I was looking at some google N-grams and noticed that the ratio between usage for gern and gerne has a strange bump in the late 1940s. In particular, gerne gained in usage but then returned to the expected curve in the late 1950s. (N-gram link.) It's not just the relative rate, you can see the same feature in the percentage data (link).

gern v. gerne

My initial guess is that it has to do with changes in publishing after WWII, or books being destroyed, or something along those lines (rather than actual change in usage).

I'm not a langauge scholar and I'm barely a German speaker, so any thoughts or insights would be greatly appreciated. (Probably this is another case of a question that should be a PhD thesis.)

  • 1
    You might be interested in this peak as well. I did suppose something similar (see point 4).
    – c.p.
    Feb 5, 2014 at 20:56
  • Thanks! Very interesting. The answer below points out that there are only 18 books in 1942 that contain "gerne". But "dass" shows up much more often, indicating that it is not the small sample size that creates this feature.
    – philshem
    Feb 5, 2014 at 21:16

1 Answer 1



If you look closer at the data, there is no noticeable decrease in the use of "gern", but a substantial peak in the usage of "gerne".

If you go even further back with the statistics, you can also see that the usage pattern change again in the 18th and early 19th century. At least in 1946 and 1947, several reprints of books and material from authors dating back to that period were issued, e.g. from Anna and Johann Heinrich Pestalozzi, Friedrich Hölderlin, Georg Büchner, Jeremias Gotthelf and Adalbert Stifter. I am not aware of any apparent conflicts between the nazi regime and these authors, making their books likely victims for the book burning processes or similar. Any persecution by the nazis could of course have explained why these books had to be reissued after the end of WW2.

For the years 1941 to 1949 (except 1946), Google knows less than 30 books a year, in which the word "gerne" is used. Considering the small data set, it is really not unlikely that the usage peak is only a random deviation.

  • Super answer, thanks very much. How can you tell the number of total books in the year, for a given corpus?
    – philshem
    Feb 5, 2014 at 20:13
  • 2
    I don't know if there is a better way, but you can search for the books in a specific period (each year) and count the number of hits, e.g. google.com/… will find 18 books.
    – jarnbjo
    Feb 5, 2014 at 20:31
  • Clever. Thanks! edit: with smoothing off then it's mostly a sharp peak in 1946.
    – philshem
    Feb 5, 2014 at 20:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.