I’ve long had an interest in languages.  I adored my English classes in high school, and one of my majors as an undergraduate was German.  As a result I get particularly excited when I see stories having to do with the intersection of computation and language.  A recent post on a blog associated with MIT discusses a new algorithm developed by John Kleinberg and others that measures power differences between individuals using the pattern of words they use.  Their observation is that in conversations (oral or written) the person with less power will tend to echo the linguistic style of the person with more power.  The blog article discusses it in more detail, including giving some possible applications of the approach.  I am a bit skeptical of the suggested applications, but I think it’s a really cool idea.

Advertisements