Pronouns Raising and Emerging

This dissertation revisits the question of the syntactic and semantic status of pronouns, incorporating new syntactic, sociolinguistic, and pragmatic data to support an analysis of n-to-D head movement in the nominal domain. The support for pronouns originating in n comes from predicative pronouns, including pronominal relative clauses (1) and depronominalizations (2). 

(1) he who is without sin
(2) That person is a she. 
 

I compare predicative pronouns with variable grammatical restrictions on singular 'they' using data from two sociolinguistic studies that I conducted. I show that there is an effect of speaker age on production and perception of definite, specific uses of singular 'they' (dsT) as in (3), while definite generic (4) and epicene uses (5) are more broadly accepted. 

(3) Jayden forgot their homework.
(4) The ideal student never forgets their homework.
(5) Every student should do their homework. 
 

I take the sociolinguistic variability in singular 'they', particularly the differences related to age, as evidence of an ongoing change in the grammar of English towards increasing use and acceptance of the type shown in (3). The inclusion of dsT in the grammar predicts intraspeaker sociopragmatic variation in pronoun use, and this prediction is borne out. Through variable rankings of pragmatic constraints I show that dsT enables speakers to include or exclude gender features from pronominal choices in order to achieve strategic discourse goals in various contexts. 

The n-to-D head movement analysis that I propose accounts for predicative pronouns, as well as differences in grammaticality of dsT and its related discourse-sensitivity, by separating pronouns into sub-classes depending on how far head raising proceeds. For predicative pronouns, external determiners (overt or covert) block head movement completely, and pronouns stay in n. For epicene pronouns like (5) and definite generic antecedents like (4), the pronoun raises from n to an intermediate functional projection (Num) but is merged with a variable D. Finally, referential (specific) pronouns like (3) are formed through movement from n through Num to D, where the pronoun combines with a phase head D that is linked to a discourse referent. Because phase edges are sensitive to discourse context, it is only when a pronoun moves to D that it is evaluated for context-appropriateness relative to the referent picked out by D. 
 

This system of context-appropriateness necessitates analyzing the natural gender features of pronouns as less like noun classes and more like honorifics, in that they signify social relationships rather than grammar-internal categories. An honorific analysis of gendered pronouns more robustly explains the sociopragmatic variation found in natural language use, and is more generalizable cross-linguistically.