Phys. org: Texts as networks: How many words are sufficient to identify an author?. “People are more original than they think—this is suggested by a literary text analysis method of stylometry proposed by scientists from the Institute of Nuclear Physics Polish Academy of Sciences. The author’s individuality can be seen in the connections between no more than a dozen words in an English text. It turns out that in Slavic languages, authorship identification requires even fewer words, and is more certain.”
I am a big fan of language preservation, which is why I was so happy to read this article on a new site that will publish news in Lakota. There are only a couple thousand Lakota speakers left. “Other media outlets provide news of interest to the community, but in English. Woihanble.com’s local news content comes from two area weeklies that mostly focus on Native American issues. The site has an agreement with those weeklies to translate stories into Lakota, with links back to the original articles. And in recognition that many people who speak the language well do not read it easily, news stories include audio clips in Lakota.” I went to the site to hear an audio clip. If you go to http://www.woihanble.com/#!pipeline/f2l81 it seems like the audio clip might be the whole story… I didn’t listen to it all the way through, but I listened to half of it.