[kelseyfrog]

> What ends up in the university libraries used for the Google n-gram data varies with trends in Google’s book-inclusion policy, editorial practices, library policies, and popularity of genres. As none of those effects can be excluded it is important that we find the same trends for word use in the New York Times.

The interesting line of thought here is that NYT articles have fairly clear authorship (and to a lesser degree editorship). Do individual authors follow a similar trend (implying that people are changing) or do authors stay the same (implying that the NYT composition is changing)? In short, how much is personal change, and how much is churn?