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

Brian Dillon

Second Advisor

Lyn Frazier

Third Advisor

Adrian Staub

Fourth Advisor

Marcel den Dikken

Fifth Advisor

Kyle Johnson

Subject Categories

Psycholinguistics and Neurolinguistics | Semantics and Pragmatics | Syntax


This dissertation answers a deceivingly simple question: why can her in Hermione talked about her refer to the sentence subject in Romanian, but not in English? The Romanian facts, which are surprising for both classic and competition-based accounts of the Binding Theory over the last 40 odd years, bring us to the following overarching question: what are the constraints on pronominal reference? To address these main questions, I carry out a psycholinguistic investigation of Romanian pronouns and argue that the distribution and interpretation of pronominal forms is jointly determined by pragmatic and morphosyntactic constraints.

I discuss evidence from four experiments, two on language production and two on language comprehension, which focus on intra-sentential pronominal reference. I provide an overview of previous Binding Theory accounts, both classic and competition derived, and, based on English data, I define the terms of competition of pronominal forms and then show how a pragmatic constraint I propose, BE CLEAR!, can account for Condition B effects for both English pronouns, as well as for Romanian clitic pronouns.

The two production experiments in Romanian bridge the theoretical and psycholinguistic literatures by extending ambiguity avoidance studies to the intra-sentential domain. These experiments also provide evidence that BE CLEAR! is active in both coreference and bound variable contexts, in contradiction to previous seminal accounts in the Binding Theory literature. Lastly, based on the production data, I refine the proposal and introduce the syntactic economy constraint BE SMALL!.

The data from the comprehension studies show that complex reflexives like and regular pronouns do compete, contrary to existing assumptions of syntactic economy based competition accounts of Condition B effects. Furthermore, the comprehension data also serves as evidence against accounts which claim that there is a processing advantage of bound variable logical forms, as well as evidence against purely pragmatic accounts of disjoint reference.

Instead of divorcing pragmatic and syntactic competition based approaches, which is often the case in the literature, I propose, by virtue of the experimental data, that pragmatic and morphosyntactic considerations are both responsible for what we typically refer to as Condition B effects.