Find common phrases with your words in Netspeak

A colleague’s student just introduced us to Netspeak, a project of Bauhaus Universitat of Weimar, Germany. The point of this website is to help writers select words and phrase structures when writing in English. Netspeak uses a huge Google database (corpus) of 1-word to 5-word phrases or “NGrams” that appear frequently in publicly-available web pages in English.

You can input any word or phrase, including “wildcard” blanks, and will get back a list of phrases with your word(s) ranked in order of frequency in the reference corpus. Some really helpful searches are possible:

  • word-1 ? word-2  (What word turns up in the middle most frequently?)
  • word-1 ? ? word-2 (What two-word phrases turn up in the middle?)
  • word-1  … word-2 (What are the most common 1-3 word phrases that turn up in the middle?)
  • word-1 word-2 [word-3 word-4] (Which word is more common in this phrase, 3 or 4, or neither?)
  • And a huge bonus–every time you see one of those little plus signs, you can click on it to see a range of examples in context from the corpus.
  • {word-1 word-2 word-3} (What is the most common word order, 123, 132, 213, 231, 312, or 321?)
  • word-1 … #word-2 (What are the best synonyms for word-2 in this phrase structure?)

The data behind the phrase finder is described on the Bauhaus Universitat website.