Green Dam Girl: policing Internet content

With the July 1 deadline for installation of the Green Dam-Youth Escort content filter fast approaching, Internet users inside and outside China continue to pick apart the software's security and the scope of its filtering.

Analysts have uncovered lists of sensitive words (some are available on WikiLeaks) that concern the software's stated target, vulgar and pornographic content, as well as other sensitive areas such as politics and the Fаlun Gοng.

Lists of sensitive words used by other software programs and websites on the mainland frequently contain some fairly strange terms, and Green Dam is no exception. The software even seems to want to filter out articles critical of topics frowned upon by the Chinese government.

Fang Zhouzi, an anti-fraud activist who has written extensively against Fаlun Gοng, posted some interesting results in a blog post, translated below:

Computer security experts at the University of Michigan unlocked Green Dam's list of filtered words. The list is primarily made up of sensitive words related to sex and politics, and although it's not as strange as the other sensitive word list that has been circulated online, there are still a few places that I don't understand.

1. The list includes common terms like "essence" (本质), "fallacies and heresies" (歪理邪说), "Cat-III" (三级), "naked" (裸露), and "homosexuality" (同性恋). "Fallacies and heresies" is likely a politically-sensitive term, and the three at the end are sexually-sensitive terms, but I can't even imagine what "essence" counts as. I've had these words appear in my non-sexual, non-political writing, so does that mean that the websites that host those articles will be zapped? And Green Dam monitors word processing in addition to Internet access, so does this mean that if these common terms appear in a Word document, the draft will suddenly get deleted without so much as warning you to save it? Does this mean that from now on, the word "essence" can no longer appear in primary and secondary school essays, and that these terms must be removed from textbooks and dictionaries?

2. The list also includes the strange term fanyu (梵欲), which I've never even heard of. I thought it might be some new sex term, but after Googling it I found only a few pages that contained it, a little over 100. The majority of those were Buddhist scriptures, and not even a single one was sexually- or politically-sensitive. So why filter this peculiar term? If it was a typo, then isn't that just a little bit careless for a piece of software that's going to go out to hundreds of millions of users?

3. To no one's surprise, most of the filtered words on the list concern cult organizations, but many of the listed terms criticize or attact the cult and its leader. For example, "The anti-science character of Fa XX", "The anti-society character of Fa XX", "The fallacies and heresies of Fa XX", "Fa XX foments hatred", "Fa XX is an evil cult", "an illegal organization like Fa XX", "Li XX is an exceptionally greedy man", "Li XX's henchmen", "Li XX cheats his followers", "Exposing Li XX's true form", and "Li XX, who misleads the people with heresies". If a website preserves those old articles issued by the government criticizing cults, will it be zapped? Why don't they want everyone in the country to be able to read articles criticizing cults and their leaders? Could the 'lun have been involved in drawing up the list of filtered words?

Fang's blog post uses XX to obscure sensitive words whose presence would probably have it deleted from Sina and other blog portals on the mainland, but you can find the full terms on the uncensored XYS forums, which are hosted overseas.

The image in this post comes from a series featuring the Green Dam Girl, a crusader who carries a rabbit (the software's mascot), wears a River Crab badge, and holds a bucket of paint to wipe out online filth. See more of the images at Hecaitou's blog.

Until the entry was deleted, Green Dam Girl was briefly part of Baidu's encyclopedia (screen shot) under the name "Greendam" (格林达姆), or GD. The entry included the line, "Reports indicate that Greendam may have foreign ancestry," in reference to allegations that Green Dam lifted some of its code and wordlists from Solid Oak Software's CyberSitter filtering program.