When are prepositional phrases sibling to leaf nodes?

Question

In syntax trees in English, can prepositional phrases, modifying either verbs or nouns, ever be sibling to the verb or noun itself for example and not a verb phrase/noun phrase?

I've heard the correct description of the constituents of a sentence like "The phone in Jack's pocket rang" is

(S (NP (NP (Det The) (N phone))
       (PP (P in)
           (NP (Det Jack's) (N pocket))))
   (VP rang))

as corroborated by the parser https://www.link.cs.cmu.edu/link/submit-sentence-4.html that generated the skeleton of the above tree (I added the N, Det, P, leaf nodes).

As you can see, the subject noun phrase "The phone in Jack's pocket" consists of two constituents, a noun phrase "The phone" and a prepositional phrase "in Jack's pocket". This is despite what might be (in my opinion) another reasonable syntax tree, breaking the noun phrase "The phone in Jack's pocket" directly into a Det "The", an N "phone", and a PP "in Jack's pocket". Additionally "The phone" seems to fail the replacement test for a noun phrase: "It in Jack's pocket rang" doesn't sound grammatical to me anymore.

What's the correct hierarchy of NPs or VPs that contain PPs and why?

Answer 1

It is reasonable to think that the locative adjunct "in Jack's pocket" restrictively modifies a part of the noun phrase that contains the noun but not the determiner; the determiner picks out the uniquely identifiable "phone in Jack's pocket". (Compare, "The new phone in Jack's pocket was right next to the old one".) Sometimes people distinguish a level N' intermediate between N and NP; N' would contain a complement of N (as in "picture of his mother") but not an adjunct.

In order to prevent *"it in Jack's pocket" you have to prevent the locative PP adjunct from attaching to the pronoun. This might not be a structural fact, but if you wanted to, you might be able to argue that the pronoun doesn't project N'; then a rule allowing attachment of adjunct NPs to N' wouldn't apply to the pronoun.

Some people (e.g., Abney 1987 MIT PhD dissertation) argue that pronouns are category D, like determiners, and that the determiner is the functional head of the noun phrase, which is then a DP. If there is a lexical noun, then the DP also contains an NP. In that case, the locative PP can adjoin to NP, below the determiner, and a pronoun like "it" doesn't contain any NP at all that could host the PP adjunct.