Can and Patton (2010) provide a quantitative analysis of twentieth-century Turkish literature using forty novels of forty authors. Using weighted least squares regression and a sliding window approach, they show that, as time passes, words, in terms of both tokens (in text) and types (in vocabulary), have become longer. They indicate that the increase in word lengths with time can be attributed to the government-initiated language "reform" of the 20th century. This reform aimed at replacing foreign words used in Turkish, especially Arabic- and Persian-based words (since they were in majority when the reform was initiated in early 1930s), with newly coined pure Turkish neologisms created by adding suffixes to Turkish word stems (Lewis, 1999).