The present study examines the effects of code-switching (CS) attitudes in Acceptability Judgment Tasks (AJTs) among early Spanish/English bilinguals in the United States. In doing so, we explore whether negative attitudes towards CS result in lower/degraded ratings, and, likewise, whether positive attitudes result in higher acceptability ratings. Fifty Spanish/English bilinguals completed a survey that comprised a linguistic background questionnaire and a set of monolingual and code-switched sentences featuring two sets of stimuli, pro-drop (Sande, 2015) and pronouns (Koronkiewicz, 2014), that they rated on a 1–7 Likert scale; additionally, the survey included a final component that gathered information about the speakers’ attitudes towards CS. The pro-drop and pronouns code-switched stimuli gave rise to a total of four conditions. Results from a Linear Mixed Model revealed that all participants, regardless of attitude, distinguished between all Conditions. Furthermore, an effect for attitude was found for two of the conditions, such that the more positive the attitude, the higher the rating given on the AJT. In fact, these two conditions were composed of the CS structures that were rated higher by participants in Sande (2015) and Koronkiewicz (2014). No effect for attitude was found for CS structures that were rated low in the original studies. Thus, this investigation suggests that the attitudes that bilingual speakers have towards CS play a role in the ratings that they provide in AJTs, but in a manner that highlights, rather than obscures, the rule-governed nature of CS.

This study examines whether the foreign language effect mitigates reactions to value-inconsistent sociopolitical content. We examined 69 English–Spanish bilinguals and 31 Spanish–English heritage bilinguals, half of whom did the experiment in their native language and half in their second language. Participants were administered a survey in which trial emotiveness was manipulated by using the quantifiers some and all (e.g., Some Trump supporters are racists vs. All Trump supporters are racists). The some-types (n = 30) served as a baseline for the all-types (n = 30). After each target, participants rated their willingness to be prosocial (e.g., holding the door for a stranger) on a scale of 1–7, 1 being totally agree and 7 being totally disagree. Our results suggest that processing emotional information in a second language is less emotional than in a first language and that such a decrease in emotionality results in the neutralization of offense taken. However, individual differences in linguistic profiles across participants, as well as contextual framing, lead to discrete value judgments. Proficiency, learner type, political affiliation, and context type affect willingness to engage in prosocial behavior. As a group, the bilinguals showed no decrease in their willingness to engage in such behaviors, regardless of context type; speakers of higher proficiency and stronger political values increase prosocial sentiment; and lower proficiency and weaker views lead to neutral prosocial sentiment.

No abstract