefficient Bundling”), but having started a new quarter of antitrust today, I am now starting to use GBS more systematically. That use has quickly raised a number of questions regarding how we will work with public domain works and with government works.
Tomorrow I am teaching excerpts from the Supreme Court’s 1911 opinion in the Standard Oil case. That case resulted in the break-up of Standard and killed off the literal reading of Section 1 of the Sherman Act announced by the Court in Trans-Missouri in 1897.
The Standard Oil opinion describes the development of the English monopoly law of forestalling, regrating and engrossing. Don’t know what those are? Me either, really, but that’s where GBS comes in.
So go to books.google.com and search on “engrossing edward VI” (a search suggested by the Supreme Court’s opinion in Standard). The first item should be the “The Penny Cyclopædia of the Society for the Diffusion of Useful Knowledge” described by GBS as a public domain work from 1843.
The second item listed is Adam Smith’s The Wealth of Nations, or, more precisely, a 2000 edition of books IV and V available from Penguin Classics for $15.00. Smith has a terrific discussion of why Edward VI was wrong to limit engrossing—described by Smith as wholesaling or the buying of corn with the intent of reselling it—and Charles II’s steps towards greater freedom for middlemen.
But in GBS, we can only see five pages of Smith’s work easily. This particular edition of The Wealth of Nations is available through the Google Books Partner Program, and the material bears the legend “Copyrighted Material.” That brings with it a two-pages forward and backward limit.
Of course, the publication of Smith’s work was the second great event of 1776 (or was it the first?), so it would be surprising if the work remained subject to copyright. And indeed, if you flip over to the second page of the search results, we find an 1869 edition of The Wealth of Nations listed as in the public domain and we have full, unfettered access to that book. For now, GBS users will have to anticipate this possibility, but Google might think about linking over from restricted works to unrestricted versions of the work.
Switch from the public domain to governmental works. The fourth item to appear in the search is Trust Laws and Competition, a 1916 work authored by the U.S. Bureau of Corporations. We see only snippets from this book as presumably the Bureau of Corporations hasn’t consented to broader use.
Actually, the Bureau of Corporations doesn’t even exist anymore and hasn’t existed for more than 90 years. It came into existence while Theodore Roosevelt was president—he hoped to use the bureau to exert greater executive control over the economy—and was folded into the new Federal Trade Commission when it was created in 1914. (Dennis Carlton and I have a draft paper for a NBER volume in which we lay out some of this history.)
We can talk generally about the difficult of tracking down authors and what that means for an undertaking like GBS, but we should start with the low-hanging fruit and Trust Laws and Competition is exactly that. Under the current U.S. copyright statute, works of the U.S. government don’t enjoy copyright protection and instead become part of the public domain. I am not enough of a copyright maven to know what the rule was in 1916 (perhaps we will find out in the comments), but if nothing else, the feds should consent to the use of these works.
Perhaps the U.S. Copyright Office can jump in and offer guidance as to how this type of consent should be implemented.
Section 7 of the 1909 Copyright Act (codified in 1947 at 17 USC §8) provided that there shall be no copyright "in any publication of the United States Government." Therefore, it's hard to imagine how "Trust Laws and Competition" would ever have enjoyed copyright protection if it was a publication of a US government agency.
Posted by: David Carson | January 03, 2006 at 09:31 AM
The problem may be at Google - maybe they should have (by their own terms) allowed full access, not realizing that this should be in the public domain?
Posted by: ADR | January 03, 2006 at 09:40 AM
My experience with Google Books is similar. It is as if Google Books has set up the system in a manner most friendly to publishers and retailers, and especially Amazon. When we search in Amazon we are not surprised that Amazon sometimes fails to direct us to less expensive copies of what we seek. (Sometimes it does so by listing paperbacks and used books directly beneath first entries, but sometimes not, and certainly not when Amazon is not offering the alternative for sale.) The "as if" may be intentional. perhaps Google's initial design aims to put forth the best case possible in litigation. As presently constituted, Google Books may well provide a windfall to Amazon, because my searches may cause me to buy more books that I previously would have. Vendors and publishers might still sue to try to get a bigger piece of the new pie, but even if they lose they may be thrilled with the development of Google Books. But the more efficient it becomes, the more existing copyright holders will lose out.
Posted by: slevmore | January 04, 2006 at 08:44 AM
Interesting post. I spoke with people at Google once about goverment works. They, as I recall, pointed out that Goverment works, while obstensibly in the public domain, often have claimed copyrights once republished by private parties, who claim copyright in formatting decisions, etc.
However weak those claims may be, given any uncertainty, any potential claim of copyright and Google usually treats the book as a copyrighted work. And since they need to do so in algorithmic fashion leading to lots of false positives (for protection).
I agree with your suggestion Randy. But how much work would it be for the copyright office to rule on what was definitely public domain? Seems like more work that google book search itself, no?
Posted by: Tim Wu | January 09, 2006 at 01:13 AM
I believe that when the CORDS program was first proposed at the Copyright Office, it was conceived as a system that would allow not only electronic filing and submission of deposit copies but also, ultimately, the digital licensing of the Library's collection. (David can correct me if I'm wrong, but it appears now to have been reconceived, as a result of technological and budgetary constraints, as principally a business process reengineering project that will save administrative costs when ultimately implemented.) But that system, or one like it, could offer a method for identifying and permanently tagging public domain material.
CORDS and the DOI Foundation's system, which is used by the publishing industry, are based on the Handle System, which was developed by one of my clients. It is a system of unique persistent identifiers of digital objects that permits the association of metadata with the object, which can include information about public domiain status, history, and license requirements, if any.
The Copyright Office's budget will presumably not permit it to consider digitizing and tagging the Library's public domain collection. Some further evolved set of technological developments and economic incentives might well lead someone like Google to consider doing so. But Google's current operation appears to depend too heavily on the cooperation of publishers who see it as in their interest to provide sales-boosting access to portions of their current copyrighted works. As a result, Google doesn't seem likely to decide to provide free online access to the arguably public domain portions of the new editions those same publishers are still selling. And isn't it worth buying a new edition of Adam Smith's works from Amazon, just to have all those "f's" translated into "s's?"
Posted by: John Stewart | January 10, 2006 at 03:07 PM