I encountered a difficult question for syntax.

The authors discuss a special case of coordination called “right node raising”. When both conjuncts have the same string (e.g. typically an object or a complement, i.e. underlined strings) (i.e. (1a), (2a) and (3a)), the shared string (i.e. wavy underlined strings) can appear only once at the end of the clause or “right node raised” (i.e. (1b), (2b) and (3b)).

a. They play unusual music, and I listened to unusual music.
b. They play and I listened to unusual music.

a. Everyone claims that John lied, but Mary does not believe that John lied.
b. Everyone claims but Mary does not believe that John lied.

The question is:

Based on the coordination test results in (1b)—(2b), discuss which structures above the data support.

Paul claim that the well-formedness of (1b), (2b) implies an unexpected structure involving the verb and modal.discuss why the test results in Part A are unexpected. Whenever appropriate, include relevant examples in your answer.

  • 1
    To be subject to RNR, the coordinated constituents have to end in the same constituent (not just the same string). There is a Wikipedia entry for RNR, and McCawley has extensive discussion of it in his text The Syntactic Phenomena of English. (I don't understand what is being asked in the exercise you're working on.)
    – Greg Lee
    Feb 19 '17 at 10:01
  • The exercise is quite interesting but what is your own question? What exactly are you struggling with?
    – Alex B.
    Nov 11 at 14:10

Coordination is supposed to be a test for constituency. Based on this test, it seems that in 1b "they play" and "I listen to" are constituents - precisely because they are coordinated. Same for 2b. Here the coordination test suggests that "everyone claims" and "Mary does not believe" are constituents.

These results are unexpected because (I think) the mainstream view is that in English the subject + verb cannot form a constituent without also including the object. And yet that is what we seem to be seeing here.

  • 1
    That's one reason why RNR is an unusual rule; it doesn't behave normally with respect to constituency.
    – jlawler
    Feb 24 '17 at 17:43
  • 2
    @jlawler An alternative perspective is that the received notion of constituency is wrong, and e.g. "they play" is a possible constituent. That would be the position of many categorial grammarians. Mar 3 '17 at 12:50
  • 1
    Same thing at a different level. It's a matter of what you want to pin the irregularity on. For many speakers they play is a constituent; for others, it's not. That's the way language changes. These distinctions don't have sharp edges as a rule; it's only the prototype cases that act that way.
    – jlawler
    Mar 3 '17 at 14:33

Your Answer

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

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