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. )
)
______________________________________________________)
EXPERT REBUTTAL REPORT
OF KAREN G. SCHNEIDER
I have reviewed the report of David Burt submitted on behalf of the Loudoun County Library Board of Trustees on this case. For the reasons detailed below, I strongly disagree with Mr. Burt's analysis. The opinions expressed below are based upon my review of his report and upon my professional experience as described in my initial expert report. Additional reference works that I consulted are listed at the end of this report.
Comparing Internet Filtering to Library Acquisition and Selection Policies
1. Mr. Burt erroneously states that the use of filtering software is fully consistent with the theoretical and ethical framework of public librarianship. Contrary to Mr. Burt's report, the locus of librarianship is access, not exclusion. This is captured in S.R. Ranganathan's First Law of Librarianship: "books are for use"1/ (Gorman 1997). Just as librarians would not tell a patron that they may not sit at a table and read a book they have brought in to the library, so too it is inconsistent with librarianship to limit patrons' access to the Internet. Furthermore, as a profession, librarianship is inclusive and ecumenical, and supportive of the right to read and the right to privacy.
2. Mr. Burt's claim that librarians select materials on the basis of content does not demonstrate a clear understanding of public librarianship. First, a brief summary of several key documents used in public librarianship makes it clear that the theoretical and ethical framework of traditional materials selection is use-based and need-based. It is true that contents plays a role in material selection. In addition to selecting materials based on popularity, librarians actively seek materials of high value to the community--such as accurate reference work, novels with glowing reviews, and children's picture books with award-winning art. However, librarians do not assume that our sincere attempts to tailor information resources in paper formats to community needs and interests are perfect, absolute, or binding. It distorts the nature of library work to suggest that librarianship encourages the creation of collections exclusive of materials librarians find controversial or offensive.
3. The flaws in Mr. Burt's report are clearly revealed by reference to the core materials used in public librarianship. Since 1987, the guiding document for public library management has been Planning & Role Setting For Public Libraries: A Manual of Options And Procedures, written by a team of librarians led by Charles R. McClure, now a Distinguished Professor at the School of Information Studies, Syracuse University and published by the American Library Association (ALA) (McClure 1987a). A companion document of powerful influence, published the same year, is the second edition of Output Measures For Public Libraries: A Manual Of Standardized Procedures, also published by ALA by a team of librarians including Dr. McClure and several others who had helped write Planning and Role Setting (McClure 1987b). Planning and Role Setting helps public librarians decide which roles in the community they can best fulfill; Output Measures provides tools to help public librarians measure their services and to provide consistent measurements across librarianship.2/
4. The eight major roles suggested for librarians in Planning and Role Setting are community activities center, community information center, formal education support center, independent learning center, popular materials library, pre-schooler's door to learning, reference library, and research center. Librarians are expected to select and prioritize roles based on assessments of their community's needs. The traditional paper-based collection is just one of the tools for meeting these needs. Nowhere in Planning and Role Setting is it suggested that traditional material selection should "discriminate on the basis of content." Just the opposite is true: according to the guidance of Planning and Role Setting, the collection is driven by the roles the library assumes--in other words, by the needs and interests of the library's community. For example, the guidance for collection development in the popular-materials role is that "the collection includes current and popular materials in a variety of formats, with sufficient duplication to meet demand." Furthermore, the suggested output measures are also content neutral. Turnover (the number of times a book is checked out in a year), browser's fill rate (how successful browsers are at finding books they want to read) and circulation per capita are just three of the content-neutral criteria used to measure excellence in library service. Reviews of popular literature in librarian review sources such as Booklist and Library Journal reflect this emphasis on the projected popularity of material; comments such as "popular" and "buy where demand warrants," and references to book signings and television tie-ins, are common.
5. Output Measures elaborates on the measurement of success, and strongly emphasizes the importance of measuring outcomes in terms of how well the library has met its role and served its communities. The formulae for determining material use and patron satisfaction, including end-user surveys, underscore the professional emphasis on relevant services. It is a truism in librarianship that the very largest urban libraries compete among themselves on the basis of total circulation.
6. Nowhere is the content-neutral ethos of collection development more obvious than in weeding, the process of removing materials (sometimes known as deselection or deacquisition). The CREW Method, popularly known as the CREW Manual, is the standard guide for weeding library collections (Boon 1995). CREW is an acronym for Continuous Review, Evaluation and Weeding, and the CREW method is intended to keep library collections relevant to the library mission. Weeding, as the CREW manual emphasizes, is essential for "a good, useful community library," as it saves space, improves the collection, finds books that need repair or replacement, and saves the time of the patrons by increasing the potential that on any given shelf a browser will find a book she or he wants.
