The most common method used by academics to map or track what bloggers are talking about in various countries is by counting the use of various keywords and putting them into categories, then figuring out how the various conversations - tagged by subject matter - seem to cluster. The Chinese Internet presents a special problem for this kind of research, because in order to avoid censorship, people frequently talk about one thing when all their peers know they're talking about something completely different.
Take, for example, Chinese bloggers' recent obsession with pushups: People created a website and a forum dedicated to pushups; somebody photoshopped a naked man doing pushups at various famous Chinese tourist sites; people created all kinds of flash mashups celebrating pushups... huh?
What these people are actually doing is expressing their frustration about the fact that many BBS forum conversations and blog posts talking about the recent Weng'an riots were censored. For very detailed coverage and translations of a variety of media reports, see Roland Soong's blog. In a nutshell, a young girl turned up drowned in the river in Weng'an county, Guizhou province. Family members suspected foul play and word quickly spread that the girl, Li Shufen, had been raped and murdered by boys who were probably related to people in the Public Security Bureau - resulting in protests by 30,000 people and the burning of the local police station. Three autopies were performed on the girl in which the coroner declared no foul play, but locals didn't believe it. It remains unclear what really happened, but at any rate four local officials have been sacked for "severe malfeasance." Li Shufen's godfather was also arrested for inciting riots and spreading rumors on the Internet. So where do the push-ups come in? There were three young people with Li Shufen when she died, and according to the police interrogation report they say that she committed suicide suddenly while one of the boys was doing push-ups on the bridge.
In the wake of the riots, Internet chatrooms and forums have been heavily censoring discussions about Wengan. Some bloggers came up with a clever online tool to convert text from left-to right sideways (as modern Chinese is written) into right-to-left vertical (as classical Chinese was written) - in an effort to get around keyword censors. But it was still difficult to hold in-depth exchanges discussing all the ins and outs of Li Shufen's death and reasons for the Weng'an unrest. So people just gave up and started joking about pushups instead...calling on their friends to write about pushups as a kind of protest.
A number of people have written very insightfully on this incident, including Roland and Jonathan Ansfeld. The Wall Street Journal declared the sacking of four officials and calls for more media transparency a victory for China's bloggers. However, from what I can tell it seems like Chinese journalists may be the bigger winners from this whole incident.
I'm in the middle of conducting a fairly extensive research project on how Chinese blog hosting services censor their users. My team and I are posting a variety of content on 16 different blog services and documenting what gets censored, by whom, and how. I'll be writing up the overall findings for an academic paper later on. But meanwhile as I come up with interesting findings I'm sharing them along the way and am interested in people's feedback. Over the past week I posted two items about the Weng'an riots on 16 different Chinese blogging systems, plus one item about how the term "push-up" has become a censored word. Both of the Weng'an articles were censored by the same six blog hosts included in the tests: Baidu, iFeng (run by Phoenix TV), Netease, Tianya, Yahoo China, and MySpace. The latter two are American Internet brands. Tianya receives investment from Google. So far, none of the other ten services have taken down the Weng'an related posts I published - I'm not going to name those ten here because I'm concerned that the 6 might use this information to get the 10 in trouble with authorities, as is known to happen.
On Netease, you can save the post privately, but when anybody else (who isn't logged in as the author) tries to view the post, only an error message appears, see screenshot at left (click to enlarge).
Now here's the really interesting part: while it's impossible for a citizen-blogger to write about Weng'an or push-ups on a Netease blog, the Netease news portal has extensive coverage of the Weng'an situation, including this long article about why one of the boys with Li Shufen was doing pushups on the bridge when she allegedly jumped into the river.
Similarly, when I tried to post about Weng'an or push-ups on iFeng, the blog hosting service run by the Hong Kong-based Phoenix TV (aimed mainly at a Chinese mainlander audicence), the post is censored. Yet the iFeng news portal has a whole special coverage page about Weng'an.
"Weng'an" and "push-up" are NOT being blocked by the "great firewall" - I mean specifically the filtering mechanism that causes your browser to turn up an error message if you try to visit a site showing the offending words. If you search "Weng'an" and "push-up" in Google.cn, Baidu, and Yahoo China you'll get lots of results - albeit with reports overseas dissident and human rights websites taken out.
This is quite different than the way the Yilishen incident was handled last fall - when large numbers of people protested about having been gypped in a pyramid scheme. If you scroll down to the bottom of my post written at the time, you'll see screenshots of Baidu, Google.cn, and Yahoo China, all of which gave ZERO results for searches on "Yilishen." Very little reporting about Yilishen appeared in the Chinese domestic media.
