Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier



Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program


Year Degree Awarded


Month Degree Awarded


First Advisor

Thomas Roeper

Second Advisor

Jeremy Hartman

Third Advisor

Rajesh Bhatt

Fourth Advisor

Jill de Villiers

Subject Categories

First and Second Language Acquisition | Psycholinguistics and Neurolinguistics | Syntax


This dissertation examines the syntax of Free Relatives (FRs) in English at different stages of first language acquisition. The goal is to provide a theory of Free Relatives that explains phenomena in adult and child FRs, is feasibly learnable by a child, and reflects principles expressed in theories of Universal Grammar based on the Minimalist Program (Chomsky 1993, 1995, 2005).

The central empirical concern is the difference between the distribution of Wh expressions in FRs vs. Wh questions in English, the difference in grammaticality between Charles wondered dish what Sebastian made and *Charles ate what dish Sebastian made (*Wh-NP). To explain this and related phenomena I develop an analysis of FRs building on recent work by Caponigro (2003), Donati (2006), and Cecchetto and Donati (2011, 2015); I propose that *Wh-NP is a consequence of FRs being derived from question-like Wh clauses by Head Movement of the Wh word to a higher D head. This movement is motivated by a requirement that definite D always be overtly realized; *Wh-NP arises because Left Branch movement is barred in English. I describe this in terms of a feature [±Overt].

Further, I show that while the basic form of FRs is acquired early by children (Flynn and Lust 1980, Guasti and Shlonsky 1995), experimental given here data shows that the syntax deriving *Wh-NP is developed considerably later. I explain this in terms of devel- opment of the functional structure of definite descriptions, in particular requirements on overt Determiners. I compare this with the development of other types of nominals in English. I then discuss recent insights on the nature of Parameter Setting by Bieberauer and Roberts (2012) and describe ways in which a simple Acquisition Device can deal with learning FRs by generalization of features of the nominal domain.

Finally, I discuss the issue of Wh-ever FRs, which do not show *Wh-NP. I give experimental evidence from adults and children, as well as naturalistic data, that this is because Wh-ever FRs are not definite descriptions, and so do not require the movement of the Wh word to a higher D head.