7. In the CREW manual, the rule of thumb is X/Y/MUSTIE, where X equals the date of copyright (variable, depending on the Dewey class), Y is "maximum permissible time without usage" within Y years, and MUSTIE is an acronym referring to five factors than can influence the decision to weed: a book may be Misleading, Ugly, Superseded, Trivial, Irrelevant, or Elsewhere (the material "may be obtained elsewhere through Interlibrary loan or reciprocal borrowing"). For example, the guidelines for Dewey 640 are 5/3/MUSTIE: a book in the 640 class that is at least five years old, and has not circulated for at least 3 years, or meets at least three of the criteria in MUSTIE, is a candidate for weeding. The CREW manual comments, "be strict with old sewing and grooming books in which styles change rapidly; however, keep cookbooks unless little used; replace worn popular titles." (The CREW Manual applies similar guidelines to tangible media housed by the library such as videos, but is silent about the Internet.) The CREW manual actually proscribes weeding based on discrimination; the general advice for the 300s--where many topics related to sexuality and family life are found--is to "see that controversial issues are represented from all viewpoints and that information is current, accurate, and fair."
8. There are two observations from this discussion that apply to the issue before us. First, the driving factors in weeding are the utility and potential circulation ability of the books, and the significant confounding factor is that books are expensive to keep and maintain. Far from encouraging librarians to use weeding as an opportunity to bring library collections in line with their personal biases, the CREW manual actually encourages adding "controversial" materials to library collections to ensure multiple viewpoints on issues are available. Second, book-based collections have severe physical limitations not found with other types of materials available in libraries, and the guidelines for managing paper-based collections are peculiar to the format. It is with no small angst that librarians weed books they are personally fond of that have nonetheless never circulated or are seriously outdated in their advice. Nevertheless, we do not have the luxury of allowing our personal bias to interfere with the greater good of library service to the people in our communities.
9. Mr. Burt quotes Carol Tenopir, an academic librarian, who suggests that it is telling that librarians provide special policies for electronic media. Yet libraries rarely have omnibus policies addressing all areas of public service. Instead--perhaps because librarians, by habit, try to make information as accessible as possible--most libraries divide larger policies into small, easily digestible chunks of specific guidance, for the mutual convenience of library employees and library patrons. That is why libraries will have policies on meeting rooms, which technically could be covered under facilities, or circulation, which could be covered under the larger category of technical services. Librarians, as a group, are sensitive to Fourth Law of Librarianship (save the time of the reader); we do not send our patrons on fishing expeditions in cumbersome policy manuals to determine what our guidelines are. An example of this typical approach can be found on the Waltham, Massachusetts Public Library website, which lists policies for faxing, use of the piano, and use of the study rooms, as well as use of the Internet (of which the Library observes, "due to the nature of the Internet, it is impossible to implement selection criteria as the Library has done for its circulating and reference collections") (Library, W.P. 1997).
10. The 125 Internet policies Mr. Burt has collected on his library's website are largely directed at patron behavior, not content restrictions. Library policies are directed at patron behavior, not to prevent patrons from reading information which librarians disapprove of. Some examples of problem patron behavior include pushing and shoving, selling drugs or holding conversations via cellular phone in quiet areas (Schneider 1998a). A standard handbook for problem patrons, Patron Behavior in Libraries, repeatedly emphasizes addressing the issue of behavior--such as noise, smoking, theft, drinking, mutilating books, and so forth (McNeil 1996). If librarians as a profession were concerned about what patrons read versus what patrons do, it would be routine practice to check bookbags on entry, not departure, to ensure compliance with library mission statements, and to patrol library tables, peering over shoulders to determine if reading material complied with library guidelines.
11. Certainly patrons are capable of misbehaving with respect to the Internet; they are capable of misbehaving with all other formats we carry, and typically misbehave in ways that have nothing to do with searching and retrieving information. If a patron walked up to a librarian and despite the librarian's objections insisted on reading aloud ostensibly titillating portions of Ulysses, Lolita, or The Other Side of Midnight, the librarian could rightfully point out that the patron was creating a disturbance. With computers, however, to read is not necessarily to display, and implicit in the Library Bill of Rights is the concept that patrons are entitled to their personal space until they make it their business to intrude on others. The Intellectual Freedom Manual notes, "choice requires both a varied selection and the assurance that one's choice is not monitored" (Freedom 1996). If a librarian walked up to a patron, examined the book he or she was reading, and concluded it did not meet the library's guidelines and should not be read in the library, nearly all librarians understand that this would be a gross violation of the patron's right to read and right to privacy.
12. There are at least 5,000 public libraries offering public access to the Internet (Bertot 1998);3/ Mr. Burt notes only 125 policies and cites only five. Of the five policies cited on pages 10 and 11 of his report, only one makes reference to "pornographic materials," and only one makes reference, without definition or explanation, to proscribing the viewing of materials which are "obscene." Mr. Burt discusses at length the Internet policy of Noblesville-Southeastern (IN) Public Library. However, as he points out, the policy is directed very specifically at behavior. (Library N.-S.P. 1997). The Library--like many other libraries included in Mr. Burt's policy archives--is pointedly silent about what the patrons may or may not read. The overriding guidance any sensible patron could extract from the Noblesville-Southeastern policy is to exercise discretion in a public environment (just as one should not push, shove, or speak loudly on a cellular phone) and that, as their policy states, "the responsibility for determining what is offensive or inappropriate lies with the potential user or, in the case of minors with the parents or guardians." Yet Mr. Burt goes on to say that "this library acknowledges that this also means that certain materials are not within the scope of the ALA statements, and are not appropriate to the Library mission." If this is what the Noblesville Southeastern Public Library acknowledges, that information is not available in the policy itself.
