Those who have visited my blog this year will know that I have put up some “filtered French”, such as a list of the most common one-word sentences in French classic literature, and sentences that fit the highly constrained vocabulary of my comic books. After musing on language acquisition, in particular how babies learn, not to mention our experience of picking up a few words and phrases in a foreign language by ear, I thought I’d try a different approach. This has resulted in producing a book (with more volumes to come) where I filter Les trois mousquetaires, and add vocabulary one word at a time based on which word will complete the most sentences. Using a combination of manual and automatic filtering, I have created extracts that have sufficient repetition in their vocabulary for people to become familiar with the words.

It has been fascinating to see what happens as I add each new word. The algorithm tends to find dialogue first, gradually increasing in average sentence length, then short non-dialogue sentences – after the 93rd word of vocabulary was added.

Anyway, if you’d like to have a look, it’s on Amazon, with a substantial preview.

(Affiliate links in this post.)