3

Why is the sentence

John wants to read.

normally analysed as consisting of 2 clauses? (John wants, PRO to read)

I understand the idea of PRO but why must to read be a completely different clause?

What does the traditional syntax say about such examples? Does it also claim that there are 2 clauses?

But the sentence

John wants a book.

is only analysed as 1 clause. (Or am I mistaken?)

Why is there a difference?

  • This has always bugged me too. Either I never fully grokked it or other analyses may be possible that some of us might find more intuitive ... – hippietrail Nov 22 '14 at 4:11
1

It depends who you talk to, but I think there's some logic in saying that a clause is simply another name for a verb phrase. (Ignoring non-verbal clauses.) Two verbs mean two clauses.

Infinitive arguments can have their own set of independent arguments and adjuncts. Here is a complex example, but I think it illustrates that there are two clauses quite clearly:

Last night at the bus stop the son wished for the father to read a story to him in bed when they next went to grandma's house.

Two locative adjuncts: at the bus stop and in bed
Two temporal adjuncts: Last night and when they next went to grandma's house
Two subject arguments: the son and the father And the infinitive verb has its own set of object arguments which are not arguments of wish: a story and him

Strip all that back and you get

The son wished to read.

Rather than propose that it has a different structure I think it's best to say that it has the same structure: two clauses.

| improve this answer | |
  • Downvote: The second sentence is definitely not what the first one is about. The first one shows a desiderative construction, where the son wants somebody else (=his father) to read something. The second sentence shows a volitive construction: the son wants/wishes to do something himself. In volitive constructions, the subject argument of the volition equals the subject argument of the activity, in desideratives, subject arguments are distinct. – Thomas Gross Nov 22 '14 at 9:36
  • @ThomasGross I wasn't implying the sentences were equivalent and I don't know why you'd think I was. I know the subject of the infinitive changes. – curiousdannii Nov 22 '14 at 10:11
0

Clauses are usually made up of finite verbs in predicates and additional dependent information. "John wants to read" and "John wants a book" are more-less equivalent. Non-finite verbs, such as "to read" (a full infinitive) are usually parts of a verb chain/verb catena. The difference between those two is just that there is a noun in the second sentence and a non-finite verb in the first. Both are mono-clausal. You might've gotten the bi-clausal bit of data from some book; it is probably mistaken (at least as far as I know). It might have a different definition of clause.

| improve this answer | |
  • 1
    If you're couching your answer within a particular framework (in this case dependency gramar), do make it explicit. – P Elliott Nov 11 '14 at 19:37
  • Sorry, I'm using dependency grammar mostly here. EDIT: Just saw you mentioned it's dependency; double facepalm insues – Darkgamma Nov 11 '14 at 19:38
  • No worries, I just think it's good practice. Just to satisfy my own curiosity, how to non-finite interrogative clauses fit into this picture (Dependency Grammar)? Are they mono- or bi-clausal? "John knows where to go". – P Elliott Nov 11 '14 at 20:07
  • The thing that goes here is that those bits of verb chains are nonfinite clauses/"defective" clauses (as I've somewhere read). That's a relative nonfinite clause introduced by an interrogative pronoun; I'd probably not go wrong to say it's bi-clausal. It differs from "wants to X" in that it's made up of a full nonfinite clause while "to X" is bare. In general, non-finite verbs are limited as to what kinds of clauses they can form. – Darkgamma Nov 11 '14 at 20:17
0

This is the analysis of generative grammar, there is a D-structure for a sentence, which is usually much different from the S-structure of the sentence. By D-structure, you can know why, to understand D-structure, you have to know a lot of reasons or evidences that support those structures.So, it seemingly is not consistent with traditional analysis.see Chomskyan works for reference. BTW, I don't think clause is an important notion of generative grammar.

| improve this answer | |
  • 2
    Why did you make a second answer? This is essentially the same as your first one. You should have just edited it. – curiousdannii Nov 13 '14 at 0:37
  • 1
    @curiousdannii about the flag... mods can't merge posts. It's up to the original authors to make all such modifications. – prash Nov 13 '14 at 13:33

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.