Project Bait and Switch

Introduction

"Bait and Switch" is an experiment to find out whether small, personal home pages and Web sites of large organizations get identical treatment from blocking software companies in deciding what to block.

Most censorware products attempt to block "hate speech", with "hate speech" usually defined to include derogatory statements based on sexual orientation. (The definitions used by the different companies are usually published on the company Web sites; current definitions at the time of the experiment are collected here)

We collected some of the anti-gay statements from the home pages of four well-known conservative sites: the Family Research Council, the Focus on the Family, the Official Dr. Laura web page and Concerned Women for America (none of these sites are currently blocked by any of the programs that we tested). We then created one different "bait" Web page for each of these organizations, with the "bait" page consisting of quotes taken from the organization's Web site, without telling the viewer where the quotes came from. The "bait" pages were submitted for review to each of the blocking companies (through anonymous HotMail accounts so that the companies wouldn't know the submissions were coming from Peacefire).

In all cases, the blocking companies agreed to block the pages we submitted in their "hate speech" categories. We then contacted the blocking companies to ask if they would block the organizations whose Web pages were the sources of our anti-gay quotes.

So far, all of the censorware companies in our experiment have been back-pedaling since then, saying that they will not block the pages which were the original sources for the anti-gay quotes. Naturally, Peacefire does not advocate censoring these pages, but only because we are against censorship in general. If blocking software claims to block sites which "denigrate people based on sexual orientation" -- as almost all censorware companies claim to do, in their published definitions of "hate speech" -- then the sites that we listed clearly meet those criteria.

The "bait" pages

We have set up local mirrors of each of the following "hate sites" since most of the free home page services currently hosting the pages have policies against "hate" -- so now that the following pages are going to get a lot of attention, they stand a good chance of getting deleted from the free page servers. Our local mirrors are identical to the originals, except for banner ads and cookie-setting scripts that have been removed. Under "Quote attributions", we have also included the original location of each quote that appeared on the "bait" pages.

http://members.tripod.com/bryan_cody/
Comprised of quotes taken from the Focus on the Family site at http://www.family.org/
Local mirror | Quote attributions

http://www.geocities.com/terence_mason/
Comprised of quotes taken from the Concerned Women For America site at http://www.cwfa.org/
Local mirror | Quote attributions

http://www.angelfire.com/mt/firstjohn1976/
Comprised of quotes taken from the Family Research Council web site at http://www.frc.org/
Local mirror | Quote attributions

http://members.theglobe.com/plexicraft/
Comprised of quotes taken from Dr. Laura at http://www.drlaura.com/
Local mirror | Quote attributions

Why were some blocking programs left out of the experiment?
We submitted sites to SurfWatch, Cyber Patrol, Net Nanny, Bess, WebSENSE and SmartFilter. Most of these products comes with a secret database of sites that are blocked by the software (Net Nanny works off of a combination "bad word" and "bad site" list that is viewable to the user, although there is no way to tell in advance what sites will get blocked by the "bad word" filter). Most other products, such as WebChaperone and Cyber Sentinel, block sites solely on the basis of keywords that appear on the page and do not operate off of a blocked-site list, so the experiment would have been meaningless with them since they have no direct control over what sites get blocked. One product, I-Gear (made by Symantec), was left out of the experiment because we never got any response to the emails we sent suggesting sites to be blocked. CYBERsitter was excluded since long-standing legal issues between CYBERsitter and Peacefire precluded us from contacting them.

References: