Will Filtering Software Ever Be Good Enough?

Currently filters don't work well, but can we hope that sometime in the future, technology will advance enough that filters will be reliable? The ALA says no. The ideal filters would keep all constitutionally protected speech, and only block those sites that are legally obscene, such as child pornography.¹⁸ In other words, the ALA thinks that no robots or software company employees should be in the business of deciding which sites to block. Unless courts decide to spend their time evaluating every Internet site for obscenity, filters will never be consistent with the ALA's commitment to First Amendment principles.

Pro-filtering advocates can respond, however, that librarians have always made decisions about excluding obscene materials from collections. The professional practice of library book selection has always limited patrons' access to materials considered obscene.

There's a big difference, though, between book selection and Internet filtering: Who does it. In neither case is it a court of law. But when it comes to Internet filtering, a software employee or even an automatic computer program makes the decision to filter out a particular website. Thus by filtering the Internet, librarians give up local control over a crucial part of their library 'collection', and not only that, but neither librarians nor patrons can easily monitor the collection. They lack detailed knowledge of how the filter works. Librarians can, of course, unblock sites they or patrons want to see. But librarians cannot be expected to catch all the sites that should be unblocked (or blocked, even).

Even more than the arguments about censorship and selection, local control could be a bigger reason why we should be reluctant to filter. Would patrons rather have community members and parents decide for themselves what is acceptable to view, or would they like the decision to be made by anonymous software engineers who create the filters? The New York Times reports that commercial filters often base settings and stoplists on objections submitted to software companies by their customers across the country: "So they tend to reflect the standards of the more cautious communities. Conservative, Bible-Belt communities like Conway, SC, are helping to set the standards for what students in more cosmopolitan places like New York are allowed to see."¹⁹

A similar problem could arise when libraries buy filters from companies with diverse clientele. A library's filtering standards could be very different from the standards desired by school districts (which have a mandate to inculcate values and to act In loco parentis) and by private households. But if a company sets its filtering standards by the desires of the majority of its customers, libraries could be at a disadvantage unless they buy filters en masse (And even then, librarians in liberal communities may not want filtering standards set by librarians in conservative communities).

When librarians argue that Internet filters don't work, they could prepare by doing at least three things: They should have up-to-date examples of filters' flaws, representative of the best filters, because problems with older filters might have been solved by later versions and brands. Second, librarians might choose to emphasize the non-pornographic sites, with educational value, that get blocked by filters. Third, librarians might decide whether they think the problem with filters is not just technical, but a matter of principle, and prepare their arguments accordingly. For the ALA, the solution isn't to fix buggy software, it's to protect the patron's freedom of choice and local control.

Next:
Should Filters Be Used In Children's Rooms?

Previous > Table of Contents