13. Mr. Burt's statement that libraries do not select hard-core pornography confuses the issue. There are many materials that librarians do not select, depending on their communities, their budget limitations and the roles they are trying to fulfill. Many librarians operate under severe fiscal constraints. In 1998, my library is allocated $10,000 to select books, recorded books and other tangible media for my patrons in a town of 11,000. This works out to about $200 per week--or about 13 adult and 4 juvenile books . Equally seriously, according to my best estimates, my library will run out of shelf space on August 15, 1999, despite my recent weeding of our collection by discarding nearly every book that had not circulated in two years. My library is obviously not able to buy everything we believe we should acquire to meet our mission statement--and the funding issues are far more complex than the cover price of a book, magazine or video.
14. Furthermore, it is difficult to understand Mr. Burt's discussion about the amount of "pornography" in public libraries. Mr. Burt uses a term-- pornography--which he does not define, does not describe his search methods, does not mention which module of OCLC he was searching, or explain why he limited his search to the international bibliographic utility OCLC (versus expanding his search to WLN and RLIN, two other major bibliographic utilities, and then beyond, to online bibliographic records not included in any major utility). It is perplexing that Mr. Burt would say he found four (in his words) "pornographic" movies in the catalog--then instantaneously exclude the results by saying that "none of these four could really be considered 'hard-core"'--a new term he introduces but does not clarify in this section of the report. He quickly introduces the term "soft-core," but does not explain what this term means, either. All told, Mr. Burt's enquiries are poor bibliometrics, relying as they do on undefined and unexamined terms, search strategies, and results, and border dangerously on legal advice, which librarians are expressly taught not to offer.
15. It is not meaningful to state, as Mr. Burt does in his report, that the Internet allows access to information that might not be acquired as part of a library's paper collection. Above, I discussed core documents such as Planning and Role Setting, Output Measures and the CREW Method. These documents, written in the "pre-Internet" era, very specifically speak to one medium: the paper-based book. (Videos, microfiche and similar new media are briefly discussed, but the overwhelming emphasis is on the book.) It is one of the excellent added-value features of newer tools available to libraries, such as the Internet, that we can have access to materials representing topics or authors we have not added to our physical, in-house collection.
16. The sharp distinctions between paper and electronic media are nowhere more apparent than with fee-based databases which vendors deliver to libraries over the Internet. Major vendors, such as Ebsco and UMI, offer full-text magazine databases for public use in libraries. A public library will pride itself on offering as many paper-based magazine titles as shelf space, budget and staffing can afford--within the constraints of its mission statement and selected roles. With the advent of tools such as Ebscohost (cf.<http://www.ebsco.com>), ProQuest Direct, OCLC's FirstSearch, and similar Internet-based proprietary databases, vendors can now use the Internet as a backbone to provide the library affordable, instantaneous access to far more articles and magazine titles than the library would ever consider for its own collection. These databases are enhanced, not diminished, by the range of titles out of scope of the library's roles and mission, and the vendors compete with one another on this basis. Title lists for Ebsco's Masterfile 1,000--a title which refers to 1,000 full-text magazine titles, far more titles than most small or medium public libraries could aspire to in paper format--include journals rarely found in most public libraries, such as the Advocate, African Arts Review, Playboy, Plastics World, and the Pittsburgh Business Review. Libraries licensing commercial databases buy them in their entirety. Librarians do not advise vendors to provide only those titles that match the library's roles: just the opposite--the pressure on vendors is to provide breadth, which they market by making their title lists globally available on their websites; vendors I have worked with have been known to toss in entire databases to "sweeten the deal." I would not ask Ebsco to exclude the Pittsburgh Business Review from the title list, though I would not purchase the paper counterpart unless demand warranted (which is highly unlikely, in rural upstate New York), nor would I write a policy directing library patrons away from this resource or advising them that use of this resource was a waste of their tax dollars, nor would I ask a patron to stop reading the Pittsburgh Business Review because other patrons were waiting to use the database to seek material I personally felt was more suitable. I would, instead, consider it a significant added value that my patrons had access to this resource.
