![Sample our Information Science journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5940/TOC-Subject-Sample-2021-Communication-Studies.jpg"})
Abstract
This article describes and explains the shift in the database industry's treatment of downloading. Downloading began as an unwanted by-product of new technology and became a product feature. The authors explain this shift in terms of shifts in “use-regimes,” or changes to market practices, legal rules, user expectations, and technology-based tools that shape the use of intellectual and cultural property. In the early 1980s, citation database users did not have the right to “download,” or save, citations from bibliographic databases, but by the early 1990s, citation database publishers had partnered with bibliographic citation software developers (e.g., ProCite) to make easy downloading of citations a product feature. In this article, the authors tell both the lost story of the pre-Internet downloading controversy and how and why the meaning of downloading changed over a twenty-year period. In doing so, they present a theoretical framework that is useful for analyzing changes in use rights for a variety of types of intellectual and cultural goods. Finally, the authors compare lessons from this historical case study to contemporary use right debates in the intellectual and cultural property literature.
Acknowledgments
This article has been improved by the helpful comments of many people, including Andrew Dillon, Art Elias, Eugene Garfield, Tona Hangen, Nathan Riley Johnson, Bonnie Lawlor, John Regazzi, Victor Rosenberg, the attendees of a 2009 Rob Kling Center for Social Informatics symposium (especially Ron Day, Kalpana Shankar, and Jeffrey Hart), the attendees of a 2008 Wisconsin Holtz Center for Science and Technology Studies brownbag who suggested the Farjoun article, and the insightful comments of the TIS reviewers.
Notes
1. By “bibliographic citations” we refer to information about a journal article, such as author(s), title, year, volume/issue, and page numbers. In this article we do not address the issue of abstracts. This period saw a separate debate about copyrights associated with abstracts and compensation of primary publishers for secondary publishers’ use of abstracts. Readers interested in this debate should see CitationKoch (1987).
2. For a helpful comparison of Hilgartner's and Kling's uses of regimes, see CitationMeyer (2005).
3. For example, Hilgartner defined scholarly communications regimes as consisting of producers and consumers, distribution systems, incentives, funding structures and copyright rules (CitationHilgartner 1995). Kling analyzed six elements of a sociotechnical interaction network: gatekeepers, business models, speed of information sharing, mobilization and motivation of authors, and the communication infrastructure (paper vs. e-mail) and copyright ownership (CitationKling et al. 2004).
4. This description simplifies differences between primary and secondary publishers.
5. Later in the 1980s vendors like CompuServe (1980), AOL, and Prodigy (1985) entered the market to offer online services including access to some databases (CitationLevinson 1995).
6. Early CMS included ISI's Sci-Mate (released in 1983), Personal Bibliographic Software (PBS) developed by Victor Rosenberg of the University of Michigan (released in 1982), the Cuadra STAR system (introduced in 1980), Reference Manager by Research Information Systems of LaJolla, CA, and End Note, developed by Niles and Associates. Numerous other commercial and in-house systems were also developed; we limit our discussion to those most important to our story.
7. Journal sources include Against the Grain, CD-ROM Professional, Database, Information Services and Use, Information Today, Journal of the American Society for Information Science, Learned Publishing, Library Journal, Publishers Weekly, Publishing Research Quarterly, Online, Searcher, Serials, Serials Librarian, and Serials Review.Conference proceedings include Annual Meeting of the American Society for Information Science, the Onlineconferences, and the International Online Information Meeting.
8. Handbooks include So You Want to be a Profitable Database Publisher(CitationRosenau and Chase 1983), Guide to Database Distribution(CitationBremner, 1994), and Contracts in the Information Industry I, II, and III(CitationMarx 1988; Citation1990; Citation1995).
9. Copyright protection requires that a work be “original” to an “author.” “Original” in this context means that the work owes its “origins” to the “author” and that the work possesses a minimal level of creativity (Harper & Row Publishers, Inc. v. Nation Enterprises 1985; United States Code, Title 17, §102(b)).
10. The Feist decision in 1991, during the midst of the online industry's expansion, “rocked the database world” (CitationBramson 1991). The Feistruling was the first to require originality in the arrangement of information in order for a database to qualify for copyright protection. The originality ruling was troublesome to an industry whose products often organized information in straightforward (or user-friendly) ways. As one industry lawyer described, “an entire class of valuable and desirable information products is now devoid of any substantial copyright protection …it will become open season on directories and databases that lack the requisite originality” (CitationHusick 1992).
