Language Proficiency and Code Switching in English–Spanish Autistic Bilingual Speakers

PURPOSE: Evidence that bilingual exposure does not negatively affect language outcomes in autistic individuals has expanded over the past 15 years. However, research on bilingualism in autism has been limited to focusing on one language or each language separately, which does not capture natural bilingual dynamics such as code switching. This study examined how English-Spanish autistic speakers use their languages to communicate in semistructured settings. METHOD: Thirty English-Spanish autistic bilingual individuals (age range: 5.33-42.33 years) participated in two semistructured interactions with either a caregiver or a bilingual examiner. In each context, participants were instructed to speak either English or Spanish. Video recordings were transcribed and coded for code switching. Language proficiency in each language was measured using mean length of utterance in morphemes and number of different words. From the transcripts, we extracted rates of target language utterances (matching the instructed language), intersentential code switching (using the noncontextual language), and intrasentential code switching (mixing languages within an utterance). RESULTS: Participants were, on average, more proficient in English than in Spanish. Code switching occurred primarily when participants were instructed to speak Spanish, with English-dominant individuals and those who were less proficient in Spanish switching back to English more frequently. Code switching was also more frequent during interactions with a familiar caregiver than an unfamiliar examiner. CONCLUSIONS: The language of English-Spanish autistic individuals reflects expected patterns of bilingual use. They showed greater proficiency in their societal than in their heritage language and seemed to rely on code switching primarily to compensate for limited proficiency in their nondominant Spanish. These findings add to the growing body of evidence that bilingualism should not be viewed as detrimental in autism.