17. Ranganathan's Fifth Law of Librarianship states, "the library is a living organism." With respect to the freely-available resources on the Internet, the tremendous advantage of this new tool, Internet access, becomes evident. It is a very powerful added value of the Internet that patrons can access resources electronically that for whatever reason are not in the traditional paper-based collection. The Internet allows us to realize as we have not before both the Second and Third Laws of Librarianship, "every book his reader," and "every reader his book." Consider a patron who wants to read a comic strip not syndicated in newspapers held by his or her public library. When librarians, as they must, prioritize with limited budgets, selecting a subscription to a newspaper solely for its comic strips is not justifiable from a pecuniary perspective, particularly if demand is limited to one patron. Even where newspapers are free, they require labor and facility space to acquire, house, circulate, inventory and ultimately deselect the item when it has outlived its use or utility. Nevertheless, we do not prevent library users from accessing comic strips over the Internet. In fact, government-funded, librarian-created indices such as the Librarians' Index to the Internet encourage users and librarians alike to pay extended visits to several sites devoted to cartoons and comic strips on the Internet, using words such as "excellent resource," "must visit" and "excellent" to describe these resources (Leita 1998).
18. As with other resources, access to the Internet can make available sexually oriented materials in libraries that might not have been selected for a paper collection. Such materials can cover a wide range, including fine art photographs, advocacy, safe sex and health-related information and certain material that Mr. Burt colloquially labels as pornography. I disagree with Mr. Burt's assumption that filtering software can distinguish "valuable" material in this category from that which in his personal judgment 'lacks value." Moreover, to whatever extent the potential availability of such materials in libraries causes actual behavior problems (as opposed to the many imagined problems discussed in Mr. Burt's report and elsewhere), existing content-neutral policies are adequate to the task.
Library Cost Issues
19. Internet access in public libraries, as Mr. Burt acknowledges and as I have explored at length, is a finite and precious resource, with a specific (though difficult to quantify) dollar value, and is usually allocated in time sessions at computers (Schneider 1998d). As with popular best-sellers, librarians cannot keep up with demand, even with the most stringent resource-allocation controls. Consider that librarians allow patrons to access comic strips over the Internet, even if these comic strips do not meet the more traditional selection priorities of the library. Furthermore, librarians not only allow but encourage people to visit websites that are misleading, ugly, superseded, trivial, or outside the scope of our stated mission. Public libraries also do not have policies advising patrons that they can only look at Internet sites where the material is an exact match for materials in the book collection. The same rationale underlies why librarians do not tell people they cannot sit at library tables and read material that librarians have not selected: the ability to access material outside the scope of the paper-based collection is not a limitation, but one of the great strengths of the Internet as a library tool. Librarians can offer their patrons materials far beyond the scope of the collection, and for that matter far beyond the scope of the librarians' awareness, and in this way librarians can do their job--which is not to exclude, but to facilitate access to information. It is predicted that by the year 2000, there will be 90 million Internet hosts--not websites, not files, but domains, to be further subdivided into a mindboggling quantity of intellectual units (Rutkowski 1998). The half-hour library patrons spend at the computer is theirs to explore this burgeoning information universe, far beyond what libraries can offer in their paper-based collection. The librarian's job is to provide, to educate, to help, to guide, and to promote use of these resources--but, also, ultimately, to stand back and let patrons experience the world as their wisdom and interests direct.
The Hustler Challenge
20. Mr. Burt claims to have put his theories to "empirical test" by offering libraries a free copy of Hustler magazine, with subsidized support to cover maintenance costs (a point he initially overlooked in the challenge). He claims that an unidentified library director informed him that Mr. Burt could offer $10,000 to subsidize the cost of selecting Hustler, and he would still not get any takers." Mr. Burt also observed that library science has addressed collection development in the area of gift policies.
21. One of the first rules of gift policies is to examine the source. As Mr. Burt points out, you need not accept a gift "because it is free" (Evans, 1987, quoted in Burt, 1998). Wood and Hoffman specifically caution that the temptation to accept a gift must be balanced against the benefactor's emotional attachment to the donated materials" (Wood 1996). Furthermore, the Trojan horse of many potential donations is not whether you accept them, but what you next do with them; quite often, a donor has "an inflated idea of the donation's] value" and "a desire to impose certain restrictions on the transaction" (Wood 1996).
22. Mr. Burt's "Hustler Challenge" includes very specific guidance with respect to offering, promoting, maintaining and displaying Hustler magazine, including making it available to children, and he has also made it clear that his offer of Hustler magazine is directly tied to his work as a filtering "advocate." Accepting Hustler magazine under these conditions would establish a pattern of directed donations which most collection development specialists, including Evans, Wood and Hoffman, specifically advise against, would countermand any other time/place/manner restrictions--including age-based--directed at use of magazines in the library, and would also leave the library open to the perception that it was endorsing Mr. Burt's activities and beliefs. Most libraries would not accept any gift with such built-in restrictions and problems.
