COPYRIGHT AND COPY-RELIANT TECHNOLOGIES Matthew Sag* Abstract This article studies the rise of copy-reliant technologies – technologies such as Internet search engines and plagiarism detection software that, although they do not read, understand or enjoy copyrighted works, necessarily copy them in large quantities. This article provides a unifying theoretical framework for the legal analysis of topics that tend to be viewed discretely. Search engines, plagiarism detection software, reverse engineering and Google’s nascent library cataloging effort, are each part of a broader phenomenon brought about by digitization, that of copy-reliant technologies. These technologies raise two novel, yet central, questions of copyright law. First, whether a non-expressive use that nonetheless requires copying the entirety of a copyright work should be found to infringe the exclusive rights of the copyright owner. Second, whether the transaction costs associated with copy-reliant technologies justify switching copyright’s default rule that no copying may take place without permission to one in which copyright owners must affirmatively opt-out of specific uses of their works. Keywords: Internet, Copyright, Non-expressive use, Opt-out, Fair use, Transaction costs. Word-count: 34,987. * Visiting Assistant Professor, University of Virginia School of Law & Assistant Professor, DePaul University College of Law. Please address comments to [email protected]. Thanks to Tonja Jacobi, Adam Mossoff, Dotan Oliar, Glen Robinson, and Christopher Sprigman and to participants at the American Intellectual Property Law Association Annual Meeting, and the Intellectual Property Scholars Conference for their comments and suggestions. Special thanks to the University of Virginia School of Law Library staff for their invaluable assistance. This limited circulation draft is dated July 23, 2008 and is subject to revision by the author. Citation to this version of the document may be premature. A revised version of this document will be posted to SSRN.com in late August, 2008. COPYRIGHT AND COPY-RELIANT TECHNOLOGIES CONTENTS Introduction......................................................................................................................... 3 I. Copy-Reliant Technologies and The Internet............................................................. 5 A. New Technologies, Copyright Markets and Copyright Law.................................. 6 B. Four Case-studies of Copy-reliant Technology .................................................... 10 II. The Doctrinal Implications of Non-Expressive Use................................................. 17 A. The Principle of Non-Expressive Use................................................................... 17 B. Doctrinal Incorporation of Non-expressive Use................................................... 28 C. Fair Use and Non-Expressive Use ........................................................................ 32 III. The Doctrinal Significance of Transaction Costs ................................................. 43 A. Transaction Costs and Copy-Reliant Technologies.............................................. 43 B. Transaction Costs and Property Rights................................................................. 52 C. The Significance of Opt-outs in Fair Use Analysis .............................................. 58 Conclusion ........................................................................................................................ 64 2 COPYRIGHT AND COPY-RELIANT TECHNOLOGIES INTRODUCTION Although we have been living in the Internet age for more than a decade now, its implications for copyright law and the fair use doctrine are only just beginning to manifest.1 By expanding the breadth, diversity and sheer number of copyrighted works in existence, the Internet has fundamentally changed the nature of copyright markets. This change is most significant in the context of what I term “copy-reliant technologies” — technologies that copy expressive works for non-expressive ends. Copy-reliant technologies, such as Internet search engines and plagiarism detection software, do not read, understand or enjoy copyrighted works, but they necessarily copy them in order to process them as grist for the mill, raw materials that feed various algorithms and indices. The copyright implications of Internet search engines, plagiarism detection software, reverse engineering of software and the emerging Google Book Project controversy have been separately considered by other scholars.2 This article is the first to provide a unifying theoretical framework for the analysis of the issues raised, largely because it is the first to recognize them as sub-parts of a broader phenomenon. Copy-reliant technologies tend to interact with copyrighted works by copying them routinely, automatically and indiscriminately. These technologies are vital to the operation of the Internet, but they are vulnerable to claims of copyright infringement at key stages of their operation. Copy-reliant technologies typically display three significant traits: the copying of expressive works for non-expressive uses; a high volume of transactions; and the use of technologically enabled opt-out mechanisms to reduce transaction costs. 