Bilingual individuals seem to easily speak in just one language, and switch back and forth between languages, suggesting they have powerful mechanisms for controlling activation of their two languages. A prominent theory suggests that cognitive control, and specifically inhibition of the non-target language, enables successful switching. We used behavioral and neuroimaging methods to study Spanish-English bilinguals to determine: 1) if college-aged bilinguals show an advantage in general task-switching ability relative to monolinguals given bilinguals’ extensive practice with language switching; 2) if cognitive control regions are recruited in bilingual language comprehension, and 3) if an aging deficit in inhibitory control affects older bilinguals’ (age 65+) ability to switch languages. In Study 1 (n = 80 per group; Stasenko et al., 2017) bilinguals exhibited more efficient task-switching, but only when participants had longer preparation time, and the advantage dissipated quickly. These findings suggest that although bilingualism improves the efficiency of task switching, this advantage might be more related to preparing to switch than to switching per se. In Study 2 (n = 24; Stasenko et al., 2020), bilinguals recruited fronto-parietal brain regions (i.e., right frontal inferior gyrus, bilateral middle frontal gyrus, and left supramarginal gyrus) when switching relative to not switching languages even in silent reading of mixed language paragraphs (without producing any switches in their speech). These results suggest that although reading comprehension seems to be passive, it recruits brain regions known to support cognitive control, possibly reflecting a modality-general switch mechanism. Study 3 (ns = 48 and 25; Stasenko et al., submitted) revealed a reversal of language dominance in mixed-language testing blocks, and a transfer of inhibition from a repeated set of items to a new set of items (that was introduced halfway through the task). Both effects were found only in younger but not in older bilinguals. Overall, these findings support the role of domain-general cognitive control and inhibition as an important mechanism in bilingual language control that spans across production and comprehension and exhibits decline in healthy aging.

Objective

Method

Results

Conclusions

Three sets of studies explore lexical regulation in bi- and trilinguals. Chapter 1 examines the foreign language effect (disproportionate interference between non-native languages) by conducting two experiments in which Dutch-English-French trilinguals monitor phonemes in picture names. Results show evidence of a foreign language effect in this task, and further posit that the possibility that such a phenomenon is driven by language of instruction (the language from which a bilingual learns a third language). Chapter 2 explores this theory with two experiments where Spanish-English bilinguals learned Hebrew from one of their two languages before performing a language switching task between these languages. Results suggest the presence of a language of instruction effect in this task and further explore the mechanics that drive it. Finally, Chapter 3 explores lexical regulation among known languages in a picture word interference task. Spanish-English bilinguals named pictures that had distractor words superimposed. These two experiments show that task strategies are inadequate in explaining translation facilitation effects of this nature, and reveal the translation facilitation effects to be highly robust. Taken together, these three sets of studies establish a new explanation for the acquisition of lexical regulation mechanisms (language of instruction), and explore the nature of bilingual control mechanisms in current theories of bilingual lexical access.

Different languages have different sentence structures—i.e., rules and information that guide the assembly of words into sentences. How are the sentence structures from two languages with very different word orders organized in a bilingual’s mind? Chapter 1 aimed to disentangle whether structural representations are shared or separate-and-connected by using cumulative cross-language structural priming, which does not involve frequent language switching (unlike standard cross-language structural priming). Contra the rich evidence from standard cross-language structure priming, results from Chapter 1 suggest separate-and-connected representations of sentence structures from two languages with different word orders. By measuring production time of each word in a phrase using an extended picture-word interference paradigm, Chapter 2 examined whether bilinguals access sentence structures from both languages even when only speaking one. Results suggest language-specific structural activation for phrases that have different linear word order across languages, even when there are frequent language switches. Finally, using acceptability judgment and memory-recall paradigms, Chapter 3 investigated whether the word order information from a second language (L2) influences the representation and use of the first language (L1). Results suggest that L1 structural representations can change after L2 immersion, but not such that L1 sentences resemble L2 structures. Instead, the L2 immersion seems to be associated with “noisier” L1 representations. Together, this dissertation demonstrates the mostly separate structural representations of two languages that use different word orders.