Does it matter how you present it? Modality, formality, and speaker type in Spanish-English CS acceptability judgments

Co-authored paper presented at the UIC Bilingualism Forum, Chicago, Illinois. [PDF]

This study examines whether the way code-switched sentences are presented in acceptability judgment tasks influences how bilinguals rate them, a question that matters given that such tasks typically rely on written, formal, decontextualized stimuli even though code-switching is often associated with informal speech. Thirty-five early Spanish-English bilinguals rated sentences with grammatical and ungrammatical switches presented in four formats, written in a formal or informal style and spoken aloud by either a heritage bilingual or a late L2 speaker. Participants reliably distinguished grammatical from ungrammatical switches across all formats, with no effects of modality or formality and only a limited, structure-specific effect of speaker type. These null results suggest that bilinguals’ judgments of code-switching are robust to presentation format, giving researchers considerable flexibility in how they design such tasks.

Proficiency, experience, and dominance as predictors of code-switching grammaticality sensitivity in late L2 bilinguals

Co-authored paper presented at the Second Language Research Forum (SLRF), Montréal, Canada. [PDF]

This study examines whether late L2 bilinguals acquire structural constraints on Spanish-English code-switching, such as the restriction against switching between a subject pronoun and its verb or between an auxiliary and its participle. Over 200 L1-English learners of Spanish completed either an acceptability judgment task or an elicited production task, along with measures of L2 proficiency, code-switching experience, and language dominance. All three factors predicted sensitivity to the constraints on their own, but when considered together, code-switching experience showed the most consistent effect across tasks, proficiency was the strongest predictor of acceptability judgments, and dominance was largely subsumed by the other two. The findings show that late L2 bilinguals can acquire these constraints and that how this knowledge surfaces depends on both the task and the linguistic experience each bilingual brings.

Task complexity effects on writing processes and correlations with text features

Co-authored article published in the International Journal of Applied Linguistics.

Research on task complexity (TC) effects on second language (L2) text features is abundant and has contributed critical insights. Comparatively, less is known about how L2 writers respond cognitively to diverse TC conditions or about potential effects of such processing activity on the characteristics of the texts produced. To expand on previous research, this study examined (i) the effects of TC on L2 writers’ processing activity (operationalized as on-line writing behaviors and macro writing processes) globally and across three time intervals; and (ii) any potential relationship of processing activity with text features (as measured by complexity, accuracy, and fluency indices). In an intra-subject design study, 45 advanced L2 English users completed two argumentative writing tasks (in time-compressed conditions) that varied in cognitive complexity, operationally defined as the provision of ±content support. Processing activity was captured with the keystroke logging software Inputlog, together with a self-rating post-task questionnaire. Results provided little evidence of TC effects on online writing behaviors but showed clearer differences in self-reported macro-writing processes, especially in the complex task condition. The analyses also indicated little evidence of TC effects on accuracy and syntactic complexity, together with selective associations between writing processes and text features across conditions.

*She parle beaucoup de langues: The acquisition of code-switching constraints in L2 French

Co-authored article published in the International Journal of Applied Linguistics.

The present study investigates the acquisition of code-switching (CS) constraints in late L2 bilinguals in relation to their L2 proficiency and their experience with CS. L1-English speakers who are acquiring L2-French as adults completed an acceptability judgement task or an elicited production task to reveal whether they accepted or produced intra-sentential CS in two syntactic positions where previous research has deemed it unlikely for bilinguals to code-switch: PRONOUN-VP and AUXILIARY-VP. Acceptability data were analyzed using cumulative link mixed-effects models, and production data were analyzed using mixed-effects modeling. Regarding acceptability, despite little exposure to English-French CS, the participants in the present study demonstrated having broadly acquired these two CS constraints at all levels of proficiency. Moreover, as L2 proficiency increased, sensitivity to the constraints against AUXILIARY-VP and PRONOUN-VP switches increased, suggesting that proficiency plays a crucial role in acquiring these constraints. Regarding elicitation, neither L2 proficiency nor CS experience was a significant predictor in constraint production, suggesting that this task measures a different part of bilingual competence. More broadly, the findings inform models of bilingual grammatical competence and have implications for bilingual education and assessment, supporting the view that CS is an integrated component of multilingual development.

Task complexity effects on L2 writing processes and correlations with text features

Co-authored paper presented at the American Association for Applied Linguistics (AAAL) Conference, Chicago, Illinois. [PDF]

This study examines whether presentation variables influence bilinguals’ acceptability judgments of intrasentential code-switching. Using a within-participant Acceptability Judgment Task (AJT), it tests the effects of modality (written vs. oral), writing style (formal vs. informal), and speaker type (heritage bilingual vs. late L2 speaker) on judgments of grammatical and ungrammatical code-switched sentences. The participants are early Spanish-English bilinguals from New York who evaluate the same structures across four presentation formats. Analyses show no reliable effects of modality, formality, or speaker background on acceptability ratings. The findings suggest that, within AJTs, code-switching grammatical intuitions can remain stable across varied presentation styles, supporting the robustness of this method while leaving open questions about ecological validity beyond the task context.

