Gender assignment: Monolingual constraints contribute to a bilingual outcome

Abstract

Aims and objectives: This study explores the well-researched topic of gender assignment to English nouns in Spanish discourse through a usage-based framework. The goal is to elucidate the relative impact of both previously studied and novel constraints on the variable application of feminine determiners. Methodology: A variationist analysis of English nouns surrounded by Spanish discourse in the spontaneous speech of bilinguals. Data and analysis: Data come from the New Mexico Spanish–English Bilingual Corpus. Tokens ( N = 707) were coded for independent variables and submitted to a logistic regression. The goodness of fit was determined via the area under the receiver operating characteristic (ROC) curve method. Findings: All independent variables were selected as significant by the logistic regression model. Based on factor weight ranges, the hierarchy of constraints is the following, from the most to the least impactful: Analogical Gender, Phonological Shape, Syntactic Role, and Determiner Definiteness. These results suggest that bilinguals utilize a variety of constraints in gender assignment, as opposed to a single default strategy. Originality: While previous studies have tested and found similar results for constraints such as analogical gender and phonological shape, none have offered a unified analysis explaining findings from a usage-based approach. The originality and utility of this approach is most apparent in the discussions of prototypicality and schematicity. Significance/implications: A corpus-based approach and usage-based theory is shown to bring new insight to a topic of interest in many other linguistic sub-fields. The discussion reinterprets previous conclusions about gender assignment using a framework not proposed in previous research, despite similar overall results.