IN THE UNITED STATES DISTRICT COURT
FOR THE EASTERN DISTRICT OF VIRGINIA
Alexandria Division
______________________________________________________
)
MAINSTREAM LOUDOUN, LOREN )
KROPAT, MARY C. DUCHATEAU, )
and JOHN S. WHITE, )
)
Plaintiffs, )
)
v. ) Case No. 97-2049-A
)
)
BOARD OF TRUSTEES OF THE )
LOUDOUN COUNTY LIBRARY )
)
Defendant. )
)
______________________________________________________)
)
THE SAFER SEX PAGE; BANNED )
BOOKS ON-LINE, owned and operated )
by JOHN OCKERBLOOM; AMERICAN )
ASSOCIATION OF UNIVERSITY )
WOMEN MARYLAND; ROB MORSE; )
BOOKS FOR GAY AND LESBIAN )
TEENS/YOUTH PAGE, owned and )
operated by JEREMY MEYERS; )
SERGIO ARAU; RENAISSANCE )
TRANSGENDER ASSOCIATION, and )
THE ETHICAL SPECTACLE, )
THE SAFER SEX PAGE, et. al., )
)
Plaintiff-Intervenors, )
)
v. )
)
BOARD OF TRUSTEES OF THE )
LOUDOUN COUNTY LIBRARY )
)
Defendant. )
)
______________________________________________________)
PLAINTIFFS EXPERT WITNESS REPORT
OF KAREN G. SCHNEIDER
1. My name is Karen G. Schneider and I am a library director, writer and lecturer. I am currently the Director of the Garfield Library of Brunswick, New York. From 1996 until January 1998 I was a government contractor serving as Director of the United States Environmental Protection Agency Regional Library, Region 2, in New York City. In this capacity I developed the library website, and created and staffed a position of Web Content Librarian. From 1993-1994 I was the Electronic Resources Librarian for the Newark Public Library in Newark, N.J., and from 1992-1993 I served as the Children's Librarian/Electronic Resources Librarian of the Queens Borough Public Library in Jamaica, New York.
2. I am a member of the American Library Association ("ALA"), the largest librarians' organization in the world, and a Councilor-At-Large for the Association. I am also a member of the ALA Outsourcing Task Force. I am a member of the Library and Internet Technology Association ("LITA"), and from 1993-1996 served on LITA's Internet Room Steering Committee. I am a co-moderator of PUBLIB, an Internet-based discussion list for public librarians.
3. I write and lecture extensively in the area of library services and Internet technology. I am the author of A Practical Guide to Internet Filters (Neal Schuman 1997) and The Internet Access Cookbook (Neal Schuman 1996). Since 1995, I have written a monthly column for American Libraries called "The Internet Librarian." A complete list of my publications is contained in my curriculum vitae, which is attached to this report.
4. I received a B.A. in English from Barnard College in 1982, and an M.S. in Library and Information Science from the University of Illinois, Urbana, in 1992.
5. The Plaintiffs in this case have asked me to render an opinion on the use of filtering software in the library setting and to assess the effectiveness of such technology in blocking and/or allowing access to information. I am not being compensated for my time in this case.
6. I became interested in the effectiveness of Internet blocking software in the fall of 1996, while I was a government contractor managing the Region 2 Library for the Environmental Protection Agency. I had read the book Our Stolen Future, which examined the effect of downstream water pollution on the genitalia of alligators in Florida. While following up with an Internet search on the impact of estrogenic disruptors on aquatic animals, I realized that had I attempted an Internet search using a terminal with blocking software installed I may have been unable to locate any information on the topic because it involved reproduction and sexual organs.
7. In April of 1997 I began The Internet Filter Assessment Project ("TIFAP") as a volunteer project designed to assess Internet filter programs which block sites and/or keywords. The project resulted in the publication of A Practical Guide to Internet Filters in late 1997. The project arose following a growing concern expressed by many librarians over the use of Internet filters in library systems. The principal objective behind the formation of TIFAP was to critically evaluate the software available for blocking or restricting Internet access. Central to the concerns of many librarians involved in this project is the delegation of the library systems' deselection and selection decisions to a third party, such as a filter company, which makes such decisions based on content in determining what and when to block Internet sites or keywords, and then hides this information in encrypted lists and databases which cannot be viewed by those who license or use the software. A second objective of TIFAP was to identify variations in how librarians perceive information. Most filters use central databases, but communities vary widely in their perceptions of what is prurient or appropriate, and also vary widely in what is appropriate for specific age groups, such as adult, teen, and child. Because this is a more subjective issue, the outcomes of these perceptions were much less clear.
8. TIFAP was conducted in three phases. Phase One involved the development of a survey instrument that we believed would assess adequately the effect of filtering software on the Internet as a research resource and the questions to be used to evaluate the filters. Over 30 librarians and information specialists volunteered in the assessment phase of the project. Phase Two involved active testing of various filtering products using the survey that we developed in Phase One. Phase Three involved a further survey instrument by librarians of websites that were either blocked or let through to the browser during the Phase Two testing. We also adjusted filters for Phase Three toward their "lightest" settings.
