Small Town Takes Stand Against Internet Censorship.

MOST big-city types probably figure they know more about the Internet than anyone living in a small, rural town like Holland, Mich.

But Holland residents recently showed a savvy beyond their census size by rejecting an initiative to clamp software filters onto Net-connected computers in the town's library.

Filters are fine for parents who don't have time to police their own children's time online. But libraries are run by governments, not Mom and Dad. And governments make lousy parents, particularly when they subject their citizens' First Amendment rights to the whims of a software company.

The Feb. 23 vote was the first of its kind in the nation. Right-wing Christian groups sponsored the measure in hopes that voters in this town of 32,000 people would voluntarily restrict the reach of their library's computers.

The proposal would have cut off the library's funding unless it installed filtering software to block access to sexually explicit sites. Free-speech activists campaigned against the proposal, but right-wing groups outspent them by a 14-1 margin, The New York Times reported.

As it turns out, Holland residents didn't want anyone telling them what to do -- either at the ballot box or in their library. This conservative town acted in the truest sense of that word, rejecting a proposal that would have expanded the reach of government into the private realm of parenting.

And that power wouldn't have been exercised in an open, democratic process. Indeed, the real problem with public use of software filters is that censorship is exercised through the closed corporate process of a government contractor.

Software filtering

Filters compare the Web addresses users request to a long list of sites that, for one reason or another, are found to violate the software company's standards. If a requested site turns up on the list, the filter won't allow it to be downloaded.

Most filters let users block content from various categories like pornography, hate speech or violent images. But only the software company knows for sure which sites fit in each category.

What one filter calls "obscene" another might deem merely "suggestive."

Some consider any discussion of homosexuality to be sexually explicit content, while others block mammograms as pornography.

Knowing this, parents and community leaders might want to compare the "block lists" of various filters to determine which company's standards most closely match their own. But here's the rub: Most companies claim these lists are corporate secrets and refuse to release them.

So whenever a library decides to install filters on a public computer, it subjects patrons to the inscrutable standards of some out-of-town corporation. Sure, people won't know what they're missing -- but that's the point.

Software filtering companies are essentially government contractors that can never be audited. And if that sounds like a good idea to you, I've got a multimillion-dollar consulting contract you're just going to love.

One reason filtering companies want to hide their block lists is because they generally contain sites that shouldn't be censored. It's impossible to accurately assess the ever-changing contents of every page on the Web, particularly when you're using subjective standards. Employees of filtering software companies race to review as many sites as possible, and they're prone to making mistakes. Even if they accurately rate a site, its content may change after their visit.

Making mistakes

Anti-filter activist Bennett Haselton recently cracked the pornography block list used by X-Stop, a filter made by Log-On Data Corp. of Orange. He checked the first 50 dotedu sites on the list -- sites posted by universities or their students -- and found that 34 contained nothing remotely pornographic.

"No blocking software company seems to think it would be worth paying someone to search their list for mistakes," Haselton said. "They only pay people to add stuff!"

Log-On spokeswoman Kelly Rodnon suggested those innocent dot-edu sites might have hosted pornography previously, though the students Haselton contacted said that wasn't the case. She also said her company, like other filtering software manufacturers, is happy to correct its list when notified of an apparent error.

But public library patrons shouldn't have to petition some company for permission to visit a site that never should have been blocked in the first place. This is particularly true if they want information about a sensitive subject, like homosexuality or breast cancer, that they'd rather not have to ask to see.

The vote against filters in Holland was the first of its kind, but it probably won't be the last. Filtering software has already made its way onto computers in an estimated 15 percent of U.S. public libraries, and more libraries are considering filters every day.

But if those towns put the issue to a public vote, they should expect results similar to what happened last month in Holland.

In towns big and small, people don't take kindly to secretive censorship in a public place.