search

nert-nlp / cgel

CGEL trees.
Creative Commons Attribution 4.0 International
6 stars 3 forks source link

"look around at how lucky we are to be alive right now" #108

Open nschneid opened 4 months ago

nschneid commented 4 months ago

An exclamative clause within a PP. Is "to be alive right now" a complement of "lucky"? How does that work?

BrettRey commented 4 months ago

Nice find!

There are subordinate exclamatives, though I'm not sure I had really internalized that fact.

  1. I’ll tell them what a good player she is. (p. 62)
  2. I was amazed what a fuss he made. (p. 545) [could equally be amazed at or amazed about]
  3. I realise what a bargain it is. (p. 854) etc.

Other prepositions:

  1. It has reminded me of how important it is to maintain control of copyrights.
  2. It makes me reflect on how important it is to be careful.
  3. you'll be impressed by how easy it is to use
  4. critics have pointed to how difficult it is to establish the adaptive function
  5. Nobody can prepare you for how difficult it is to move out
  6. I've been frustrated with how difficult it is to do things

In terms of the structure, the canonical counterpart is We are (very) lucky to be alive right now and the exclamative phrase is fronted, so yes, the infinitival is licensed by lucky.

nschneid commented 4 months ago

What I'm not sure about is the whether only a part of the AdjP has been targeted for fronting:

[Clause Prenucleus:[how lucky]i we are [AdjP Head:__i Comp:[to be alive right now]]]

Or whether the whole thing is fronted and then there is postposing of the heavy complement phrase:

[Clause     Head:[Clause Prenucleus:[AdjP Head:[how lucky] Comp:__j]i we are PredComp:__i]     Postnucleus:[to be alive right now]j ]

BrettRey commented 4 months ago

So, not just a problem within the PP. Would you say it's similar to:

  1. What time should I arrive that would be better?
  2. Which one did he choose of the three you offered?
  3. Consider the knowledge he had about the past.
nschneid commented 4 months ago

Yeah I suppose so. Things that would be nonprojective in a dependency parse.

nschneid commented 3 months ago
image
nschneid commented 3 months ago

^ This suggests the postnucleus is extracted from inside the WH-phrase prenucleus, which is weird but (it seems to me) less weird than the reverse.

BrettRey commented 2 months ago

The reverse being extracting the prenucleus from the post nucleus? That hadn't occurred to me and strikes me as very weird. The other option I was wondering about, but not seriously, is somehow leaving the complement in the are VP and extracting only the head. But your solution seems much better than that.

BrettRey commented 2 months ago

Can you paste the LaTeX for this?

nschneid commented 2 months ago

Pushed to the datasets/oneoff directory: https://github.com/nert-nlp/cgel/blob/main/datasets/oneoff/lookaround.cgel https://github.com/nert-nlp/cgel/blob/main/datasets/oneoff/tex/lookaround.tex https://github.com/nert-nlp/cgel/blob/main/datasets/oneoff/pdf/lookaround.pdf

BrettRey commented 2 months ago

Thanks!