申論題

Part 2.(25%)
A syntactic constituent is a string of words that functions as a unit. Syntacticians use various tests to determine whether a word string is a constituent. One such test is substitution. If the test string can be replaced with the appropriate pronoun without making the sentence ungrammatical, then it is a constituent. For example, in John had a cat with blue eyes, John is a constituent (John had a cat with blue eyes → He had a cat with blue eyes) but cat with blue eyes is not (John had a cat with blue eyes → *John had a it, where * indicates ungrammaticality). For a verb phrase, we can test its constituency by substituting the appropriate form of do so. For example, in John hit the ball, hit the ball is a constituent (John hit the ball → John did so) but hit alone is not (John hit the ball →*John did so the ball).

Question 2-2. Sometimes a test may give false negative results. That is, a word string is a constituent, but the results suggest otherwise. Therefore, syntacticians usually apply multiple diagnoses. In addition to substituting pronouns, we can use one. Consider this sentence: I met the guy with brown hair from the US, but not the one from the UK. Here one is substituted for guy with brown hair, indicating that guy with brown hair is a constituent. Apply this one substitution test to the sentence in the above question. Does this test reveal any more constituents than what you have just identified? Why or why not? (5%)