23. A far more serious omission from the Hustler Challenge is guidance for patrons carrying magazines into libraries, since this analogy is far closer to Internet use than the concept of a library purchasing a paper-based magazine. Consider a patron who brings a copy of Hustler magazine into the library and reads it at a library table. This is identical to what the patron may be doing at an Internet workstation: occupying finite and costly space, time and equipment to read material the library has not selected. Should the library interfere? If the job of the librarian is to ensure compliance with library mission and roles, why should the librarian stop at Hustler magazine--and who determines what adult patrons may or may not read in the privacy of their cubicle or the 12 inches between their nose and the paper on the table? How would librarians know what the patron is doing unless the library saw fit to patrol and spy on patron reading habits? If a patron is not causing a disturbance or otherwise attracting attention to his or her reading, who would or should care what the patron is reading? Seating space in libraries is generally calculated at 5 seats per 1,000 residents, with a sliding scale to decrease this ratio as the population rises, and each reader is allocated 30 square feet (Dahlgren 1998). The cost per square foot for library construction ranges from $50 to well over $100 (Carow 1998). The patron could easily be sitting in $3,000 worth of space, exclusive of maintenance and personnel costs associated with that seat--and rarely if ever do libraries require that patrons get up after half an hour to give their seat to another patron. What is the appropriate economic model for determining that the patron's First Amendment rights must be violated in order to ensure compliance with Mr. Burt's personal standards of acceptable library materials?
The Analogy Of ILL Applied To Viewing Hard-Core Pornography In The Library
24. Mr. Burt states that he did "a survey of the public library holdings of popular pornographic videotapes" by consulting the Online Computer Library Catalog (OCLC"). However, Mr. Burt did not clarify if he was searching the OCLC ILL module, or what his search strategy was. Furthermore, he did not explain why he chose to stop the ILL request with an OCLC search, rather than consulting with an ILL specialist, as most librarians would do. It may well be that very few libraries have holdings for the movies Mr. Burt cites. However, only about 3,300 public libraries have holdings (records for books and other materials) in OCLC's database (OCLC 1998).4/ The 29-library public library system I belong to, for example, is a "selective" member of OCLC, meaning that we belong for resource sharing purposes but do not have holdings in OCLC. Our collective library holdings totaled nearly 1.9 million items in 1996 (System 1997).
Outsourcing Of Selection In Libraries
25. As a library manager, and as a member of the Outsourcing Task Force of the American Library Association, and as senior editor of the joint document (still in press, but in second draft) of the Task Force written in response to widespread professional concerns about privatization of library services, I was surprised to see Mr. Burt reduce the discussion of outsourcing to a few words (Schneider 1998c).
26. Filtering and approval plans are hugely dissimilar. An approval plan is a librarian's aid designed to help the library receive a type of material (such as children's series titles, like Nancy Drew or Goosebumps) without having to reorder or individually order items--the "sure bets," usually in high demand, that are a labor drain to order individually. Sometimes libraries will use approval plans to build "opening day collections," designed to help new libraries get started. Blanket orders are used for the even more certain "sure bets," such as a library that knows that it will purchase all of Fodor's United States travel guides. An example of a very small approval plan is the contract Baker & Taylor provides for a library to provide all books recommended by Oprah Winfrey (Taylor 1998). As the ad states, "No more placing separate orders, no more worrying about title availability--the book will arrive at your door in time to meet increasing patron demand." Librarians profile the material they will request, and receive these materials on approval; if they do not meet selection guidelines, materials may be returned. (In the case of blanket orders, you are committed to what you asked for.) In large systems, approval plans and blanket orders may easily accommodate half of all purchases--but that leaves the rest of the budget to account for.
27. Selected outsourcing is an accepted and even encouraged practice in librarianship. Wholesale outsourcing is not. The one major experiment with outsourcing all book purchases in a library system was an unmitigated disaster by all accounts. In 1986, the Hawaii State Library outsourced the selections of Hawaii's public library collections to a major vendor, Baker & Taylor. This decision was uniformly considered in the library community as a usurpation of a core library activity by a commercial company, and it was vigorous opposed by professional librarians in Hawaii and elsewhere, who pointed out that Baker & Taylor was supplying resulted in the purchase of many overpriced books, titles already held, and titles irrelevant to Hawaii's communities. Ultimately, the State Librarian broke the contract with Baker & Taylor. In practice, this debacle proved that approval plans cannot replace selection. (Schneider 1998c).
28. An Internet content filter outsources weeding to a third party, then hides this decision in encrypted sites inaccessible to the person licensing the product or the end user. It is as if people crept into the library in the middle of the night, stole books from the shelves, then hid all evidence that the books were missing, and encrypted the library catalog so that it could not be searched for the stolen items. The filter company is not working with librarians to help make content decisions, or with lawyers to make legal decisions; the filter company is making the decisions itself--and keeping this knowledge hidden from everyone but the company employees making the content decisions.
29. The librarian who elects to filter the Internet is not simply outsourcing a part of his or her work, but delegating all content decisions to commercial companies. As a library director, I believe this raises fundamental governance issues. The primary accountability for commercial services is with internal stakeholders, whose motives are commercial viability, not public service.