Today, Chinese journalists are being allowed all over the Weng'an story. Roland Soong just translated a long investigative article published by Southern Weekend late last week, and the respected Caijing - famous for pushing the edges - has its own special report section on Weng'an.
So what's going on here? Why do some web service companies ban blogs from talking about Weng'an while at the same time running extensive news coverage about it? We'll have to see whether this pattern holds in future, but if it does, that would point to a growing sophistication in the Chinese government's strategy for managing online media - both professional and amateur. The strategy would appear to be: give the professionals more rope to report while censoring the amateurs more heavily. Let Chinese people searching on the internet for information about unrest incidents read about them primarily from the state-sanctioned media, not from bloggers repeating things they got from chatrooms repeating things that people heard on the street.
You then combine this with what Paul Denlinger calls the Chinese government's astroturfing strategy, with a few hundred thousand web commentators who are paid to write pro-government comments on blogs and in chatrooms. These people are known as the "fifty-cent party" because at least some of them get paid 50 Chinese cents per post. My colleague David Bandurski describes the system in detail in the latest issue of the Far Eastern Economic Review. He describes how the rage of Chinese cyber-nationalists against CNN's Jack Cafferty was fueled by 50-cent party postings.
Put all of these things together and once again, it's clear that there's a lot more than censorship going on: in addition to censorship there's information management, message management, and "astroturfing." At last month's Chinese Internet Research Conference, Chinese journalist and academic Li Yonggang talked about how we should view the Chinese government's efforts to control or manage the internet like a water management system. Roland Soong picked up on that idea in a recent analysis in which he compares the government's online information management strategy to hydrological engineering:
Yes, HYDROLOGICAL ENGINEERING! Many of the current crop of central government leaders are technocrats with engineering background. As such, they must understand that public opinion is water that can carry the ship as well as turn it over. The point about hydrological engineering is not to build dams to hold the water back because there will be a catastrophic dam break one day that might bring down the entire system. Instead, the point should be about controlling and redirecting the awesome power of nature in less harmful ways down selected channels.
In the case of the Weng'an mass incident, the major portals were deleting the related posts as quickly as possible. At Tianya Forum, it was estimated that a Weng'an-related post has an average lifetime of 15 seconds before being deleted by the administrators. That was supposed to be a record speed. The same thing was happening at Sina.com, Sohu.com, Baidu, etc. So this was building massive dams all over the map which builds up a tremendous pressure. Where was the pressure release point? You may be amazed that it was over at the Xinhua Forum. The webmasters posted the official Xinhua news story on the forum. That does not help in itself because Chinese netizens think that this Xinhua story was vague and misleading. However, the webmasters allowed the comments to run freely. This meant that the Xinhua posts became the meeting points of all those who want to talk about the Weng'an incident but could not do so elsewhere. Although that post did not contain any news information (such as photos and videos), it was a place for people to vent their outrage. As a result, Xinhua got a record-setting number of visitors who were very appreciative. Is this the plan for the future? You'll find out at the next mass incident (and there will be many).
The system continues to learn and evolve. The immediate beneficiaries are likely to be Chinese journalists, who have been chafing at their short leash for quite a long time now. Giving journalists a longer leash results in more credible, complex reporting while at the same time the propaganda authorities can still exert some control to prevent certain things from being reported. Independent bloggers like Zola who traveled down to Weng'an, who are not being paid by a news organization, are much harder to control by means other than direct censorship, blocking, and when necessary physical threats (as Zola experienced last Fall). If there's a news blackout on something, bloggers can become a vital conduit for information about what's going on. But when there is a decent quantity of professional news reporting on an event or issue, the role of the blogger as citizen reporter is weakened unless they have some truly unique material or insights. It's very difficult for a blogger like Zola traveling down to Weng'an to compete with a seasoned investigative reporter from Caijing or Southern Weekend: the reporters get interviews with many of the principal actors in a situation, as well as all the relevant officials, while a blogger like Zola only gets to talk to townspeople who have lots of opinions but little first-hand knowledge.
The Internet buzz about Weng'an led to public outrage, which in turn created pressure for the government to clean house in Weng'an and open up the story to greater media coverage. But the outcome may not be increased power or respect for China's bloggers. And just because the journalists get a longer leash doesn't mean that the Chinese information environment won't still be heavily manipulated. As we know in the U.S., you can even call yourself a "free press" and still be manipulated by your government. We're starting to see early signs that China's Internet and media regulators are becoming a bit less Leninist in their techniques and a little more Rovian.