SOME PRACTICAL APPLICATIONS OF THE LAWS OF INFORMATION 1

Jamail Center for Legal Research Tarlton Law Library The University of Texas School of Law SOME PRACTICAL APPLICATIONS OF THE LAWS OF INFORMATION 1 by Harry S. (Terry) Martin Henry N. Ess III Librarian Emeritus & Professor of Law Harvard Law School Visiting Professor of Law & Interim Director of Research Tarlton Law Library, Jamail Center for Legal Research University of Texas School of Law Address at the Fall Convocation, School of Information, University of Texas at Austin, 6 December 2008 In order to renovate Langdell Hall, the Harvard Law School closed its library; nobody cared. The most heavily used legal materials - court decisions and legal periodicals - were available full-text online. Bibliographic references for everything else were also available on the desktop. Bradford's Law states that the literature most significant to a field is found within a relatively small group of publications. This is a particular example of the well-known Pareto Principle, which states that a small number of causes is responsible for most of the effect - usually a 20% to 80% ratio. Mooers's Law states that an information retrieval system will tend not to be used whenever it is more painful and troublesome for a customer to have information than not to have it. The amount of the core legal literature available online now exceeds 20%. Online legal information is so easy to access that walking to the library to retrieve a document now seems too inconvenient. What does this imply for the continued development of large collections of law books? What does this imply for the use of legal materials not available online, particularly for historical materials or for materials from developing countries? Thank you for the opportunity to be here today. To all the degree recipients, my heartiest congratulations. The years ahead will be exciting ones for information professionals. Those of us currently in the trenches welcome your arrival. To all of your supporters friends, family, bankers my thanks as well. As a law librarian, I thought it appropriate to weave some law into my remarks today. But I ve decided to put copyright, privacy, censorship and other topics aside and look at some laws to which I wish I had paid more attention as a library student some of the laws of bibliometrics 2 that I have seen in practice over the years. I am speaking of Mooer, Bradford, and, of course, Murphy. And I am speaking from experience. For the 1996-97 academic year, in order to renovate Langdell Hall, the Harvard Law School closed its library; nobody noticed. 1 2 Presented initially at the Institute of Advanced Legal Studies, Charles Clore House, University of London, 4 December 2000. A revised and expanded version was presented at the Workshop on Technology, Legal Information, and Legal Knowledge sponsored by the Baldy Center for Law and Social Policy, State University of New York at Buffalo Law School, March 14-15, 2002 William Gray Potter, 'Of Making Many Books There is No End': Bibliometrics and Libraries. 14 J. ACAD. LIBRARIANSHIP 238a-238c (September 1988) gives simple explanations of Lotka's Law, Bradford's Law, and Zipf's Law aimed at a general library audience.

Information Laws 2 Perhaps I exaggerate. The legal historians certainly noticed when we closed our Special Collections Department, removing access to 250,000 rare books and manuscripts. The students noticed when the 800 seats in Langdell were replaced by only 100 places in the School s only large multi-purpose room. 3 But the removal of 500,000 books from the open stacks did not seem to cause much inconvenience. There are many reasons for this. Students and faculty were willing to sacrifice in order to see the main library building refreshed and modernized, with air-conditioning throughout and power and data at every seat. With plenty of advance notice, regular library users 4 were able to negotiate alternative working arrangements. Library staff organized an aggressive set of services to replace the reduced collection; for example, many staff members from special collections were temporarily assigned to document delivery duties. The core collection of American law was kept accessible in a local warehouse, with deliveries to campus twice a day. 5 The separate International Legal Studies Library remained open, so over 250,000 volumes of current foreign law were directly accessible to the all the lawyers and scholars in the Boston area who relied on this collection. Access to commercial legal databases and to the Internet played a large role. The fact that current news was available on the Web and that statutes, court decisions, and legal periodical articles were available on Lexis and Westlaw meant that the library s temporary reading room could be restricted to a reference collection of 35,000 volumes, the equivalent of the book collection then typically found at a large Wall Street law firm. 6 A decade further on and those law firm libraries have now shrunk to 90% of their former size. The library's online catalog contained records for 95% of the entire collection and indexes to periodical articles and government documents were also online. If the full text of the core collection was not available on the desktop, its bibliographic information was. For this fourteen-month period, those arrangements proved sufficient. Of course, it was a relief to have the common law collection return from storage and to unseal our rare books; but the remarkable fact is that the level of grumbling during this academic year was quite low. In fact, the Harvard Law Record, the school s student newspaper, gave the Library an "A" for service in its annual year-end report card on the School. 