Is discussion of WikiLeaks censored by Twitter?

Here's an interesting development in the 'internet and free speech' debate. Twitter is apparently censoring discussion of the WikiLeaks release by not showing #wikileaks, #cablegate or #assange as trending topics. Tech tweeter Dave Winer noticed this earlier this morning and this well-researched blog post has also suggested censorship on the part of Twitter, demonstrating the evidence with a graph of 'true' trending phrases.

It will be fascinating to see if there is any truth behind the suggestion of wilfull censorship. Twitter seem to be denying it so far (see also; Mashable), but it seems likely that another bit of murky water is about to flow into the torrent of WikiLeaks news stories, especially now that Assange has been arrested.

If Twitter is censoring trending topics, or has done in the past - on any subject - then that is a major infringement on what Twitter claims to be. That is, a platform for free speech and discussion. Clandestine censoring is even worse, but of course any public announcement of this as 'official policy' would hardly have been taken well by Twitter users.

It would also be interesting to know who specifically took the decision to censor, and where pressure to do so originated, should the allegations prove true.

However, one additional point to consider, in support of Twitter's statement that their algorithm has remained true in accurately showing tweet diversity and volume in the trending list, it is possible that those who have commented on this 'apparent censorship' feel the way they do because of a skewed sense of perspective. Not everyone is talking about WikiLeaks, after all, but social media and journalism buffs have been obsessing about it for two weeks now. Interestingly, 'Assange Arrested' - which refers more generally to a news story - has entered the trending topics today.

Without clear evidence to suggest such skewed perspectives, however, (for which we currently have to trust in algorithms and the motives of those who deploy them...) there is no empirical way to know for sure.