You made a lot of strong points here, Mitch. I agree that China censorchip, like most other forms of censorship, have workarounds. Once you figure it out, you can keep doing it and hope that the ones in charge won't patch the 'loophole' so you can keep at it.

As for this part: "And you have to wonder what those people could be doing if they turned that energy to something that was actually productive."

I couldn't agree more. They should be doing something more.

Good point, Michael. China and censorship have been talked about by many for years, but the ones who really know what's going on are the ones who are in China, and they probably won't be able to tell us themselves.

I'd like to remain optimistic though and take these results and studies with a grain of salt. If it's true, then good for them. Hopefully this will spin off into something bigger, for the sake of its people.

I've been reading over the past year or so that sentiment analysis was having a similar problem, because it couldn't tell whether people were being sarcastic, and so they might attribute the wrong feelings to a particular post.

Twitter, in particular, due to its length limitations, lends itself to lots of verbal shortcuts, code words, and references (like the people who know the jokes so well that they just tell them by number: "467!" "You didn't tell it right.") Look at how many popular culture references there are in Tweets; if you weren't familiar with the back culture, you'd have no idea. It's like the Star Trek TNG episode Darmok (one of my favorites, actually), where their entire language is metaphor. 

And then there's "family dialect," or the code words based on shared experiences that families develop and use without even realizing it. People online develop that regionally but also based on whatever 'tribe' they're in, programmers or whatever. How many references do you see to xkcd or Oatmeal comics that you wouldn't recognize if you didn't know?

When the show Seinfeld was on the air, I decided one night to watch it. It happened to be the "Yadda yadda yadda" episode, now considered to be a classic. And I was astonished by how many coworkers made indirect reference to it during the day -- something I wouldn't have noticed at all if I hadn't happened to have seen it.

I hear a lot about China and censorship, but all opinions. The only ones who truly know how successful are the ones doing it.

Define successful. 

Chinese censorship is easy to get around for people with a modicum of technical skills. But most people don't have that modicum, so most forbidden content remains hidden. In that respect, an imperfect censorship regime proves to be good enough. 

But censorship always fails in that societies that practice censorship lose out to free societies. For nations to excel, they require a robust exchange of ideas, not timid discussions where all parties are looking over their shoulders for the cops. 

You have to admire the ingenuity of people who have imported Cockney rhyming slang from half a planet and more than a century away. And you have to wonder what those people could be doing if they turned that energy to something that was actually productive. 

According to the authors: "... our data collection process involves locating,  accessing, and downloading posts from many web sites before Internet content providers or the government reads and censors those that are deemed by authorities as objectionable..."

then: "... revisiting each post frequently enough to learn if and when it was censored..."

Read the full paper here: http://gking.harvard.edu/files/censored_0.pdf

I was wondering how researchers or anyone know if or if not the censorship is successful?