9. We tested the first TIFAP survey instrument for over two weeks before we began submitting forms in Phase Two. We first ensured that the questions could be answered on the Internet. This was essential because the core question we were answering on the survey was, "did the filter block information you needed to answer a question?" This is a question any librarian with education and training in information services can understand. It is the equivalent of "is the book on the shelf?" We wanted to know, primarily, would the filters interfere with the ability to answer a reference question, and what kind of information the filters were blocking.
10. Phase Two of the project involved approximately 12 librarians and other participants testing 9 filters with the over 100 sample reference questions developed during Phase One. Filters tested included Bess, Cybersitter, Cyber Patrol, Cyber Snoop, Library Channel Net Nanny, Net Shepherd, Surfwatch, and Websense. We asked that testers, who were primarily, though not exclusively, reference librarians (who work primarily in question-answering services) to "search like a librarian." In other words, the guidance was to attempt locating information in a manner typical to librarians with their specialized background in information sciences and first-hand knowledge of Boolean (keyword) searching techniques, and not try to emulate inaccurate, simplistic searches such as casual library users often attempt. Responses were submitted through the question form on the TIFAP website. Three hundred and seventy seven completed responses were processed.
11. In Phase Three of the study, I culled Internet addresses of sites blocked or let through in Phase Two from the 377 forms and asked the testers to test these URLs with their filters. Any currently valid URL identified in Phase Two was included in Phase Three (and all invalid addresses were rejected). These URLs were personally tested by me and my work was double-checked by three of the librarian testers to ensure we had working URLs. For Phase Three, we designed a new survey which focused on whether or not the site was accessed, what kind of information it presented, and the librarian's assessment of the value of the information. The guidelines were to test the filters with keyword blocking disabled. Five filters, Bess, Cyber Patrol, I-Gear, Smart Filter, and Websense, were configured to only block sexually-graphic information, according to the categories used in the filters.
12. The conclusions drawn by the TIFAP study were made after taking the results of the questionnaires completed by the participants and inputting them into a software program. All programs prevented access to non-prurient, non-controversial information. All filters blocked information that librarians felt was non-prurient and inappropriately blocked. In Phase Two, with filters fully configured, 35 percent of the time, filters blocked information librarians needed to answer a question. Even with the strong bias included toward finding information, through the tester guidance, "search like a librarian," keyword blocking and site selection presented formidable obstacles to information retrieval.
13. For purposes of my testimony in this case, I purchased a copy of the X-Stop filtering program directly from Log-On Data Corporation of Anaheim, California. I also received a copy of an X-Stop program disk that I understand was provided to the parties by Michael Bradshaw at his deposition dated May 27, 1998. I personally performed an analysis of the X-Stop program (both disks) based on the TIFAP methodology.
14. Based on my experience as a librarian, my analysis of and experience with Internet filtering software in general, and my analysis of X-Stop filtering software performed in connection with this case, it is my opinion that filtering software is an inaccurate tool that fails to provide the type of "protection" asserted by its proponents and that censors valuable material that is central to the library's mission. In particular, X-Stop blocks access to a wide variety of websites that contain valuable, protected speech, yet fails to block many arguably
"pornographic" websites that the software company claims are blocked. For example, I found that the following websites were blocked: "Safe Sex - The Manual," <http://animafest.hr/filmvi/a5.html>, winner of the "Education through Humor" award at the 11th World Festival of Animated Films; "Jewelry by Ponce," <http://www.gayweb.com/113/ponce.html>, a commercial site for gay-themed jewelry; Black Iris, <http://www.blackiris.com/pride/> a commercial site for gay-themed jewelry and clothing; an educational guide to safe sex, <http://www.gai.com/text/aids22.htm>; Rainbow Mall, http://www.rainbow-mall.com/search/search.htm>, an index to gay-related sites; a historical discussion of the Stonewall uprising of 1969, <http://qrd.tcp.com/qrd/www/Stonewall25.html>; Bisexual Action on Sexual Health, a safe-sex information site, <http://theory.doc.ic.ac.uk/~kcl/bash.html>; Arrow Magazine, an electronic journal for homosexual men in committed relationships, <http://www.arrowmag.com>, which is also self-rated with the Recreational Software Advisory Council as having no sex, violence, or nudity. However, some of the sites I was able to access with X-Stop enabled included <http://www.patches.net/home2.html>, a site displaying naked women urinating, in some cases on other people; <http:/layla.net/new.htm>,a site displaying photographic images on the main page of a naked woman engaging in various sexual acts; Absolute Anal Porn, <http://www.absoluteanalporn.com/apmain.html>, a commercial sex site, the "guest area" of which displayed naked women engaging in graphic sexual acts; and <http://www.xxxnakedwomen.com/index5.html>, a site with images on the main page of naked women and men engaging in a variety of sexual encounters. In short, such software blocks much information that is far afield from any concerns about the availability of sexually explicit information, and it permits access to many websites that might be characterized as "pornographic." Additionally, reliance on such software removes local control over information resources on the Internet. In addition, neither X-Stop nor other filtering programs can filter out or permit access to websites based on local community standards.