7 There is no doubt that careful planning, constant communication, and very hard work by the entire library staff were major reasons everything went so well. But it is also clear to me that the renovation of Langdell would have been impossible a decade earlier, when the state of legal information technology was not as advanced. Online access to Lexis and Westlaw, the maturing of the World-Wide Web, the availability of important collections on CD-ROM, all gave us significant alternatives to collections of printed books. And the technology of document delivery gave us rapid 3 4 5 6 7 The Ropes-Gray Room normally serves as the Law School s multi-purpose room, hosting alumni dinners, theatrical reviews, religious services, workshops, receptions, bazaars, and the like. I ve always been intrigued by the fact that librarians are one of only three professions who call their customers users; the other groups are information technologists and drug dealers. We were fortunate to be able to rent a floor in the former Athenaeum Press building. With a live-load capacity of 1,000 pounds per square foot capable of supporting printing presses, neither our 180,000 volumes nor the 450,000 volume-equivalents of the heavier microform collection caused posed any problem at all. In fact, I tried to get our students to think of this year as a major clinical experience. In the HARVARD LAW RECORD, Vol. 104, No. 11, p. 2, May 2, 1997 issue, the editorial Spring Report Card grades several professors and HLS administrators; the entry for the Library reads: "Who would have thought the largest law library in the world could be moved to an East Cambridge warehouse without major hassles?"

Information Laws 3 access to our off-site collections of those books. Photocopy and fax machines were heavily used, as were automated interlibrary loan systems. HLS Renovation Collection Arrangement Open Stacks Temporary Reading Room 35,000 1.7% Athenaeum House 180,000 8.7% International Legal Studies 250,000 12.0% 22.4% Media Microforms 415,000 20.0% Digital sources 150,000 7.2% 27.2% Closed stacks (accessible) Local storage 250,000 12.0% Remote storage 345,000 16.6% 28.6% Inaccessible Rare books 250,000 12.0% Remote storage 200,000 9.6% 21.6% Total 2,075,000 We learned much from renovating Langdell Hall. The lesson I wish to expand on today is the distinct preference on the part of law students for digital information, available at the desktop. This preference has only strengthened in the last decade and has significant implications for our future roles as information professionals. Preference for Desktop Access. Librarians soon learn that library users have preferences when it comes to the format and accessibility of materials. When a student comes to the reference desk with a citation and asks where to find a particular document, different answers evoke different responses. When told that a particular volume is available on open shelves on the next floor, the normal response is a cheery thank you. 8 When told the book is not in our library but can be obtained on interlibrary loan, the response is often a disappointed sigh, a pause, and then a glum, "Never mind". The librarian has just seen Mooer's Law in action. Future accessibility gap New accessibility gap Old accessibility gap Preferred Information Sources 1. Desktop 2. Books on open stacks 3. CD-ROMs on a local network 4. Local books in storage 5. Stand alone CD-ROMs 6. Microfilm 7. Books available on interlibrary loan Calvin Mooers's 1 st Law states: In an environment in which it is more painful and troublesome for a customer to have information in hand than for him not to have it, an Information Retrieval system will tend not to be used. 9 When information is not directly accessible - because it is located 8 9 Oddly enough, after a year of having off-site books delivered to the circulation desk, a few students became disgruntled at not having staff retrieve books now on open stacks. Calvin N. Mooers, Mooers Law; or why some retrieval systems are used and others are not. (Zator Technical Bulletin 136). Cambridge, MA: Zator Company, 1959. Editorial of same title in 11(3) AMERICAN

Information Laws 4 elsewhere or is stored in a format that requires intermediation - that information must be more directly relevant to our student than other more accessible documents in order for the student to keep pursuing it. The apparent need for the particular document must overcome what might be called an "accessibility gap". Only a few years ago, most of the materials in a law library's collection fell above the accessibility gap; if it was on the premises, students would use it, though the grumbling might increase if microfilm was involved. If it was off-site but could be ordered through the online catalog, it would be recalled, though the chances of actually being picked up at the library were only 90%. But travelling to or getting items from another library involved bridging the accessibility gap. Now that more legal materials are available on the desktop, the accessibility