This question on NLP asks for the how of parsing, and the given answer shows how the state-of-the-art isn't at that level. But whether this is a hard problem or not is not my question: what I ask is, is it a possibility to unambiguously parse sentences for levels of nesting?

  • 1
    That seems to depend on what kind of input you use. Written sentences are almost impossible to parse unambiguously, as this paper pointed out long ago. Humans using real (spoken) language are much better because they're not limited to written input without intonation, rhythm, gaze direction, gestures, etc, and we use them all. Writing is neolithic technology. English writing is paleolithic. – jlawler Apr 10 at 16:53
  • I don't think I understand the question. Do you require the parsings to be correct? If so, of course you need to know how human speakers of the language actually do parse them. A beginning can be made by examining anaphoric relationships among the members of 3 or more coordinated items, which McCawley does in The Syntactic Phenomena of English. – Greg Lee Apr 11 at 10:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.