In 2005, the Seigenthaler incident happened on English Wikipedia (https://en.wikipedia.org/wiki/Wikipedia_Seigenthaler_biography_incident). This probably would have just been another incident of routine vandalism, except for the fact that John Seigenthaler, the victim of the vandalism, was the founding editorial director of USA Today. As it happened, the incident caused a huge amount of PR damage to Wikipedia, and Jimmy Wales had to take drastic measures to address the problem. One of these measures was to disable anonymous page creation on English Wikipedia as an experiment (https://lists.wikimedia.org/pipermail/wikien-l/2005-December/033880.html, https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2005-12-05/Page_creation_restrictions).
This "experiment" became the de facto state of affairs and was not addressed again until 2007, when Gregory Maxwell proposed re-enabling anonymous page creation, also as an experiment (http://article.gmane.org/gmane.science.linguistics.wikipedia.english/84182, https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2007-10-29/Page_creation). Tim Starling stated that such a change would need consensus from the community, so an RFC was created (https://en.wikipedia.org/wiki/Wikipedia:Requests_for_comment/Anonymous_page_creation). No consensus was reached in the RFC, so anonymous page creation remained disabled (https://en.wikipedia.org/wiki/Wikipedia:Wikipedia_Signpost/2007-11-12/Page_creation_redux), and has been that way ever since.
Actually Thomas Dalton said that consensus would be required. Ray Saintonge responded by saying that more than consensus is required -- someone would have to actually do it. I responded to Ray by saying: just file a bug, we'll do it, don't worry. I didn't mean to block the change.