Stable intuitions across presentation styles: Effects of modality, formality, and speaker type on code-switching judgments

Co-authored paper presented at the Bilingualism in the Hispanic and Lusophone World Conference, Reading, United Kingdom. [PDF]

This study examines whether presentation variables influence bilinguals’ acceptability judgments of intrasentential code-switching. Using a within-participant Acceptability Judgment Task (AJT), it tests the effects of modality (written vs. oral), writing style (formal vs. informal), and speaker type (heritage bilingual vs. late L2 speaker) on judgments of grammatical and ungrammatical code-switched sentences. The participants are early Spanish-English bilinguals from New York who evaluate the same structures across four presentation formats. Analyses show no reliable effects of modality, formality, or speaker background on acceptability ratings. The findings suggest that, within AJTs, code-switching grammatical intuitions can remain stable across varied presentation styles, supporting the robustness of this method while leaving open questions about ecological validity beyond the task context.

Adverb position in heritage and L2 Spanish-English code-switching

Article published in Borealis: An International Journal of Hispanic Linguistics.

This study examines adverbs in intraclausal code-switching among late L2 Spanish bilinguals and heritage speakers of Spanish. Spanish allows adverbs pre-verbally and post-verbally, while English restricts adverbs to pre-verbal position. Previous research has shown that the language of the adverb predicts acceptability in code-switching contexts, with English adverbs allowed only pre-verbally and Spanish adverbs allowed in both positions. This study extends this work to L2 bilinguals, an understudied group in terms of code-switching acquisition. Using a written acceptability judgment task, 73 bilinguals rated code-switching stimuli involving verb-adverb switches, as well as monolingual versions of those sentences. Results showed that all groups preferred pre-verbal adverbs in English, regardless of whether the sentence was switched or completely in English. As for Spanish, participants accepted complement-type adverbs in either position, while specifier-type adverbs patterned like English, being accepted only preverbally. Crucially, there were no significant differences found between the heritage speakers and the two bilingual groups, which varied in terms of L2 Spanish proficiency. These findings suggest that the language of the adverb drives acceptability, and that all bilingual groups follow similar structural constraints. Thus, the acquisition context (heritage vs. L2) does not significantly impact adverb positioning in Spanish-English code-switching.

Code-switching acceptability in Spanish-English bilinguals: The impact of language environment

Paper presented at the Cognitive Linguistics in the Year 2025 Conference, Kraków, Poland. [PDF]

This study examines how Spanish–English bilinguals judge the acceptability of code-switching, with a focus on whether the order of language acquisition influences sensitivity to structural constraints. Using an acceptability judgment task, the study compares heritage speakers, L1-English L2-Spanish bilinguals, and L1-Spanish L2-English bilinguals on their evaluations of pronoun-switch constructions. Results show that while all groups distinguish between grammatical and ungrammatical switches, late bilinguals differ: those who learned English after Spanish were less likely to reject ungrammatical pronoun switches than those who learned Spanish later. Additional mixed-effects modeling demonstrates that both participant group and bilingual proficiency interact to shape acceptability patterns. Overall, the findings suggest that acquisition order may play a meaningful role in how late bilinguals internalize code-switching constraints.

Acquiring inalienable possession in Spanish: A study of heritage and L2 bilinguals

Paper presented at the Congreso Internacional Nebrija en Lingüística Aplicada a la Enseñanza de Lenguas, Madrid, Spain. [PDF]

This study examined how different groups of Spanish-English bilinguals—heritage speakers, L1-English L2-Spanish learners, and L1-Spanish L2-English bilinguals—judge inalienable possession constructions in Spanish. The results show that all three bilingual groups most consistently rated the canonical (clitic + definite determiner) form as acceptable, while hybrid and English-like forms were judged more variably—especially by heritage and L2 Spanish speakers, who were more flexible than the L1-Spanish group. Across all participants, variables like Spanish proficiency, dominance, and language exposure correlated with acceptability, but subgroup analyses showed these effects were not uniform: many correlations disappeared or shifted when groups were analyzed separately, revealing distinct within-group patterns, particularly among heritage speakers. These results suggest that bilingual type crucially shapes how language background factors influence judgments of inalienable possession, underscoring the importance of analyzing bilingual groups separately rather than assuming a single bilingual profile.

Task complexity effects on L2 writing pausing and revision processes

Co-authored paper presented at the Conference on European Second Language Acquisition (EuroSLA), Tromsø, Norway.

This study investigated how task complexity, manipulated through the presence or absence of content support, influenced L2 writers’ pausing and revision behaviors as well as their perceptions of task difficulty and engagement in writing processes. While keystroke logging data revealed no significant differences in pausing or revision behavior between simple and complex tasks, participants in the complex condition reported spending more time planning and structuring their writing and revising language features afterward. The study offers new empirical insights into task complexity effects on writing and highlights methodological considerations for aligning process data with writers’ self-reported experiences.