30. Finally, a library's standards are impossible to enforce with filters, given the hidden and remote nature of the decisions. It is impossible to have accountability in a tool that provides no quality control measures. Internet content filters have an imbalance of control over information decisions. The reality is that from day to day, none of us know what X-Stop is doing, and the company can block or unblock material on a whim. The Log-On Data Corporation could decide to reduce blocks to a minimum prior to a court trial, then block most or all of the Internet later on. Lack of accountability and lack of quality control are serious problems with all Internet content filters, which I discuss in my book, A Practical Guide to Internet Filters (Schneider 1998b). The phenomenon of watching a filter block--unblock--and then block a site again is something I observed as project manager of The Internet Filter Assessment Project, and may only have been observable by someone who spent six months observing filter behavior. Mr. Burt is correct; the Internet is composed of many discrete intellectual units of information. With this in mind, I was surprised to see X-Stop block sites related to nothing more controversial than jewelry. The individual decisions of the X-Stop company are not consonant with principles of open access.
Less Restrictive Means
31. I frequently consult in the area of Internet access management, and discuss methods for Internet management in my book, A Practical Guide to Internet Filters. Many methods of managing Internet content can be combined for greater effect, such as user instructions with a firm policy and parental permission for child use.
32. Mr. Burt cites three examples of "behavior problems" "caused" by "unfiltered, uncontrolled access to pornography" in public libraries. As stated earlier, there is no end of creative enterprise in the area of patron misbehavior. Libraries are public environments, strikingly open to everyone. Three incidents is a remarkably low number of problems with any material. In my experience, policies designed to deal with behavior problems when they occur are sufficient to cope with such instances without the need to resort to censorship.
33. It is my opinion that filtering in the most restrictive method for blocking the viewing of Internet content, and I do not recommend it. My extensive experience with filtering products has convinced me beyond a doubt that filtering all public-access workstations will ensure that material is blocked which patrons have the right to see. Every filter product has the same basic weakness; they all rely on private, encrypted site lists created by remote employees with little understanding of the user's right to access information. X-Stop is no exception. Under the best of circumstances, filters would ultimately block speech through error. Sometimes it appears that filters are blocking for other reasons than by error, such as X-Stop's block of http://gayweb.com, a compendium of gay and lesbian resources. Furthermore, I do not believe Mr. Burt or I are qualified to determine how well software can distinguish between "obscene" and other speech, or "illegal" and legal speech. We are not lawyers.
34. Based on my library management experience and my experience working directly with public users, I also believe it is unreasonably restrictive to expect library patrons to ask librarians to unblock Internet sites. Consider the parallels with other media. Current intellectual freedom guidelines from the American Library Association do not accommodate a scenario where books can be freely (and quietly) removed by third parties and then reinstated by libraries on request of the patrons. It is not consonant with other library behavior policies to be extremely restrictive before patrons have proven themselves to be troublesome. Librarians know patrons will turn books in late and eventually lose or damage some of them, and yet we still encourage people to walk out of libraries with armloads of things to read. Finally, librarians know anecdotally, and library research has demonstrated, that under the best of conditions patrons are reluctant to approach librarians (Swope 1972) (Childers 1997) (Massey-Burzio 1998). Patrons are not going to be more likely to approach librarians when searching sensitive data, when they know the library has made a decision to block material the library has suggested is not appropriate, or when the patron is confronted with a computer screen advising the patron, as X-Stop does, to "contact supervision."
35. "Tap on the shoulder" is a misnomer for policies that explain library guidelines for patron behavior. Behavior policies, claim Mr. Burt, do "not prevent the display of pornography." Disregarding the problem of defining "pornography," no policy or law is ever 100% effective in preventing misbehavior. Mr. Burt has no basis for concluding that behavior policies are not overall effective-- or less effective than other policies--in communicating guidelines in a way that patrons understand. With over 5,000 public libraries now offering public Internet access, even Mr. Burt's report demonstrates that there have been very few Internet related problems.
36. Patron instruction is classic librarianship at its best, evoking the "library militants" of the last fin de siecle who pushed up their starched muslin sleeves and threw open their library doors to the immigrants pouring into the United States" (Williams 1988). Patron instruction allows librarians to set the tone for Internet use at the library while providing library patrons with much-needed (and greatly appreciated) instruction in navigating the choppy waters of the Internet. This user instruction in public libraries may be practiced on the job while working with walk-in or telephone patrons, such as described on the website for Morris County Library, New Jersey, or it may be offered in a formal, classroom setting, such as the "Wandering the Web" programs offered to patrons at the William K. Sanford Public Library in Colonie, New York (Weissman 1996) (Kraus 1998).
37. Mr. Burt claims that "privacy booths" do nothing to "prevent the display of obscenity." I rarely recommend privacy booths per se, as they interfere with group work and absorb quite a bit of library floor "real estate," but I have discussed privacy screens, desks, and computer positioning with a number of clients. Privacy screens slip over the front of a monitor and obscure the view from angles other than straight-forward, while privacy desks place monitors in recessed wells with glass tops (Schneider 1998a). Computer positioning, where architecturally feasible, simply means placing some computers where adults can read in private. Some software programs can return the browser to a default page after a period of inactivity, which both protects patron privacy and aids in user navigation.
