> The company might want to take some steps to detect specific examples of search result censoship, such as when items that appear when searching yahoo for A B

> The company might want to take some steps to detect specific examples of search result censoship, such as when items that appear when searching yahoo for A B C D do not appear at all in a search for A B C without D, and fix them. To me this does not sound like censorship, but SEO on behalf of D. My guess would be that D paid some web presence management company to cultivate a strong presence of pages containing A B C but not D (since they want to hide their name after all), and that's what the engine is picking up. There could even be an innocent explanation for this - lots of UK discussion about A B C that does not mention D since they're not allowed to, and very few outside the UK who care. So the {A,B,C}\{D} pages would naturally rise to the top of search results. This particular case aside, Stallman's test looks like misuse of statistics to me - searching for "washington" gets me lots of pages that don't mention the town in England, whereas "washington england" finds it. Conspiracy? Censorship? No.