gap has moved into our own collections. Soon, I fear, materials not deliverable to the desktop will become generally neglected. The preference for all things digital is reinforced by another law of information science. Bradford's Law 10 states that the literature most significant to a field is found within a relatively small group of publications. This principle lies behind the development of the JSTOR online collection, which cherry picks the leading journals in each field. For people outside of a specialty, the JSTOR core collections provide effective access to the literature of a field. Bradford s Law is a particular example of the well-known Pareto Principle, 11 which states that a small number of causes is responsible for most of the effect - usually a 20% to 80% ratio. Conservative estimates of the volume-equivalents of legal online databases exceed 40,000 volumes - 20% of a collection of 200,000 volumes, though 2/3 ds of those books may duplicate each other. 12 More generous estimates of the size of online databases would mean that a majority of law school libraries now provide desktop access to 20% of their collections. Law firm libraries, of course, now provide virtually all of their core collections digitally and now use books simply to supplement online databases for specialized purposes or for convenience. Citation studies indicate that the vast majority of sources cited in legal scholarship are in fact available online. 13 Most law libraries have seen the piles of law reviews and court decisions stacked DOCUMENTATION i (July 1960); reprinted in 11(2) THE SCIENTIST 10 (Mar 17, 1997). Note that Mooer s Law can apply not simply to access problems but also to retrieval problems; systems that produce too much (poorly sorted) information are no better than systems that make the information difficult to retrieve. Mooer s 2 nd Law states: In an environment in which it is absolutely critical for a customer to have information, an Information Retrieval system, no matter how poorly designed, will tend to be used. Mooer s 3 rd Law states: 10 11 In an environment in which the trouble of having information versus that of not having it are fairly evenly balanced, system design and performance tend to be the deciding factors in whether or not an Information Retrieval system will be used. Samuel C. Bradford, Sources of information on specific subjects, ENGINEERING: AN ILLUSTRATED WEEKLY JOURNAL, No. 137, pp. 85-86, 26 January, 1934; reprinted in 10(4) J. INFORMATION SCI. 173 180 (October 1985). The 80/20 rule was first suggested by Vilfredo Pareto in his 1896 study of wealth distribution - see G.H. Bousquet & G. Busino, eds., 1 OEVRES COMPLETES DE VILFREDO PARETO (1064) and popularized by Joseph Juran fifty years later see Joseph M. Juran, QUALITY CONTROL HANDBOOK (1951). 12 Penny A. Hazelton, How Much of Your Print Collection Is Really on WESTLAW or LEXIS-NEXIS? 18(1) LEG. 13 REF. SERV. Q. 3 (1999). Simon Canick, Availability of Works Cited in Recent Law Review Articles on LEXIS, WESTLAW, the Internet, and Other Databases, 21 (nos. 2/3) LEG. REF. SERV. Q. 55 (2002) found 77% of the 1,984 citations in the articles reviewed were available online.

Information Laws 5 by their photocopy machines diminish. Complaints about printing have replaced complaints about photocopiers. Though much legal publishing remains exclusively in print major multi-volume treatises are just beginning to make a successful electronic appearance and the e-law book is still largely a concept 14 - the trends in legal publishing seem clear: a critical mass of current US law is now available in digital form and a critical mass of European law should be available in a few years. But this development, combined with the preferences of our students for desktop access, raises some concerns. What about legal materials from developing countries? What about historical sources? How accessible from the desktop will they be? 15 Information Technology in Libraries. Of course, information technology has transformed libraries. Automated processing, computerized acquisitions & serials control, and online cataloging through bibliographic utilities have evolved to the point where we could not manage our collections without computers. Likewise, library users have so taken to online catalogs, digital indexes and electronic reference works that even those who will not read online would not want to return to the days of manual searching. 16 Digital information has many advantages. It provides tremendous access. It is easy to share. It can store data in a highly compressed state. It is easy to manipulate, though this is also a concern. Powerful searching of reference and bibliographic tools and keyword searching of text has transformed traditional techniques of research. The new information technologies are popular and effective, though some may worry they are too popular. But digital information also has disadvantages. It is not eye-legible. It must be stored in a computer. It must be indexed by people or by smart programs. It must be formatted for different kinds of output: display on a computer monitor, printed on paper, or downloaded for editing. Many of these formats are proprietary. There is fast turnover in equipment and software, which poses problems of cost, investigation, and training. Furthermore, preserving digital media is a major undertaking. The storage media