I'm currently reading a Ph.D. thesis concerning the structure of English adjectives. These generative trees are used to illustrate the structure of the phrase "someone tall". The author states that Picture 185 is illicit since it violates HMC and Picture 186 is the correct one. So what is HMC and why it doesn't allow the N-to-D movement in this situation? I'm confused.
1 Answer
This stands for "Head Movement Constraint". It was introduced by Lisa Travis in her 1984 MIT dissertation Parameters and Effects of Word Order Variation, p. 131:
Head Movement Constraint: An X0 may only move into the the [sic] Y0 which properly governs it.
(Note that X0 is notation for an X head.)
In your (185), someone could not move from N to D, because the A head intervenes. For this reason the AP must be taken as an adjunct of the NP, so that the D does properly govern the N.
-
I'm not a student of generative grammar, but am wondering why someone is regarded as a determiner there and not a pronoun with tall an ascription. Someone could do it. Someone tall could do it. Or is it some that is the determiner there, with one as the pronoun? Does generative grammar split words like someone into two constituents? Some one tall could do it.– TimRCommented Aug 9, 2023 at 12:09
-
@TimR I don't think the two ("regarded as a determiner ... and not a pronoun") are mutually exclusive. Pronouns are in complementary distribution with e.g. definite articles (*the they, likewise *the someone), so they are typically seen as determiners. I am not familiar with the term ascription. It is possible to analyze someone as an internally complex head (just as in front of can be seen as a single, internally complex prepositional head), but I'm sorry, I don't see how this is related?– KeelanCommented Aug 9, 2023 at 12:35
-
I was just trying to understand the trees and identify the heads. I take some there as the determiner in the DP, with one as the N of the NP in the DP, and tall to be predicated of one not a modifier of one : Some one [who BE] tall, not some tall one.– TimRCommented Aug 9, 2023 at 16:31
-
@TimR The trees in the question do not give an analysis of the internal make-up of someone. They analyze someone as the head of the DP. In (185), the DP governs an AP with head tall, which governs an NP from which someone has been moved. In (186), the DP governs that NP directly, and the AP modifies the NP.– KeelanCommented Aug 9, 2023 at 20:08
-
So an AP is always assumed to have a slot for an NP complement?– TimRCommented Aug 9, 2023 at 20:25