Partitives/subsets

[nschneid]

Perhaps I missed it, but I don't think the guidelines address examples like:

• some of the books are missing
  • Note that this presumes a relevant set of books. It entails, but is not equivalent to, "some books are missing".
  • "All of the books are missing" does not necessarily entail "all books are missing".
• half/a third/20% of the books are missing
• 4 of the 5 books are missing
• the rest of the books are missing

Cf. SNACS analysis under Quantity and Whole.

[nschneid]

Suggestion based on today's discussion in class:

When referring to a subset of items, if two quantities or sets are given and only one of them has a descriptor with it, use a remote to associate it with the other:

• [[most_Q (books)_C]_C of_R [the_E books_C]_E]_A are missing
  • Contrast: [most_Q books_C]_A are missing
• [[4_Q (books)_C]_C of_R [the_E 5_Q books_C]]_A are missing
  • Note that there may be a singular/plural mismatch between the primary and remote descriptor: [[1_Q (books)_C]_C of_R [the_E 5_Q books_C]]_A is missing
• TODO: set proportions (half/a third/20%/the rest)—can't say "a third books of the books", so a remote feels weird
• TODO: rates: "4 in/out_of 5 books are worthless", "I read a book a day", "I read 2 books per day"
• TODO: non-set wholes: "most of the foliage is green", "most of the land is unplotted"
• TODO: collective/framal wholes: "most of the group/company"
  • Does this warrant an implicit unit??? "Most of the company" = most people of/in the company
• TODO: composite measures: "a third of a meter long"

[nschneid]

Most of the examples are plain quantities rather than subquantities of a definite NP superset. "Portions" had the_F rest_Q [of_R the_F cake_C]_C. I added some_Q [of_R the_F cats_C]_C and 80%_Q [of_R women_C]_C. Also, elsewhere, the example "one of Selznick’s productions" appears. This solution collapses the two kinds of quantities, which is probably a good idea (though SNACS distinguishes them).

There may be unresolved TODOs above, but I'm not going to worry about them now (they may be rare anyway).

[omriabnd]

OK. We should add rare examples to the interesting examples appendix to prevent the guidelines from over-inflating.

On Fri, Nov 23, 2018 at 8:17 AM Nathan Schneider notifications@github.com wrote:

Most of the examples are plain quantities rather than subquantities of a definite NP superset. "Portions" had the_F rest_Q [of_R the_F cake_C]_C. I added some_Q [of_R the_F cats_C]_C and 80%_Q [of_R women_C]_C. Also, elsewhere, the example "one of Selznick’s productions" appears. This solution collapses the two kinds of quantities, which is probably a good idea (though SNACS distinguishes them).

There may be unresolved TODOs above, but I'm not going to worry about them now (they may be rare anyway).

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/UniversalConceptualCognitiveAnnotation/docs/issues/26#issuecomment-441162831, or mute the thread https://github.com/notifications/unsubscribe-auth/AIG866bjOR4xrLUSK8gLzh6pyLBYRh5Uks5ux5MQgaJpZM4WoZOs .