1 I use the term Internet age here to refer to the period from 1994 to the present – the period in which the Internet was popularized and commercialized. Technically, the first packet-switching node of what would later be called the ARPANET went live on October 29, 1969. The first TCP/IP-wide area network was operational by January 1, 1983, when the United States' National Science Foundation (NSF) constructed a university network backbone that would later become the NSFNet. 2 On search engines, see Urs Gasser,Regulating Search Engines: Taking Stock and Looking Ahead, 9 YALE J. L. & TECH. 124 (2006); James Grimmelmann, The Structure of Search Engine Law, 93 IOWA L. REV. 1 (2007). On reverse engineering, see, Pamela Samuelson & Suzanne Scotchmer, The Law and Economics of Reverse Engineering, 111 Yale L.J. 1575 (2002).On Plagiarism, see, Samuel J. Horovitz, Two Wrongs Don't Negate A Copyright: Don't Make Students Turnitin If You Won't Give It Back, 60 FLA. L. REV. 229 (2008). On Google Book, see e.g. Siva Vaidhyanathan, The Googlization of Everything and the Future of Copyright, 40 U.C. DAVIS L. REV. 1207 (2007); Oren Bracha, Standing Copyright Law on Its Head? The Googlization of Everything and the Many Faces of Property, 85 TEX. L. REV. 1799 (2007); Emily Anne Proskine, Google's Technicolor Dreamcoat: A Copyright Analysis of the Google Book Search Library Project, 21 BERKELEY TECH. L.J. 213 (2006); Elisabeth Hanratty, Google Library: Beyond Fair Use?, 2005 DUKE L. & TECH. REV. 10 (2005); Kinan H. Romman, The Google Book Search Library Project: A Market Analysis Approach To Fair Use, 43 HOUS. L. REV. 807 (2006); Hannibal Travis, Google Book Search and Fair Use: Tunes for Authors, or Napster for Books?, 61 U. MIAMI L. REV. 87 (2006); Steven Hetcher, The Half-Fairness Of Google's Plan To Make The World's Collection Of Books Searchable, 13 MICH. TELECOMM. TECH. L. REV. 1 (2006); Michael R. Mattioli, Opting Out: Procedural Fair Use, 12 VA. J.L. & TECH. 3 (2007). 3 COPYRIGHT AND COPY-RELIANT TECHNOLOGIES The rise of copy-reliant technologies exposes seemingly novel questions. First, should a non-expressive use that nonetheless requires copying the entirety of a copyright work be found to infringe the exclusive rights of the copyright owner? Our historical intuition is that when a work is copied it is copied to communicate at least some part of the work’s original expression: books are copied to be read, not to serve as paper weights; compact discs are copied to be played, not to function as drink coasters. This Article concludes that because the copyright owner’s exclusive rights are implicitly defined and limited in reference to expressive communication to the public, acts of copying which do not communicate the author’s original expression to the public should not be held to constitute copyright infringement. The second important question raised by copy-reliant technologies springs from an empirical observation. The architects of many of the copy-reliant technologies surveyed in this Article have chosen to build in technologically enabled opt-out mechanisms that preserve the autonomy of the copyright owner, but switch the default rule from ‘no copying without permission’ to one in which copyright owners must affirmatively opt-out of specific uses of their works. Accordingly, the second question whether this switch in copyright’s ordinary default rule is justified from either a doctrinal or a utilitarian perspective. The questions that come into focus in this study of copy-reliant technologies are to a large extent questions about fair use. Technically, the fair use doctrine renders certain otherwise infringing actions relating to copyrighted works non-infringing.3 More generally, fair use allows the use of copyrighted works without permission; as such it performs a vital function in the modern copyright system by establishing limits on the otherwise expansive rights of copyright owners.4 Because of the fair use doctrine’s pivotal role in adapting copyright law to new technology, it is inevitable that any examination of copyright and new technology becomes a reflection on the nature of fair use. This Article makes a significant contribution to our understanding of the fair use doctrine by explaining its application