7

I'm not terribly familiar with the syntax of PPs, but I've noticed that a handful of them do not seem to satisfy any constituency tests, particularly 'than'. 'Than' is listed in dictionaries as both 'preposition' and 'conjunction'. The former is likely referring to cases where 'than' precedes - and, supposedly, governs - an NP. For example:

  • *Than the ant, the elephant is larger.
  • What is the elephant larger than? ?Than the ant.
  • *Than what is the elephant larger?
  • *It is than the ant that the elephant is larger.
  • I can't think of any word that can substitute 'than + NP'.

Moreover, 'than' may precede clauses as well:

  • He's better than you think.
  • I've eaten more salt than you've eaten rice. (OK, this is a joke translation of a Cantonese phrase.)

That raises the question of whether it constitutes a coordinating conjunction or a subordinating conjunction (which, depending on the conjunction, are subsumed under prepositions and complementisers in modern times). Since '*Than you think, he's better.' is not grammatical, the prepositional analysis seems unlikely (compare as long as, because). While this doesn't undermine its preposition status when preceding NPs, it seems to make it even less likely now.

Coordinating conjunction actually seems possible. We say 'He is tired but happy' and 'He is more tired than happy'. We say 'He looks confident and you think he is (confident)', and likewise we can say 'He looks more confident than you think he is (confident)'.

The other alternative is that 'than' is a complementiser, but that cannot capture the cases where 'than' precedes an NP. Besides, that would imply constituency, which again seems not to be satisfied (e.g. *than I have, John has eaten more.)

My questions are:

  • Is there any good way to defend the constituent status of than-phrases?
  • If not, have there been any alternative accounts in the literature? Could 'than' be a coordinator?

Thanks.

  • What's strange is "more beautiful than someone" like structures. Like, we can assume that {more} and {-(i)er} particles are simply adverbs if we try on empirical ground. Then, we assume a Adv AdjP situation which is too odd. I will try to draw a tree on this. – Eray Erdin Jun 2 '16 at 13:52
  • I failed. Tree seems too odd to me. – Eray Erdin Jun 2 '16 at 14:01
  • 1
    It's part of the comparative construction, and it serves to introduce the baseline for the comparison. You can call it anything you want, POSwise, but it doesn't behave like anything else because it's specific to comparatives and never occurs outside them. – jlawler Jun 3 '16 at 4:47
2

The problem with your tests is that the than-phrase is a complement. The PPs that you compare them with are adjuncts (read adverbials). We expect to freely front adverbials, but we don't expect to readily be able to front complements, which usually occur to the right of the word(s) that license them.

It seems very unlikely that then could be a coordinator, because no words seem to take a single coordinate as a complement!

| improve this answer | |
3

In answer to your first question, acting as antecedent or deletee in anaphoric deletion is sometimes taken as evidence for being a constituent. And "than NP" does that:

  • "I like cookies better than cake, but Moira likes pie better (0 = than cake)."
| improve this answer | |

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.