38. Privacy screens, privacy desks and computer positioning are not cure-all tools that ensure all patrons will behave appropriately in libraries, any more than "no smoking" signs in restrooms ensure patrons will never light up. However, these tools and techniques, though they have some limitations, do far more than address the issue of viewing potentially objectionable material: they are helpful for improving patron confidentiality in an electronic environment. Privacy equipment and techniques support both sides of the patron confidentiality equation--the reader and the passers-by--by establishing a privacy "buffer zone" that restores the private reading experience that can be lost on large computer monitors in public places. Among other advantages, these tools can be an important equalizer for those who rely on libraries for their Internet access, so that they can send email, credit card information, seek medical information and so forth with a measure of privacy.
39. Go-lists (sometimes, though less preferably, called "whitelists") are sites created to point people toward high-quality resources on the Internet. Non-restricted go-lists, like user instruction, are excellent tools for guiding patrons towards high-quality resources and for demonstrating the best parts of the Internet. The Librarian's Index to the Internet (LII), for example, is supported in part by federal Library Services and Technology Act (LSTA) funding and administered by the California State Library. LII is a "searchable, annotated subject directory of more than 3,600 Internet resources selected and evaluated for their usefulness to the public library user's information needs" (Leita 1998).
40. The comparison with filters is instructive. LII is an inclusive tool that makes quality information easier to find; filters are exclusive tools that deny access to content. LII offers an index page where all of its citations are listed; filters hide their site lists through encryption. One file on LII is devoted to explaining how the team of 32 librarians add material to the LII, with clear pictures illustrating the process; filter companies consider such information proprietary. Finally, LII is funded with government money, and is ultimately accountable to the public interest (LSTA grant funding, as a federal program, has particularly stringent reporting requirements); filters, such as X-Stop, are commercial products, ultimately accountable only to private interests.
Testing of X-Stop
41. The one certainty about filters is that they inevitably block information people have a right to see. Mr. Burt's assessment of X-Stop was primarily directed at determining how much and what type of "pornography" (his term) it blocked relative to other content filters. Throughout his report Mr. Burt makes reference to that which is "pornographic" or "likely to be obscene" or "sexually explicit" without any clear or consistent definitions of these terms. Mr. Burt's subjectivity in this regard was starkly revealed by his claim that http://www.gayweb.com is a pornography site. I revisited that site several times (with X-Stop disabled, of course) and am at a loss to understand Mr. Burt's assessment, unless he routinely classifies non-graphic homosexual-oriented material as "pornographic," "obscene," or "sexually explicit."
42. Given my broad experience with filters, which includes leading a team of librarians in a project to evaluate nine filters over the course of six months, the mismatches between Mr. Burt's results and mine were predictable. As I discussed above, filters have different results on different days, and control of what filters block is out of range of the person licensing or using the product. It is entirely possible that someone notified the company of the blocks I had experienced before, during or after my tests and Mr. Burt's tests, or that the company itself chose to remove the blocks in the database, and furthermore, there is no way to know otherwise.
43. The results of The Internet Filter Assessment Project are far more systematic and reliable than Mr. Burt's personal assessment of Internet content filters. With at least two librarians evaluating every product, we found that even librarians rarely agree on assessments of material on the Internet, and that "The strongest predictor of filter performance...was... the person using the computer" (Schneider 1998b). As I say elsewhere in my book, filters are mechanical tools wrapped around subjective judgment. You cannot predict how two people will interpret the same site any more than you can predict what a filter will block (or not block) on any given day.
Conclusions
44. In 1931, S. R. Ranganathan,a former mathematician turned librarian, wrote Five Laws of Librarianship intended to address tendencies in the library profession to treat libraries like mausoleums and patrons like intruders (Ranganathan1931). The first rule was "books are of use."Even in a distributed, digital environment, this rule continues to resonate its truth. The locus of librarianship is access to information. One definition of a librarian is someone who believes you can solve problems by throwing information at them. Our ethos is inclusive, and new media such as the Internet have allowed us to adapt our services to be more, not less, supportive of that ethos. Mr.Burt's claims that librarians seek to exclude material from their collections on the basis of content is a gross misunderstanding of the necessities of resource allocation, and is belied by the facts, as demonstrated by commercial magazine databases and by freely-available information on the Internet, which are that librarians, when given the opportunity to do so, provide as much access to as much information with as few impediments as possible.
45. Filters are antithetical to librarianship. They block information people have a right to see. In the short run, delegating authority to a commercial third party may seem easier than the more complicated approach of policy, user instruction, go lists, and similar tools. In the long run, however, the unintended consequences of delegating material selection to an offsite commercial enterprise with no quality control or accountability over the final product could set a precedent for information delivery that is not congruent with principles of free speech and privacy.
