The accustomed methods of government censorship begin to creak under the possibilities in computing contemporaries, who can successfully apply appropriate statistical methods and probabilistic. With these techniques it becomes possible to reveal certain types of censorship, such as those obtained by deleting some words in official documents before they are made public. Claire Whelan of Dublin City University has just put a technique that has enabled it to decode what had been removed from a confidential document to the U.S. president dating shortly before 11 September 2001 and made ​​public in April 2004. As reported in an article in Le Monde , the researcher has completed its task indentificando, in the order of the text font used and size of the deletion, and comparing the results with the rules of grammar and context, resulting in less of a ten possible choices from which to choose the most sensible. Although, by his own admission, the technique becomes imbroba for three or more consecutive words and useless for entire paragraphs as it proves the 'omission', which in the American documents are blackened, may be revealed by mathematical analysis of the text. Which also opens new avenues for the study of historical documents that could reveal unspoken truth so far.