search

delph-in / erg

English Resource Grammar
MIT License
17 stars 3 forks source link

the limits of ERG interpretations of wrongly chosen prepositions? #38

Open arademaker opened 1 year ago

arademaker commented 1 year ago

a charge in the shape of a small ring

http://delph-in.github.io/delphin-viz/demo/#input=a%20charge%20in%20the%20shape%20of%20a%20small%20ring&count=10&grammar=erg2018-uw&mrs=true&dmrs=true

image

Why I didn't get a reading with _in_p_state? OK, from the ERG source, I know that. 

terg/etc/hierarchy.smi
364:  _in_p_dir < _in_p_loc.
365:  _in_p_loc < unspec_loc.
366:  _in_p_state < _in_p_loc.
367:  _in_p_temp < temp_loc_sp.

So loc is a kind of supertype for _in_p_state... but still, we read location, right? Here the preposition in is more about with or X has the shape ....

Maybe this WordNet definition should be best written as one of the following paraphrases.

  1. a charge with the shape of a small ring
  2. a charge that has the shape of a small ring

Does it make sense? I mean, it is to expect so much from ERG to give us all our desired interpretations for a poorly chosen NL sentence.

(This is all thoughts from a non native English speaker)

guyemerson commented 1 year ago

I think "in" is a perfectly good choice of preposition in this sentence.

I suppose this comes down to how we are supposed to interpret _in_p_loc. If this just means "non-temporal" (because it contrasts with _in_p_temp), then this seems okay: "in the shape of" means something spatial, not temporal.

arademaker commented 1 year ago

But still, no reason for not having a reading with _in_p_state, right?

Although, _in_p_state is actually a subtype of _in_p_loc ... so maybe that is the reason, no deep constraint from the noun to more fine-grained interpretation of the preposition.

guyemerson commented 1 year ago

I think that's right.

(Although I'm struggling to think of an example of an in-PP modifying a noun with a directional sense... I keep wanting to say "into" instead.)