Dated: July 30, 1998
Signed__________________________
Karen G. Schneider
Notes
1/ S.R. Ranganathan coined the Five Laws of Librarianship in the 1930s. The laws are: 1. Books are for use. 2. Every book its reader. 3. Every reader his book. 4. Save the time of the reader. 5. The library is a living organism. A current interpretation of these laws, and Gorman's own suggested updating for them, is in Gorman, 1997.2/ A new ALA planning document, Planning for Results, published just months before this report and not yet in widespread use, is "not a drastic change," according to my own examination, anecdotal reports and correspondence with Public Library Association President Christine Hage, director of the Rochester Hills, Michigan Public Library (Hage 1998). According to Ms. Hage, Planning for Results "refines but does not alter" the guidance in Planning and Role Setting and Output Measures. Ms. Hage added, "PLA supports and will continue to support local control so that the public library can help maintain the rich cultural diversity of our nation."
3/ For the reporting year 1997, 60% of all public libraries offered public access to the Internet. There are over 8,900 public libraries in the United States. (Bertot 1998).
4/ OCLC is one of three major electronic bibliographic utilities for library holdings. Approximately 3,300 public libraries have holdings in OCLC (OCLC 1998) and professional material is often uncataloged since it is kept in library offices for professional reference, or is purchased centrally at consortium offices for reference. A search of OCLC alone does not demonstrate the presence (or absence) of a given book in public libraries. For example, an OCLC search shows only 930 holdings for Planning and Role Setting and 997 holdings for Output Measures (Schalow 1998). Even after excluding other type holdings for those documents, such as library schools, the actual count of public libraries owning these documents is likely much higher. For example, 4 copies of each title are held by the Upper Hudson Library System (telnet://uhlan.uhls.lib.ny.us ) just one of the many library systems whose holdings are not available in OCLC.
Reference Materials
Bertot, J.C., et al. (1998) 1997 Survey of U.S. Public Libraries and the Internet
Boon, B. (1995). The CREW method: expanded guidelines for collection evaluation and weeding for small and medium-sized public libraries / revised and updated by Belinda Boon. Austin, TX, Texas Library Association.
Carow, J. (1998). Re. Thanks for the package.
Childers, T. (1997). "Using public library reference collections and staff." Library Quarterly 67(2): 155-173.
Dahlgren, A. C. (1998). Public Library Space Needs: A Planning Outline, State of Wisconsin Department of Public Instruction.
Freedom, O. f. I. (1996). Intellectual Freedom Manual. 5th Edition, American Library Association, Office for Intellectual Freedom.
Gorman, M. (1997). Our Singular Strengths. Chicago, American Library Association.
Hage, C. (1998). RE: confidential request.
Kraus, S. (1998). William K. Sanford Town Library, Kraus, Stephanie.
Leita, C. (1998). Librarians' Index to the Internet, Leita, Carole.
Library, N.-S. P. (1997). NSPL Electronic Access Policy. Noblesville-Southeastern, Indiana Public Library.
Library, W. P. (1997). Waltham Public Library Rules, Waltham, Massachusetts Public Library.
Massey-Burzio, Virginia (1998). From the Other side of the reference desk: a focus group study. The Journal of Academic Librarianship, 24:3, 208-215
McClure, C., et al. (1987a). Planning and Role Setting for Public Libraries: A Manual of Options and Procedures. Chicago, American Library Association.
McClure, C., et al. (1987b). Output Measures for Public Libraries: A Manual of Standardized Procedures, American Library Association.
McNeil, B. (1996). Patron behavior in libraries: a handbook of Positive approaches to negative situations. Chicago, American Library Association.
OCLC (1998). OCLC at a glance. Online Computer Library Catalog.
Ranganathan, S. R. (1931). The Five laws of library science. Bombay, Asia Publishing House.
Rutkowski, A. M. (1998). Internet Trends, Center for Next Generation Internet.
Schneider, K. G. (1998a). Shining a Light on Filters in Libraries, Schneider, Karen G.
Schneider, K. G. (1998b). A Practical Guide to Internet Filters. New York, Neal Schuman.
Schneider, K. G. et al. (1998c). Outsourcing Task Force Initial Background Document. Albany: 10.
Schneider, K. G. (1998d). The Tao of Internet costs." The Bottom Line 11(2): 52-64.
Swope, M. J. (1972). "Why don't they ask questions?" RQ 12(2): 166-172.
System, U. H. L. (1997). Annual Report. Albany, Upper Hudson Library System.
Taylor, B. (1998). Oprah Alert, Baker & Taylor.
Weissman, S. (1996). Minimum Internet Standards, Morris County Public Library.
Williams, P. (1988). The American Public Library and The Problem of Purpose. Westport, CT, Greenwood Press.
Wood, R. J. a. H., Frank (1996). Library Collection Development Policies: A Reference and Writer's Handbook. Lanham, Maryland, Scarecrow Press.