11. Courts had held that compilations such as street directories, telephone books, and railroad timetables were copyrightable before the Second World War (Triangle Publ., Inc. v. New England Newspaper Pub. Co. 1942). Similarly, the National Commission on New Technological Uses of Copyrighted Work (known colloquially by its acronym “CONTU”), established by the 1976 Copyright Act, concluded that databases were protected, but did so without any indication that some databases might not be protected because of a lack of originality (CONTU 1978, at 38).
12. As we described earlier, there was much debate about what constituted the database publisher's legitimate “market,” because access was usually charged based on “connect time,” so anything that might reduce “connect time” could arguably be viewed as impacting the publisher's “market” (CitationBeard 1985).
13. Publishers were also extremely concerned about downloading to create competing information products (e.g., CitationHearty and Polansky 1987). We leave this aspect of reuse aside, however, because this aspect of use rights has not changed—users still are not permitted to download large numbers of citations to create competing products.
14. While we focus on bibliographic citation databases, it is important to note that downloading policies also varied by type of information contained in the database. Numerical database vendors, who aimed their products primarily at business markets, encouraged downloading by providing manipulation and display software. Their content—stock quotes, business reports—lost value quickly, so long-term reuse was not a concern (CitationHearty and Polansky 1987). However, citation database publishers tended to want to control downloading as citations could be reused for some time prior to losing value (CitationBremner 1994; CitationHemmings 1985).
15. Today, such arguments would be harder to make, since subsequent cases have enforced a contract not to copy commercially published data (ProCD, Inc. v. Zeidenberg, 86 F.3d 1447(7th Cir. 1996)) and a contract forgoing one's “fair use” rights (Bowers v. Baystate Techs., Inc., 320 F.3d 1317(Fed. Cir. 2003)). Those cases, however, were well in the future, and it was by no means certain at the time that a contract prohibiting certain types of downloading would be enforced.
16. Because MEDLINE included citations from other publishers’ material, it warned users that they might download some material “on which claims of copyright have been made” (CitationBenson and Weinberg 1985).
17. Farjoun's 1971–94 analysis of pricing structures of leading online vendors shows a drift away from reliance on connect-time pricing models. By 1994, the flat rate had become the new dominant model, and flat rate was at least an option for most vendors. Farjoun argues that the drift of pricing models away from connect time resulted from technical innovations in telecommunications that reduced the amount of time one needed to be connected in order to run a search. Further, publishers and online vendors also believed that flat-rate pricing was the best way to break into a wider consumer end-user market for information services. Past experiments had shown that these new end users would not sign up for consumer online services that employed connect-time pricing or high up-front subscription costs (CitationElias 1990; CitationO’Connor 1995; CitationO’Leary 1991; CitationO’Leary 1996).
18. Some users need to see descriptive information (such as bibliographic citations) before deciding whether or not to “use” the information, and charging for descriptive information might alienate these users. But for other customers, descriptors are the information product, and giving them away to these customers would mean giving away potential revenue (CitationHawkins 1996).
19. Searches of the literature found the following examples of metered CD-ROM products: automotive repair manuals, a financial debt and equity reporter, Harvard Business School Publishing, and a Harvard Medical Library project to use metered CD-ROMs to provide developing nation hospitals access to the medical literature (Information Today 1996; CitationHawkins 1996; CitationNathans 1996).
20. The Mosaic browser was released in 1993, and the World Wide Web was “presented” to the 1993 Online Publishing Conference (http://www.w3.org/History.html).
21. Commercial use of the Internet was still controversial, and some argued that it should not be allowed (CitationHart et al. 1992). In the early 1990s many publishers and online providers were confused about the early NSF rules restricting commercial use of those portions of the Internet funded by NSF. Some feared that it meant that customers on NSF-funded connections could not access their services or conduct commercial transactions (CitationCronin 1994).
22. The conference report envisioned a distribution system that would reduce the transaction costs associated with downloading. The system was to automatically generate a price for the download based on how the information would be used.
23. This is more of an issue for databases that serve commercial industries with significant research and development functions that may seek to avoid database subscriptions and instead conduct searches through occasional visits to a local university library.
24. Scholars conducting data mining or other types of research that requires access to large sets of citations may consider the use rights question unresolved in that they must still ask for special permission to gain access to and use the citations they need.
25. SciFinder Scholar Terms and Conditions Statement End User License Agreement (2008).
26. A simple Google search of “database cancellations budget cuts” produces a large number of recent library announcements of canceled database resources, in addition to extensive journal title cancellations.