How will a Do-So test help in identifying if a PP is an adjunct or complement to the verb? Example : I ran to work. I need two pieces of evidence to prove the status of PP in a sentence.


Recall that in English, adjunction is believed to look like this:

Syntax tree \VP \V' \V' \V ran\ \ \PP to work\ \ \

If the lower V' is a constituent, then we should expect it to be substitutable by 'do so' in the standard test, but we can't say that:

#I did so to work. (meaning 'I ran to work')

By contrast, you can substitute the whole thing 'ran to work' with do-so:

I went to work on Thursday.

I did so on Thursday.

This suggests that 'on Thursday' is an adjunct and 'to work' is the complement, i.e. this structure is appropriate:

The other tree

  • Caution: Identification of adverbial phrases is ambiguous (unless marked by a comma). I ran, [in order] to work [out]. I worked on Thursday['s edition of the newspaper]. – amI Mar 29 '17 at 21:39
  • @amI But if the meaning's evident then the structure is not ambiguous. – Araucaria - he him Mar 29 '17 at 22:49
  • Many grammarians use "licensing" to determine whether an element is a complement or an adjunct/modifier. Essentially, an element has to be "licensed" by an appropriate head to qualify as a complement. – BillJ Mar 31 '17 at 7:15
  • @BillJ I agree that's a useful test, but the OP asked for do-so, so I... lowers glasses did so. – WavesWashSands Mar 31 '17 at 15:34
  • The 'do-so' test merely shifts the tense to an auxiliary and replaces the verb with a pronoun (so, it, that, ...). It is only good for identifying the direct object, since we only expect un-case-marked objects after an actual verb. Thus 'I ate the cake' becomes 'I did so to the cake'. "I did so to work" is not ungrammatical; it is merely clunky because 'to' is poly-semantic. – amI Apr 6 '17 at 20:44

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.