are inherently unstable. Keeping data fresh and encoded in readable file formats is expensive, labor-intensive, and currently not standardized. Conversion of traditional materials to electronic formats is also expensive and sometimes damaging to originals. The preservation of digital objects is not adequately supported by the existing copyright regime. Just look at the problem with orphan works.17 Libraries have been the traditional preservation agents. Book publishers do not keep back stocks of their publications and the microform publishers have very poor track records in storing their masters. But scholars could rely on research libraries to preserve these formats over the long term. 14 An early exception has been Peter W. Martin, MARTIN ON SOCIAL SECURITY [http://www.law.cornell.edu/socsec/martin/]. See also Learning about Social Security [http://www.law.cornell.edu/socsec/spring01/syllabus.htm#topic09] 15 In predicting a preference for a digital future, I cannot help noting a current irony that research libraries are forced to acknowledge: computers have been very powerful tools for producing more paper publications. At Harvard, the annual rate of book acquisitions has been increasing: from 18,000 books in the 1980 s to 24,000 books in the 1990 s. In some recent years Harvard added nearly 30,000 newly printed volumes. A fully digital library might seem far in the future; but the preference many of our users have for desktop access is still something we must acknowledge. 16 Pace Nicholson Baker. See Nicholson Baker, Annals of Scholarship, Discards, THE NEW YORKER, April 4, 1994, p. 17 64. Register of Copyright, REPORT ON ORPHAN WORKS (2006). http://www.copyright.gov/orphan/orphan-reportfull.pdf

Information Laws 6 Born-digital resources present a new preservation challenge, one research libraries are just beginning to address. Unlike print or even microforms, born-digital resources must be preserved early in their economic life cycle. Preservation requires storing copies of material that will not come into the public domain for decades. Moreover, libraries typically lease rather than own commercial digital resources. Digital publishers are understandably wary of letting libraries preserve, redistribute, and expand access to their digital publications before all profit has been squeezed from them. In addition, libraries are still learning how to preserve digital information. When it comes to print materials, libraries know how to acquire, catalog, shelve, circulate, repair, and generally maintain a work for future generations. Readers know how to find and use books. For digital sources, acquisition, storage, cataloging, access, and preservation are all different. How to use a digital resource is not always self-evident. Seduction of Desktop Access. The laws of bibliometrics demonstrate that a core number of publications meet the great majority of information needs. They also prove that information that is easy to access is most likely to be used. The commercial information providers have concentrated on offering this core literature to the legal profession in the most accessible way - direct to the desktop. The rest - historical materials, legal materials from developing countries, documents useful to legal services offices - must become increasingly marginalized unless those materials, too, can be made available at the desktop. Indeed, without aggressive digitization, the bar for entry to the specializations of legal historian, comparative lawyer, or poverty lawyer could be raised, further reducing the ranks of new recruits. Converting existing print collections to digital form is neither easy nor cheap. Most law libraries will avoid major conversion projects, putting increasing pressure on those willing to do something. You might think that Google will solve this problem. I certainly had such hopes after first meeting with the folks who do no evil. Google has now completed its pass through the Harvard Law School collections. After removing works in copyright, rare books, works unbound or bound with another, works too fragile to survive scanning, works too large or too small for the Google workstation, the total scanned was seven percent of the book collection. Seven percent is not a solution. Murphy s Law is not specifically a law of bibliometrics, but it makes its appearance even in libraries. Collecting born digital materials is yet another challenge. Thinking globally but acting locally, every law school should make an effort to make its own publications and the publications of its scholars accessible to other desktops. Schools can then select a few particular areas where they will host digital archives, harvesting and accepting deposit of documents from other institutions and individuals. You have your work cut out for you. The record in print must be made digitally available. It must be preserved. Born digital records must be collected and authenticated. They must be preserved. The overwhelming total must be sorted and indexed in ways useful to great numbers and wide varieties of people. The total must be available at affordable prices. The seduction of desktop access is too powerful to ignore, too right to decry, and too late to change. Your challenge is to make sure it does not shape the future of research in ways we would not like to see. Otherwise, Murphy s Law will prevail. Good luck and thank you.