A Tale of Two Tiananmens

Compare two image searches, one in Google.com, the other in Google.cn: this vs. this. While it should come as no surprise in the censored circumstances, seeing the difference so starkly in images is striking.

[Courtesy of Jerry]


  1. Michael Robinson says:

    As much as I would like to join the high-fiving “busted” faction, I should point out that if you run the patented Google-exclusive PageRank algorithm on websites worldwide, and run the algorithm again on websites in China, you’ll get pretty much the results shown.
    In other words, the two tales of Tiananmen differ so much not because Google is censoring the images, but because they are running an index against a population of websites that are censoring the images.
    Conclusion: bad bust.

  2. Censorship Visualized

  3. But how are they deciding the “population of websites”? For instance, http://blogs.msdn.com/junfeng/archive/category/3238.aspx, is one of the image pages for the .cn search, so it’s not by an obvious filter (like .cn).

  4. I don’t think censorship is at hand here…; you’ll begin seeing similiar tank images on page 5 of the Chinese image search.
    For example, a Google text search for “Banda Aceh” in the US returns a lot of talk about the tsunami aftermath: http://www.google.com/search?hl=en&q=banda+aceh
    However, the Indonesian Google text search doesn’t contain nearly as many tsunami related results: http://www.google.co.id/search?hl=id&q=Banda+Aceh&meta=
    It certainly appears to me that the results are “localized.”

  5. Perhaps a better example of possible censorship, try “Falun Gong”
    In this case, you’ll find that the Chinese results are nearly all are part of propaganda articles about the mistakes of the group while the US one centers around torture and protests.