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

Hispanic Literatures & Linguistics

Year Degree Awarded


Month Degree Awarded


First Advisor

Luiz Amaral

Second Advisor

Marcus Maia

Third Advisor

Meghan Armstrong-Abrami

Fourth Advisor

Jeremy Hartman

Subject Categories

First and Second Language Acquisition


Grammatical gender is an inherent lexical property of nouns that categorizes them into two or more classes. Spanish and Portuguese have a binary gender system in which all nouns are masculine or feminine; this, along with number produces morphosyntactic agreement relationships between nouns, determiners, and adjectives. Conversely, when it comes to morphosyntactic agreement, English only produces agreement for number. The feature distinctions between Spanish and Portuguese on the one hand and English on the other can be illustrated using type hierarchies in HPSG, where the gender feature in Spanish and Portuguese has the same distribution in the hierarchy; however, the gender feature in English is limited to animate referential contexts. The aim of this dissertation is to analyze how L1 and L2 Spanish speakers assign, retain, and process novel noun gender taking into account their L1 typology. L1 Spanish speakers, L1 BP speakers, and L1 English speakers participated in three experimental tasks that manipulate novel noun gender and morphophonological shape. The first task presents speakers with 18 short stories, introducing two of the same novel item, differing along a single attribute, indicated by a gender-inflected adjective. Participants respond to a question about each story, necessarily producing the nonce noun and adjective. The second task is a description task after every six stories to investigate participants' gender retention. The third task investigates processing with a Self-Paced Reading paradigm where reading times are collected for nonce nouns and an anaphoric null nominals. The results indicate that all three speaker groups assign gender differently. L1 Spanish and L1 BP speakers rely most heavily on syntactic cues to assign gender, but L1 BP speakers rely more heavily on morphophonological cues than L1 Spanish speakers. L1 English speakers rely most heavily on morphophonological cues on the nonce noun. All speakers have more difficulty assigning feminine gender compared to masculine gender. This is taken to be due to the unmarked status of the masculine gender and suggests that Spanish gender feature values are [+/- fem] rather than masculine/feminine. These results also suggest that a theory of feature reassembly may more adequately describe the SLA process, accounting for prolonged instances of non-target optionality.