15. My analysis of X-Stop is consistent with the reports of librarians and information specialists (of which there are few) who have contacted me about the use of this program in their libraries. Although Log-On Data Corporation claims that the product it markets to libraries blocks only obscenity, bestiality and child pornography, colleagues that have used X-Stop in libraries have reported to me that it blocked Planned Parenthood, a safe-sex website, several gay advocacy sites, and sites that would be considered risque, but not obscene, let alone felonious. Another librarian in a library system in the Midwest reported to me that her library had used X-Stop for a year and it was not functioning as advertised -- that it blocked much more than "bestiality," despite the vendor's claims to the contrary.
16. Because of my in depth study of Internet blocking software, I have concluded to the best of my knowledge that presently there is not any software program available that could be utilized and relied upon to block out only prurient materials while still allowing access to appropriate resources in a library setting. Problems with blocking are not eliminated when keyword blocking is disabled. Many of the resources not available when filters are installed are blocked because filter company employees have intentionally selected these sites to be blocked. Disabling keyword blocking improves filter performance, but does not prevent non-prurient, non-controversial information from being blocked.
17. Such software cannot legitimately be compared to "collection development," "book acquisition" or "interlibrary loan" policies. All filters have categories which are proprietary and are not consonant with the universal organizing schemes known to libraries, such as the Library of Congress, the Dewey Decimal or Superintendent of Documents (SuDoc). Determining what the categories mean varies from filter to filter, which means there are no standard guidelines for determining how to block, for example, only that speech which is not protected by the U.S. Constitution. Hidden site lists make it impossible for librarians to comply with Policy 53.1.11 of the Policy Manual of the American Library Association, which states that "access to all materials legally obtainable should be assured to the user . . ." For this and other reasons, it is not accurate to characterize the installation of a software filter as an "acquisitions policy" or as a "collection development policy." Nor is it accurate to analogize blocking software to an interlibrary loan policy. The issue of viewing what the filters block is important to librarians because it displaces work we are educated and paid to do. In all of the libraries I have worked in as a professional librarian, I intentionally selected resources for our traditional paper collection. This task was always written in my job descriptions and, when I worked as a government contractor, specified as a deliverable (required task) in the statement of work submitted by the federal government to detail the services I was to provide. If, in any of these libraries, we had used a filter to block Internet content, I would then be paying a company to not only make all of my decisions for me but hide the outcomes of these decisions so I could not see them. This is antithetical to current business practice in most libraries, and also runs counter to the practices recommended by the ALA. In this respect, relying on filtering is similar to the wholesale outsourcing by a library of all of its acquisitions decisions.
18. With my experience as a library manager responsible for supervising, training and developing policy for Interlibrary Loan (known in the profession as ILL), I know that ILL is a labor-intensive service designed to fill exceptional gaps in the library's collection. It is a service of last resort, relied on due to the limits of any one collection, and a library would, if it could, provide services in-house, rather than requesting them long distance. ILL is not a content-based service. The focus of the provision of ILL is to meet the patron's needs and interests. The role of librarians in ILL is not to intercept and examine ILL requests to see if they comply with the librarian's personal sense of appropriate reading, but to negotiate the request so it can be filled as quickly and as efficiently as possible. Furthermore, much professional guidance on ILL and on patron reading habits in general focus on the patron's right to privacy. Ia addition, librarians do not arrange ILL services so that patron's choice of reading material is on display for the world to see. Just the opposite is true. Materials, as they are processed in and out of the library, are treated with a focus on patron confidentiality.
19. In librarianship, we are careful to heed the Interpretation to the Library Bill of Rights on Labeling which cautions librarians to avoid using "prejudicial labels" on library resources. This Interpretation, written in 1951, was referring to the then-current practice of labeling books as "Communist," but it is fresh today as we consider the implications of a non-prurient website labeled by error or intent as unfit for viewing by a library user. For this and other reasons, policies that permit unblocking certain websites on request are ineffective. Patrons tend to be stigmatized by prejudicial labels associated with filtering software. Moreover, such unblocking on request policies cannot overcome the problems caused by the use of secret blocking lists that are continuously updated.
20. Another problem of relying on filtering is the adverse impact on access to information caused by technical problems with software programs. Errors can occur at the software company or at the local library that limit access to websites that were never intended to be blocked. Such errors can be difficult, if not impossible, to detect, and can cause the restriction of valuable information for extended periods of time. Such errors are not a risk for libraries that utilize less restrictive means of providing for and regulating Internet usage.
21. To the extent Internet access in public libraries presents the possibility of exposure to potentially offensive information, a number of measures are available that are consistent with established principles of librarianship and are less restrictive of informational access than filtering software. Such alternatives include the use of acceptable use policies, education of Internet users in proper search techniques, privacy screens, and other options.
22. At this time, I do not anticipate using any exhibits as a summary of my opinions.
23. I have not provided expert testimony in any cases within the past four years.
Dated June 18, 1998 Signed /s/ Karen G. Schneider